ampa safety report
TRANSCRIPT
A Safety Review and RiskAssessmentin Air MedicalTransport
Supplement to the Air MedicalPhysician Handbook
November, 2002
A Safety Review and RiskAssessmentin Air MedicalTransport
November, 2002
To the Air Medical Community,
Safety is a primary concern for all medical transport professionals. Regardless ofwhether transporting in helicopters, ground ambulances, or fixed-wing aircraft,or if transporting from scenes to hospitals, between hospitals or within a hospi-tal, the movement of patients raises both the risk of medical complications andthe risk of transportation accidents and incidents. A great deal of attention hasbeen focused on helicopter incidents and accidents. While improvements inequipment, protocols, training, and operations have occurred, air medical heli-copter transport is not as safe as we would like.
In the attached safety report, Dr. Ira Blumen and his team from UCAN providea fresh perspective on medical helicopter safety statistics and an assessment ofrisk. They look at the existing data from a variety of viewpoints and haveundertaken a unique research project to fill in many of the statistical gaps thatpreviously existed. They compare Helicopter EMS with other aviation data,other routine risks, and provide a basis for future work.
The authors are to be commended for this significant effort. AMPA is proud tomake this report available to everyone in the air medical community, and theAMPA Board thanks the many contributors who made this possible.
All of us associated with air medical care should read this report, discuss it withtheir peers, friends and family, and then read it again. Then they should go towork motivated to improve safety even further. When the second edition of thisreport is produced, wouldn’t it be wonderful to demonstrate lasting improve-ments in safety, for our community and our patients?
Sincerely,
Kenneth A. Williams, MD, FACEP Harry E. Sibold, MD, FACEPPresident, AMPA President, AMPA2000-2002 2002-2004
A I R M E D I C A L P H Y S I C I A N A S S O C I A T I O N
383 F Street
Salt Lake City, Utah 84103
voice: 801-534-0829
fax: 801-534-0434
www.ampa.org
A Safety Review and Risk Assessment
in Air Medical Transport
Ira J. Blumen, MDand the
UCAN Safety Committee
Supplement to the Air Medical Physician Handbook
November, 2002
AIR MEDICAL PHYSICIAN HANDBOOK
Air Medical Physician Handbook
2002
A publication of:
Air Medical Physician Association
Copyright © 2002 by Air Medical Physician Association.
Reprints Are Encouraged. In the interest of safety, this Supplement to the Air Medical Physician Handbook may be repro-duced in whole or in part. All uses must credit the authors and the Air Medical Physician Handbook.
The research, opinions, and suggestions expressed in this publication belong to the authors and are not necessarily endorsed byAMPA or by the supporting organizations, corporations or companies.
Cover design by: Erin Strother, Studio E Design, Escondido, CA
Mailing Address:383 "F" Street
Salt Lake City, Utah 84103
Telephone: 801-534-0829Fax: 801-534-0434
www.ampa.org
TABLE OF CONTENTS
Executive Summary ........................................................................................................................................iIra J. Blumen, Jeffrey Coto, Charles L. Maddow, Michael Casner, Craig Felty, Karen Arndt
A Safety Review and Risk Assessment in Air Medical TransportIra J. Blumen and the UCAN Safety Committee
INTRODUCTION ............................................................................................................................................................1
SECTION 1: AIR MEDICAL ACCIDENTS AND INCIDENTS ................................................................................2An Overview of HEMS Accidents: 1972 to 2002 ....................................................................................................3Air Medical Accidents: A 20–Year Review ..............................................................................................................4National Transportation Safety Board Safety Study: Commercial Emergency Medical Service Helicopter Operations – 1988 ..................................................................................................................................10Flight Safety Foundation: Human Error as Major Cause of U.S. Commercial EMS Helicopter Accidents ........11Emergency Medical Service Helicopters Incidents Reported to the Aviation Safety Reporting System ..............12Air Medical Accident Analysis ................................................................................................................................15HEMS Accidents and Incidents, 1998 to 2001 ......................................................................................................18HEMS Accident and Fatal Accident Rates ............................................................................................................18
SECTION 2: A COMPARISON OF HEMS TO OTHER TYPES OF AVIATION....................................................24Helicopter EMS vs. All Helicopter Accident Data: 1990–2000 ............................................................................24Factors Related to Occupant Crash Survival in EMS Helicopters..........................................................................28Helicopter Accident Analysis Team ........................................................................................................................29HEMS vs. Other Aviation Operations ....................................................................................................................31Single- vs. Twin-Engine Helicopter Accident Rates................................................................................................37Helicopter Accidents, 2001 ......................................................................................................................................38
SECTION 3: A COMPARISON OF RISK....................................................................................................................38HEMS: Analyzing the Population at Risk ..............................................................................................................39Comparing HEMS to Other Risks ..........................................................................................................................41
SECTION 4: SAFETY and RISK MANAGEMENT in HEMS....................................................................................47Principles of a Safety and Risk Management Program............................................................................................47Risk Management ....................................................................................................................................................48Air Medical Resource Management ........................................................................................................................51
SECTION 5: CONCLUSION ........................................................................................................................................53
Attachment 1: HEMS Accident Summary, 1998–2002 ................................................................................................55
Editorials:Risk Management, Benefit and Cost
Ed MacDonald ......................................................................................................................................................64Can We Have Benefits Without Risk?
Thomas Judge, CCT-P ..........................................................................................................................................68Angels of Mercy or Angles of Death
Michelle North, Ph.D. ............................................................................................................................................69
ISpecial Supplement
EXECUTIVE SUMMARY
INTRODUCTION
In 1998, helicopter EMS (HEMS) saw the beginning of an alarming accidenttrend. That year there were eight accidents, followed by ten in 1999, and twelve in2000. That was a significant increase considering there had been only one accidentin 1996 and in 1997 there were a total of three. Questions were raised—and theindustry, as well as individual programs, began to look for answers. Is HEMSunsafe? Are HEMS crewmembers at a significantly higher risk for injury or death?Is the information available to answer these questions?
In general, the lack of readily available data made it difficult, if not impossible, toanswer these questions adequately. This prompted the University of ChicagoAeromedical Network (UCAN) Safety Committee to begin its own investigationand research in the fall of 2000.
There are many ways to review the safety of air medical transport and to assessthe relative risk to its crewmembers. For more than a decade, when our industry orindustry observers (e.g., FAA, media, general public) looked at HEMS accidentinformation, the only available information was the number of accidents and fatali-ties—commonly referred to as the raw data. No one was able to determine if theincrease in accidents was simply related to an increase in the number of hoursflown, or whether HEMS had indeed become more dangerous.
Our primary investigation began with an extensive review of accident and inci-dent data specific to the HEMS industry. Based on this review, a comprehensiveresearch model was developed to estimate and project exposure data that the airmedical industry has stated “does not exist”. Without this exposure data (numberof transports and/or the number of hours flown by HEMS), it would be impossibleto calculate annual HEMS accident rates and fatality rates or to draw any meaning-ful conclusions or comparisons.
The study focus next shifted to a comparison of HEMS accident rates and fatalityrates to other forms of air travel. Finally, a comparison of air medical transport toother occupations or “routine” risks is made to contrast the fatality rates and theodds of death. To accomplish this, the “population at risk” in HEMS would need tobe determined if we were to attempt to make these unique comparisons. This datahas never been tracked or even estimated in the HEMS literature.
Upon analyzing HEMS-related risks and assessing the risks faced by air medicalcrewmembers, this report also identified ways to enhance program (and industry)safety with the hope of reducing our level of accident exposure. Every occupationhas inherent risks. Medical professionals and transportation professionals are nodifferent and are exposed to various risks every day.
AIR MEDICAL PHYSICIAN HANDBOOK
Grant support to fund the publication of this report has been received in part from: Foundation for Air MedicalResearch, Bell Helicopter Textron; CJ Systems Aviation Group; American Eurocopter; Fitch and Associates;Association of Air Medical Services (AAMS); National Flight Paramedic Association (NFPA); National EMSPilots Association (NEMSPA); Air Methods; Sikorsky; Golden Hour Data Systems; Air and Surface TransportNurses Association (ASTNA); Petroleum Helicopters, Inc. (PHI); Metro Aviation; Turbomeca EngineCorporation; Heli-Dyne Systems; Agusta Aerospace Corporation; University of Chicago Aeromedical Network(UCAN); Illinois Association of Air and Critical Care Transport (IAACCT); Keystone Helicopter; InnovativeEngineering/AeroMed Software; and National Association of Air Medical Communication Specialists (NAACS).
This research was supported in part througha grant from the Foundation for AirMedical Research and by the University ofChicago Aeromedical Network (UCAN).
A SAFETY
REVIEW AND
RISK ASSESSMENT
IN AIR MEDICAL
TRANSPORT
Ira J. Blumen, MD
Jeffrey Coto, RN
Charles L. Maddow, MD
Michael Casner, MD
Craig Felty, RN
Karen Arndt, RN
Geoff Scott, RN
Copyright © 2002Ira J. Blumen and Air Medical Physician Association
II AIR MEDICAL PHYSICIAN HANDBOOK
Being safe does not eliminate risk—itreduces it. This paper provides valuableinformation regarding risk that will hope-fully help each and every air medical pro-gram and air medical professional betterunderstand their day-to-day risk. Moreimportantly, they will have informationthat should help guide the coordination,implementation and evolution of a safetyand risk management program, therebyreducing the number of accidents andenhancing survivability.
HEMS DATABASE
This investigation began with a com-prehensive analysis of various databasesto identify individual HEMS accidentsand their corresponding fatalities andinjuries. Our research model thenenabled us to estimate flight hours,patients flown, and other variables forevaluation. Since 1972, HEMS hasflown an estimated 3.0 million hourswhile transporting approximately 2.75million patients. In 31 years (throughSeptember 2002) there have been 162accidents involving dedicated medicalhelicopters and four accidents involvingdual-purpose helicopters in the UnitedStates. There have been 67 fatal acci-dents with 183 fatalities, including 144crewmembers. Since 1998, there havebeen a total of 50 accidents—nearly 31%of all the accidents we have experiencedover three decades.
In the early and mid-1980s, during theHEMS industry’s most rapid growth, weexperienced an alarming number of acci-dents. From 1980-1987, there were 54accidents, averaging 7.7 accidents peryear. The late 1980s to mid-1990sshowed considerable improvement. From1987-1997, dedicated HEMS averaged4.9 accidents each year. From 1998-2001, however, we have seen that aver-age more than double to 10.75 accidentsper year. Not since the four-year periodof 1983 to 1986 have we seen such alarge number of accidents.
The data is not uniformly grim, howev-er. Of the 44 accidents from 1998 to2001, 15 (34%) resulted in at least onefatality. For the 1980s, our accidentdatabase revealed 39% of all accidentsresulted in at least one fatal injury, whilefrom 1990-1997, that rate had increased
to 47%. Despite the increase in accidentsover the past four years, the percentageof fatal accidents has declined by nearly30% compared to the early 1980s. Inaddition, the last four years have seen thelowest consecutive 4-year fatality rate inHEMS history since the early 1980s.The percentage of fatal injuries has alsodecreased in the 1998-2001 accidentsand we see a higher percentage ofcrewmembers and passengers who sus-tained no injuries. The fact remains,nonetheless, that since 1990, there hasbeen an average of 2.5 fatal accidentsannually, taking the lives of 5 to 6crewmembers each year.
RECURRENT FACTORS INHEMS ACCIDENTS
This report summarizes several notablepublications, investigations and presenta-tions that analyze and identify key factorsrelated to HEMS accidents. Included inthese studies were evaluations regardingincident, operational, and human factorvariables. The analyses included theseverity of injury, when HEMS accidentsoccur, condition of light, phase of flight,purpose of flight, the cause of HEMSaccidents, pilot experience, aircraft dam-age and occurrence of post-impact fires.Two studies were also included that ana-lyzed the chain-of-events and problemsthat led to past accidents. These reportspropose interventions intended to pre-vent future accidents.
A disproportionate number of HEMSaccidents occurred during night opera-tions. While an estimated 38% of allHEMS flights were at night, 49% of theaccidents over 20 years (1978-1998)occurred during night operations. It wasalso found that more accidents occurredduring cruise (36%) than any otherphase of flight.
Scene transports also accounted for adisproportionate number of HEMS acci-dents. Since 1988, the percentage ofscene response flights has averaged 31%,while a total of 42% of the accidents dur-ing patient-related missions occurred onscene flights. If all missions had equalrisk, then 31% of the accidents shouldhave been on scene missions.
Pilot error was attributed as the director indirect cause of HEMS accidents
nearly three times more often thanmechanical failure. The studies we ana-lyzed have associated human factors with65-76% of the HEMS accidents studied.Of the fatal accidents studied, humanerror was associated with 84%. Humanerror was a factor in more than two-thirds of the en-route accidents, morethan 80% of the accidents during takeoff,and approximately 90% of the accidentsduring approach and landing.
In 1988, the NTSB concluded thatpoor weather poses the greatest singlehazard to EMS helicopter operations.Subsequent publications found that inthe 1980s, 22% of the accidents weredetermined to be weather-related. In theearly to mid-1990s, 32% were related toweather. Tragically, while the total num-ber of accidents went down, the percent-age of weather-related accidents hadincreased by 10%. Since 1998, however,it appears that this trend has dropped sig-nificantly to less than 15%. In addition,88% of the weather-related HEMS acci-dents occurred at night. Approximately75% of all weather-related HEMS acci-dents resulted in fatalities and nearlytwo-thirds had no survivors.
In-flight collision with an object(CWO) has occurred with 25 HEMSaccidents and there has been a dramaticincrease in the number of CWO acci-dents and incidents over the past fewyears.
More than 25% of the accidentsinvolving CWOs resulted in fatalities.Sixteen (64%) of these CWO accidentsoccurred during scene response missions.Once again, if all missions had equal risk,31% of the CWO accidents should havebeen on scene missions. Instead thatpercentage rate has more than doubled.It was also identified that more than 40%of all the approach and landing accidentsand 50% of all takeoff accidents wereCWOs. Unexpectedly, weather has notbeen identified as a factor in CWOaccidents.
Also unexpectedly, pilot fatigue andtotal hours of flight time do not appear tobe significant factors in HEMS accidents.Analysis of HEMS incidents suggests thatan IFR rating and currency may be pro-tective in overcoming situations andavoiding accidents. In addition, commu-nications problems, time pressures, anddistractions are frequently identified as
IIISpecial Supplement
contributing risk factors in HEMS acci-dents and incidents.
The magnitude of injuries and aircraftdamage is significant in HEMS accidents.HEMS accidents are more likely to resultin fatalities or serious injuries than otherhelicopter accidents. One study foundthat main cabin occupants in EMS heli-copters have nearly 4.5 times the risk ofserious injury (especially back injuriesand head injuries) or death in survivablecrashes when compared to a comparablepopulation of occupants in the maincabin of non-EMS air taxi helicopters.For front seat occupants, there was nosignificant difference in injury riskbetween the two groups. This seemed tosupport the author’s premise that EMSaircraft modifications, which are general-ly limited to the main cabin, were direct-ly associated with the risk of injury andmay contribute to occupant injury anddeath in otherwise survivable crashes.
Different studies reported the inci-dence of post-impact fires between 2-15%. Nearly half of the accidents result-ed in the destruction of the helicopter.No conclusions, however, can be maderegarding single- vs. twin-engine aircraft.
HEMS ACCIDENT ANDFATAL ACCIDENT RATES
An important aspect of this study wasthe determination of HEMS accidentand fatal accident rates for the definedstudy period (1980-2001). Calculationsare based upon estimated exposure data,which has been determined through sev-eral industry-wide surveys, various calcu-lations and several assumptions, whichare stated.
An extensive review of the air medicalliterature was conducted to determinewhat data was available. Total flighthours and total patients transported wereunavailable for more than 50% of theyears reviewed. For various years, pub-lished data included: average flight hoursper program, total patients transported,average patients transported per program,loaded miles, and average flight hours perpatient transport. Unfortunately, therewas no data regarding the number ofHEMS programs or dedicated helicoptersin operation since 1992.
Considering the available vs. unavail-
able information, it was concluded that ifthe number of programs and helicoptersin operation could be determined, thenthe total flight hours and patients trans-ported could be estimated for the year.Furthermore, an assumption was madethat the growth in the HEMS industryhas been fairly constant since 1992. Itwould then be possible to estimate thenumber of programs and helicopters forthe years lacking data.
To determine a fairly accurate numberof dedicated HEMS programs and heli-copters, several steps were performed.First, a state-by-state survey was postedon the Flightweb listserve. The resultinginformation was supplemented by theAAMS Membership Directory and the“Directory of Air Medical Programs”published in AirMed. To further supple-ment these results, information wasobtained from the helicopter manufactur-ers and a survey of five of the largestHEMS operators was conducted.
It was also determined that ourresearch model should include total flighthours in this evaluation. In HEMS, aconsiderable number of accidentsoccurred on non-patient missions—including public relations (PR) flights,refueling, maintenance, training, and soon. Personal correspondence with sever-al HEMS programs and operators suggeststhat non-patient flight time may rangefrom 5 to 15%. PR flights made upapproximately half of this for many pro-grams. In an effort to more accuratelydetermine the average number of totalflight hours per program, additionalinformation was obtained from the sur-veyed HEMS operators.
The survey results identified 231 dedi-cated HEMS programs, operating a totalof 377 helicopters (excluding back-up air-craft). Subsequent discussion with severalindustry leaders and evaluation of infor-mation from the aircraft manufacturerssuggests that, if anything, these numbersmay be slightly underestimated. The hel-icopter manufacturers estimated thatthere were 462 medical helicopters (dedi-cated and backup) in the United States,not including dual-purpose helicopters.With an average of one backup for every7.1 dedicated helicopters in the combinedoperators fleets, there are an estimated 53backup aircraft yielding a total of 430 hel-icopters. This represents a variation of
approximately 7% fewer aircraft com-pared to the number of helicopters fromthe manufacturers’ survey. This may alsoindicate a slight discrepancy in the num-ber of HEMS programs resulting from theInternet survey. In using the lower state-by-state survey results of 377 dedicatedhelicopters and 231 programs for our cal-culations, it is realized that the proposedexposure data (flight hours and number ofpatients transported) may be underesti-mated. As a result, the calculated acci-dent and fatality rates could be overstatedby an estimated 7-10%. This difference,however, does not impact the overalltrends identified in HEMS accidents northe comparison with other aviation operations.
The calculated results show a dramaticdecrease in the HEMS accident ratesince the mid-’80s. As the raw datawould predict, the accident rate since1998 has steadily increased. However,despite this increase, the rate remainsroughly one-third of what was experi-enced in the early to mid-1980s due tothe overall increase in flight hours.
Looking at an average accident rate for1992-2001, (3.78 accidents per 100,000flight hours), the average HEMS programflying 911 hours per year, would have oneaccident over 29.1 years of flight time. Ifa program flies less, the number of yearswould presumably increase, and if a pro-gram flies more, the time frame woulddecrease. Another way to propose thelikelihood of an accident would be tocompare the number of accidents to thenumber of programs (or helicopters).Again, using the most recent ten years,there has been an average of 7 accidentseach year. With 231 dedicated HEMSprograms, and assuming all things beingequal, we would find a similar predictionof one accident per program every 33years. If you base this comparison on thenumber of dedicated helicopters estimat-ed for 2001 (400) rather than programs,the margin now goes up to nearly 57.1years. Looking at the average number ofHEMS accidents (9) for the past fiveyears, we would expect one accident perhelicopter every 44.4 years.
The second comparison looks at thefatal accident rate per 100,000 flighthours. Here too, a dramatic improve-ment is identified since the early andmid-1980s. Our current rate, despite
IV AIR MEDICAL PHYSICIAN HANDBOOK
having gone up slightly over the past fewyears, is approximately 75% less than ourworst years. Once again if we take anaverage flight program and an averagefatal accident rate for the past ten years(1.38 fatal accidents per 100,000 flighthours), we would predict one fatal acci-dent while flying for over 79.3 years.When we focus on the average fatal acci-dent rate for 1997-2001 (1.69) this figuredrops to 64.8 years.
The final normalized comparison eval-uates the accident rate per 100,000patients transported.
With a 10-year (1992-2001) averageaccident rate of 3.89 accidents per100,000 patients transported and the typ-ical HEMS program flying 882 patientsin a year, a program would have one acci-dent while transporting an estimated25,700 patients over 29.2 years.Calculations for 1998-2001 (4.79 acci-dents per 100,000 patient transports)resulted in an estimate of one accidentwhile transporting nearly 21,000 patientsover a 23.74 year period.
A final annual comparison of HEMSaccidents considers the percentage ofHEMS programs and helicopters thathave sustained an accident. These calcu-lations do not take into account the pos-sibility that an individual program mayhave suffered more than one accident—which has occurred. Overall, the 22-yearaverage annual percentage calculates to5.8% of the programs having had acci-dents between 1980 through 2001. Ifone considers only the past five years, anaverage of 4.1% of the programs havehad an accident each year. In 1982, anestimated 16.3% of the HEMS programs(8 accidents, 49 programs) were involvedin accidents. The safest year was in1996, when an estimated 0.5 % of theprograms had an accident (1 accident,207 programs).
Calculating the percentage of helicop-ters that were involved in HEMS acci-dents each year finds a high of 12.9% ofthe HEMS aircraft in 1982 (8 accidents,62 helicopters). In 1996, there was 1HEMS accident during a year when anestimated 309 dedicated medical helicop-ters were in operation, for a total of0.3%. The average percentage over 21years calculates to 4.4% of the HEMSfleet. Over the past 5 years, this percent-age has averaged 2.5% for each year.
HEMS COMPARED TOOTHER AVIATIONOPERATIONS
Raw data and normalized statistics areavailable from the FAA/NTSB for thedifferent types of aviation operations.With the estimated accident and fatalityrates per 100,000 flight hours for HEMS,it is now possible to compare varioustypes of aviation in a more meaningfulway.
The accident rate for HEMS was dra-matically higher than for all other avia-tion operations during the early and mid-1980s. Beginning in 1987, we see a sharpdecline in the HEMS accident rate,which has remained consistently belowthe accident rates for both general avia-tion and all helicopter aviation. In addi-tion, from 1987 through 1997, theHEMS accident rate was lower than theoverall accident rate for all Part 135 non-scheduled flights 6 of the 10 years. Since1998, however, the HEMS accident ratehas surpassed that of the non-scheduledPart 135 operations each year
Comparing the average accident ratesfor the past 20 years (1982–1999), 10years, and 5 years for the five types ofaviation operations, helicopters, andHEMS yield some interesting findings.Even with the high accident rates of the1980s, the 20-year average for HEMS isbelow all helicopter operations and gen-eral aviation. For the 10-year average,the HEMS accident rate is less than 50%the rate of helicopters and general avia-tion. For the past 5 years, the averageaccident rate for HEMS has gone up, butremains significantly lower than all heli-copter operations and general aviation.
Initially, the fatality rate for air med-ical helicopters was equal to or dramati-cally higher than all other aviation oper-ations. In 1990, however, there were nofatal HEMS accidents. From 1992 to1997, HEMS was consistently below bothgeneral aviation and all helicopter opera-tions in fatal accidents. Since 1998, theHEMS fatality rate has been consistentlyhigher. Looking at the average fatal acci-dent rate for the past 20 years, 10 years,and 5 years for the various aviation oper-ations, HEMS has a higher fatal accidentrate than all other aviation during the20-year period. However, over the past10 years HEMS has averaged a lower
fatality rate than helicopters and generalaviation (Part 91). For the past 5 years,however, the average HEMS fatality rateonce again exceeded all other aviationoperations.
A COMPARISON OF RISK
In order to assess and compare risk, therelevant figure needed must be in theform of a ratio, fraction, or percentage.To arrive at these figures normalization ofthe data had to occur and we needed toknow two primary numbers. The numer-ator of the fraction tells us how manyindividuals doing a particular activitywere either injured or killed over a givenperiod of time. The denominator repre-sents how many people were engaged inthat activity—the population at risk. Byreducing all risks into ratios by utilizingthis format we can begin to compare dif-ferent types of activities and the relativerisks.
If one were to try to compare air med-ical transport to other occupations or“routine” risks to determine either theodds of death in one year or the fatalityrate per 100,000 (the most commoncomparison) we would need to know twothings. The first is the number of HEMScrew fatalities per year. The secondwould be the number of people engagedin HEMS transport (i.e., the number ofHEMS pilots and medical crewmembers)for each year. The number of fatalities isknown, but the number of crewmembersin HEMS has never been tracked.
For the purpose of this study, the aver-age number of crewmembers per helicop-ter is estimated to be 22 persons (4 pilots,6-8 nurses as the primary caregivers, and10-12 second medical crewmembers). In2001 there were an estimated 400 dedi-cated medical helicopters, producing anestimated population at risk of approxi-mately 8,792. Having estimated thenumber of helicopters for each year inthis study, the number of crewmemberscan be approximated for each year.
It is important to realize that in esti-mating exposure in this method, it doesso for the average crewmember and theaverage flight program. When the rawdata is normalized, it does not take intoaccount the amount of exposure for anindividual during the year. For this por-
VSpecial Supplement
tion of the study, calculations are basedon the average program, which in 2001transported approximately 882 patients,flying an estimated 957 hours over thecourse of the full year, and all crewmem-bers (pilots and medical crewmembers)flying an equal amount of time.
Fatality and Death Rates
Fatality statistics for HEMS personnelare presented in several different formats.In each case, the number of crew fatali-ties was determined for each year. To beconsistent with statistics from theNational Safety Council (NSC), the datawas normalized to produce a death rateper 100,000 population at risk for eachgiven year. Over the 21 years reviewedfor this portion of the study (1981-2001),the HEMS population has grown fromapproximately 858 to 8,792. While thisgrowth seems impressive, this is still avery small sampling to translate to a ratioper 100,000. With such a small popula-tion base, each fatality has a significantimpact on the fatality rate. In thisdesign, the range for the HEMScrewmembers fatality rate is from 0 to699 per 100,000. With such a widerange, a 22-year average was calculatedand used in the various comparisons.The average annual death rate over the22 years is 196 per 100,000 crewmembers.
Another relationship used to comparethe annual number of HEMS crew fatali-ties is in terms of “odds.” For example,in 2001there were only two crew deathsout of an estimated crew population of8,792. Looking solely at the numbers,the odds to an individual crewmembersuffering a fatal accident that year wouldbe considered to be 1 in 4,396.Contrasting this with what could be con-sidered our riskiest year (1980), therewere 6 crew fatalities out of an estimated858 crewmembers industry-wide. Thiswould correspond to fatality odds of 1 in143. Excluding 1990 when there wereno fatalities, the average odds per yearover the 22-year period are 1 in 1,158.
To further illustrate the risk related toHEMS transport, we can compare andcontrast the above numbers with otheractivities, other types of accidents, andother causes of death. Taking into con-sideration the wide range of fatality ratesand odds that we have estimated for each
year in HEMS, the calculated averagesare used in subsequent comparisons.
In 1998, the death rate per 100,000 forall accidental deaths was 36.2. Motorvehicle accidents were the highest in thiscategory with 16.1 deaths per 100,000individuals. When one considers theaverage annual death rate (192) forHEMS crewmembers over the 22 yearsreviewed, HEMS is surpassed only byheart disease (268.2 deaths per 100,000)and cancer (200.4 deaths per 100,000)when the data is normalized. The aver-age one–year odds of 1 in 1,158 forHEMS crewmembers is also higher thanthe one-year odds for all deaths due toinjury (1 in 1,796) and all accidentaldeaths (1 in 2,762).
A final comparison looks at the esti-mated exposure that will produce a 1 in1,000 risk of death.
In the 22-year HEMS study period anestimated 3,002,176 total hours havebeen flown. Adding together the esti-mated number of crewmembers each yearyields a total of 105,922. This corre-sponds to an average exposure of 28.3hours of flight time producing the esti-mated odds of 1 in 1,158. Adjusting theratio to 1 in 1,000, we get an averageexposure of 32.9 hours.
HEMS: The Risk to the Patient
There is some level of risk related toall aspects of healthcare. The results oftwo comprehensive studies have suggest-ed that between 44,000 and 98,000Americans die in hospitals each year as aresult of medical errors. Normalizingthese numbers produces a death ratebetween 131 and 292 per 100,000patients due to medical errors. TheNSC, however, provides statistics on“complications of surgery/medical care”that is based upon the reported numberof deaths compared to the entire U.S.population. Their finding of 1.2 deathsper 100,000 individuals is dramaticallydifferent.
While air medical transport is not amedical treatment and aviation accidentswould not be considered a medical error,some could argue that these accidentsrepresent an adverse event in the health-care environment. In our 22-year study,we estimate a total of 2,745,207 patients
have been flown by HEMS. Over thissame time period, 21 patients have losttheir lives in HEMS accidents. This cor-responds to a death rate of 0.76 per100,000 patients flown. Based upon thesefigures, it would appear that there is a fargreater risk to the patient of dying froman adverse event while hospitalized thanfrom an accident aboard a medical helicopter.
Occupational Risks: Deathsand Injuries in theWorkplace
The NSC reports an average death rateacross all industries at 3.8 per 100,000workers, with mining (21.2) and agricul-ture (22.5) having the highest rates.With a death rate of 192 per 100,000,the HEMS death rate is approximatelynine times greater than the riskiest indus-tries. However, it must be pointed outthat this comparison is greatly distortedwhen you consider the small HEMS“population.” Even in 2001, with ourlargest estimated number of crewmem-bers, 2 fatalities resulted in a death rateof 23 per 100,000.
RISK MANAGEMENT IN HEMS
A HEMS accident is not caused by asingle event, but by a chain of events. Inmost accidents, numerous risk factors canbe identified. Acting on any of theserisks and breaking the chain at any pointmay prevent an accident from occurring.Safety of flight requires the right attitudeand active participation. Every pilot,mechanic, medical crewmember, commu-nication specialist, and administratormust be fully knowledgeable of their roleand responsibilities. Each must be com-mitted to a safe operation and to ongoingrisk management. Nothing takes theplace of comprehensive training, profi-ciency, and sound judgment.
An important training componentshould be Air Medical ResourceManagement. The goal of AMRM is toimprove crew communications and inter-actions by addressing teamwork, commu-nication skills, decision making, work-load management, situational awareness,preparation and planning, cockpit dis-
VI AIR MEDICAL PHYSICIAN HANDBOOK
tractions, and stress management. The risks in HEMS should not be
underestimated. The cumulative effectsof multiple risk factors must be consid-ered when making decisions on each andevery transport. Risk management is amajor component of the decision-makingprocess. It relies on situational aware-ness, problem recognition, and exercisinggood judgment to reduce risks associatedwith each flight.
Accident prevention must be theobjective. But there will be “a next acci-dent.” The air medical community andeach program must be certain that theproper aircraft, systems, training, andequipment are in place to enhance thesurvivability of an accident. A study thatevaluated the use of helmets in surviv-able military crashes found that maincabin occupants with no helmet were at5 times the risk for a serious head injuryand 7.5 times the risk for a fatal headinjury than their helmeted counterparts.Another series of U.S. Army studiesfound that shoulder harnesses were an
important factor in reducing the inci-dence and severity of serious backinjuries. Yet, not every crewmemberwears a helmet, not every aircraft isequipped with shoulder harnesses andnot every person buckles their seat beltprior to takeoff.
CONCLUSION
Throughout all types of industry, thereis generally a 20/80 ratio when it comesto accidents: 20% of all accidents arecaused by machine-related problems and80% of the accidents are caused byhuman error. This seems to be accuratefor HEMS as well. The NTSB common-ly identifies “pilot error” as the probablecause in the majority of HEMS accidents.But is it truly pilot error? Or have we allfailed in some way?
There is no logical reason for theincrease in the number of accidents overthe past several years. We have regula-tions, we have safety committees, we
have standards, we have safety summits,we have AMRM, we have surveys andwe have reports. We have better aircraft,we have newer technology and we haveaccreditation. And we have 30 years ofexperience. What we also still have areunnecessary pressures, unnecessary risks,unnecessary distractions, poor communi-cations, complacency—and the same oldhuman errors. What we do not have isan excuse!
This report provides a comprehensivereview of past studies and offers consider-able new information regarding accidentrates and risks related to HEMS. Thisreport does not provide the solution to asafer industry. Only you can do that—atyour program, on your next flight, onevery flight. By providing new informa-tion and increasing awareness, we hopeto decrease future accidents, and to bet-ter educate flight program personnel ofthe potential risks at hand each andevery time the phone rings and the air-craft lifts off.
1Special Supplement
AIR MEDICAL PHYSICIAN HANDBOOK
A SAFETY REVIEW AND RISK ASSESSMENT IN AIR MEDICAL TRANSPORT
INTRODUCTIONFor nearly thirty years, safety has been a focus of committees, articles, lectures,
position statements, standards and recommendations within the helicopter EMS(HEMS) industry. Despite the safety programs and safety initiatives industry-wide and program-specific, helicopter EMS accidents continue to occur. HEMShas been credited with saving the lives of tens of thousands of critically ill orinjured persons. Tragically, however, there have also been over 150 lives lostdue to helicopter accidents in the pursuit of these life-saving missions.
Does the fact that HEMS accidents occur every year suggest that air medicaltransport is unsafe? Are HEMS crewmembers at a significantly higher risk forinjury or death? This report will try to answer these questions by reviewing acci-dent data and trends, identifying potential causes of helicopter accidents andincidents, and reviewing accident and injury rates.
Nothing we do is completely safe. In a 1972 U.S. Supreme Court decision,the court concluded that “safe is not the equivalent of risk free.” There are risks,often potentially serious ones, associated with every occupation, every mile trav-eled, every food eaten, every hobby, every investment – basically with everyaction we take. Clearly, some actions are riskier than others and the only way toeliminate risk from any activity would be to avoid participating in it completely.
As health care providers, we are constantly faced with risk. We make patientcare decisions based upon the related risks and benefits of any given treatmentor transfer. The Emergency Medical Treatment and Active Labor Act (EMTALA) requires that we “inform the individual (or a person acting on theindividual’s behalf) of the risks and benefits…” and document that the benefitsof transfer outweigh the increased risks to the individual.
In air medical transport safety must be the top priority. We must also considerthe risk related to air medical transport—not just to satisfy EMTALA—but toassure that procedures are in place to avoid unnecessary risk and that proactivesteps are taken to control risk.
This report reviews and summarizes some of the most important publications,presentations, and studies that address various aspects of helicopter EMS acci-dents. In addition, we describe an extensive research project and our results thatenable us to better understand the risks related to air medical transport.
Ira J. Blumen, MD1,2
Daniel L. Lemkin, MD1,2
Geoff Scott, RN1
Michael Casner, MD3
Karen Arndt, RN1
Michelle Bate, RN1
Sonia F. Callejas, MD4
Charles L. Maddow, MD5
Fred Ligman1,6
Clinton Pope1,6
Edward Ban1,6
Craig Felty, RN1
Mose L. Freeman, EMT-B1
Linda Burdett1
John Collins1,6
Ray McCall1,6
Dave DeFauw1,6
Author's current affiliation:1UCAN, University of ChicagoHospitals, Chicago, IL; 2Section ofEmergency Medicine, Department ofMedicine, University of Chicago,Chicago, IL; 3Department of EmergencyServices, San Francisco GeneralHospital, University of California SanFrancisco, San Francisco, CA; 4 Sectionof Emergency Medicine, Department ofSurgery, University of MarylandMedical Systems, Baltimore, Maryland;5 Emergency Medicine, University ofRochester School of Medicine &Dentistry, Rochester, NY; 6 CJ SystemsAviation Group, West Mifflin, PA.
Address for correspondence:Ira Blumen, MDProgram/Medical Director, UCANUniversity of Chicago Hospitals5841 S. Maryland Ave. MC 0952Chicago, IL 60637Phone: 773-702-9895Fax: 773-702-1993Email: [email protected]
Copyright © 2002Ira J. Blumen and Air Medical Physician Association
Grant support to fund the publication of this report has been received in part from: Foundation for AirMedical Research, Bell Helicopter Textron; CJ Systems Aviation Group; American Eurocopter; Fitchand Associates; Association of Air Medical Services (AAMS); National Flight Paramedic Association(NFPA); National EMS Pilots Association (NEMSPA); Air Methods; Sikorsky; Golden Hour DataSystems; Air and Surface Transport Nurses Association (ASTNA); Petroleum Helicopters, Inc.(PHI); Metro Aviation; Turbomeca Engine Corporation; Heli-Dyne Systems; Agusta AerospaceCorporation; University of Chicago Aeromedical Network (UCAN); Illinois Association of Air andCritical Care Transport (IAACCT); Keystone Helicopter; Innovative Engineering/AeroMed Software;and National Association of Air Medical Communication Specialists (NAACS).
This research was supported in part througha grant from the Foundation for AirMedical Research and by the University ofChicago Aeromedical Network (UCAN).
2 AIR MEDICAL PHYSICIAN HANDBOOK
REVIEWING THE DATA
There are many ways to review thesafety of air medical transport and toassess the relative risk to its crewmem-bers. One can look at the safety recordand practices of an individual program orlook at the industry as a whole. Section1 of this report looks at the accident andincident data that is specific to theHEMS industry. Another approach is tocompare the accident data with that ofother forms of air travel and other modesof transportation. This approach is takenin Section 2. Finally, Section 3 comparesthe identified risks for HEMS to otheroccupations, high-risk activities, and var-ious routine daily activities.
When looking at accident data, it iscommon to talk about raw numbers(number of accidents), percentages, oraccident rates. But what does that reallymean? How can we look at the data,compare statistics, and arrive at conclu-sions? The Office of System Safety with-in the Federal Aviation Administration(FAA) has stated that research hasshown the importance of comparing likegroups when comparing accident data orsafety performance. For example, a com-parison of one year’s HEMS accident sta-tistics with another year’s HEMS acci-dent statistics is more likely to be accu-rate and meaningful than a comparisonof general aviation accident data toHEMS data. It may also be helpful attimes to look at different categories ofaviation to try to draw comparisons andconclusions.
There is no consensus amongresearchers and participants in the avia-tion industry as to what constitutes “safe-ty data.” In addition, when interpretingcomparisons, equivalent types of data aremore likely to be accurate than compar-isons of different types of data.Therefore, we must first level the playingfield in terms of exposure to risk. It isessential that accident and incident databe “normalized.” Raw data on accidentsand incidents must be converted to acci-dent or incident rates before it can beused for drawing conclusions about safetyover time, or to compare different typesof aviation, airplanes, pilots, types ofoperations, and so on.
Comparisons based strictly upon thenumber of events that occur may not tellthe real story. To be meaningful, compar-isons must be based upon equal exposureto risk. The longer we are exposed to aparticular risk, or the more times weundertake an activity involving risk, thegreater the overall risk. However, thisalone does not determine total risk.Reduction factors such as experience,proficiency, equipment, and flight condi-tions can have a significant positiveimpact on safety.
The FAA and other organizationstrack aviation data in a number of differ-ent ways. The most common method iswith respect to flight hours and depar-tures. The data is then normalized interms of accidents (or fatalities) per100,000 flight hours or accidents (orfatalities) per 100,000 departures (i.e.,takeoffs). In most aviation, unlike
HEMS, takeoff and landings are thehigher risk periods. Therefore, short andmultiple stop flights are riskier mile-for-mile than long nonstop flights. Airlinesgenerally prefer to focus on accident ratesper mile, which makes air travel seemvery safe.
The FAA estimates the total flighthours for general aviation based on a sur-vey of a sample of aircraft owners andoperators. Scheduled air carriers, on theother hand, must report flight hours,departures, and passengers carried, sotheir accident statistics may be comparedusing either departures or flight hours. Inthe early 1980s when data was kept forHEMS, it was tracked in terms of flighthours and patients transported (which isdifferent than departures).Unfortunately, the air medical transportindustry has been unable to develop aconsistent method to track its own data –a problem that has recently drawn a greatdeal of attention. However, numerouscalculations are outlined later in thisreport to make some appropriate compar-isons with HEMS.
In much the same way that the FAAnormalizes its data, similar strategies mustbe used to compare the risk of injury ordeath. To assess the magnitude of risk,we must again normalize the data insome fashion. In this case, the relevantfigure is in terms of a ratio, fraction, orpercentage. By reducing all risks to acommon format, we can begin to com-pare different types of activities and therelative risks. The larger the percentage,the riskier the activity.
SECTION 1: AIR MEDICAL ACCIDENTS AND INCIDENTS
This section presents an overview ofHEMS accidents and reviews severalnotable investigations. The first, from aseries of articles written by Rick Frazerand published in AirMed, presents a 20-year review of air medical accidents andspecific types of accidents. Second is a1988 report of 59 HEMS accidents(1978-1986) by the NationalTransportation Safety Board (NTSB).The next study by Patrick Veillette isfrom the Flight Safety Foundation (April2001) and reviews 87 accidents from1987 through 2000. A 1994 study by theNASA-Ames Research Center is thensummarized which evaluates air medical
incidents, rather than accidents. Thefinal report in this section is the “AirMedical Accident Analysis” which ana-lyzed past accidents and identified inter-ventions to prevent future accidents.This section concludes with a review ofthe 1998-2001 HEMS accidents andfinally with a comparison of HEMS acci-dent rates, which are based upon avail-able data, research, several necessaryassumptions, and various calculations.
BACKGROUND
Air medical transport began in themilitary, which laid the groundwork for
the air transport of critically ill andinjured patients. The first civilian airmedical operations were established inthe late 1960s in the form of dual-pur-pose programs, which combined thefunction of public safety agencies withEMS missions. In 1972, the first fatalEMS-related accident occurred withinone of these dual-purpose programs.That same year, St. Anthony’s Hospitalin Denver established the first dedicatedhospital-based HEMS program.
In the 1970s and ’80s, various publica-tions and organizations tracked thegrowth of the dedicated helicopter EMSindustry. The number of dedicated pro-
3Special Supplement
grams grew slowly in the ’70s, but gath-ered significant momentum in the early1980s. In 1981 there were 45 programsand by 1986 the industry had almosttripled to 129 programs. Unfortunately,this was accompanied by an increase inthe number of air medical accidents.
In the late 1980s new trends were seen.More programs added a second helicopterto their operation, while other programsclosed down. In many areas competitionwas fierce. By 1990 there were 174 pro-grams operating 231 helicopters.
In the ’90s, air medical programs andindustry-wide statistics became increas-ingly difficult to track. While new pro-grams started and helicopters were addedto many established programs, othersmerged operations and still more closed.Fixed-wing (airplane) air medical trans-port continued to grow and became anintegrated component of the industry.During this time, the industry did notmaintain an accurate census. In addi-tion, industry-wide data was not kept asto the total number of patients transport-ed or the total number of hours flown.The importance of this data will becomeevident later in this report.
AN OVERVIEW OF HEMS ACCIDENTS: 1972 TO 2002
This research study, developed bymembers of the UCAN SafetyCommittee, began with a comprehensivereview of numerous databases to identifyHEMS accidents and their correspondingfatalities and injuries. Since the intro-duction of civilian HEMS in 1972, theUnited States has experienced 162 acci-dents involving dedicated medical heli-copters and four additional accidentsinvolving dual-purpose aircraft. Of theseaccidents, 67 have resulted in at leastone fatality. Figure 1–1 shows the totalnumber of HEMS accidents and fatalaccidents (an accident where at least oneoccupant died) broken down for eachyear since 1980. As the lines in grayshow, until the last few years, the highestnumber of accidents occurred in the mid-1980s—the same time that the industryexperienced its most rapid growth. Theearly and mid-1990s showed an improve-ment, but unfortunately, 1998– 2001showed a steady increase in accidents.
Fortunately, the number of fatal acci-dents did not rise at the same rate.
The NTSB defines an aircraft accidentas “an occurrence associated with theoperation of an aircraft which takes placebetween the time any person boards theaircraft with the intention of flight andall such persons have disembarked, andin which any person suffers death or seri-ous injury, or in which the aircraftreceives substantial damage.” An inci-dent means “an occurrence other than anaccident, associated with the operation ofan aircraft, which affects or could affectthe safety of operations.” The NTSBalso classifies the severity of variousinjuries that may occur from an accidentas fatal (“any injury which results indeath within 30 days of the accident”),serious, minor, or none.
When reviewing any accident databaseor report, it is essential to know theinclusion and exclusion criteria for thedata that is used. There are many agen-cies, organizations, and individuals whomaintain and track data regarding HEMSaccidents and incidents. There is notalways agreement on what should beincluded in a report or study. Some stud-ies look at only patient missions. Otherswill look at non-patient missions as well.Some reports include fixed-wing (FW) aswell as rotor-wing (RW) and some data-bases include incidents as well as acci-dents. An accident database may includeonly dedicated air medical services, whileothers might include dual-purpose (e.g.,
police, fire) aircraft. There is no right orwrong way, but it is essential to knowwhat is included in each statisticalreview.
The accidents listed in Figure 1–1 arethe end product of a detailed review ofvarious personal databases, publications,and Internet websites. In numerous cases,available information was insufficient todetermine inclusion vs. exclusion intoour own database. To clarify the acci-dent information, direct communicationby telephone or email was also employed.Included in this table are dedicated med-ical helicopter accidents that occurred oneither patient or non-patient missions.Also included are known accidentsinvolving dual-purpose helicopters (atotal of four accidents and nine fatalities)that crashed during medical missions.Fixed-wing air medical accidents, mili-tary accidents and international acci-dents have not been included in ourdatabase or statistical analysis.
Our review found 162 HEMS accidentsin which we identified a total of 183 fatalinjuries. Accidents have taken the livesof pilots, nurses, physicians, paramedics,respiratory therapists, patients, policeofficers, fire fighters, and observers.There were also 60 individuals who suf-fered serious injuries and 73 with minorinjuries. A total of 195 people were notinjured. Figure 1–2 shows a breakdownfor injuries and fatalities for each year inour study. We have also included year-to-date information for 2002.
Figure 1–1: HEMS Accidents and Fatal Accidents, 1972–2002*(*through 9/30/02)
0
2
4
6
8
10
12
14
72-79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 '00 '01 '02Year
Acc
iden
ts
Total Accidents Fatal Accidents
4 AIR MEDICAL PHYSICIAN HANDBOOK
Figure 1–3: Accidents and Fatal Accidents, 1978––1998
Figure 1–4: Fatalities per Year
Figure 1–2: Fatalities and Injuries, 1980–2002* (*as of September 30, 2002)
Adapted from: Frazer, AirMed, Sept/Oct 1999
Figure 1–5: Severity of Injuries –- Air Medical Accidents, 1980––1998 (n=397)
F at al 42 %
S er i ou s 13 %
M i n or 15 %
N o I n j u r y 30 %
Adapted from: Frazer, AirMed, Sept/Oct 1999
Adapted from: Frazer, AirMed, Sept/Oct 1999
AIR MEDICAL ACCIDENTS: A 20-YEAR REVIEW
Perhaps the most complete analysis ofair medical accidents is represented in aseries of AirMed articles by Rick Frazer.The first of the series was published in theSeptember/October 1999 issue. The arti-cle “Air Medical Accidents—A 20 YearSearch for Information” presents a review104 rotor-wing, 15 fixed-wing, and threepublic EMS accidents for a total of 122accidents between 1978 and 1998.Included were only accidents thatoccurred to dedicated medical transportservices, not to private or public aircraftthat may also perform an occasional med-ical transport.
This study included only accidents asdefined by the NTSB. In follow-up tothis article, Frazer looked at specific typesof accidents—those related to weather,collisions with objects, and maintenance.
Accidents and Fatal Accidents
While the number of HEMS accidentsis a critical consideration, even moreimportant is the high number of fatalaccidents seen in HEMS. In the 1980s,42% of all accidents resulted in at leastone fatal injury. Between 1990 and 1998,this percentage rose to 56%. Figure 1–3compares the number of accidents to fatalaccidents for 1978 through 1998—theyears included in Frazer’s report. It isimportant to note that these numbers do
not necessarily correspond to the number ofaccidents previously presented since thisstudy also included over a dozen fixed-wingaccidents and several accidents that did notmeet our inclusion criteria.
Severity of Injury
Frazer reported that 406 individuals wereinvolved in HEMS accidents. There were168 fatalities, 50 serious injuries, and 61minor injuries. There were 127 who sustained no injury. Figure 1–4 shows thetrend over the years with regard to fatalitiesand Figure 1–5 summarizes the severity ofinjuries for all the reported accidents.
0
5
10
15
20
25
30
35
40
45
50A
ccid
ents
None
Minor
Serious
Fatal
0
2
4
6
8
10
12
14
16
18
Fat
aliti
es
0
2
4
6
8
10
12
14
Year
Acc
iden
ts
Total Accidents Fatal Accidents
5Special Supplement
When Do HEMS Accidents Occur?
When an accident occurs is an impor-tant consideration when looking at thesafety of flight and may provide insight tothe severity of injuries and the highnumber of fatalities. The time of day(i.e., day vs. night), type of mission (i.e.,the purpose of the HEMS flight), and thephase of flight are all-important factors toevaluate.
Condition of Light
It would appear in Figure 1–6 thatEMS accidents showed no significantpreference for day vs. night operations.However, the 1999 AirMed TransportSurvey indicates that only 38% of allHEMS flights were at night. Since 1988,the range has been between 35 to 42%,with an average of 38%. In contrast,49% of the HEMS accidents over 20years occurred during night operations.Of interest, Frazer’s report identifies threenoticeable peak accident times: noon to1pm, 6–7pm, and 10–11pm.
Phase of Flight
The phase of flight that the aircraftwas in at the time of the accident isreported in the NTSB reports. Thisincludes takeoff, landing, cruise, hover-ing, maneuvering, or taxiing. Data forboth helicopter and fixed-wing accidentsare provided in Figure 1–7.
As previously mentioned, in most avia-tion operations, aircraft are more proneto incidents or accidents during takeoffsand landings. This was not found to bethe case with HEMS accidents. Moreaccidents occurred during cruise (36%)than any other phase of flight. Accidentreports frequently identify entry intoinadvertent instrument meteorologicalconditions (IMC) as a contributing fac-tor for many of the accidents (i.e., flyinginto poor weather conditions).
Purpose of Flight
The purpose of flight for HEMS acci-dents was tracked by Frazer in his article.He divided flights into two major cate-gories: patient-related and non-patientrelated. He further divided the patientflights into scene vs. interhospital; and
finally determined if the aircraft wasenroute to the patient, had a patient onboard, or if the aircraft was returningfrom a patient flight (i.e., the third “leg”of a transport). Evaluating these compo-nents of HEMS transports might indicatea particular interval when more accidentsseemed to occur. Figure 1–8 depicts thevarious purposes of flight and the differ-ent legs of patient transport.
Trying to determine the most danger-ous flight or leg of a patient transportfrom Figure 1–8 may be difficult. Moreaccidents occur on the transport to thepatient than on either remaining leg.However, if the helicopter was transport-ing the patient back to its base facility,there would be no third leg of a trans-port. The pie charts in Figures 1–9 to1–12 look at the percentage of accidentsthat occurred on the various types of mis-sions and then each leg of the transport.
In general, it appears that approxi-mately 50% of all accidents occurenroute to the patient and 50% on theway back. Of note is the fact that 85 ofthe 104 HEMS accidents that occurredwere on patient-related missions, ofwhich 36 were scene responses. Thiscorresponds to 42% of the accidents dur-ing patient-related missions. Referring
again to the “2000 Annual TransportStatistics,” scene missions accounted foronly 28% of all patient transports. Since1988, the percentage of scene responseflights has ranged from 25% to 36%, withan average of 31%. If all missions hadequal risk, then 31% of the accidentsshould have been on scene missions.Scene missions have always been per-ceived as the most difficult and poten-tially the most dangerous in HEMS andthis data would seem to confirm this.However, the fact that a higher percent-age of the accidents occur after thepatient has been picked up rather thanenroute to the scene (56% to 44%) fur-ther confounds this theory.
N i gh t 49 %
D ay 46 %
D aw n 1%
D u sk 4%
Figure 1–6: Day vs. Night HEMSAccidents, 1978–1998 (n=107)Adapted from: Frazer, AirMed, Sept/Oct 1999
10
2
39
3
6
10
9
28
0
7
0
2
2
0
1
0
2
1
Approach
Climb
Cruise
Descent
Hover
Landing
Maneuvering
Takeoff
Taxiing
Number of Accidents
Fixed-wing
Rotor-wing
Figure 1–7: Phase of Flight Adapted from: Frazer, AirMed, Sept/Oct 1999
6 AIR MEDICAL PHYSICIAN HANDBOOK
Adapted from: Frazer, AirMed, Sept/Oct 1999Figure 1–8: Phase and Purpose of Flight
49
26
15
8
36
16
19
1
19
4
4
7
1
2
1
28
14
9
5
14
6
8
0
4
0
1
2
1
0
0
0 10 20 30 40 50 60
TOTAL: Interhospital Accidents
Enroute to Patient
Patient on board
Returning from transport
TOTAL: Scene Accidents
Enroute to Patient
Patient on board
Returning from transport
TOTAL: Non-patient Accidents
Fuel
Maintenance
PR
Training
Other
Unknown
Number of Accidents
Non-Patient Accidents
Scene Accidents
Interhospital Accidents
The Cause of HEMS Accidents
It is rare that a single isolated eventcauses an accident. Instead, it is gen-erally agreed that a set of contributingfactors and circumstances usually leadto a final event that results in the acci-dent. The NTSB investigations, how-ever, usually conclude that an accidentoccurred as a result of one of two prob-able causes—either pilot error ormechanical problem. The NTSB mayalso list the case as “Unknown” or “Tobe Determined.” At times, the NTSBmay list more than one cause for a par-ticular accident.
Identifying and specifying the causeof an accident may not be easy, evenafter a thorough and detailed investi-gation. In one HEMS accident, afteran engine problem developed, thepilot shut down the wrong engine.The result was a fatal accident.Should this accident be attributed topilot error or a mechanical failure?
The two main categories—piloterror and mechanical failure—aredivided into more specific causes asdetermined in the final report of theNTSB. In his report, Frazer listed thecauses of 104 helicopter accidents. Atotal of 68 (65%) were pilot error and26 (25%) were mechanical failure.The remaining accidents wereunknown (3) or still to be determined(9). Figure 1–13 lists the identifiedcauses of the rotor-wing accidents.
Weather-related Accidents
Weather-related accidents remain anall too familiar theme in HEMS.Frazer’s 1999 report listed 23 such acci-dents, of which 17 were fatal. In theMay/June 2000 issue of AirMed,Frazer’s article “Weather Accidentsand the Air Medical Industry” updatedhis weather-related accident data.With this article, Frazer’s accidentdatabase has increased from 122 to 136helicopter and fixed-wing accidents,with 121 final reports available. Thereare now 31 weather-related accidentsin his database, equal to 26% of theaccidents with final reports. Of these31 accidents, 25 were rotor-wing and 6were fixed-wing.
Figure 1–12: Accidents During SceneTransports (n=36)
S c e n e 3 5 %
I n t e r f a c il it y 4 7 %
N o n - p a t ien t 1 8 %
E n r o u t e t o
p a t ien t 5 3 % P a t ien t
o n B o a r d 3 1 %
R e t u r n f r o m
t r a n s p o r t 1 6 %
P a t ien t o n B o a r d
5 3 %
E n r o u t e t o
p a t ien t 4 4 %
R e t u r n f r o m
t r a n s p o r t3 %
E n r o u t e t o p a t ien t
4 9 %
P a t ien t o n B o a r d
4 0 %
R e t u r n f r o m
t r a n s p o r t 1 1 %
Figure 1–9: All HEMS Accidents (n=104) Figure 1–10: Accidents During AllPatient Transports (n=85)
Figure 1-11: Accidents DuringInterhospital Transports (n=49)
Adapted from: Frazer, AirMed, Sept/Oct 1999
Adapted from: Frazer, AirMed, Sept/Oct 1999 Adapted from: Frazer, AirMed, Sept/Oct 1999
Adapted from: Frazer, AirMed, Sept/Oct 1999
7Special Supplement
In the 1980s, there were a total of 73accidents, of which 16 (22%) were deter-mined to be weather-related. In the1990s the number of accidents (withfinal reports) had decreased to 44, with14 (32%) related to weather. Tragically,while the total number of accidents wentdown, the percentage of weather-relatedaccidents increased by 10%.
Looking at when these weather-relatedaccidents occur is also of importance.Previously it was noted that nearly half ofall HEMS accidents occurred at night(53 of 107). In comparison, Frazer’s arti-cle in 2000 found that 22 of 25 (88%)weather-related HEMS accidentsoccurred at night. Fixed-wing weather-related accidents remained at 50%, evenat night.
Of the 25 weather-related helicopteraccidents, 17 (68%) of the accidentswere interhospital and 8 (32%) werescene missions. This correlates well tothe overall percentage of scene missionsfor HEMS program, which was previouslynoted at 31%. Of interest, there is anoted decrease in the percentage of sceneaccidents when comparing weather-relat-ed accidents (32%) to all patient-relatedmission accidents (42%).
The outcome of weather-related acci-dents is also very dramatic. Nineteen ofthe 25 HEMS accidents (76%) had fatal-ities and 64% had no survivors.Compared to all accidents in Frazer’s1999 study, only 45% (55 of 122) of allaccidents resulted in at least one fatality.In addition, 76 of 103 (74%) of the soulson board sustained fatal injuries, 14 hadserious injuries, 8 suffered only minorinjuries, and 5 had no injury. Figure1–14 shows the relative comparison (per-centage) of the severity of injuries in theweather-related accidents reported in2000 to the total accidents reported byFrazer in 1999.
In general, the cause of the weather-related accidents does not appear to be apilot’s disregard for established weatherminimums at takeoff. Instead, it is thepilot’s encounter with instrument mete-orological conditions (IMC) en route. Inthe narratives of the 25 helicopter acci-dents, Frazer noted that 10 pilots wereturning around, one was circling, six con-tinued into IMC and one was on an IFRflight plan. The pilots’ actions wereunknown in 7 cases.
All Accidents
Weather-Related
Fatal Serious Minor No Injury
6
1
16
9
2
3
12
5
4
3
4
17
4
3
3
2
2
1
2
3
5
Weather
Spatial Disturbance
Hit Obstacle
Loss of Control
Fuel Starvation
Other
Foreign Object into a/c
Engine: Wrong Shutdown
IFR: Failure to Follow
Engine
Flight Controls
Improper Maintenance
Unknown
To Be Determined
Total Number of Accidents
Mechanical
PilotError
Figure 1–13: HEMS Accidents by Cause Adapted from: Frazer, AirMed, Sept/Oct 1999
Figure 1–14: Severity of Injuries–All Accidents vs. Weather-related, 1982–1998(All accidents: n=397; Weather-related accidents: n=103)Adapted from: Frazer, AirMed, Sept/Oct 1999 and Frazer, AirMed, May/June 2000
Condition Area Ceiling VisibilityDay Local 500' 1 mile
Day Cross Country 1000' 1 mile
Night Local 800' 2 miles
Night Cross Country 1000' 3 miles
Background Information: Weather Limitations
Subpart D of the FAA Part 135 regulations (sections 135.201 to 135.205)outlines the FAA’s operating limitations and weather requirements, includingminimum altitudes and visibility for “air taxis.” In addition, the FederalAviation Regulations (FARs) under Part 135 require that Air Taxi Operatorsset their own weather minimums in Chapter 4 of their Operations Manual.The weather minimums describe the minimum ceiling (lowest cloud heightabove the ground that covers 5/8 or more of the sky) and the minimum visibili-ty the pilot must have to accept a flight.
The ceiling and visibility requirements generally vary for day vs. night flightand for local vs. cross-country flights. The definition of “Local” may also varyfrom 50 to 100 miles or more. Many state regulatory agencies and organiza-tions may also set minimums for ceiling and visibility. The Commission onAccreditation of Medical Transport Systems’ weather guidelines are:
8 AIR MEDICAL PHYSICIAN HANDBOOK
There are three types of weather-relatedaccidents:
• Controlled Flight Into Terrain (CFIT)refers to an event that normally occursin IFR conditions or at night. Loss ofsituational awareness is apparent in allCFIT accidents.
• Loss of Aircraft Control corresponds toan event in IFR conditions when thepilot is unable to maintain control ofthe aircraft by reference to the flightinstruments. Spatial disorientation isthe primary cause and is often the resultof continued VFR flight into IFR conditions.
• In-flight Collision with an Obstacle isan event that normally occurs in condi-tions of restricted visibility when thepilot is unable to see an obstacle or theterrain in time to avoid a collision.This is the most common weather-relat-ed accident for helicopters.
Collision with Objects
An in-flight collision with an object(CWO) may occur in various conditionsand settings. Frazer’s 1999 study identified19 HEMS accidents that resulted from aCWO. In his 2001 report “Air MedicalAccidents Involving Collisions withObjects,” the database increased from 122to 150 fixed-wing and rotor-wing acci-dents, 27 of which involved a CWO.This represented a dramatic increase inthe number of CWO accidents, as therehad been a total of 8 such accidents dur-ing the 9-year period from 1990 to 1998,while in 1999 and 2000 alone, there were7 CWO accidents. Frazer also identified21 CWO incidents over the past 5 years.
Twenty-five of the CWO accidents inFrazer’s report involved rotor-wing air-craft. Frazer observed that weather wasnot considered to be a factor in any ofthese. Sixteen (64%) of the accidentsoccurred during scene response missions—all occurring at or near the landing zone.Eight (32%) were during some phase offlight during interhospital transports, fourof these at or on the hospital helipad.The remaining CWO was during a main-tenance flight. Of the 24 patient mis-sions, 12 occurred on the way to pick upthe patient, 9 occurred with the patienton board, and 3 were returning after a
patient flight. As previously presented,scene response flights account for an aver-age of 31% of all patient transports. If allmissions had equal risk, 31% of the CWOaccidents should have been on scene mis-sions. Instead that percentage more thandoubles.
Figure 1–15: Type of Mission—CWOAccidents (n=25)Adapted from: Frazer, AirMed, May/June 2001
S c e n e 6 4 %
I n t e r h o s p it a l 3 2 %
M a int e n a n c e 4 %
Wires are the most common objectswith which helicopters collide. Of the 25CWO accidents, 9 were wire strikes. Frazerclassified four of the CWO accidents as“other,” which included the ground, rocks,and in one incident, a barge. There werethree CWOs with trees, three with groundobstacles (lamp posts, fencing, etc.), twowith support cables to towers and one eachwith a building, hangar, hospital helipad,and tower.
As expected, collisions with objectsoccur most commonly on takeoff and land-ing. The time of day also varies, but yield-ed some interesting results as seen inFigure 1–16.
Figure 1–16: Phase of Flight and LightingConditions
2
9
4
8
4
Landing
Takeoff
Cruise
Total Number of Accidents
Day Night
Accordingly, only 7 of the 25(28%) RW accidents resulted in afatality. Of the 4 cruise accidents, 3were fatal. All the cruise CWOaccidents identified by Frazer wereRW, all were during the day in clearweather, and 3 were on the returnflight home.
Frazer concludes his 2001 studywith recommendations for compre-hensive training and procedures forflight personnel to enhance safetyand reduce the likelihood of CWOaccidents. “Medical personnelshould be taught how to spot obsta-cles, and detailed procedures shouldbe in place on how to communi-cate—down to the exact words thecrew uses—an immediate threat ver-sus an obstacle in the distance.” Inaddition, Frazer suggested the disas-sociation of Landing Zone (LZ) andsafety training from marketing andemphasize the importance of ongo-ing safety education as a separate,budgeted activity.
Maintenance-relatedAccidents
The data presented in Figure 1–13represent Frazer’s 1999 study, whichshowed 26 maintenance-related acci-dents (MRAs) out of 122 total acci-dents. In the May/June 2002AirMed, Frazer examined the “AirMedical Accidents Attributed toMaintenance.” A total of 34 suchaccidents were identified in the 2002report which now includes 143 rotor-wing and 18 fixed-wing MRAs.Thirty of the MRAs were rotor-wingand four were fixed-wing. Excluding11 accidents that lacked NTSB finalreports, Frazer concluded that 23%of all accidents were maintenance-related.
Frazer separated the probable causeof the MRAs into five different cate-gories. Over the 24-year period, therewere a total of 17 engine-related acci-dents, 15 of which involved rotor-wing aircraft. Twelve involved single-engine aircraft and three involvedtwin-engine aircraft. Even thoughthe engine-related accidents representhalf of all MRAs, Frazer observed thatengine-related incidents happen much
Adapted from: Frazer, AirMed, May/June 2001
Since most of the CWO accidentsoccur during takeoff and landing, it fol-lows that the aircraft are at a lower alti-tude and slower speed.
9Special Supplement
more frequently. Fortunately, due to the skills ofthe EMS pilots, the vast majority of these inci-dents do not result in accidents.
In 11 of the final NTSB reports, Frazer found“inadequate or improper maintenance” as a fac-tor in the accident. In addition, there were fivereports that identified “manufacturer design” or“inadequate aircraft component product/design”as a factor.
E n g ine r e lat e d 5 0 % M a in r o t o r d r iv e
/ m a in t r a n s m is s ion
2 4 %
T a il r o t o r d r iv e s y s t e m
1 3 %
Fl igh t c o n t r o l s y s t e m
1 3 %
Figure 1– 17: Probable Cause ofMaintenance-related HEMS Accidents (n=30)
Adapted from: Frazer, AirMed, May/June 2002
2
1
2
3
4
22
1
1
1
2
4
9
16
Unknown
PR
Maintenance
Fuel Run
Scene Response
Interhospital
TYPE OF MISSION
Hover
Descend
Climb
Landing
Approach
Takeoff
Cruise
PHASE OF FLIGHT
Number of Accidents
Figure 1–18: Maintenance-related Accidents—Phase of Flight and Type of Mission Adapted from: Frazer, AirMed, May/June 2002
As Figure 1–18 shows, nearly 50% (16 of 34)of all rotor-wing and fixed-wing MRAs occurredduring cruise. Interhospital transport accountedfor nearly two-thirds of the accidents.
Pilot Experience
Experienced pilots don’t make mistakes orhave accidents. Or do they? Unfortunately, fly-ing with “high-time” pilots does not necessarilyguarantee that an accident won’t happen. Muchof the HEMS industry requires a minimum of2,000 hours of flight time before assuming com-mand of a medical helicopter. In addition, mostoperators, organizations, and even some statesdictate a specific number of flight hours in type-specific aircraft before a pilot can fly medicalmissions.
Figure 1–19 shows the flight experience of thepilots involved in HEMS accidents.Unfortunately, there is no data available regard-ing the average number of flight hours for allHEMS pilots or the percentage of all HEMSpilots that would fall into each category listed inFigure 1–19. Of the pilots involved in HEMSaccidents, the lowest total hours were 1,432,while the highest was 14,000. More importantlyperhaps, in 27 of the 122 accidents (22%), thepilot had fewer than 200 hours of flight time inthe make and model of the aircraft they wereflying at the time of the accident. Eighteen(15%) had fewer than 100 hours and one hadonly three hours.
1
6
8
16
23
13
8
7
10
3
11
0-1432
1433-2000
2001-3000
3001-4000
4001-5000
5001-6000
6001-7000
7001-8000
8001-9000
9001-10000
>10,000
Tot
alF
light
Hou
rs
Number of Accidents
Figure 1–19: Pilot Experience Adapted from: Frazer, AirMed, Sept/Oct 1999
Engines
A frequent debate with regard tosafety is whether “two engines arebetter than one.” Of the 17 acci-dents attributed to engine problems,12 were single-engine and 3 weretwin-engine. However, this does not
help resolve the debate, knowing thatin the early years of HEMS, most helicopters were single-engine.Therefore, more of the accidents werein single-engine aircraft. In anothersection of this report is accident datacollected by the HelicopterAssociation International (HAI)
10 AIR MEDICAL PHYSICIAN HANDBOOK
comparing single- vs. multi-engine heli-copters which shows a lower accidentrate overall for multi-engine helicopters.
Aircraft Damage and Post-Impact Fire
In the NTSB accident reports, aircraftdamage is classified as either destroyed orsubstantial. Of the 122 air medical trans-port accidents, 70 (57%) aircraft weredestroyed and 51 (42%) were reported ashaving suffered substantial damage. Atthe time of Frazer’s report, one was listedas unknown. Of note, when lookingonly at the weather-related accidents,84% of the helicopters and 100% of thefixed-wing aircraft were destroyed.
Post-impact fire is often perceived asanother major concern with regard toHEMS accidents. Frazer’s data, seen inFigure 1–20 shows that the vast majorityof HEMS accidents do not result in apost-impact fire. While fixed-wing air-craft accidents resulted in fires 67% ofthe time, only 15% of helicopter acci-dents sustained post-impact fires. Therewere, however, six accidents that had anin-flight fire.
NATIONALTRANSPORTATION SAFETYBOARD SAFETY STUDY:COMMERCIAL EMERGENCYMEDICAL SERVICEHELICOPTER OPERATIONS– 1988
Frazer’s 20-year review provides a com-prehensive look at the HEMS industryand our accident history from 1978 to
1998. This next report by the NationalTransportation Safety Board (NTSB),however, looks at a narrower time frame.When published in 1988, it became thehallmark safety study at the time of ourindustry’s highest accident and fatalityrate.
During the early 1980s, the increaseduse of helicopters as air ambulances cameat a high price. While the number offlight programs more than tripled from1981 to 1986, the NTSB began to identi-fy a significant rise in the number of acci-dents. In 1984 there were 7 HEMS acci-dents. The next year, there were 11 andin 1986, there were 14 accidents investi-gated by the NTSB. These 14 accidentscorresponded to 9 percent of the totalcommercial HEMS industry operatingthat year. As a result, a formal safetystudy was undertaken by the NTSB andpublished in 1988.
The NTSB studied 59 commercialHEMS accidents that occurred between1978 and 1986. Nineteen of these acci-dents resulted in fatalities, taking thelives of 53 people (19 pilots, 28 medicalpersonnel, and 6 patients). A total of 47accidents were on patient mission flightsand 12 while on other activities (refuel-ing, PR, training, etc.). However, in thecalculation of accident rates, the NTSBonly included the 47 “mission” accidents.In addition, this study did not includeany public-use aircraft.
According to the NTSB report, from1980 through 1985, HEMS had an esti-mated accident rate of 12.34 accidentsper 100,000 hours of flight, nearly doublethat of nonscheduled Part 135 (“airtaxi”) helicopter operations(6.69/100,000 flight hours). They alsodetermined that the fatal accident rate
for HEMS was 5.40—nearly 3.5 timeshigher than the 1.60 determined forother nonscheduled Part 135 helicopteroperations.
The NTSB identified four major fac-tors in the 59 HEMS accidents studied.Human error (i.e., pilot error) was attrib-uted as the cause, directly or indirectly, ofthe majority of these accidents (68%).Weather was the second most commoncause of these HEMS accidents (30%),followed by mechanical failure (25%),and obstacle strikes (20%). The weath-er-related accidents accounted for 61% ofthe fatalities and the NTSB concludedthat poor weather poses the greatest sin-gle hazard to EMS helicopter operations.
The NTSB report identified severaldisturbing trends involving HEMS opera-tions. They were concerned that therapid increase in the number of programsand resulting competition could result ina focus on transport volume instead offlight safety. Pilots might feel self-imposed or externally imposed pressure(i.e., by management) to accept andcomplete flights despite marginal operat-ing conditions. In addition, pilot trainingwas often deficient in interpretation ofweather conditions and in instrumentflight procedures. They also found thatmodified EMS interiors and various pro-gram practices often compromised crash-worthiness standards, resulting in anincreased risk of injury and death.
The NTSB report ended with specificrecommendations—some to the FAAand some to the American Society ofHospital-Based Emergency Air MedicalServices (ASHBEAMS, which changedits name in 1989 to the Association ofAir Medical Services). Included in theserecommendations were: improved interi-or modifications that would not compro-mise crashworthiness; the use of shoulderharnesses and protective clothing (e.g.,helmets, flame-resistant suits, protectivefootwear); the development of programsafety committees; improved training in anumber of areas, including marginalweather operations, emergency proce-dures, pilot-crewmember coordination,and communications.
In reviewing the NTSB data an inter-esting observation is made with regard toinclusion criteria and selection. A 1986Chicago accident is included that did not
7 8
1 6
6
7
5
1 0
0
0
0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 0
N o n e
P o s t I m p a c t
I n f li g h t
U n k n o w n
N u m b er o f A c c i d en t s
F i x ed - w i n g
R ot or - w i n g
Figure 1–20: Air Medical Accidents with Fires Adapted from: Frazer, AirMed, Sept/Oct 1999
11Special Supplement
involve a medical helicopter, but a cor-porate helicopter that suffered a tail rotorstrike upon departing a hospital helipadafter dropping off a passenger. TheNTSB, however, chose to include this asa medical helicopter accident eventhough, in this author’s opinion, the heli-copter did not meet the NTSB inclusioncriteria (aircraft dedicated to the EMSmission, has trained medical personnelon board, and the pilot is employed pri-marily to fly dedicated EMS missions).
The NTSB report identifies a total of88 accidents that occurred during theirdesignated study period (1978 through1986). Based upon their inclusion crite-ria, the study group was reduced to the47 “mission” accidents. The “IndustryReported EMS Accidents” came from sixdifferent reporting sources and theirrespective databases: American Societyof Hospital-Based Emergency AirMedical Services (ASHBEAMS, nowAAMS), Aviation Safety Institute,Hospital Aviation Magazine, FAAAccident/Incident Data System,National Emergency Medical ServicesPilots Association (NEMSPA), and theNTSB Accident database.
FLIGHT SAFETYFOUNDATION: HUMAN ERROR AS MAJOR CAUSE OF U.S.COMMERCIAL EMSHELICOPTER ACCIDENTS
Patrick Veillette, Ph.D., studied a totalof 87 HEMS accidents and 56 incidentsfrom January 1987 through December2000. Of interest, Veillette’s 2001 studyand his database begin where the 1988NTSB report concluded.
Much like Frazer’s studies, this compre-hensive report analyzes many commonfactors related to HEMS accidents,including phase of flight, weather-relatedaccidents, collisions with obstacles, andmechanical-related accidents. Veillette,however, also presents his findingsregarding human error as a major factorin HEMS accidents. His discussions alsoinclude the findings of numerous othersafety reports, FAA Advisory Circulars,and publications.
Veillette concluded that human errorwas associated with 66 of the 87 (76%)HEMS accidents studied. Of the fatalaccidents studied, human error was asso-ciated with 84% (27 of 32). Recurringhuman error factors identified and thenumber of accidents for which they werecited is shown in Figure 1–21.
Veillette also reviewed the effecthuman error had on the various phases offlight accidents. He found that humanerror accounted for 68% of the en routeaccidents, 91% of the accidents duringapproach and landing, and 82% of theaccidents during takeoff. Of theapproach and landing accidents, 41%were due to a collision with an obstacle,
while 50% of the takeoff accidents werecollisions with obstacles. Figure 1–22shows the most common phase of flightaccident categories.
As mentioned, Frazer and Veillettehave reviewed many of the same charac-teristics surrounding HEMS accidents.Although their findings are similar, sev-eral differences are noted. Since theNTSB database was the major source forboth studies, the variations must takeinto account the different years eachauthor included in their study and thefact that Frazer also included fixed-wingaccidents. Veillette’s review of1987–2000 excluded many of the yearswith the highest number of HEMS acci-
17
15
14
8
7
7
7
In-Flight Decision Making
Preflight Planning
Risk Taking
Inadequate Crew Coordination
Failure to Follow SOPs
Delayed Remedial Actions
Misinterpretation of Environmental Cues
Number of Accidents
41
22
22
28
20
18
21
3
3
En Route
Approach andLanding
Takeoff
Number of Accidents
Fatal Accidents
Human Error Accidents
Total Accidents
Figure 1–21: Human Error Factors in HEMS Accidents, 1987–2000Adopted from: Veillette, Flight Safety Digest, April/May 2001
Figure 1–22: Phase of Flight Accidents, 1987–2000Adopted from: Veillette, Flight Safety Digest, April/May 2001
12 AIR MEDICAL PHYSICIAN HANDBOOK
dents (1982–1986) that were included inFrazer’s study. However, Veillette mayhave been able to identify any significanttrends that may have changed after thehigh number of accidents of the early’80s.
Frazer’s 2001 study found that 18% ofthe accidents were the result of collisionswith obstacles, while Veillette’s reportidentified 31%. Another differencenoted was with regard to aircraft damageand post-impact fires. In Frazer’s study of122 aircraft accidents, 57% weredestroyed and 42% had substantial dam-age. Veillette’s data is nearly reversed,with 41% destroyed and 59% sustainingsubstantial damage. Frazer also foundthat 15% of the helicopter accidentsresulted in a post-impact fire, whileVeillette study found less than 6% result-ed in a fire.
Veillette identified 47% of accidentsoccurring en route compared to 36% inFrazer’s study. This difference, however,could be due to the number of categoriesidentifying the various phases of flight.Veillette listed only 5 categories (takeoff,en route, maneuvering, approach andlanding, ground), while Frazer listed 8(takeoff, cruise/en route, maneuvering,approach, landing, hover, descent,climb). Finally, as previously noted,Veillette found 76% of the accidents tobe associated with human error, whileFrazer’s study found this to be 65%.
Veillette made several references to thecrew in his analysis of the accident data-base. He found that inadequate crewcoordination was cited in eight accidents.Each of these accidents involved a colli-sion with obstacles that occurred duringtakeoff or landing. Other factors includ-ed incorrect or untimely information anddistracting comments or movements bythe medical crew during a critical phaseof flight. He also cited four accidentsand six incidents that were caused bycowlings and panels that separated fromthe helicopter. In half of these events,the medical crew had closed the cowlingsimproperly prior to flight.
Unique to this study, Veillette reportson his personal observations made duringmore than 400 HEMS transportsbetween 1995 and 2000. His observa-tions were categorized by the type offlight—scene response (128), interhospi-
tal transfer (58), and repositioning (247).It was not mentioned if these observationflights were all with the same program orwith different programs. Veilletteobserved that during none of the trans-ports was the medical crew wearing hel-mets, while the pilots wore helmets morethan 70% of the time. The author didnot mention whether helmets were avail-able to the medical crew or to all of thepilots.
Most surprising was his observationregarding the use of seat belts and shoul-der harnesses by the medical crew. In100% of the repositioning flights, themedical crew was wearing seat belts dur-ing takeoff, but only 45% were wearingshoulder harnesses. However, on takeofffrom scene responses, only 11% wereobserved to be wearing seat belts andonly 4% had shoulder harnesses fastened.The percentage was even lower on take-off from interhospital transfers, with seatbelts buckled only 6% of the time andshoulder harnesses only 4%. While thiswould appear to be a serious breech ofroutine safety protocols, Veillette onlycomments that “because of their in-flightmedical duties, medical crewmembers fre-quently are not seated in energy-absorb-ing seats with their seat belts and shoul-der harnesses fastened.”
Veillette’s article concludes with acomplete listing of the accidents andincidents included in his database. Inthis same issue of Flight Safety Digest,Veillette offers a second comprehensivestudy that specifically addresses EMS air-plane accidents. This second article,which spans 1983–2000, may be the mostcomplete analysis of fixed-wing air med-ical accidents available. Both of thesearticles provide excellent informationand detailed insight into various aspectsof air medical accidents.
EMERGENCY MEDICALSERVICE HELICOPTERSINCIDENTS REPORTED TOTHE AVIATION SAFETYREPORTING SYSTEM
All too often, helicopter accidentsinclude pilot fatalities. Many of these
accidents do not provide investigatorswith adequate and complete informationas to the chain of events that led to theaccidents. With the high number of fatalHEMS accidents, information from alter-native perspectives could be helpful toidentify potential problems and preventfuture accidents. One such perspective isreports of aviation incidents that did notresult in accidents. The U.S. NationalAeronautics and Space Administration(NASA) Aviation Safety ReportingSystem (ASRS) has the world’s largestdatabase on aviation incidents and servesas an important resource for this alterna-tive perspective.
The NASA-Ames Research Centersearched the ASRS database for reportsrelated to EMS helicopter incidents.From 1986 through 1991, 68 of 81HEMS incident reports were consideredrelevant and were included in their study.These reports, which were voluntarilysubmitted by EMS helicopter pilots, airtraffic controllers, and pilots of other air-craft, (i.e., anyone who observes an inci-dent) often included the crucial “chainof events” and the successful resolutionsof the incidents. The benefit of thesereports was that they enabled the pilot toreport what he/she was thinking at thetime of the incident.
The objectives of the Ames study wereto: (1) identify the types of safety-relatedincidents reported to ASRS in EMS heli-copter operations; (2) describe the oper-ational conditions surrounding theseincidents, such as weather, airspace,flight phase, and time of day; and (3)assess the contribution to these incidentsof selected human factors, such as com-munication, distraction, time pressure,workload, and flight/duty impact.
The type of information obtained fromthe ASRS incident reports and their nar-ratives is very different from the type ofaccident data already presented. Thedata were evaluated according to inci-dent variables, operational variables, andhuman factor variables.
Incident Variables
The Ames report found that non-adherence to the Federal AviationRegulations (FARs) was identified in53% of these reports. This included vio-lations of flight/duty limitations, mainte-
13Special Supplement
nance requirements, and so on.Airspace violation ranked second(23%). Conflict or near-midair collision(NMAC) and in-flight encounter withinstrument meteorological conditions(IMC) were reported in 14% of thereports. It should be noted that 22% ofthe reports identified problems thatcould not be specifically addressed byother incidents. Figure 1–23 lists theanomalies reported in the ASRS reports.
Operational Variables
This category includes phase of flight,weather conditions, time of day, andtype of airspace involved at the time ofthe incident. Most commonly, HEMSincidents occurred during cruise andduring good weather. The time of daymost often involved was from 1201 to1800 hours, which generally correspondswith the busiest time of day for flightprograms. The reported incidentsoccurred in all types of controlled anduncontrolled airspace.
In comparing their findings with the1988 NTSB report, the Ames reportfound several similarities and one signif-icant difference. The Ames report
found that in comparing the two studies,the weather conditions (i.e., unplannedentry into IMC), airspace, phase offlight, and experience levels of the pilotswere similar. In addition, the qualityand interpretation of weather informa-tion was noted as a concern in bothstudies. In the Ames report, they foundthat pre-flight weather briefings hadbeen obtained in 80% of the incidents,but 75% of the briefings did not matchthe actual weather conditions the pilotsencountered in flight.
One difference noted between theNTSB and Ames report was IFR curren-cy. In the ASRS study, 68% of theHEMS pilots had an instrument ratingand 66% were IFR current at the time ofthe incident. In contrast, in the NTSBreport 86% of the pilots were IFR-rated,while only 6% were IFR current. TheAmes report concluded that “this find-ing appears to be a compelling reason toadvocate IFR currency for EMS pilots,although additional research is necessaryto reach this conclusion because of thelimitations of the ASRS data.” The narratives in these ASRS incidentsreported that IFR rating and currencywere very helpful, if not invaluable.
53%
23%
22%
14%
14%
9%
9%
9%
6%
6%
6%
4%
4%
1%
1%
1%
Nonadherence to FARs
Airspace Violation
Other
Near midair collision
Encounter with IMC
VFR in IMC
Aircraft Equipment
Nonadherence with Clearance
Nonadherence to Procedures
Nonadherence to Other
Airborne Confict
No Specific Anomaly
Inflight Encounter
Separation Conflict
Runway Transgression
Altitude Deviation
Percentage of Reports
Figure 1–23: Frequency of ASRS Reported Anomalies HEMS, 1986–1991 (n=68)Adapted from: Connell, Flight Safety Foundation, 1995.
Background Information:Instrument Rating,
Currency, and Proficiency
Rating: The FAA defines “rating”as “a statement that, as a part of a(pilot’s) certificate, sets forth specialconditions, privileges, or limitations.”To achieve an instrument rating, apilot must pass a written test and aflight test given by the FAA. Uponpassing these tests, the instrumentrating becomes part of the pilot’s cer-tificate (license) and the pilot is nowlegal to fly under instrument condi-tions for the next six months.
Currency: To maintain currency,a pilot must have flown, within thepreceding six months, at least six (6)instrument approaches in a helicop-ter, completed holding procedures,and intercepted and tracked coursesthrough the use of navigation sys-tems. The flight experience must berepeated at least every six months tomaintain IFR currency.
It is important to note that an IFRflight must be to a location (e.g., anairport or helipad) that has anauthorized instrument approach.IFR flight does not facilitate travel tothe scene of an accident or to mosthospitals. Recent technology, how-ever, has enabled dozens of hospitalsto develop these instrumentapproaches.
Proficiency: If a pilot does notmeet the instrument experience out-lined in “currency” within the pre-scribed time (i.e., the preceding sixmonths) or within 6 calendarmonths after the prescribed time,he/she will not be able to fly underIFR conditions until he/she passes aninstrument proficiency check. Thismust be in an appropriate aircraft(i.e., for helicopter pilots, it must ina helicopter) and must be given byan examiner, a company check pilot,an authorized flight instructor, or anindividual authorized by the FAA toconduct instrument practical tests.Once the pilot has passed this checkhe/she may fly under instrumentconditions for the next 6 months.
14 AIR MEDICAL PHYSICIAN HANDBOOK
Human Factor Variables
Concerns related to communications,time pressure, and distractions werereported at very high rates as seen inFigure 1–24. In addition, workload andflight/duty conditions were also identifiedin the ASRS reports.
Communications
Of the communications incidents, 60%involved pilot-air traffic control (ATC)communications. Another 13% werecommunications problems between pilotsand weather services (i.e., poor or inac-curate weather information that becamea major contributor to in-flight encoun-ters with IMC). Communication diffi-culties between pilots, HEMS dispatchersand ground personnel (e.g., police, fire-fighters, paramedics, ground crew, main-tenance) were also reported as a frequentproblem, especially if they interfered withATC communications.
Time Pressure
Time-related pressures were cited as afrequent contributor to the ASRS inci-dents. These pressures centered aroundfour different considerations: patient con-dition, rapid mission preparation, flightto the patient pick-up location, and lowfuel. Patient condition was reported 44%of the time and was the most importantcontribution to time pressure. The criti-cal condition of a patient could create asense of maximum urgency. As a result,preflight planning may be inaccurate orpreflight inspections and checklist maybe hurried and incomplete. Other reportscited such oversights as not stopping forrefueling; failure to obtain or review cor-rect charts; overflying scheduled aircraftmaintenance; inadequate or less-than-thorough weather briefings; and inade-
quate evaluation of weather briefings pre-ceding the go/no-go decision. The Amesreport found that time pressure associatedwith the patient’s condition seemed to bepresent regardless of whether the patientwas already onboard the aircraft or thepilot was en route to patient pick-up.
Most programs strive to isolate thepilot from knowing any medical informa-tion so that their flight decisions aremade objectively. Unfortunately, we maynot be as successful as we would hope.HEMS pilots are well aware that theirservices are generally requested whenthere is a critically ill or injured person inneed of transport. In addition, the pilotmay be faced with a sense of urgencyfrom both verbal and/or nonverbal sig-nals from the medical crew. One ASRSreport stated, “No flight is so importantthat the lives of the flight crew should bejeopardized due to incomplete or inaccu-rate preflight planning.”
Distractions
Distraction from the primary task offlying the aircraft was reported in manyof the ASRS incidents. External factorscreated many distractions that were citedin the reports. These included in-flightaircraft equipment problems, the need tomonitor multiple radio frequencies, traf-fic avoidance in high-density trafficareas, interruptions, radio frequency con-gestion, poor visibility due to haze ornight operations, marginal weather, noisefrom on-board medical equipment, andimpending low-fuel situations. Many ofthese distractions could also lead to time-pressure situations.
Internal factors were also reportedwhich led to significant distraction. Thisincluded personal or family-related con-cerns, anxiety in the current situation,disorientation, involvement in patient
condition, confusion about procedure,and general inattention.
Distractions can lead to accidents. Inour daily activities, it is common to try todo multiple things at the same time—tomultitask. While driving, it seems tomake perfect sense to get something elsedone. Many people use their handheldcellular telephones. There are severalrecent studies that indicate a strong cor-relation between the use of cell phonesand the increased probability of an autoaccident due to the distraction.
Aviation is no different. The idea ofthe sterile cockpit began in commercialairlines and has been around for years.By eliminating unnecessary talking dur-ing critical stages of flight, we can reducedistractions and improve our safetyrecord significantly. Medical crewmem-bers, patients, passengers, and even otherpilots, despite good intentions, can besignificant distracters.
Workload and Flight/DutyConsiderations
While workload (12%) and flight/dutyconsiderations (4%) were reported in theASRS incidents, the Ames study con-cluded that they were not a significantcontributor to any HEMS incident.
However, workload, flight/duty length,crew rest, and the number of duty dayscan influence many factors, includingjudgment, error recognition, concentra-tion, forgetting tasks, fatigue, and ulti-mately can lead to aviation incidents.
An unexpected finding in the HEMSASRS reports was that cruise flight was acommon time for HEMS safety incidents.Airspace violations and near-midair colli-sion most frequently occurred in cruiseflight and in VFR weather. In-flightweather encounters were also reported asoccurring most often in cruise flight.During cruise, it would be anticipatedthat cockpit activity would be low—unlike takeoff and landing. However, itappears that the HEMS pilot might beattending to tasks inside the cockpit,rather than watching for conflicting traf-fic, low clouds, or airspace boundaries.These cockpit activities might includeproviding position reports to dispatch,coordinating with the medical center,programming navaids, or communicatingwith other EMS personnel.
78%
64%
60%
12%
4%
Communications
Time Pressure
Distraction
Workload
Flight/Duty Conditions
Percentage of Reports
Figure 1–24: Human Factors in HEMS Incident Reports (n=68)Adapted from: Connell, Flight Safety Foundation, 1995.
15Special Supplement
The Ames report focused on theunique demands placed on the HEMSpilot that led to distraction and timepressure. It concluded that thesedemands could compromise good com-munications, thorough planning, cooper-ative teamwork, and safe flight duringpatient transport. It recommends thatsteps need to be taken to improve com-munication, decrease distraction,decrease time pressure to realistic levels,and assist in workload management.
Ames proposed that Crew ResourceManagement (CRM) was not just formajor airlines or big companies. Effectivecommunications among all HEMS teammembers—pilots, flight nurses, para-medics, doctors, administrators, andcommunication specialists—are vital ifthe HEMS team is to perform its dutiesefficiently, successfully, and safely. CRMwill be discussed in more detail inSection 4 of this report.
AIR MEDICAL ACCIDENT ANALYSIS
In April 2000, an Air Medical SafetySummit was convened to address the ris-ing number of air medical accidents.From this Summit of industry leaders andsafety experts came the Air MedicalSafety Advisory Council (AMSAC), theAir Medical Service Accident AnalysisTeam and several other initiatives.
The Air Medical Service AccidentAnalysis Team was created to study pastaccidents, analyze the root causes ofthese accidents, and identify effectiveand feasible interventions that wouldprevent future HEMS accidents.Chaired by Richard Wright, Jr., of theHelicopter Association International,the main focus of the Team was humanfactor accidents. It was felt that identi-fied interventions in this area mighthave the greatest impact on accidentprevention and safety.
The Team identified 20 HEMS acci-dents between November 1993 andNovember 1999 whose Final NTSBAccident Reports included extensivedata for review and evaluation. Theprocess used to examine these accidentsis referred to as “root cause analysis.”
Event SequenceA timeline of events was developed for
each flight from the available accidentreport. This included all aspects of thetransport, from prior to the flight, duringthe flight, and ending with the accidentitself.
Problem StatementsAny and all “problems” that could
have contributed to the accident werethen identified. Among the 20 flights, atotal of 56 individual problems wereidentified. They were then classified as:pilot performance issues (23), aircraftissues (9), infrastructure issues (9), envi-ronmental issues (6), landing zone issues(5), and corporate and/or program man-agement issues (4), as identified inFigure 1–25.
Intervention Strategies and Ratings
Interventions were next identified foreach problem to determine what mighthave prevented the problem, potentiallyaverting the accident. A total of 65unique interventions were proposed,which were categorized as TrainingInterventions, Equipment Interventions,Air Traffic Management Interventions,Regulatory or FAA-sponsoredInterventions, National Airspace (NAS)or Infrastructure Interventions, orMiscellaneous Interventions.
Each intervention was evaluated andscored by the Team for its effectiveness,yielding a ranking, or Effectiveness Score.The combined score was then dividedinto thirds to group the interventionsthat rated High (21–17), Moderate(16–13) and Low (12–8) Effectiveness.
The next step was to evaluate andscore each Intervention to determine it’stechnical feasibility, financial feasibility,regulatory feasibility, and operational fea-sibility. A range from 48 to 13 wasobtained for the 65 interventions. Asbefore, the combined scores were thendivided into thirds, to rank the interven-tions as High (48–37), Moderate (36–32),and Low (31–13) Feasibility.
Recommendations
The results were combined into amatrix, with interventions classified intonine different categories. It was theTeam’s recommendation that AMSACfocus efforts within the air medicalindustry to develop implementationstrategies for those interventions thatranked highly effective and highly feasi-ble. Recommended for considerationwere those interventions that were iden-tified as highly effective but moderatelyfeasible, highly feasible but moderatelyeffective, and those that were moderatelyeffective and moderately feasible.Interventions classified as low effective-ness and/or low feasibility were not rec-ommended by the Team for implementa-tion or further pursuit. Figure 1–26 liststhe nine categories and the Team’s rec-ommendations to AMSAC.
This document is an excellent resourcethat takes a detailed and comprehensivelook into the specific problems andevents that have led to previous HEMSaccidents. The Team has also provided acomprehensive list of specific technolo-gies, training, regulations, and opera-tional enhancements and rankings oftheir effectiveness and feasibility as inter-ventions that can improve the safety ofair medical transport and reduce thenumber of accidents. However, in mak-ing their recommendations, the Teamdivided the scores into thirds to yieldtheir High, Moderate, and Low ratings. Indoing so, it is inevitable that some inter-ventions will miss a cut-off by a merepoint or two. For example, full motionsimulators, improve safety programs andimprove safety cultures were each classi-fied as “High Effectiveness” but fell onepoint short of Moderate Feasibility.Following the specific recommendationsof the Team, these highly effective inter-ventions would not be pursued. Clearly,the most highly ranked interventions inboth effectiveness and feasibility shouldbe investigated very carefully for possibleimplementation. However, operators,manufacturers, associations, and programsshould review this entire list of possibleinterventions to identify opportunities,such as full-motion simulators, that arewithin their capacity for improvementand enhanced safety.
16 AIR MEDICAL PHYSICIAN HANDBOOK
Figure 1–25: Consolidated Problem StatementsAdopted from: Air Medical Accident Analysis, 2001
Pilot Performance Issues:
• Loss of situational awareness• Poor aeronautical decision making• Limited experience in make/model• Flight check not conducted in operational type of aircraft• Pilot disregarded company policies• Inadequate preflight planning• Pilot failed to obtain weather briefing• Pilot ignored weather briefing• Pilot not wearing helmet• Pilot continued VFR flight into IMC conditions• Pilot descending to avoid IMC• Pilot fails to maintain safe altitude• Pilot fails to conduct area recon• Pilot fails to conduct pre-departure briefing• Improper response to inflight emergency• Inadequate Nr (rotor RPM) control• Pilot failed to recognize and avoid power settling• Improper pilot technique• Pilot took off with sun in eyes• Destination position not entered in navigation equipment• Pilot failed to use aircraft searchlight to detect wires• Pilot failed to hear or respond to ATC special VFR clearance• Pilot’s attention is diverted to inside the cockpit
Environmental Issues:
• Night VFR operations• Night IMC operations• Reduced visibility• Mountain operations• High altitude operations• Featureless terrain
Aircraft Issues:
• Aircraft not IFR certificated• No autopilot or second pilot• Poor configuration of navigation equipment • Pilot unable to determine altitude above LZ• Pilot unable to detect weather• Pilot unable to detect wires• Misleading/inaccurate fuel quantity gauge• Aircraft flotation inadequate for existing sea conditions• Uncrashworthy fuel tank
Infrastructure Issues:
• ATC unclear regarding pilot’s request• Inadequate vector by ATC to intercept localizer• Pilot unable to obtain ATIS (Automated Terminal
Information Service) information• Airport uncontrolled• Airport congested, requiring landing on ramp• Helipad small• Helipad surrounded by obstacles• Powerlines did not meet marking criteria• Powerlines not depicted on aeronautical charts
Landing Zone Issues:
• Difficulty identifying landing zone• No landing site supervisor• Incorrect/inadequate obstacle information on LZ• Congested landing zone• Obstacle-rich environment
Corporate/Management Issues:
• Corporate pressure to complete the mission• Personal pressure to complete the mission• “Ready Aircraft” change required equipment transfer• Preflight preparations rushed
AIR MEDICAL ACCIDENT ANALYSIS: CONSOLIDATED PROBLEM STATEMENTS
17Special Supplement
High
• Enhance the training for night flying operations
• Enhance the training for mountain flying operations
• Equip aircraft with Terrain AvoidanceWarning Systems (TAWS)
• Equip aircraft with radar altimeters• Provide aircraft with mission-essential
equipment• Improve the content of weather
briefings
• Conduct/enhance annual IFR proficiency checks
• Conduct/enhance training to improve the understanding of weather briefings
• Enhance overall training: recurrent, professional knowledge, etc.
• Conduct/enhance training in ADM• Establish an integrated and structured
Pilot Training Program• Conduct/enhance mission-oriented
training• Conduct/enhance CRM training• Equip aircraft with Moving Map
Displays to provide weather, obstacle, and terrain data
• Equip aircraft with avionics to provide avertical awareness display or warning
• Standardize cockpits of similarmake/model used in similar operations
• FAA to enhance/improve contents ofannual IFR proficiency checks
• Establish national criteria for the marking of wires and towers
• Horizontal Awareness From Terrain • Synthetic vision• Heads-up display• Night vision devices• Full-motion simulators• Enhance visibility/detection of wires
and towers• Change corporate mind-set• Improve safety culture• Improve safety program
Low
• Readily available crew/passengerbriefing cards
• Fuel flow indicators• Simplify calling FSS• Publish a mountain flying
advisory circular• Publish a “flat light/whiteout”
advisory circular• Require flight plans• Provide more UNICOM
frequencies
• Improve pilot handbooks• Data link technology• Require annual calibration of fuel
quantity gauges
• Increase dual-pilot time prior tosolo PIC
• Increase time requirements for “mission certification”
• Obstacle database• Enhanced ice detection equipment• Raise minimums for night
instrument approaches• Require ATC monitoring of
instrument approaches• Prohibit night VFR• Update FAR Part 135 requirements• Require crashworthy fuel tanks
for certification
Moderate
• Enhance the awareness of accident causes• Improve physiological training• Improve training with avionics
equipment: usage, capabilities, etc.• Improve weather radar• Encourage greater utilization, interaction
with and assistance from Air TrafficManagement
• Improve/enhance training of ATC personnel in rotorcraft operations and capabilities
• FAA to enhance training elements ofBiennial Flight Reviews and Pilot Training Standards
• Operators to enhance training for Biennial Flight Reviews and PilotTraining Standards
• Develop helicopter-specific, mission-specific computer-based emergency procedures simulators
• Develop satellite-based Communications,Navigation and Surveillance (C/N/S)technology
• Increase the rate of commissioning of new AWOS/ASOS (Automated WeatherObserving System/Automated SurfaceObserving System) facilities
• Improve aeronautical charts (symbology, data, etc.)
• ADS-B (Automatic DependantSurveillance-Broadcast) Technology
• Automated voice call-outs• Over-bank warnings• Excess terrain closure warnings• improve equipment with state-of-the-art
technology• Prohibit night flying by non-IFR rated
pilots• Require human factors/ergonomics in
cockpit designs
EFFECTIVENESSFEA
SIB
ILIT
Y
Adopted from: Air Medical Accident Analysis, 2001
Hig
hLo
wM
od
erat
e
18 AIR MEDICAL PHYSICIAN HANDBOOK
HEMS ACCIDENTS ANDINCIDENTS, 1998 TO 2001
The past four years has seen an alarm-ing increase in the number of HEMS acci-dents across the nation. No one is certainwhy there have been 43 accidents (plusone dual-purpose medical helicopter acci-dent) in four years. From 1987–1997,dedicated HEMS has averaged 4.9 acci-dents per year. The past four years aver-aged 10.75 per year. Not since the fouryear period of 1983 to 1986 have we seensuch a large number of accidents.
While a few of the reports that wereviewed included accidents from 1998 to2000, none of these studies isolated theseyears from the rest of their accidents.Therefore, with no published review ofthe most recent series of accidents, wedetermined that it was necessary to do ourown preliminary review of these accidentsto see if any trends could be identified.
Several resources and databases werereviewed in the development ofAttachment 1. This table presents thepertinent data on 44 HEMS accidentsbetween 1998 and September 30, 2002.
Attachment 1 does not include anyfixed-wing EMS. In addition, we havenot included an accident that killed threecrewmembers in England, a non-EMS hel-icopter that crashed on a hospital helipad,or an “as needed” service that crashedafter picking up three firemen to respondto a medical emergency in Hawaii. Finally,two additional accidents were not includ-ed which involved medical helicoptersthat had been “out-of-service” for at leastseveral days and were undergoing post-maintenance test flights.
Of the 44 accidents from 1998 to 2001,15 (34%) resulted in at least one fatality.For the 1980s, our accident databaseshowed 39% of all accidents resulted in atleast one fatal injury. From 1990–1997the rate increased to 47%. Despite theincrease in accidents over the past fouryears, the percentage of fatal accidents hasdeclined by nearly 30%. Figure 1–27depicts the percentage of fatal accidentssince 1980. As the figure shows, the lastfour years has had the lowest consecutive4-year fatality rate in HEMS history sincethe early 1980s.
Looking at the breakdown of all injuries(Figure 1–28), there is also a significantdifference noted in the percentage of
crewmembers who sustained no injuries inHEMS accidents. Figure 1–28 shows fourdifferent comparisons: the entire 21-yearstudy period; the 1980s; 1990–1997; andfinally 1998–2001. The percentage offatal injuries has decreased in 1998–2001and we now see 47% of the crewmembersand passengers sustained no injuries.
Since 1998, a slightly higher percent-age of HEMS accidents occurred at night(52%). More accidents occur during thecruise phase of flight (16) than any otherphase and we now see a similar number ofaccidents occur on takeoff and landing (8each). More accidents (45%) are takingplace during scene missions than previous-ly noted (35%). Finally, our preliminaryreview of the HEMS accidents seems toshow a decrease in the percentage ofweather-related accidents. Frazer had
noted an increase in weather-related acci-dents from 22% in the ’80s to 32% in theearly and mid-’90s. From 1998 to 2001,there appear to be 6 weather-related acci-dents, dropping the percentage to 14%.
20%
67%
25% 25% 25%17%
50%
31%
100%
25%
57%
0%
57%
29%
67% 67%
14%
100%
67%
38%30%33% 31%
0%
20%
40%
60%
80%
100%
Year
Fat
alA
ccid
ents
(%)
34% 34%
41%
30%
11%9%
12%15%15%
18%
14%
9%
39% 38%
34%
47%
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
Fatal Serious Minor No Injury
Figure 1–27: Percentage of Fatal Accidents per Year
Figure 1–28: Severity of Injuries for HEMS Accidents, 1980 – 2001 (1980–2001, n=494; 1980–1989, n=247; 1990–1997, n=116; 1998–2001, n=131)
HEMS ACCIDENT ANDFATAL ACCIDENT RATES
The UCAN Safety Committee felt thatan important aspect of our research wouldbe to determine HEMS accident and fatalaccident rates for our entire study period(1980–2001). In order to normalize theraw accident data for a meaningful com-parison of dedicated HEMS accidents, it isnecessary to have two key elements—thenumber of accidents (total and fatal) andthe number of flight hours flown or thenumber of patients transported for the
19Special Supplement
entire industry each year. We havealready reviewed the accident data that isreadily available from several sources.Unfortunately, as previously mentioned,a major obstacle is our lack of accurateHEMS exposure data (e.g., flight hours,patients transported) making it impossi-ble for a meaningful year-to-year compar-ison. Dual-purpose medical helicopteraccidents are not included in this portionof the report.
Methodology
Our research model required that anextensive review of the air medical litera-ture be conducted to determine whatdata was available that could be used tonormalize the HEMS information. Wefound that in 1982, Hospital Aviationbegan to track and publish statistics andaccident rates that were based on thenumber of patient transports (rather thanflight hours) for the air medical industry.In 1988, Hospital Aviation also looked atrevenue flight hours as well as patienttransports. Their survey found that theaverage patient transport corresponded to1.05 revenue flight hours. Because ofthis finding, accident rates continued tobe tracked based upon patient transports.No data were collected on total hoursflown (training, repositioning, mainte-nance, PRs, etc.). In the journals, the“per 100,000 patients transported” acci-dent rates were found through 1992 andresurfaced again in a 1997 article thatprovided a graph for the years 1978through 1995.
“Annual Transport Statistics,” theresults of another industry-wide survey,was published first in Hospital Aviation,and then in the Journal of Air MedicalTransport, the Air Medical Journal; andthen in AirMed. This analysis providesinformation on averages per flight pro-gram (as well as highs and lows) for theyear, region-by-region, and across theUnited States. This included informa-tion on the average number of patientstransported, interfacility missions, scenemissions, night flights, and loaded miles.In 1994, they also started to includeaverage flight hours in the survey results.Personal correspondence with Bill Rau,who had authored the transport surveyarticles since 1996, stated that the surveyasked for “total annual flight hours
excluding PR flights.”A final source of HEMS statistics was
the “Mid-Year Report” that identified(among other related data) the numberof hospital-based programs and helicop-ters. This data was found from 1984through 1989 in Hospital Aviation andsubsequently in the Journal of Air MedicalTransport. Several additional articles andthe 1988 NTSB report were alsoreviewed to provide raw data.
Total flight hours and total patienttransports were unavailable for morethan 50% of the years reviewed. Fromthe “Annual Transport Statistics” we didhave the average number of patientstransported per program per year and theaverage flight hours per program.Unfortunately, we lacked any estimate as to the number of HEMS programs inoperation since 1992. In order to normal-ize the HEMS data, the missing informa-tion would need to be obtained or estimated.
In an effort to estimate the number ofHEMS programs, several assumptionswere made. If we could determine thenumber of programs and helicopters cur-rently in operation (November/Decem-ber 2000), we could estimate the totalflight hours and patients transported dur-ing the year. In addition, unlike theearly 1980s, we assume that the growthin the industry has been fairly constantover the past eight years. We would thenbe able to estimate the number of pro-grams and helicopters for the years lack-ing data.
To determine a fairly accurate numberof dedicated HEMS programs and heli-copters, we undertook a review of theAssociation of Air Medical Services(AAMS) Membership Directory and theDirectory of Air Medical Programs, pub-lished in the May/June issue of AirMed.Unfortunately, neither resource is com-plete. Not all air medical programs areAAMS members and neither directoryincludes the total number of aircraft. Tosupplement this data, a survey was postedon the Internet’s Flightweb listserve. Thesurvey requested state-by-state informa-tion on the number of dedicated HEMSprograms (hospital-based, independent,etc.) and the number of dedicated helicopters.
Knowing the flight hours is the nextconsideration in an attempt to normalize
the HEMS data. The NTSB documentreported flight hours for 1980 through1985, but did not specify if their figurescorrespond to revenue flight hours ortotal flight hours. In addition, theAirMed surveys provide average flighthours per program from 1993 to 2000. Ifwe are successful in estimating the totalnumber of HEMS programs in operationfrom 1993 to 2000, we will be able toestimate the total flight hours.
Total flight hours is an important con-sideration in our evaluation. In HEMS,a considerable number of accidentsoccurred on non-patient missions—including PRs, refueling, maintenance,training, and so on. The average flighthours per program, as published in theAirMed surveys, should have included allflight hours except PRs. The vast major-ity of HEMS flight hours are patient mis-sions. Personal correspondence with sev-eral HEMS programs and operators sug-gests that non-patient flight time mayrange from 5 to 15%. PR flights made up approximately half of this for manyprograms.
In an effort to more accurately deter-mine the average number of total flighthours per program, a survey of five of thelargest HEMS operators was conducted.These operators were chosen with theknowledge that they account for themajority of the air medical programs inthe country. The operators were askedfor total flight time, total number of pro-grams and total number of helicoptersthey operated for the years 1998 through2000. The goal was to compare theseflight hours with the information pub-lished in AirMed for 1998 and 1999. Inaddition, a correlation could then bemade comparing revenue flight hours and total flight hours for other years.
Results
In review of the available publishedstatistics, several limitations and prob-lems were identified that make accurateyearly comparisons very difficult. Mostimportant, the “Annual TransportStatistics” report only on hospital-basedHEMS programs. Over the years, moreand more HEMS operations have deviat-ed from this original, yet still dominant,model. In 1989, the journals began send-ing surveys to non-hospital helicopter
20 AIR MEDICAL PHYSICIAN HANDBOOK
14 Year 72–79 1980 1981 1982 1983 1984 1985 1986 1987 1988
15 HEMS Accidents 5 3 4 8 8 6 12 13 4 8
16 Dual-Purp. Acc.
17a Accident Rate (TFH) 14.46 14.25 21.74 17.69 10.62 16.71 14.20 3.47 6.50
17b Accident Rate (-PR) 14.46 14.25 21.74 17.69 10.62 16.71 15.41 3.76 7.05
17c Difference 0.00 0.00 0.00 0.00 0.00 0.00 1.21 0.29 0.55
18a % programs w/ accidents 9.4% 10.8% 16.3% 12.9% 7.9% 11.9% 10.1% 2.8% 5.2%
18b % helicopters w/ acc. 7.7% 8.9% 12.9% 10.7% 6.6% 10.1% 8.6% 2.2% 4.1%
19 Fatal Accidents. 1 2 1 2 2 1 6 4 4 2
20 Fatal Dual-Purp.
21 Fatal Acc Rate 9.64 3.56 5.44 4.42 1.77 8.35 4.74 3.76 1.76
22 % Fatal Acc. 0.20 0.67 0.25 0.25 0.25 0.17 0.50 0.31 1.00 0.25
23 per 100,000 pts 16.57 17.16 15.99 24.98 19.47 11.57 17.47 14.89 3.81 6.62
Figure 1–29: HEMS Program, Helicopter and Accident Statistics (Published data is in bold italics)
Key to Lines:2. Survey results are routinely published in the following year. 3. 83–99, published data. 80–82, calculated (total pts flown / # of programs) 4. 83–99, published data.5. 93–99, published data. 86–92, calculated (ave # pts transported per program X 1.05).6. 88, published data. 93–99, calculated (ave hrs per program / ave pts flown per program) 7. 98–2000, operator survey. 86–97, calculated (line 5 X 1.085)
1 Data for Year 72–79 1980 1981 1982 1983 1984 1985 1986 1987 1988
2 Year Data was Published [1984] [1985] [1986] [1987] [1988] [1989]
3 Avg Pts Flown / Program 546 676 654 553 570 590 623 698 697
4 Avg Loaded Miles 61 63 61 61 61 58
5 Avg Hrs Flown / Program See total hours below (Line 8a) 654 733 732
6 Avg Flt Hr / Pt Transport 1.05 1.05 1.05
7 Avg Total Hrs/Prog 710 795 794
8a Total Flight Hrs (-PR) 20,750 28,071 36,794 45,233 56,516 71,831 84,385 106,271 113,437
8b Total Flight Hrs 20,750 28,071 36,794 45,233 56,516 71,831 91,558 115,303 123,079
9 # of Programs 32 37 49 62 76 101 129 145 155
10 # of Helicopters 39 45 62 75 91 119 151 184 195
11 Total Pts Flown 30,168 17,483 25,013 32,027 41,097 51,855 68,694 87,299 105,000 120,900
programs, but the data was never includ-ed in the published average statistics. In1989, the total number of patients trans-ported by these non-hospital programswas referenced. However, these numbersincluded some dual-purpose helicopterprograms as well.
Another problem was timing of thedata collected. Some statistics regardingthe number of helicopter programs andtotal number of helicopters were pub-lished in Hospital Aviation in a mid-year(July) report, while other program and
helicopter statistics were based on thecalendar year. The Annual TransportSurvey was based on the calendar yearuntil 1993 when it was switched to theacademic (July to June) year for annualstatistics tabulation. In general, annualaccident statistics were based on the cal-endar year.
Since 1986 the journal surveys weresent to both helicopter and fixed-wingprograms. The percentage of surveysreturned ranged from a high of 96% to alow of 33%. In general, the yearly data
was fairly consistent. The one exceptionwas 1994, which had significantly higheraverage flight hours per program andmuch lower loaded miles than other years.
The results of the literature search andthe following calculations can be foundin Figure 1–29. The data obtained fromthe various publications is presented inbold italics while the calculations andsurveyed data is in plain type.
Helicopter EMS Programs andHelicopters. Data was available for thenumber of HEMS programs from 1972 to
21Special Supplement
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
[1990] [1991] [1992] [1993] [1994] [1995] [1996] [1997] [1998] [1999] [2000] N/A N/A
705 709 869 876 805 812 813 845 827 796 880 832 882
60 58 57 55 49 33 47 47 47 55 58 50.76 51
740 744 912 920 876 1159 785 823 823 814 821 813.22 911
1.05 1.05 1.05 1.05 1.09 1.43 0.97 0.97 1.00 1.02 0.93 0.98
803 808 990 998 950 1258 852 893 893 854 921 841 957
122,141 129,534 162,416 169,141 166,245 226,778 158,221 170,727 175,573 178,447 184,839 187,854 210,396
132,523 140,545 176,221 183,518 180,376 246,054 171,670 185,239 190,497 187,216 207,327 194,271 217,584
165 174 178 184 190 196 202 207 213 219 225 231 231
213 231 225 242 259 276 293 309 326 343 360 377 400
125,200159,027 154,682 161,087 152,771 158,881 163,865 175,291 176,427 174,501 198,098 192238 203,772
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
7 1 7 7 3 6 7 1 3 8 10 12 13
2 1 1
5.28 0.71 3.97 3.81 1.66 2.44 4.08 0.54 1.57 4.27 4.82 6.18 5.97
5.73 0.77 4.31 4.14 1.80 2.65 4.42 0.59 1.71 4.48 5.41 6.39 6.18
0.45 0.06 0.34 0.32 0.14 0.21 0.35 0.05 0.13 0.21 0.59 0.21 0.20
4.2% 0.6% 3.9% 3.8% 1.6% 3.1% 3.5% 0.5% 1.4% 3.6% 4.4% 5.2% 5.6%
3.3% 0.4% 3.1% 2.9% 1.2% 2.2% 2.4% 0.3% 0.9% 2.3% 2.8% 3.2% 3.3%
4 0 4 2 2 4 1 1 2 3 3 4 4
1 1 1
3.27 0.00 2.46 1.18 1.20 1.76 0.63 0.59 1.14 1.68 1.62 2.13 1.90
0.57 0.00 0.57 0.29 0.67 0.67 0.14 1.00 0.67 0.38 0.30 0.33 0.31
5.59 0.63 4.53 4.35 1.96 3.78 4.27 0.57 1.70 4.58 5.05 6.24 6.38
8a. 72–85, published data. 86–92, 93–99 calculated ( # of program X Avg flight hrs per program).8b. 72–85, published data. 86–2000, calculated (line 7 X line 9)9. 72–92, published data. 2000, Flightweb survey. 93–99, calculated assuming consistent growth for 92–0010. 81–91, published data. 2000, Flightweb survey. 92–99, calculated assuming consistent growth for 91–00.11. 80–90, published data. 91–99 calculated (# of programs X Avg # of pts transported each year).17a. 80–99, calculated. (100000 X number of accidents)/total flight hours for the year17b. 80–99, calculated. (100000 X number of accidents)/total flight hours for the year, excluding PR
1992 (Figure 1–29, Line 9) and informa-tion on the total number of helicopterswas available for 1981 to 1991 (Figure1–29, Line 10). From the Flightweb sur-vey and personal follow-up, informationwas obtained from 45 of the 50 states(plus the District of Columbia). Programdata for the remaining states was estimat-ed from the AirMed directory. The surveyresults identified 231 dedicated HEMSprograms, operating a total of 377 heli-copters (not including back-up aircraft).
Additional follow up with severalindustry leaders and aircraft manufactur-
ers suggests that, if anything, these num-bers may be slightly underestimated.From a telephone survey of the five air-craft manufacturers, it was estimated thatthere were a total of 462 dedicated andbackup medical helicopters in the UnitedStates. This did not include dual-purposehelicopters. From the operators’ survey,there was an average of one backup heli-copter for every 7.1 dedicated helicopterin their combined fleets. Using this ratio,it would appear that our state-by-statesurvey of 377 helicopters would have anestimated 53 backup aircraft yielding a
total of 430 helicopters. This represents avariation of approximately 7% fewer air-craft compared to the number of helicop-ters from the manufacturers’ survey. Thismay also indicate a slight discrepancy inthe number of HEMS programs resultingfrom our Internet survey. In using thelower state-by-state survey results of 377dedicated helicopters and 231 programsfor our calculations, we realize that wemay be underestimating our flight hoursand number of patients transported. Thisalso results in higher accident and fatalityrates.
22 AIR MEDICAL PHYSICIAN HANDBOOK
Our Flightweb survey results wereentered into an Excel spreadsheet for theyear 2000 along with the published datathrough 1991/1992. The estimated num-ber of programs and helicopters for1992/1993 through 1999 were then cal-culated assuming a consistent growth rateeach year. The results are found inFigure 1–29, Lines 9 and 10.
Flight Hours. Total flight hours forthe HEMS industry were available for1972 to 1985 (Figure 1–29, Line 8a and8b). In addition, the average total flighthours per program, excluding PR flighttime, were published for 1983 to 1999(Figure 1–29, Line 5). Total flight hours(less PR flights) for 1993 to 1999 couldnow be estimated (number of programs Xaverage flight hours per program).
For 1986 to 1992, we had to first esti-mate the average number of flight hoursper program. We relied upon Collett’spreviously documented average of 1.05flight hours per patient flight in 1988when the average loaded flight was 58miles. To test this hypothesis, we calcu-lated the average flight hours per patienttransport from 1993 to 1999 (Figure1–29, Line 6). Over these seven years,the average was 1.058 flight hours perpatient transport—despite the fact thatthe loaded miles had decreased.Multiplying 1.05 times the average num-ber of patients transported per programper year for 1986 to 1992, we were ableto estimate the flight hours needed tonormalize our data.
To determine the accuracy of the cal-culated total flight hours, we comparedAirMed’s average flight hours per programto the results of our HEMS operator sur-vey. For the year 2000, the companiesoperated 160 programs and 269 dedicatedhelicopters. Comparing this to ourFlightweb survey results (231 programsand 377 helicopters), the operators repre-sent 69% of the HEMS programs and71% of the helicopters. In 1999, theaverage total flight hours per programwas 925 hours for our five operators.This represents a 12% increase over the821 hours published in AirMed. Basedupon our operators’ survey, the 1998average was 854 hours. This correspondsto a 5% increase compared to the AirMedsurvey for that same year. Taking intoaccount our earlier estimates regarding
non-patient flight time and PR flighttime, these calculated increases (an aver-age of 8.5%) seem accurate. Figure 1–29,Line 7 lists the average total flight hoursper program from our operators surveyresults (1998 to 2000) and the calculatedvalues for 1986 to 1997. This is done bymultiplying the previously recorded aver-age flight hours per program (Figure1–29, Line 5) by 1.085. Total flighthours for each year can now be moreaccurately estimated, as documented inFigure 1–29, Line 8b.
Patients Transported. The journalspublished data on the total number ofpatients transported each year from 1980to 1990. With the number of programsnow estimated for each year, the totalnumber of patients transported annuallyfrom 1991 through 1999 could also beestimated (number of programs X theaverage number of patients transportedeach year).
Accident Data. The yearly accidentand fatal accident data (Figure 1–29,Lines 15 and 19) were obtained and sum-marized from several resources. The ref-erences included the NTSB and NASAwebsites, the CONCERN Network, andpersonal databases from several industryleaders.
Year 2000. For the year 2000,AirMed no longer published their“Annual Transport Statistics,” whichincluded the average number of patientstransported and hours flown. In order toestimate these numbers, the AirMed sur-vey results for the previous five yearswere averaged. It is interesting to notethat our operators’ survey shows adecrease in the average flight hours perprogram to 841 hours, a decrease of 9%compared to 1999 figures.
Year 2001. In order to include the2001 HEMS accidents in our calcula-tions, additional information wasobtained from our aviation operators andaircraft manufacturers. Recognizing thatthe operators represented approximatelytwo-thirds of the programs and helicop-ters in our earlier analysis, this seemed anappropriate perspective for year 2001projections.
Operators were asked for their 2001
total flight time, total number of pro-grams, and total number of helicopters operated. The combined results identi-fied 159 programs and 286 dedicated hel-icopters. Compared to 2000, this was adecrease of one HEMS program, while 17additional helicopters were placed inservice—an increase of 6%. Follow upwith the aircraft manufacturers, however,showed a one-year increase of 13% forthe total number of EMS helicopters.Combined total flight hours according tothe operators increased at the same rate,showing a gain of 13% over the previousyear. These values were used to calculate2001 data for our comparisons.
Accident Rates
The necessary data is now available tonormalize the HEMS data and comparewhat has occurred each year. Accidentrates were calculated using all of the acci-dents that were included in Figure 1–29.This included patient and non-patientmissions. The only medical helicopteraccidents excluded were several dual-pur-pose helicopter accidents, two mainte-nance flights where the aircraft were notin service to the HEMS program at thetime of the accident, and a training flightof a newly hired pilot preparing for hischeck ride.
The first comparison for the HEMSaccidents is a determination of the acci-dent rates per 100,000 flight hours(Figure 1–29, Line 17a). The formula([100,000 X number of accidents]/totalflight hours for the year) used the totalflight hours as calculated with the 8.5%increase (Line 8b) to account for allflight time (patient and non-patient mis-sions). A second accident rates per100,000 flight hours was calculated(Figure 1–29, Line 17b) using the totalflight hours before the increase. The dif-ference between these two accident ratesfor each year ranged from 0.05 to 1.3accidents per 100,000 flight hours withan average difference of 0.39 accidentsper 100,000 hours. When these twoaccident rates were graphed together,they virtually overlapped. As a result, weare only including the calculations fromLine 17b in our graphic analysis, whichdepicts the slightly higher rate.
Figure 1–30 shows that there has beena dramatic decrease in the accident ratesince the mid-’80s. As the raw data
23Special Supplement
0
2
4
6
8
10
12
14
16
18
20
22
24
YearA
ccid
ents
per1
00,0
00fli
ghth
ours
Figure 1–30: Accident Rates for HEMS Operations, 1980 – 2001
0
2
4
6
8
10
12
Year
Fat
alac
cide
nts
per
100,
000
fligh
thou
rs
Figure 1–31: Fatal Accident Rates for HEMS Operations, 1980 – 2001
HEMS data in much the same manner aswas done in the ’80s—comparing theaccident rate per 100,000 patients trans-ported (Figure 1–29, Line 23). With ahigh correlation of flight time per patienttransport, Figure 1–32 is very similar tothe Figure 1–30.
We would expect similar results if weagain look at our average flight program.With a 10-year (1992–2001) averageaccident rate of 3.89 accidents per100,000 patients transported and the typ-ical HEMS program flying 882 patients ina year, a program would have one acci-dent while transporting an estimated25,700 patients over 29.2 years.Calculations for 1997–2001 (4.79 acci-dents per 100,000 patient transports)
resulted in an estimate of one accidentwhile transporting nearly 21,000 patientsover a 23.74 year period.
It may be of interest to note a 1990study by Rhee et al., that compared theHEMS accident rate in the United Statesto that of the Federal Republic ofGermany. For 1982–1987, Rhee’s calcu-lations found an accident rate in the U.S.of 11.7 per 100,000 flight hours. Thecalculated accident rate for WestGermany was found to be comparable at10.9 per 100,000 flight hours. The fatalaccident rates were also found to be simi-lar. The U.S. rate was 4.7 fatal accidentsper 100,000 flight hours while the WestGerman rate was 4.1. In contrast, usingour own database and calculations for
would predict, the accident rate since1998 has steadily increased. However,despite this increase, the rate remainsroughly one-third of what we experi-enced in the early to mid-1980s due tothe overall increase in flight hours.
Looking at an average accident ratefor the past 10 years 1992–2001 (3.53accidents per 100,000 flight hours), theaverage HEMS program flying 911 hoursper year, would have one accident over31.1 years of flight time. Changing thiscalculation to include the accident ratesfor the past 5 years, (1997–2001), we seea moderate change. We now have a 5-year average of 4.56 accidents per100,000 flight hours and the predictionfor the average program decreases to oneaccident over 24.1 years. Naturally, if aprogram flies less, the number of yearswould go up, and if a program flies more,the time frame would decrease. Anotherway to propose the likelihood of an acci-dent would be to compare the number ofaccidents to the number of programs (orhelicopters). Again, using the mostrecent ten years, there has been an aver-age of 7 accidents each year. With 231dedicated HEMS programs, and assum-ing all things being equal, we would finda similar prediction of one accident perprogram every 33 years. If you base thiscomparison on the number of dedicatedhelicopters estimated for 2001 (400)rather than programs, the margin nowgoes up to nearly 57.1 years. Looking atthe average number of HEMS accidents(9) for the past five years, we wouldexpect one accident per helicopter every44.4 years.
The second comparison is looking atthe fatal accident rate per 100,000 flighthours (Figure 1–29, Line 21). Here too,in Figure 1–31, we see a dramaticimprovement since the early and mid-1980s. Our current rate, despite havinggone up slightly over the past few years,is approximately 75% less than our worstyears. Once again if we take an averageflight program and an average fatal acci-dent rate for the past ten years (1.38fatal accidents per 100,000 flight hours),we would predict one fatal accidentwhile flying more than 79.3 years.When we focus on the average fatalaccident rate for 1997–2001 (1.69) thisfigure drops to 64.8 years. The final nor-malized comparison will look at the
24 AIR MEDICAL PHYSICIAN HANDBOOK
copters before 1980. Figure 1–33 showsthe wide range for our results. In 1982,an estimated 16.3% of the HEMS pro-grams (8 accidents, 49 programs) wereinvolved in accidents. The safest yearwas in 1996, when an estimated 0.5 %of the programs had an accident (1accident, 207 programs). Overall, theaverage annual percentage over 22 yearscalculates to 5.8% of the programs having had accidents between 1980through 2001. If you consider only thepast five years, an average of 4.1% ofthe programs have had an accident each year.
Calculating the percentage of heli-copters that were involved in HEMSaccidents each year finds a high of12.9% of the HEMS aircraft in 1982 (8accidents, 62 helicopters). In 1996,there was 1 HEMS accident during ayear when an estimated 309 dedicatedmedical helicopters were in operation,for a total of 0.3%. The average per-centage over 22 years calculates to 4.4%of the HEMS fleet. Over the past 5years, this percentage has averaged2.5% for each year.
SECTION 2: A COMPARISON OFHEMS TO OTHERTYPES OF AVIATION
Having reviewed HEMS-specific data,this report will now compare HEMSaccident and incident data to othertypes of aviation. This section firstlooks at two reports and then focuses onvarious aviation industry statistics.Finally, the accident rates for helicopterair medical transport is compared toother aviation operations.
HELICOPTER EMS vs. ALLHELICOPTER ACCIDENTDATA: 1990–2000
During the 2000 Air MedicalTransport Conference (AMTC), SandraHart of the NASA-Ames ResearchCenter presented a study that comparedcharacteristics of helicopter EMS acci-dents with those of all helicopter acci-dents (EMS and non-EMS) over the
Figure 1–32: HEMS Accident Rate per 100,000 Patient Transports, 1972–2000
0
2
46
8
10
1214
16
18
20
2224
26
28
Year
Acc
iden
tRat
epe
r100
,000
Tra
nspo
rts
this same time frame (1982–1987), ourstatistics yield a much higher accidentrate of 15.58 per 100,000 flight hours anda fatal accident rate of 5.2.
Percentage of HEMSPrograms and HelicoptersInvolved in Accidents
Data is not available to accuratelydetermine what percentage of HEMS pro-grams have sustained an accident. Overnearly 30 years of civilian HEMS opera-tions, dozens of programs have closed andothers have merged operations. However,with the data that we have accumulatedto estimate the number of programs and
helicopters each year, we can determineannual percentages with some accuracy.Averaging these annual calculations, wecan estimate the overall percentage ofHEMS programs and helicopters thathave had accidents. These calculationsdo not take into account the possibilitythat an individual program may have suf-fered more than one accident—which hasoccurred.
The calculated percentage of programsand helicopters that sustained accidentseach year since 1980 is shown in Figure1–29, Lines 18a and 18b. The five acci-dents that occurred prior to 1980 are notincluded, as we do not have annual statis-tics on the number of programs or heli-
0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
20%
Year
Per
cent
age
Programs Helicopters
Figure 1–33: Programs and Helicopters with Accidents, 1980–2001
25Special Supplement
same time period. In March 2001, Hartpresented an updated report at theProceeding of the 11th InternationalSymposium on Aviation Psychology.
In her introduction at AMTC, Hartpointed out that in recent years, thenumber of HEMS accidents hasincreased. However, she emphasized thatall we have is raw numbers. Accurateexposure data is lacking regarding thenumber of transports and the number ofhours flown by helicopter EMS. Withoutthis, we do not have accurate informa-tion to determine if the increase in acci-dents is related to an increase in thehours flown or whether HEMS had got-
ten less safe. As a result, the data pre-sented in her lecture was based upon rawnumbers and percentages rather than asaccident rates.
In addition, Hart stated that “aircraftaccidents are poor indicators of safetytrends due to a low occurrence rate andlimited information about what hap-pened. Quite often, the immediate‘cause’ may have little relationship to theunderlying causes. But as risk factorsbegin to accumulate, something bad wasbound to happen.”
An analysis of 1,494 helicopter acci-dents over a 10-year period beginning in1990 was conducted by the NASA-Ames
Research Center. The database wasobtained by reviewing the NTSB acci-dent and incident reports, looking at nar-ratives, probable cause, occurrences, find-ings and coded data (e.g., pilot experi-ence, helicopter make/model, visibility,mission). A total of 58 HEMS accidentswere identified within this database—both Part 135 (patient-related transports)and Part 91 flights (repositioning, main-tenance, etc.). Of interest, 72% of theHEMS accidents were flown under Part91 and only 30% were on patient-relatedmissions. Some HEMS operations con-duct the flight to the patient under Part91 regulations, while other programs consider all legs of a patient flight to bePart 135.
Figure 2–1 illustrates the number ofaccidents that were included for eachyear in the Ames report. Several 1999and 2000 accidents lacked NTSB finalreports and were not included.
When the Accidents Occur
When comparing the data, approxi-mately 53% of the HEMS accidentsoccurred between dusk and dawn whileonly 9% of all helicopter accidentsoccurred at night. It is estimated thatonly 5% of all helicopter flights are atnight.
Hart found that nearly five times asmany HEMS accidents occurred in IMCconditions (24% compared to 5% for allhelicopter accidents), with nearly all ofthem involving inadvertent flight intoIMC conditions. This was consistentwith the previous NTSB study. Neithersnow nor rain was an identified problem,with very few helicopter accidents occur-ring with visible precipitation. Thiscould be due to the fact that fewer heli-copter flights occur during these types ofweather conditions. Another possibilitycould be that pilots are more carefulwhen they can see visible precipitation(rain or snow). In contrast, deterioratingweather conditions (decreasing ceilingsor fog developing) may be less obvious.
Of the 58 accidents, weather condi-tions were cited as a contributing factor29 times. It is important to note thatthese were factors and not the cause ofthe accidents. In some cases there wasmore than one weather factor cited with
0
1
2
3
4
5
6
7
8
9
Year
Acc
iden
ts
FAR Light Condition Weather Precipitation
72%
28%
0
53%
47%
24%
76%
12%
82%
68%
9%
21%
11%
89%
5%
95%
19%
81%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Part 91 Part 135 Other Dusk toDawn
Day IMC VMC Snow,Rain,Sleet
None
HEMS All Helicopter
Figure 2–1: Number of HEMS Accidents per YearAdapted from: Hart, Conference presentations, 2000/2001
Figure 2–2: When Accidents Occur (HEMS: n=58; All Helicopters: n=1,494)Adapted from: Hart, Conference presentations, 2000/2001
26 AIR MEDICAL PHYSICIAN HANDBOOK
regard to an accident or incident.Weather is therefore likely to add to therisk of a flight. It may not cause the acci-dent, but it may increase the likelihoodthat an accident could occur. Weatherwas cited only twice as the cause of anaccident.
With regard to the phase of flight,there are some significant differenceswhen comparing HEMS to all helicopteraccidents. For all helicopter accidentsthe highest percentage of accidents wereseen during landing (25%), maneuvering(21%), cruise (15%), hovering (11%),and takeoff (11%). This is significantlydifferent from HEMS accidents thatoccurred most commonly during cruise(24%), takeoff (19%), approach (16%),and landing (14%). Figure 2–3 showsthe distribution of accidents during thevarious phases of flight.
Pilot Experience
Experience of the pilots was carefullyreviewed by the Ames study. In generalthe HEMS pilots averaged slightly fewertotal hours than the total pilot database(6,307 vs. 6,424). However 79% of theHEMS pilots’ hours (5,010) were in heli-copters. The overall group includedcommercial fixed-wing pilots who didmuch less time in helicopter aviation(66% or 4,230 hours). HEMS pilotsaveraged fewer hours (753) in themake/model helicopter they were flyingat the time of the accident compared topilots for all of the helicopter accidents(1,273). In addition, unlike the NTSBreport, pilot fatigue was not found to be asignificant factor in these HEMS acci-dents. According to the study, the aver-age HEMS pilot had flown only 1.88hours in the 24 hours prior to their acci-dent which was less than the average(3.00) for the “all helicopter accident”group.
The Ames researchers found a signifi-cant difference with regard to instrumentratings. EMS pilots were far more likelyto have an instrument rating than all hel-icopter pilots involved in accidents.They may not have been current andthey may not have been flying helicop-ters that were IFR equipped, but theirtraining and experience was noted.According to Hart, while this additionaltraining and experience should be con-
sidered an advantage to the EMS heli-copter pilots, it may have worked as adisadvantage if the pilots felt that theirtraining and experience would allowthem to “push the envelope a little bitmore.”
Vehicle Characteristics
There were more HEMS accidentsinvolving twin-engine helicopters thansingle-engine helicopters (63% vs. 36%,respectively). This is in contrast to allthe accidents where the majority of theaircraft were single-engine aircraft (89%).It is important to note that no mentionor comparison is made with regard to thepercentage of single- vs. twin-engine hel-icopters in operation—only the compari-son of those involved in accidents.
Figure 2–3: Phase of Flight (HEMS: n=58; All Helicopters: n=1,494Adapted from: Hart, Conference presentations, 2000/2001
0% 0%
19 %
3%
24 %
7%
16 %
14 %
9%
7%
3%
1%
11 %
1%
15 %
5% 6%
25 %
21 %
11 %
0 %
5 %
1 0 %
1 5 %
2 0 %
2 5 %
3 0 %
S t a n d
ing
Tax i
Tak e
- o f f
C lim b
C r u is e
D e s c e
n t
A p p o a
c h
L a n d
ing
Man
u v e r
H o v e r
H E M S A l l H el i cop t er
EMS All Helicopter
Average Range Average Range
Total Hours 6,307 3,000–19,275 6,424 29–34,886
Helicopter Hours 5,010 27–17,380 4,230 8–25,000
Hours in Make 753 16–3,620 1,273 3–8,918
Instrument Hours 269 0–1,647 203 0–3,613
Prior 24 hours 1.47 0–6 3.00 0–15
Accident Characteristics
HEMS accidents have a much higherlikelihood of resulting in serious injuriesor fatalities than other helicopter acci-dents. Of the 58 accidents studied, 38%resulted in at least one fatality comparedto 17% for the overall database. Theseresults are similar to the previous studies.While there tended to be very few post-crash fires in either group, aircraft inboth groups were either destroyed or seri-ously damaged at very high rates (92%for HEMS vs. 97% for all accidents).HEMS aircraft, however, were destroyedat a much higher rate. Figure 2–5 showsthe accident characteristics for bothgroups.
Figure 2–4: Pilot ExperienceAdapted from: Hart, Conference presentations, 2000/2001
27Special Supplement
Figure 2–5: Accident Characteristics (HEMS: n=58; All Helicopters: n=1,494)Adapted from: Hart, Conference presentations, 2000/2001
5 % 2 %
8 8 %
3 5 %
1 4 % 1 0 %
4 2 % 4 5 % 4 7 %
5 % 3 % 1 %
1 0 %
8 9 %
1 8 % 1 4 %
2 2 %
6 4 %
2 6 %
7 1 %
1 % 2 %
0 %
1 0 %
2 0 %
3 0 %
4 0 %
5 0 %
6 0 %
7 0 %
8 0 %
9 0 %
1 0 0 %
Fire I n
- f ligh
t
Fire o
n Gr o
u n d
N o Fire
Fat a
l
S e r iou
s
Mi n o
r
N o n e
D e s t r o
y e d
S e r iou
s
Mi n o
r
N o n e
H E M S A l l H el i cop t er
Pre/Post Crash Fire Severity of Injuries Aircraft Damage
9%
2%
5%
28 %
16 %
14 %
12 %
10 %
2%
6%
13 %
1%
6%
12 %
8%
3%
17 %
27 %
3%
10 %
0 % 5 % 1 0 % 1 5 % 2 0 % 2 5 % 3 0 %
A ir f r a m e / S y s t e m Fa il u r e
Fi r e / E x p los ion
H a r d L a n d ing
I n f li g h t C o ll is ion w / Ob jec t
I n f li g h t C o ll is ion w / T e r r a in
I n f li g h t E n c o u n t e r w / W e a t h e r
L o s s o f C o n t r o l I n f li g h t
L o w E n g ine P o w e r
R o ll Ov e r
M is c . / U n k n o w n A l l H el i cop t er
H E M S
Figure 2–6: First Events (HEMS: n=58; All Helicopters: n=1,494)
Adapted from: Hart, Conference presentations, 2000/2001
Chain of Events and First Events
Hart’s report emphasized that accidentsare not caused by a single event. In mostaccidents, numerous risk factors can beidentified. An accident might have anobvious identifiable cause, but there arelikely to have been numerous risk factorsthat contributed to the event or theseverity of the event. Acting on any ofthese risks might prevent an accidentfrom happening or lessen the severity ofthe accident.
When the NTSB looks at accidentdata, they pay particular interest to the“first events”—the first obvious andmeasurable event that can be consideredan accident. It is not the first occurrencein the chain of events that leads to theaccident. Rather, it corresponds to thefirst event that would be considered evi-dence that an accident has taken place.First events are not causes. Their analysistells us what has happened, but do nottell you why it happened.
The most common first event inHEMS was an in-flight collision with ter-rain, with wires being the primary offend-er. EMS helicopters were more thantwice as likely to strike an object or ter-rain and nearly five times more likely tohave an encounter with inclementweather. Relatively few of the HEMSaccidents were caused by low enginepower, airframe, or component failurescompared to the all helicopter accidentgroup. Hart concluded that this seems tosuggest that, in general, EMS helicoptersare well maintained.
The Ames study presented a cascadingchain of events of what happens when anaccident occurs. An in-flight encounterwith weather is rarely the cause of anaccident but is often the first event. Itleads to an in-flight collision with anobject or terrain or to a loss of enginepower. Another frequent first or secondevent is loss of control in flight, whichmay follow some type of system failure(which is rare).
Cause and ContributingFactors of Accidents
When the NTSB finalizes their acci-dent reports they try to identify a cause ofeach accident. This is not necessarily theonly event that contributed to the acci-
dent. Pilot-related factors (human fac-tors) were cited 50 times in the 58 acci-dents. These factors included operatingwith known deficiencies, inadequatepre-flight planning, inability to evaluatethe weather, inadvertent flight intoIMC, failure to follow procedures, spatialdisorientation, lack of experience, and
failure to maintain proper speed, alti-tude, rate of descent or climb, or RPM.Of interest, Hart pointed out that invery few of these accidents was it identi-fied that the pilot deviated from theFARs or company regulations. This issignificantly different than the datareviewed earlier in the ASRS report.
28 AIR MEDICAL PHYSICIAN HANDBOOK
Aircraft-related causes of the 58HEMS accidents were cited 22 times,while fuel-related problems were citedtwice. Weather was cited only once.
There are a number of contributingfactors that increase the risk for an acci-dent. Many of these are related toweather (e.g., icing conditions, clouds,fog, rain, snow, sleet), which was cited 29times as a contributing factor in the 58accidents. Person-related contributingfactors were cited 25 times, whichincluded distractions, pressure felt by thepilot, as well as the items listed under“causes.” Terrain, which was cited 24times, was another consideration and fly-ing at night (dusk to dawn) was cited 18times.
FACTORS RELATED TOOCCUPANT CRASHSURVIVAL IN EMSHELICOPTERS
Much of what we have been reviewinghas dealt with the associated risks andevents surrounding accidents that haveoccurred in the past. It is also essentialto assess the relative occupational risk tothe crewmembers of EMS helicopters.Robert Dodd’s 1992 Ph.D. dissertationevaluated the incidence and seriousnessof crash-related injuries among EMS hel-icopter occupants in survivable crashes.
The study found that main cabin occu-pants in EMS helicopters have nearly 4.5times the risk of serious injury or death insurvivable crashes when compared to acomparable population of occupants inthe main cabin of non-EMS air taxi heli-copters. For front seat occupants, hefound that there was no significant differ-ence in injury risk between the twogroups. This seemed to support his prem-ise that EMS aircraft modifications,which are generally limited to the maincabin, were directly associated with therisk of injury and may contribute to occu-pant injury and death in otherwise sur-vival crashes.
For his study, Dodd reviewed 75 EMSaccidents from 1978–1983 and1983–1989 (with 241 occupants) and147 non-EMS helicopter accidents from1983–1989 with 485 occupants. Hiscomprehensive review and analysis of
available reports identified survivable vs.non-survivable crashes and occupants, asseen in Figure 2–7.
Comparing occupant location in sur-vivable accidents with specific injury pat-terns, Dodd concluded that EMS maincabin occupants were at a higher risk forserious back injuries and serious headinjuries. Figure 2–8 compares the per-centage of occupants with specificinjuries and where they were seated inthe helicopter.
Dodd’s research included written sur-veys that were sent to survivors. Twelveinjured EMS occupants indicated injuriesthat were the result of striking medicalequipment inside the helicopter duringthe crash. This equipment included thestretcher, cardiac monitor, medical panel,oxygen tanks, and portable radio.
As part of his research, Dodd evaluatednumerous variables to determine howthey had influenced injuries. The vari-ables included crash severity, post-crashfire, number of engines, helicopterweight, light conditions, use of shoulderharness, cause of the crash, age, and sex.For each of these variables, he calculatedthe relative risk of injury for EMS occu-pants compared to non-EMS occupantsin survivable crashes. Dodd found thatthere was a significantly greater risk ofinjury (significant relative risk) in HEMSaccidents compared to non-EMS acci-
dents: where there was a post-crash fire;in single-engine helicopters; with heli-copters weighing < 4,500 pounds; duringdaylight conditions; with EMS occupantswho did not wear a shoulder harness; andmechanical-related crashes. There was amarginal relative risk of injury in heli-copters that weighed > 4,500 pounds.
In contrast, he found no significant rel-ative risk of injury in: the absence ofpost-crash fire; twin-engine helicopters;during dark (night) conditions; or crashescaused by loss of control, wire strikes, orbad weather. Even though significantlymore EMS accidents (32%) occurred inbad weather compared to the non-EMSstudy group (14%), there was not a sig-nificant increase in the risk of injury.There was also no correlation betweeninjury and the age or sex of the occu-pants between the two groups.
Dodd also evaluated the severity of theaccidents and classified them as CrashSeverity Level 1 (hard landing), CrashSeverity Level 2 (hard landing with sub-stantial damage), and Crash SeverityLevel 3 (high vertical impact or cruisecollision with ground). Level 1 accidentsresulted in no injuries to EMS occupantsand only one non-EMS injury. Level 2accidents yielded a significant relative riskfor passengers in the main EMS cabin, buta non-significant risk for front-seat passen-gers. In Level 3 accidents, the significant
Serious Back Serious Head Minor Head Internal Fractures Injuries (%) Injuries (%) Injuries (%) Injuries (%) (%)
Occupant EMS Non- EMS Non- EMS Non- EMS Non- EMS Non- Location EMS EMS EMS EMS EMS
Pilot 4.5 10.8 2.3 2.4 0 3.6 0 0 2.3 7.2
Front Seat 15.4 2.7 7.8 5.4 0 2.7 11.5 0 3.8 8.1
Main Cabin 28.6 9.2 14.3 1.3 5.3 1.3 7.1 2.6 10.7 5.3
Patient 14 N/A 7.1 N/A 0 N/A 0 N/A 7.1 N/A
Total Non-survivable Survivable
Crashes Occupants Crashes Occupants Crashes Occupants
EMS 75 241 24 70 51 171
Non-EMS 147 485 20 50 127 435
Figure 2–7: Survivable and Non-survivable EMS and Non-EMS CrashesAdapted from: Dodd, unpublished dissertation, 1992
Figure 2–8: Distribution of Injuries in Survivable Crashes—EMS vs. Non-EMS CrashesAdapted from: Dodd, unpublished dissertation, 1992
29Special Supplement
relative risk for main cabin occupants waseven greater in EMS aircraft.
Dodd’s calculations of accident rateswere similar to the NTSB report that hadcome out four years earlier. He foundthat EMS helicopters had an accidentrate of 11.84 per 100,000 hours of flight.This is more than 2.5 times that of non-EMS air taxi helicopter operations thatwere found to have an accident rate of4.43 per 100,000 flight hours. WhileDodd did not compare fatal accidentrates, he did compare the percentage ofoccupant fatalities for his two studygroups. He found that 32% of the EMSoccupants suffered fatal injuries, whileonly 9% of the non-EMS occupants died—a rate that is 3.5 times greater for theEMS group.
Dodd’s report makes a convincingstatement for addressing accident survivalas well as accident prevention. EMS per-sonnel are injured more frequently andmore severely in survivable HEMS acci-dents when compared to occupants ofnon-EMS helicopter crashes.
Dodd presented a 1991 study byCrowley that evaluated the use of hel-mets in survivable military crashes.Crowley found that occupants withouthelmets were 4 times more likely to suffera serious head injury and 6 times morelikely to suffer a fatal head injury thanoccupants with helmets. Limiting thecomparison to the main cabin increasesthe risk. Main cabin occupants with nohelmet were found to be at 5 times therisk for a serious head injury and 7.5times the risk for a fatal head injury thantheir helmeted counterparts. WhileCrowley’s study was based on militarydata, he suggested that it might also beapplicable to survivable HEMS accidents.
Another series of studies for the U.S.Army concluded that back injuries werethe most common injury suffered byoccupants in survivable helicopter acci-dents. These results are consistent withDodd’s findings. The Army studies foundthat shoulder harnesses were an impor-tant factor in reducing the incidence andseverity of serious back injuries
Dodd concluded that the increased riskof injury to main cabin occupants of EMShelicopters represented an occupationalrisk that had not previously beenaddressed in the literature. He concludedthat the EMS helicopter was a very haz-
ardous place, even in a survivable crash.Dodd suggests that the use of energyattenuating seats, in combination with lapand double shoulder harnesses, and intelli-gently designed interiors could dramatical-ly improve occupant injury tolerance.
HELICOPTER ACCIDENTANALYSIS TEAM
This next study which is not limited toHEMS, reviews a series of helicopteraccidents to determine what happenedand what could be done to break thechain of events that lead to accidents.The Helicopter Accident Analysis Team(HAAT), was a cooperative effortinvolving the Department of Defense(DoD) the Federal AviationAdministration/Department ofTransportation (FAA/DoT), and theNational Aeronautics and SpaceAdministration (NASA). It began inFebruary of 1997, as mandated by theWhite House Gore Commission onAviation Safety.
The approach in this analysis is similarto the “Air Medical Accident Analysis”described in Section 1, but this studyincluded EMS and non-EMS helicopters.A balanced sample of 34 helicopter acci-dents was selected from the 1990 to 1996
NTSB database of helicopter accidentsand incidents. HEMS accidents selectedwere those that involved a patient trans-port mission at the time of the accident.
Three subgroups worked independentlyto address different aspects of these acci-dents. First they developed a sense of whathappened—the chain of events that led toeach of the accidents. Next they identi-fied the problems—issues with respect tothe aircraft, environment, pilot actions,maintenance, air traffic control, and thequality of the information in the reportitself. Finally, they brainstormed aboutwhat might have prevented the accidententirely or mitigated its severity—theinterventions. The goal of the HAATanalysis was to propose technology, train-ing, and institutional interventions thatmight have eliminated one or more “links”in those chains of events, thereby avertingan accident or decreasing the severity ofan accident that does occur.
Chain of Events
The number of events identified for anyparticular accident ranged from 5 to 33,with an average of 16 events per accident.This resulted in a total of 536 events.Five different categories of events wereidentified. Figure 2–9 lists the categoriesand examples of the events.
Category Examples of Events
Preliminary events Definition: Factors that influenced the accident but were not directly related to actions taken by those involved in the accidentExamples: Pilot’s health, pilot’s experience, adverse weather
Preflight events Definition: Events that occurred prior to departure of the accident flight that could have influenced the outcome Examples: Failing to obtain a weather briefing preflight or ensuring that the aircraft had enough fuel
Flight-related events Definition: Events or actions that occurred during the flight and were associated with the accident Examples: Continued flight into adverse weather, poor air traffic control vectoring
Emergency-related events Definition: Events that occurred during the emergency/ accident sequence or precipitated the sequence Examples: Poor landing site selection, wire strike, fuel starvation
Survival-related events Definition: Events or actions that did influence, or could have influenced, occupant survival after the accident Examples: Helmet use, delayed rescue, inoperative ELTs
Figure 2–9: Chain of Events Categories identified by HAATAdapted from: Helicopter Accident Analysis Team. Final Report, NASA-Ames Research Center, 1998
30 AIR MEDICAL PHYSICIAN HANDBOOK
Problems
The accident analysis identified thenumber of problems for each accident,ranging from a low of 3 to a high of 21.There was an average of 16 problemsper accident, resulting in a total of 442entries. Figure 2–10 illustrates the cat-egories of problems identified and spe-cific problems within each category.
Interventions
After identifying the problems,interventions were identified thatcould have prevented the accident orlessened its severity. The number ofinterventions identified for individualaccidents ranged from 4 to 25, averag-ing 13 per accident. There were a totalof 416 possible interventions identifiedacross all the accidents. Figure 2–11reviews the categories of interventionsand the specific proposals made byHAAT.
The previously presented “AirMedical Accident Analysis” evaluatedthe effectiveness and feasibility of eachof their recommended interventions.That was not done in this study.Instead, the final step of the HAATanalysis was the proposal of 26 specificSafety Investment Areas that werederived from their identified interven-tions. Rather than a single statement,each area had identified goals, back-ground, opportunities for reducingfuture fatalities, research needs, timing,related work, and the primary benefici-aries. Their emphasis was no longeron specific interventions that mighthave broken the chain of events for aspecific accident. Rather, the SafetyInvestments were more global and goaloriented. Their recommendationsincluded specific goals directed towardthe development of research and tech-nology to enhance safety and toimprove procedures and practices.Safety Investment Areas were identi-fied in helicopter design and perform-ance, situation displays, pilot aidingand automation, pilot training,improving the flight environment,crash survivability, and the improve-ment of safety data and analysis.
Figure 2–10: Problems Identified by HAATAdapted from: Helicopter Accident Analysis Team. Final Report, NASA-Ames Research Center, 1998
Problems Associated with... Problems Identified
Pre-Flight Planning Aircraft / operating limits not consideredWeather or wind not consideredMission requirements / contingencies ignoredPre-flight process inadequatePassenger safety brief inadequate
Safety Culture of the Organization Management policies / oversight inadequateSafety program / risk management inadequateHelicopter not IFR-equipped Problems with pilot’s health not addressed
Inadequate Training or Experience Emergency training inadequateSpecial operations training inadequateTraining inadequate for inadvertent IMC Pilot inexperienced with area, mission, vehicle
Maintenance Tools to detect failing parts inadequateBogus, surplus, unapproved parts usedImproper procedures/supervisionInadequate documentationComponents used not built to manufacturer’s specifications
Infrastructure Inadequate oversightIFR system incompatible with helicopter missionsPart 91 vs. Part 135 passenger-carrying operationsInadequate tower/wire markings
Pilot Judgment and Actions Sense of urgency led to risk-takingDiverted attention, distractionFlight profile unsafe for conditionsPoor cockpit resource managementPerceptual judgment errorsProcedural errorsPilot control/handling deficienciesUsed unauthorized equipment
Communications Coordination with ground personnelCoordination with ATCCoordination with other pilots
Pilot Situation Awareness Aircraft position and hazardsAircraft stateLocal and en route weather
Vehicle Part or System Failures Main rotor problemEngine failures (partial or total)Gear box failureTail rotor/tail boom failures
Post-crash Survivability Safety equipment not installed/failedPassenger/crew survival gear not usedVehicle did not withstand impactVehicle sank and/or capsizedPost-crash fireELT inoperative/damaged by impactInaccessible accident site/bad weatherNo flight following–slow to locate site
31Special Supplement
Categories Proposed Interventions
Safety Culture Solutions Adequately equip rotorcraft for mission Develop an inadvertent IMC policy Formalize passenger pre-flight briefing Develop clearly defined company policies
Training Interventions Basic training materials/syllabus Aeronautical decision-making trainingCrew resource management trainingTraining to recognize and resolve emergenciesGround personnel trainingRecovery from IMC/IFR trainingSimulation facilities for rotorcraft training Training for unique ops/maneuvers/missions
Maintenance Solutions Non-destructive inspection techniquesImproved maintenance procedures and quality control
Helicopter Design and Health and Usage Monitoring SystemsPerformance Solutions Real-time performance monitoring
Wire cutters/hardened bladesIcing protectionMiscellaneous design improvements
Helicopter Situation Ground proximity warning system for rotorcraftDisplay Solutions Electronic map/position
Obstacle detection and alertingRadar alt/distance from ground/waterEnhanced/synthetic visionWeather display and alerting
Pilot Aiding and Autorotation display/aidAutomation Interventions Attitude hold/stabilization
Automatic flight followingPC-based Pre-Flight Planner and PC-based Risk Assess System
Infrastructure Interventions Operating requirements for commercial rotorcraftRegulations/procedures for inadvertent IMCReview tower/wire marking requirementsNavigation/landing systems for rotorcraftReview training and qualification requirementsRequirements for company safety programSpecial operations regulations
Post-Crash Survival Improved crashworthinessInterventions Crash-survivable ELT
Survival equipmentRestraint systemsFlotation systemsCrash-resistant fuel systemUnderwater egress training
Improved Reporting Cockpit voice recorder/flight data recorderInterventions Improved NTSB accident forms
Improved data acquisition Data dissemination/feedback to industryInflight audio-visual recording in cockpit
Figure 2–11: Interventions Proposed by HAATAdapted from: Helicopter Accident Analysis Team. Final Report, NASA-Ames Research Center, 1998
HEMS VS. OTHERAVIATIONOPERATIONS
Background
Generally speaking, there arethree major categories of aviationregulated by the FAA. Part 135corresponds to “air taxi” and is clas-sified as Scheduled (commuter flightswith fewer than 10 seats) or Non-scheduled, which includes air med-ical transport and other on-demandair taxi services. Part 121 aviationgoverns the airlines, both scheduledand non-scheduled (charter) air-lines. The third category is GeneralAviation (Part 91), typically char-acterized by recreational (personal)flying, instructional, business, cor-porate, public use, and other vitalservices. Figure 2–12 summarizesthe number and types of aircraftthat were operated under the differ-ent regulations in 1998/1999.
As Figure 2–12 shows, helicoptersaccount for a very small portion ofaviation operations. Based upon our1998/1999 statistics, EMS helicop-ters accounted for an estimated 5%of all helicopters and approximately48% of the on-demand helicoptersin operation. In 1980 there werean estimated 20,750 HEMS flighthours. By 1990, this increased toapproximately 140,500. In 2001,HEMS hours were estimated atnearly 217,500 while all helicoptersflew approximately 2.4 millionflight hours, general aviation flew26.2 million hours, Part 135 opera-tions accounted for nearly 3.7 mil-lion flight hours, and Part 121 air-lines flew 16.7 million hours.
HEMS, a Part 135 on-demand airtaxi, is certainly a unique form ofaviation. There are some similari-ties with other 135 operations, butalso some similarities with GeneralAviation (Part 91). In fact, as Hartpointed out in her study, many ofthe HEMS accidents were operatingunder Part 91 at the time of theiraccident (e.g., ferry flight, reposi-tion, instruction).
32 AIR MEDICAL PHYSICIAN HANDBOOK
In general aviation (GA), personalflights are consistently the most danger-ous. An estimated 44% of all flying isdone for recreational or personal rea-sons and results in nearly 65% of thefatal accidents. In contrast, businessflying (i.e., business people who are notprofessional pilots), accounts forapproximately 14% of the GA flighthours, but only accounts for 6% of thefatal accidents. Instructional flying,with 22% of the flight time, results inmore than 8% of the fatal accidents.Corporate flying represents only 6% ofGA flight hours and had a fatal acci-dent rate of less than 1%. Business andcorporate pilots may be more willing toscrub a trip, may fly more reliableequipment, or may have more experi-ence. Likely, it is a combination of allthese factors.
In significant contrast to HEMS acci-dents, nearly 70% of all general avia-tion (Part 91) accidents are “fenderbenders” and result in little or no injury.Like HEMS accidents, however, themajority of accidents (more than 70%)were pilot-related. Typically, GA take-offs and landings account for less than5% of a typical cross-country flight.However, an estimated 50–70% of theGA accidents occurred during takeoffsand landings. Like HEMS, weather-related accidents were more likely to befatal than accidents with any othercause. In 2000, nearly 90% of theseweather-related accidents resulted infatalities.
Raw data and normalized statistics areavailable from the FAA for the differ-ent types of aviation operations.Figures 2–14 to 2–16 provide statisticsfrom 1982 to 2001 with regard to num-ber of accidents (total and fatal), flighthours, and annual accident rates foreach category. Figures 2–17 and 2–19graph the accident rates for side-by-sidecomparison.
In general, the FAA data for this 20-year period shows:• Helicopter and general aviation acci-
dent rates are much higher than allother aviation operations, followed bynon-scheduled Part 135 operations.
• The helicopter accident rate has fluc-tuated from a low of 6.17 accidentsper 100,000 flight hours to a high of 12.26.
Airlines On-Demand Air Taxi General Aviation(Part 121), 1998 (Part 135), 1999 (Part 91), 1999
Experimental 30 20,493Piston Single-Engine 167 652 150,081Piston Twin-Engine 44 1,607 19,469Turboprop Single-Engine 75 943Turboprop Multiengine 1,837 860 3,802Turbojet 5,108 496 6,625Helicopter 3 746 6,701Total 7,159 4,466 208,114
Figure 2–12: Aircraft in Operation, 1998/1999Adapted from: The Nall Report 2001, AOPA, http://www.aopa.org/asf/publications/01nall.pdfand The Nall Report 2000, AOPA, http://www.aopa.org/asf/publications/00nall.pdf
• General aviation has seen a generaldecrease in its accident rate.
• From 1982–1992, general aviation hadthe highest fatality rate among the var-ious aviation operations. Beginningwith 1993, helicopter operations havehad the highest fatality rate in 7 of thelast 9 years.
• Scheduled Part 135 operations have avery consistent and low accident ratefrom 1983 through 1996. The pastfive years, however, have seen a signifi-cant increase in this rate.
A Normalized Statistical Comparison
With all of this data, it is now possibleto compare the different types of aviationoperations. We have already estimatedthe accident and fatality rates per100,000 flight hours for HEMS. Theserates can now be included for a more
meaningful comparison. It should benoted that we are beginning in 1982rather than in 1980, as we did with theearlier HEMS graphs.
As the graphs indicate, accident ratesfor HEMS, all helicopters, and generalaviation have always been higher thanairline rates. There are many factors thatcontribute to this difference. In general,these three types of aviation operationsinvolve risks that are not in commonwith the airlines. These differencesinclude: • Helicopters and general aviation pilots
conduct a wider range of operations,often with less regulation and fewersupport services.
• There is a wider variance in pilot qual-ifications and training.
• There are fewer cockpit resources. Aircarrier operations require at least twopilots, while most general aviation andhelicopter operation are single pilot.
Operation Percent of Flying Percent of Total Percent of Fatal(1999) Accidents (2000) Accidents (2000)
Personal 44.3 67.4 64.4Instructional 22.1 13.1 8.2Aerial Application 4.8 6.0 5.0Business 13.6 4.0 6.1Positioning — 1.7 2.3Ferry — 1.0 0.3Public use 2.2 0.7 0.9Other work use 2.2 1.0 1.2Aerial Observation 4.1 0.4 0.3Corporate 5.5 0.4 0.9Other/unknown 1.1 4.4 10.2
Figure 2–13: General Aviation Accident Data, 1999/2000Adapted from: The Nall Report, 2001, www.aopa.org/asf/publications/01nall.pdf
33Special Supplement
U.S. General Aviation Helicopter
Year Accidents Flight HoursAccidents
per 100,000Flight Hours
Accidents Flight HoursAccidents
Per 100,000Flight Hours
All Fatal Fatalities All Fatal All Fatal Fatalities All Fatal1982 3,233 591 1,187 29,640,000 10.90 1.99 255 41 66 2,350,000 10.85 1.741983 3,076 555 1,068 28,673,000 10.72 1.94 234 35 55 2,272,000 10.30 1.841984 3,017 545 1,042 29,099,000 10.36 1.87 224 38 61 2,495,000 8.98 1.521985 2,739 498 956 28,322,000 9.66 1.75 205 36 50 2,154,000 9.52 1.671986 2,581 474 967 27,073,000 9.53 1.75 190 39 81 2,625,000 7.24 1.491987 2,494 446 837 26,972,000 9.24 1.65 180 28 44 2,283,000 7.88 1.231988 2,387 460 797 27,446,000 8.69 1.68 179 21 27 2,707,000 6.61 0.781989 2,242 431 768 27,920,000 8.01 1.53 187 30 44 2,829,000 6.61 1.061990 2,241 443 767 28,510,000 7.86 1.55 195 25 28 2,392,000 8.15 1.051991 2,197 438 799 27,678,000 7.93 1.58 170 30 51 2,756,000 6.17 1.091992 2,111 451 867 24,780,000 8.51 1.82 179 41 72 2,282,000 7.84 1.801993 2,063 400 740 22,796,000 9.05 1.75 180 37 71 1,833,000 9.82 2.021994 2,022 404 730 22,235,000 9.08 1.81 218 44 79 1,777,000 12.26 2.451995 2,056 413 735 24,906,000 8.24 1.65 161 25 45 1,961,000 8.21 1.271996 1,908 361 636 24,881,000 7.67 1.45 176 32 55 2,120,000 8.29 1.511997 1,845 350 631 25,591,000 7.21 1.36 164 28 46 2,084,000 7.87 1.341998 1,904 364 624 25,518,000 7.45 1.42 191 34 66 2,138,000 8.93 1.591999 1,906 340 619 29,713,000 6.41 1.14 198 31 57 2,171,000 9.12 1.432000 1,838 343 594 29,057,000 6.33 1.18 206 35 63 2,472,000 8.33 1.422001 1,721 321 553 26,220,000 6.56 1.22 182 29 51 2,381,000 7.64 1.22
Part135: Scheduled Part135: Non-scheduled
Year Accidents Flight HoursAccidents
per 100,000Flight Hours
Accidents Flight HoursAccidents
per 100,000Flight Hours
All Fatal Fatalities All Fatal All Fatal Fatalities All Fatal1982 26 5 14 1,299,748 2.000 0.385 132 31 72 3,008,000 4.39 1.031983 16 2 11 1,510,908 1.059 0.132 142 27 62 2,378,000 5.97 1.141984 22 7 48 1,745,762 1.260 0.401 146 23 52 2,843,000 5.14 0.811985 18 7 37 1,737,106 1.036 0.403 157 35 76 2,570,000 6.11 1.361986 14 2 4 1,724,586 0.812 0.116 118 31 65 2,690,000 4.39 1.151987 33 10 59 1,946,349 1.695 0.514 96 30 65 2,657,000 3.61 1.131988 18 2 21 2,092,689 0.860 0.096 102 28 59 2,632,000 3.88 1.061989 19 5 31 2,240,555 0.848 0.223 110 25 83 3,020,000 3.64 0.831990 15 3 6 2,341,760 0.641 0.128 107 29 51 2,249,000 4.76 1.291991 23 8 99 2,291,581 1.004 0.349 88 28 78 2,241,000 3.93 1.251992 23 7 21 2,335,349 0.942 0.300 76 24 68 2,844,000 2.67 0.841993 16 4 24 2,638,347 0.606 0.152 69 19 42 2,324,000 2.97 0.821994 10 3 25 2,784,129 0.359 0.108 85 26 63 2,465,000 3.45 1.051995 12 2 9 2,627,866 0.457 0.076 75 24 52 2,486,000 3.02 0.971996 11 1 14 2,756,755 0.399 0.036 90 29 63 3,220,000 2.80 0.901997 16 5 46 982,764 1.628 0.509 82 15 39 3,098,000 2.65 0.481998 8 0 0 353,670 2.262 - 77 17 45 3,802,000 2.03 0.451999 13 5 12 342,731 3.793 1.459 73 12 38 3,298,000 2.21 0.362000 12 1 5 373,649 3.212 0.268 81 22 71 3,553,000 2.28 0.622001 7 2 13 330,500 2.118 0.605 72 18 60 3,400,000 2.12 0.53
Source: www.ntsb.gov/aviation/Table10.htm Source: www.rotor.com/safety/stats70.01.xls
Figure 2–14: U.S. General Aviation and Helicopter—Accidents, Fatalities and Rates: 1982 through 2001
Source: www.ntsb.gov/aviation/Table8.htm Source: www.ntsb.gov/aviation/Table8.htm
Figure 2–15: U.S. Air Carriers Operating Under Part 135—Accidents, Fatalities and Rates, 1982 through 2001(Since March 20, 1997 only aircraft with fewer than 10 seats)
34 AIR MEDICAL PHYSICIAN HANDBOOK
Figure 2–17: Accidents per 100,000 Flight Hours
• General aviation and helicop-ters fly to more than 20,000landing facilities, while the air-lines serve only about 700well-lit airline-served airports.
• Many operations, such as EMS,aerial application, and lawenforcement, have special mis-sion-related risks.
• There are more takeoffs andlandings per hour, generally thehighest risk phases for generalaviation and many helicopteroperations.
Figure 2–17 shows that theaccident rate for HEMS was dra-matically higher than for allother aviation operations duringthe early and mid-1980s.Beginning in1987, we see a sharpdecline in the HEMS accidentrate, which has remained consis-tently below the accident ratesfor both general aviation and all
Part 121: Scheduled Service Part 121: Nonscheduled Service
Year Accidents Flight HoursAccidents
per 100,000Flight Hours
Accidents Flight HoursAccidents
per 100,000Flight Hours
All Fatal Fatalities All Fatal All Fatal Fatalities All Fatal1982 16 4 234 6,697,770 0.224 0.045 2 1 1 342,555 0.584 0.2921983 22 4 15 6,914,969 0.318 0.058 1 0 0 383,830 0.261 -1984 13 1 4 7,736,037 0.168 0.013 3 0 0 429,087 0.699 -1985 17 4 197 8,265,332 0.206 0.048 4 3 329 444,562 0.900 0.6751986 21 2 5 9,495,158 0.211 0.011 3 1 3 480,946 0.624 0.2081987 32 4 231 10,115,407 0.306 0.030 2 1 1 529,785 0.378 0.1891988 29 3 285 10,521,052 0.266 0.019 1 0 0 619,496 0.161 -1989 24 8 131 10,597,922 0.226 0.075 4 3 147 676,621 0.591 0.4431990 22 6 39 11,524,726 0.191 0.052 2 0 0 625,390 0.320 -1991 25 4 62 11,139,166 0.224 0.036 1 0 0 641,444 0.156 -1992 16 4 33 11,732,026 0.136 0.034 2 0 0 627,689 0.319 -1993 22 1 1 11,981,347 0.184 0.008 1 0 0 724,859 0.138 -1994 19 4 239 12,292,356 0.146 0.033 4 0 0 831,959 0.481 -1995 34 2 166 12,776,679 0.266 0.016 2 1 2 728,578 0.275 0.1371996 32 3 342 12,971,676 0.247 0.023 5 2 38 774,436 0.646 0.2581997 44 3 3 15,061,662 0.292 0.020 5 1 5 776,447 0.644 0.1291998 43 1 1 15,921,102 0.270 0.006 7 0 0 892,333 0.784 -1999 47 2 12 16,693,365 0.282 0.012 5 0 0 861,843 0.580 -2000 51 3 92 17,474,405 0.292 0.017 6 0 0 820,738 0.731 -2001 36 6 531 15,998,000 0.200 0.013 4 0 0 732,700 0.546 -
NOTE: Effective March 20, 1997, aircraft with 10 or more seats must conduct scheduled passenger operations under part 121.
Source: www.ntsb.gov/aviation/Table6.htm Source: www.ntsb.gov/aviation/Table7.htm
Figure 2–16: U.S. Air Carriers Operating Under Part 121 (Airlines)—Accidents, Fatalities and Rates, 1982through 2001
H E M S
0
5
1 0
1 5
2 0
2 5
HEMS 21.7 17.7 10.6 16.7 14.2 3.47 6.5 5.28 0.71 3.97 3.81 1.66 2.44 4.08 0.54 1.57 4.27 4.82 6.18 5.97
Helicopter 10.9 10.3 8.98 9.52 7.24 7.88 6.61 6.61 8.15 6.17 7.84 9.82 12.3 8.21 8.29 7.87 8.93 9.12 8.33 7.64
General Av iation 10.9 10.7 10.4 9.66 9.53 9.24 8.69 8.01 7.86 7.93 8.51 9.05 9.08 8.24 7.67 7.21 7.45 6.41 6.33 6.56
135 Nonscheduled 4.39 5.97 5.14 6.11 4.39 3.61 3.88 3.64 4.76 3.93 2.67 2.97 3.45 3.02 2.8 2.65 2.03 2.21 2.28 2.12
135 Scheduled 2 1.06 1.26 1.04 0.81 1.69 0.86 0.85 0.64 1 0.94 0.61 0.36 0.46 0.4 1.63 2.26 3.79 3.21 2.12
121 Nonscheduled 0.58 0.26 0.7 0.9 0.62 0.38 0.16 0.59 0.32 0.16 0.32 0.14 0.48 0.28 0.65 0.65 0.78 0.58 0.7 0.55
121 Scheduled 0.22 0.32 0.17 0.21 0.21 0.31 0.27 0.23 0.19 0.22 0.14 0.18 0.15 0.27 0.25 0.29 0.27 0.28 0.29 0.2
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 '00 '01
HEMS
35Special Supplement
helicopter aviation. In addition, from1987 through 1997, the HEMS acci-dent rate was lower than the overallaccident rate for all Part 135 non-scheduled flights 6 of the 10 years.Since 1998, however, the HEMS acci-dent rate has surpassed that of thenon-scheduled Part 135 operationseach year.
The data in Figure 2–18 representsthe average accident rate for the past20 years (1982–1999), 10 years, and 5years for the five types of aviationoperations, helicopters, and HEMS.Even with the high accident rates ofthe 1980s, the 20-year average forHEMS is below all helicopter opera-tions and general aviation. For the10-year average, the HEMS accidentrate is less than 50% the rate of allhelicopters and general aviation. Forthe past 5 years, the average accidentrate for HEMS has gone up, butremains significantly lower than allhelicopter operations and general aviation.
Figure 2–19 compares the fatal acci-dents per 100,000 flight hours for thevarious aviation operations. Theresults are similar to the total accidentrates. Initially, the fatality rate for airmedical helicopters was equal to ordramatically higher than all other avi-ation operations. In 1990, however,there were no fatal HEMS accidents.Then from 1992 to 1997, HEMS wasconsistently below both general avia-tion and all helicopter operations infatal accidents. Since 1998, theHEMS fatality rate has been consis-tently higher.
The average fatal accident rate forthe past 20 years, 10 years, and 5 yearsfor the various aviation operations isseen in Figure 2–20. The 20-yearaverage shows HEMS with a high fatalaccident rate compared to all otheraviation. However, the past 10 yearshas seen an improvement in theHEMS rate. For the 10-year average,HEMS has had a lower fatality ratethan helicopters and general aviation(Part 91). For the past 5 years, how-ever, the average HEMS fatality rateonce again exceeds all other aviationoperations.
6.81
3.53
4.56
8.53 8.838.388.47
7.65
6.79
3.6
2.622.26
1.35 1.58
2.6
0.49 0.51 0.650.23 0.23 0.27
0
1
2
3
4
5
6
7
8
9
10
20-yr Average 10-yr Average 5-yr AverageAcc
iden
tRat
epe
r10
0,00
0Fl
ight
Hou
rs
HEMS Helicopter General Aviation 135 Nonscheduled
135 Scheduled 121 Nonscheduled 121 Scheduled
Figure 2–18: Average Accident Rates for Aviation Operations
0
1
2
3
4
5
6
7
8
9
HEMS 5.44 4.42 1.77 8.35 4.74 3.76 1.76 3.27 0 2.46 1.18 1.2 1.76 0.63 0.59 1.14 1.68 1.62 2.13 1.9
Helicopter 1.74 1.84 1.52 1.67 1.49 1.23 0.78 1.06 1.05 1.09 1.8 2.02 2.45 1.27 1.51 1.34 1.59 1.43 1.42 1.22
General Av iation 1.99 1.94 1.87 1.75 1.75 1.65 1.68 1.53 1.55 1.58 1.82 1.75 1.81 1.65 1.45 1.36 1.42 1.14 1.18 1.22
135 Nonscheduled 1.03 1.14 0.08 1.36 1.15 1.13 1.06 0.83 1.29 1.25 0.84 0.82 1.05 0.97 0.9 0.48 0.45 0.36 0.62 0.53
135 Scheduled 0.39 0.13 0.4 0.4 0.12 0.51 0.1 0.22 0.02 0.35 0.3 0.15 0.11 0.08 0.04 0.51 0 1.46 0.27 0.61
121 Nonscheduled 0.29 0 0 0.68 0.21 0.19 0 0.44 0 0 0 0 0 0.14 0.26 0.13 0 0 0 0
121 Scheduled 0.05 0.06 0.01 0.05 0.01 0.03 0.02 0.08 0.05 0.04 0.03 0.01 0.03 0.02 0.02 0.02 0.01 0.01 0.02 0.13
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 '00 '01
Figure 2–19: Fatal Accidents per 100,000 Flight Hours
2.49
1.38
1.69
1.481.61
1.4
1.61.48
1.26
0.870.7
0.49
0.31 0.35
0.57
0.121 0.05 0.030.03 0.03 0.04
0
0.5
1
1.5
2
2.5
3
20-yr Average 10-yr Average 5-yr Average
Fata
lAcc
iden
tspe
r10
0,00
0Fl
ight
Hou
rs
HEMS Helicopter General Aviation 135 Nonscheduled
135 Scheduled 121 Nonscheduled 121 Scheduled
Figure 2–20: Average Fatal Accident Rates for Aviation Operations
HEMS
36 AIR MEDICAL PHYSICIAN HANDBOOK
U.S. Army HelicopterAccidents
In comparing HEMS to other aviationoperations, we have not included any formof military aviation. However, much of theresearch regarding helicopter crashes andsurvivability has come from U.S. Armystudies. It would seem a natural extensionof our own study to compare Army acci-dent rates with HEMS.
Information was obtained from the U.S.Army Safety Center at Fort Rucker,Alabama. From 1982–2000, U.S. Armyhelicopters flew an estimated 22.8 millionflight hours worldwide, ranging from a highof 1.55 million hours in 1988 to a low of753,000 hours in 1998. Total accidentsthat do not include any combat losses andaccidents rates for nine different Army hel-icopters were also provided. Of these ninetypes of aircraft, the Army uses two modelsfor medical missions—the UH-1 (“Huey”)and the UH-60 (“Blackhawk”). Additionalaccident information was provided onthese two aircraft for 1992–2000, for boththe medical (UH1-V and UH-60 MEDE-VAC) and non-medical (UH1-AC andUH-60) versions.
During the 18-year period, the U.S.Army recorded a total of 707 Class A andClass B non-combat helicopter accidents.Between 1992–2000, there were a total of212 accidents. Looking only at the UH-1and UH-60 from 1992–2000, there were 77Class A and Class B accidents—11 inMEDEVAC aircraft and 66 in non-medicalhelicopters. There were an additional 253Class C accidents reported for the UH-1and UH-60 helicopters during these 9years. A total of 44 involved medical heli-copters and 209 were non-medical aircraft.
It should be noted that the U.S. ArmyAccident Classification is very differentthan the NTSB definitions of accidentsand incidents. Clearly, Class A and Bwould qualify as accidents under the NTSBdefinitions. Class C however, seems tocross the line between the NTSB defini-tions of accident and incident.
Unfortunately, our analysis of the Armyaccident data is limited by several factors.For 1982–2000, we have total flight hours,the number of accidents and the accidentrate for the nine aircraft—but only forClass A and B accidents. For the UH-1and UH-60, we have raw numbers of acci-dents for all three classes and for the mede-
0
5
1 0
1 5
2 0
2 5
A c c
i d en
t s
p er
10 0
, 00 0
F l i g
h t H
o u
r s
HEMS 21.7 17.7 10.6 16.7 15.4 3.76 7.05 5.73 0.77 4.31 4.14 1.8 2.65 4.42 0.59 1.71 4.48 5.41 6.39
U.S. Army 6.06 4.06 3.09 3.62 2.79 3.11 2.52 2.75 2.58 4.93 2.11 3.2 2.68 2.19 2.07 2.79 2.12 3.56 1.19
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 '00
Army Accident Classification
Class A. Damage costs of $1,000,000 or more and/or destruction of an Army aircraft, missile or spacecraft and/or fatality or permanent total disability.
Class B. Damage costs of $200,000 or more, but less than $1,000,000 and/or permanent partial disability and/or five (three as of 2002) or more people are hospitalized as inpatients.
Class C. Damage costs of $10,000 ($20,000 as of 2002) or more, but less than $200,000 and/or non-fatal injury resulting in loss of time from work beyond day/shift when injury occurred and/or non-fatal illness/disability causes loss of time from work.
Figure 2–21: U.S. Army Helicopter Class A and B Accidents vs. HEMS, 1992-2000 (n=77) Adapted from: U.S. Army Safety Center, Fort Rucker, Alabama
vac vs. non-medevac aircraft. However,we lack the corresponding breakdown offlight hours for the medical flights com-pared to the non-medical flights.Knowing the specific medevac flighthours would have allowed a more mean-ingful comparison. Finally, we do nothave any fatal accident numbers.
Knowing the limitations of our com-parison, we have plotted the combinedClass A and Class B accident rates alongwith the HEMS accident rates in Figure2–21. Over this 19-year period, theArmy accident rate was as high as 6.06accidents per 100,000 flight hours in1982 and as low as 1.19 in 2000. In the’80s, the HEMS accident rate was con-sistently higher than the Class A+Baccident rate. From 1990–1997, therates are very similar and since 1998 theHEMS rates have again been higher.
While we do not know the specificaccident rates for the Army medevachelicopters, we do know that from1992–2000, there were six non-medicalaccidents for every one medevac crash
involving the Huey and Blackhawk heli-copters. However, if we were to alsofactor in the Class C accidents, our rawaccident numbers would be increasedmore than four-fold.
Of interest, it should be noted thatU.S. Army MEDEVAC missions arealways 2-pilot operations. Most flightsare “scene” flights and night missions,which account for an estimated half ofall medical missions and are aided bynight vision goggles (NVG).
Figure 2–22: UH–60 vs. UH-1 Accidents1992–2000 (n=77) Adapted from: U.S. ArmySafety Center, Fort Rucker, Alabama
U H - 6 0 5 4 %
U H - 1 V 8 %
U H - 1 A C 3 2 %
U H - 6 0 M E D E V A C
6 %
Adapted from: http://asmis.army.mil/stats/pkg_definitions.class
37Special Supplement
3. 53
1. 38
4. 56
1. 69
8. 93
1. 63
8. 28
2. 04
13 .1 9 13 .1 9
5. 43 5. 43
2. 7
0. 8
1. 94
0. 57
3. 43
1. 01 1. 57
0 0
2
4
6
8
1 0
1 2
1 4
1 0 - y r A v g . A c c ide n t R a t e
1 0 - y r A v g . Fa t a l A c c ide n t R a t e
5 - y r A v g . A c c ide n t R a t e
5 - y r A v g . Fa t a l A c c ide n t R a t e
A c c
i d e
n t s
p e
r 1 0
0 , 0
0 0 F
l i g
h t
H o
u r s
H E M S FS H e li c o p t e r A ir t a n k e r Fi x e d - w ing U S FS Ow n e d
U.S. Forest ServiceAviation Accidents
While the HEMS fatal accident ratehas been higher than that of all othertypes of aviation for the past severalyears, it is not the most hazardous fly-ing in the United States.
In August 2002, the NBC NightlyNews reported that piloting ForestService airtankers was the most dan-gerous flying in the United States.Since 1950, a total of 156 pilots hadlost their lives fighting fires. NBCreported that the accident rate forForest Service airtankers was found tobe 13 per 100,000 hours of flight time.In comparison, they reported that theaccident rate of U.S. military combatflight was 11 per 100,000 flight hourswhile that of civilian aviation was 3.6per 100,000 flight hours.
Our own research into the ForestService accident rates found that thereported accident rate of 13 per100,000 hours of flight time representsa 10-year average for airtankers from1992 to 2001. During that time, theannual range was from 0 to 51.36.During that same timeframe, otherfixed-wing aircraft (not airtankers)had an average accident rate of 2.70(range 0 to 15.13) and helicopters hadan average of 8.93 (range 0 to 24.55).The 10-year average for all ForestService aviation was 5.78 accidentsper 100,000 flight hours, with a rangeof 1.58 to 11.67.
Figure 2–23 shows the annual acci-dent rates and fatal accident rates forthe four different Forest Service avia-tion operations compared to HEMS.Figure 2–24 plots the 5-year and 10-year averages for these operations.
Within this 10-year period, HEMShad the highest accident rates andhighest fatal accidents in 1999 and2001. In addition, in 1996, whenHEMS had it’s second lowest fatalaccident rate, the U.S. Forest Servicehad a rate of zero. As the 5-year and10-year averages indicate, HEMS iswell below the accident rates for heli-copters and airtankers used under con-tract by the U.S. Forest Service, butremains above the rates for fixed-wingaircraft.
Figure 2–23: Accident Rates for U.S. Forest Service Aviation Operations vs. HEMS,1992–2001Adapted from: U.S. Department of Agriculture, Forest Service, http://www.aviation.fs.fed.us/library/fy01avsumm.pdf
’92 ’93 ’94 ’95 ’96 ’97 ’98 ’99 ’00 ’01 5-yr 10-yr Avg. Avg.
Accident rate per 100,000 flight hours
HEMS 3.81 1.66 2.44 4.08 0.54 1.57 4.27 4.82 6.18 5.97 4.56 3.53
USFS Owned 0 0 6.94 10.11 8.58 0 0 0 7.84 0 1.57 3.35
Fixed-wing 0 15.13 2.22 0 0 0 3.08 0 2.85 3.76 1.94 2.70
Airtanker 19.42 51.36 9.9 24.07 0 0 27.13 0 0 0 5.43 13.19
Helicopter 14.29 8.31 14.22 0 11.01 24.54 4.09 3.97 3.76 5.06 8.28 8.93
Fatal Accident rate per 100,000 flight hours
HEMS 1.18 1.2 1.76 0.63 0.59 1.14 1.68 1.62 2.13 1.9 1.69 1.38
USFS Owned 0 0 0 10.11 0 0 0 0 0 0 0.00 1.01
Fixed-wing 0 5.04 0 0 0 0 0 0 2.85 0 0.57 0.79
Airtanker 19.42 51.36 9.9 24.07 0 0 27.13 0 0 0 5.43 13.19
Helicopter 0 0 6.09 0 0 6.13 4.09 0 0 0 2.04 1.63
Figure 2–24: U.S. Forest Service Aviation Accidents vs. HEMS, 1992–2001Adapted from: U.S. Department of Agriculture, Forest Service, http://www.aviation.fs.fed.us/library/fy01avsumm.pdf
SINGLE- VS. TWIN-ENGINEHELICOPTER ACCIDENTRATES
Since 1990, HAI has tracked the acci-dent rate for single-engine vs. multi-engine helicopters, as well as the total hel-icopter flight hours. In general, it appearsthat single-engine flight hours have beenapproximately 3 times that of multi-engineflight hours each year. A dramatic differ-
ence is seen in the accident rate per100,000 flight hours when comparing single- vs. twin-engine helicopters. Thefatal accident rate however demonstratesless disparity. In fact, a 1999 study by theFlight Safety Foundation found the fatalaccident rate of single- and twin-enginehelicopters to be similar. Figure 2–25charts the total accident and fatal accident rates for single- and multi-engine helicopters.
38 AIR MEDICAL PHYSICIAN HANDBOOK
The Flight Safety Foundationstudy looked at turbine-engine heli-copter accidents and compared themwith all Part 135 operations and withgeneral aviation. Figure 2–26 showsthat the twin-engine helicopter acci-dent rate was lower than the acci-dent rates of general aviation air-craft, single-engine helicopters, andall Part 135 operations.
Nearly half (48%) of the twin-engine helicopter accidents werefatal. From 1993–1997, twin-enginehelicopters were involved in 23 fatalaccidents, or 1.4 fatal accidents per100,000 flight hours. Figure 2–27compares the fatal accident rates forhelicopters (single- and twin-engine)general aviation, and air taxis.
HELICOPTER ACCIDENTS, 2001
Year 2001 saw a 12% increasecompared to 2000 when looking atall helicopter accidents. During theyear, there were 217 helicopter acci-dents (see Figure 2–28), 42 of whichwere fatal, killing 105 people.Piston-engine helicopters accountedfor 108 accidents, of which 19 werefatal. There were 95 single-turbineaccidents (20 fatal) and 13 twin-tur-bine accidents (3 fatal). Of interest,this NASA “Safecopter” website listsonly one of 12 air medical helicopteraccidents as fatal.
SECTION 3: A COMPARISON OF RISK
Nothing is completely safe.Everything we do has some type ofrisk and these risks can never betotally eliminated from any situation.The issue is not one of avoiding risksall together, but rather one of man-aging risk in a sensible manner.
A 1999 Time magazine article,“Life on the Edge,” states that“America has embarked on a nation-al orgy of thrill seeking and risk tak-ing.” The article focused on the riseof extreme sports like BASE jump-
0
1
2
3
4
5
6
7
8
9
10A
ccid
entR
ate
Single Engine -- Total
Multi-Engine -- Total
Single Engine -- Fatal
Multi-Engine -- Fatal
Figure 2–25: Total Accidents and Fatal Accidents per 100,000 Flight HoursSingle- vs. Multi-Engine Helicopters, 1990–2001Adapted from: Helicopter Association International, http://www.rotor.com/safety/stats70.01.xls
8.36
6.76
4.3
2.93
0
1
2
3
4
5
6
7
8
9
A ll GeneralAviation
S ingle-EngineHelicopters
A ll Part 135Operators
Twin-EngineHelicopters
Acc
iden
tsp
er10
0,00
0F
ligh
tHo
urs
Figure 2–26: Accident Rate Comparison,1993–1997Adapted from: Harris, Helicopter Safety, Jan/Feb 1999
1.63
1.4
1.22
1.4
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
A ll GeneralAviation
S ingle-EngineHelicopters
All P art 135Operators
Twin-EngineHelicopters
Acc
iden
tsp
er10
0,00
0F
ligh
tHo
urs
Figure 2–27: Fatal Accident RateComparison, 1993–1997Adapted from: Harris, Helicopter Safety, Jan/Feb 1999
Figure 2–28: Helicopter Accident Statistics, 2001 Adapted from: http://safecopter.arc.nasa.gov/
48
37
16
7 9
21
15 18
12
6
28
7 4
1 2 3 4 5 2 1 1
12
0
1 0
2 0
3 0
4 0
5 0
6 0
P e r s o
n a l
I n s t r
u c t io
n
A e r ial
A p p
.
A e r ial
Ob s
.
B u s in
e s s
A ir T a x
i
E x t e r n
a l L o
a d
P u b li c
U s e
A ir M e d
ic a l
Tes t
Fli g
h t
Mi s c / U
n k n o
w n
M i s s i o n
A cci de n t s F at al A cci de n t s
Acc
iden
ts
39Special Supplement
ing, snowboarding, ice climbing, skate-boarding, and paragliding but mentionedother risky activities—and occupations—as well. The article pointed out thatAmericans were taking greater risks thanever before in many areas. More than30% of U.S. households owned stocks ofsome form or another, which was up from12% just 10 years before. Social behaviorhad also become more risky, with unpro-tected sex on the upswing and illicitdrugs like heroin on the increase.Finally, the article pointed out that manypeople assumed various risks in their cho-sen careers. From the MBAs who weretrying to strike it rich in the “dot-com”world; to the 14.5% who voluntarily hadleft their jobs (highest in a decade) fornew opportunities; to the options trader,neurosurgeon, fire fighter, and race cardriver—all had chosen to assume a vary-ing degree of occupational risk.
For all of these thrill seekers and risktakers, risk management requires a mini-mum of common sense and informationabout the character and magnitude of therisk taken. We must inform ourselves orbe educated about the relevant risks andthen act accordingly. Where you chooseto live, your occupation, chosen modes oftravel, recreational activities, or just stay-ing at home—all have risks of accidents,injury, and death.
Determining Risk
Statistics on accidents, injuries, andfatalities are kept for nearly every activityand occupation. With raw data avail-able, we need to be able to calculate themagnitude of a risk. This is especiallyimportant if we hope to compare differ-ent types of activities. In order to assessand compare risk, the relevant figureneeded must be in the form of a ratio,fraction, or percentage. To arrive atthese figures—to normalize the data—weneed to know two numbers. The numer-ator of the fraction tells us how manyindividuals doing a particular activitywere either injured or killed over a givenperiod of time. The denominator repre-sents how many people were engaged inthat activity—the population at risk. Byreducing all risks to ratios in this way wecan begin to compare different types ofactivities and the relative risks. The larg-er the ratio, the riskier the activity.
There are two other options to normal-ize this data which are used more com-monly by the National Safety Council(NSC). The first is to take the abovefraction and normalize it to “1 in X,” as in“the odds of something is 1 in X.” Theother option is to calculate death ratesand injury rates per 100,000 persons.
There is risk of injury and death everyhour of every day. The NSC estimatesthat while you make a 10-minute safetypresentation, two persons will be the vic-tims of unintentional deaths and approxi-mately 370 will suffer a disabling injury.On the average, there are 12 uninten-tional-injury deaths and about 2,400 dis-abling injuries every hour during the year.Figure 3–1 shows a breakdown of the fre-quency and death rate of the major clas-sifications of accidental deaths andinjuries in the United States for 2000.
An individual is nearly twice as likelyto be injured at work than in his/her car.Staying at home or going out in public iseven worse. You are more than threetimes more likely to be injured at homeor in public than in your car. However,in general, you are six times more likelyto die in your car than die due to an acci-dent at work and nearly one-and-a-halftimes more likely to suffer a fatal injuryin your car than at home.
Where you choose to live will alsoimpact your likelihood of dying from anunintentional injury. According to NSCstatistics, Massachusetts and RhodeIsland boast the lowest accidental deathrates among the fifty states. However, ifyou live in New Mexico, Wyoming, orAlaska (the states with the highest acci-dental death rate), your chances of anunintentional death are 2.5 to nearly 3times higher.
Figure 3–1: Unintentional Deaths and Injuries, 2000Adapted from: National Safety Council, http://nsc.org/library/rept2000.htm
Death RateClass Severity 2000 TotalAll Unintentional Deaths 5 minutes 97,300 35.3
Injuries 1.5 seconds 20,500,000Motor-Vehicle Deaths 12 minutes 43,000 15.6
Injuries 14 seconds 2,300,000Work Deaths 102 minutes 5,200 1.9
Injuries 7 seconds 3,900,000Home Deaths 18 minutes 29,500 10.7
Injuries 4 seconds 7,100,000Public (nonmotor-vehicle) Deaths 24minutes 22,000 8.0
Injuries 4 seconds 7,300,000
One every
HEMS: ANALYZING THEPOPULATION AT RISK
With injury and death statistics avail-able for various occupations and types ofactivities, it would be necessary to deter-mine the size of the “population at risk”in HEMS. To accomplish this—andwith no such data available—we mustmake several assumptions and do variouscalculations.
Methodology
If one were to try to compare air med-ical transport to other occupations or“routine” risks to determine either theodds of death in one year or the fatalityrate per 100,000 we would need to knowtwo things. The first is the number ofHEMS crew fatalities per year. The sec-ond would be the number of peopleengaged in HEMS transport (i.e., thenumber of HEMS pilots and medicalcrewmembers) for each year. The num-ber of fatalities is known, but the numberof crewmembers in HEMS has neverbeen tracked or even estimated in theliterature.
For the purpose of this study, we beginby estimating the average number ofcrewmembers per helicopter. We canassume that the typical flight crew foreach dedicated medical helicopterincludes 4 pilots, 6–8 nurses as the pri-mary caregivers, and 10–12 second med-ical crewmembers (often paramedics,physicians, nurses, respiratory therapists,etc., who fly full- or part-time).Therefore, the average dedicated medicalhelicopter would have 20 to 24crewmembers. For the purpose of ourcalculations, we will use an average of 22persons.
40 AIR MEDICAL PHYSICIAN HANDBOOK
A review of the air medical literatureshowed there was no documentation asto the number of dedicated EMS helicop-ters from 1981 to 1991. Our InternetFlightweb survey provided us with a fairlyaccurate number of dedicated helicoptersas well as HEMS programs for the year2000. Assuming a steady annual increasein the number of aircraft between 1991and 2000, we are able to predict thenumber of helicopters dedicated to theHEMS mission for each year. For 2001,we factored in the percent increase asdetermined from our operator and manu-facturer survey.
We now have the necessary informa-tion to determine the approximate size ofthe population at risk. In 2001, therewere an estimated 400 dedicated medicalhelicopters. Multiplying this figure by 22crewmembers, we can estimate that thepopulation at risk is approximately 8,792.Using this figure and knowing the num-ber of crewmember deaths attributed toHEMS accidents, we are now able tocompare some annual statistics in a moremeaningful manner. In addition, we willbe able to compare HEMS risk withother occupations and activities.
It is important to realize that in esti-mating exposure in this method, we aredoing so for the average crewmember andthe average flight program. The NationalSafety Council points out that predictingthe rate for potential injury or death isstrongly dependent upon the length ofexposure to a particular activity for thespecific population at risk. This wouldbe similar to comparing the potentialexposure rate for full-time vs. part-time
HEMS personnel; four pilots vs. 6–8nurses; programs that fly once a day vs.those that fly several times a day; pro-grams that fly short distances to thosethat fly much further, and so on. Whenwe normalize the raw data, it does nottake into account the amount of exposurefor an individual during the year. Forthis study, we are basing our calculationson the average program, which in 2001transported approximately 882 patients,flying an estimated 957 hours over thecourse of the full year, and all crewmem-bers (pilots and medical crewmembers)flying an equal amount of time.
Results
Fatality statistics for HEMS personnelare presented in four different formats.In each case, the number of crew fatali-ties was determined for each year. Fromthe total 171 HEMS fatalities, 21 patient
fatalities, 7 dual-purpose aircraftcrewmember fatalities and 6 other fatali-ties were removed from the appropriateyears leaving only dedicated HEMS crewfatalities for our comparison. Figure 3–2depicts the number of HEMS personnelwho have died each year since 1980.The fatalities for 2002 (as of September30, 2002) are included in this graph butare not included in any of our calcula-tions. As the graph clearly shows, 2002has had more crew fatalities than anyyear in HEMS history. Figure 3–3 showsthe various calculations and the resultsused to determine the fatality rates.
The National Safety Council routinelynormalizes fatality data to a death rateper 100,000 population at risk in a givenyear. Over the 21 years reviewed for thisportion of the study (1981–2001), theHEMS population has grown fromapproximately 858 to 8,792. While thisgrowth seems impressive, this is still a
Figure 3–3: Fatalities and Fatality Rates
1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Totals
1 # of Helicopters 39 45 62 75 91 119 151 184 195 213 231 225 242 259 276 293 309 326 343 360 377 400
2 Calc. Total Crew 858 990 1364 1650 2002 2618 3322 4048 4290 4686 5082 4950 5322 5693 6065 6436 6808 7179 7551 7922 8294 8792 185064
3a Crew Fatalities 6 5 6 4 1 11 12 9 4 11 0 11 3 3 11 3 2 4 9 10 10 2 1373b Dual-Purpose Crew Fatalities 2 2 3 74 Patient Fatalities 1 1 1 0 1 1 1 1 1 3 0 3 0 2 0 0 1 1 2 0 1 215 Other Fatalities 2 1 3 66 Total Fatalities 7 6 7 4 2 12 13 12 5 16 0 17 3 5 11 3 3 5 14 10 11 5 171
7 Crew Fatality Rate 0.70% 0.51% 0.44% 0.24% 0.05% 0.42% 0.36% 0.22% 0.09% 0.23% 0.00% 0.22% 0.06% 0.05% 0.18% 0.05% 0.03% 0.06% 0.12% 0.13% 0.12% 0.02% 0.196%8 1 in_______ 143 198 227 413 2002 238 277 450 1073 426 N/A 450 1774 1898 551 2145 3404 1795 839 792 829 4396 11589 per 100,000 699 505 440 242 50 420 361 222 93 235 0 222 56 53 181 47 29 56 119 126 121 23 196
10 No Injuries 2 6 17 9 8 11 19 0 15 8 4 9 4 0 4 13 0 5 12 18 13 18 19511 Injuries Serious 0 0 0 3 3 5 3 0 5 4 0 0 10 3 1 0 0 0 5 6 0 8 5612 Minor 0 3 0 3 8 11 10 1 6 3 0 0 4 3 3 5 0 1 0 0 7 5 7313 Total Injuries 0 3 0 6 11 16 13 1 11 7 0 0 14 6 4 5 0 1 5 6 7 13 12914 Injury Rate 0.0% 0.30% 0.00% 0.36% 0.55% 0.61% 0.39% 0.02% 0.26% 0.15% 0.00% 0.00% 0.26% 0.11% 0.07% 0.08% 0.00% 0.01% 0.07% 0.08% 0.08% 0.15% 0.16%
1980
0
65
6
4
1
1112
9
4
11
0
11
3 3
11
32
4
910 10
2
13
0
2
4
6
8
10
12
14
Year
Cre
wF
atal
ities
Figure 3–2: HEMS Crew Fatalities per Year, 1980–2002*(as of September 30, 2002)
41Special Supplement
very small sampling to translate to a ratioper 100,000. As Figure 3–4 shows, withsuch a small population base, each fatali-ty has a significant impact on the fatalityrate. In this format, the range for thefatality rate is from 0 to 699 per 100,000.With such a wide range, a 22-year aver-age is calculated that will be used whenwe compare HEMS to other risks. Theaverage annual death rate over the 22years is 196 per 100,000 crewmembers.
Another way to look at the relativerisk of HEMS transport is in the form ofa ratio—dividing the number of fatalitiesby the number of crewmembers for eachyear. Since this uses the same data, butin a slightly different equation, the graphwould look essentially the same as Figure3–4. For this annual comparison (Figure3–3, Line 7) the range is from 0.00% to0.70%, with a 22-year average of0.196%. The higher the percentage, thegreater the apparent risk in that particu-lar year.
The final relationship that is used tocompare the annual number of HEMScrew fatalities is in terms of “odds.” Forexample, in 2001 there were only twocrew deaths out of an estimated crewpopulation of 8,792. Looking solely atthe numbers, the odds to an individualcrewmember suffering a fatal accidentthat year would be considered to be 1 in4,396. Contrasting this with what couldbe considered our riskiest year (1980),there were 6 crew fatalities out of an esti-mated 858 crewmembers industry-wide.This would correspond to fatality odds of1 in 143. Excluding 1990 when therewere no fatalities, the average odds peryear over the 22-year period are 1 in1,158. Figure 3–5 illustrates the odds ofa fatality over the study time period.
COMPARING HEMS TO OTHER RISKS
To further illustrate the risk related toHEMS transport, we can compare andcontrast the above numbers with otheractivities, other types of accidents, andother causes of death. Taking into con-sideration the wide range of fatality ratesand odds that we have estimated for eachyear in HEMS, the calculated averageswill be used in subsequent comparisons.
0
100
200
300
400
500
600
700
800
Year
Figure 3–4: HEMS Fatality Rate per 100,000 Personnel, 1980–2001
No
Fat
aliti
esin
1990
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
Year
Odd
s,1
in..
...
Figure 3–5: Odds of a Fatality for HEMS Personnel, 1981–2001
Cause of Death – An Overview
Since 1921, the National SafetyCouncil has been a source of accurate,comprehensive, and objective statisticson unintentional injuries, their costs,trends, and other characteristics.Accidents and unintentional injuriesgenerally rank as the fifth leading causeof death worldwide behind heart disease,cancer, stroke, and COPD.
The NSC provides statistics looking atthe total number of deaths and the rela-tive death rate for various unintentionaldeaths per 100,000 population. Theleading five causes of fatal unintentionalinjuries (motor vehicle accidents, falls,poisoning, drowning and choking) havebeen the same from 1970 through 1998(1998 is the last year this data was avail-able). Together, these five categories of
injury accounted for nearly 80% of allaccidental deaths (77,951 of the 97,835)in 1998. Figure 3–6 identifies the lead-ing causes of accidental death, numberof deaths, and death rate for 1998. Inaddition, we have included data onother causes of death (heart disease, can-cer, and stroke) as well as a comparisonto HEMS.
As you can see the 22-year averagefatality rate for HEMS is very high.When you consider the average annualdeath rate over the 22 years reviewed,HEMS is surpassed only by heart diseaseand cancer when the data is normalizedper 100,000 persons.
What are the Odds...
The NSC reports that motor-vehicleaccidents cause more accidental deathsin the United States than any otherunintentional injury. Looking at the
42 AIR MEDICAL PHYSICIAN HANDBOOK
Figure 3–6: Leading Causes of Death, 1998Adapted in part from: National Safety Council, Injury Facts 2001 Edition
odds, an annual risk of death of 1 in amillion is equivalent to the risk of dyingin a car accident if you travel in your carone mile a week for a full year. Driving(or riding) 10 miles per week in an autoincreases your annual risk of dying in anMVA to 1 in 100,000. If you increaseyour weekly driving to 100 miles, yourannual risk of death is now 1 in 10,000,while at 1,000 miles per week our oddsof dying in a car accident is 1 in 1,000.
The NSC often fields questions about“odds.” For example, “What are theodds of dying in a plane crash or beingkilled by lightning?” In response toquestions like this, the NSC has deter-mined the “Odds of Death Due toInjury” in the United States for variousactivities. For 1998 (the most recentyear of these statistics) the one-year oddswere calculated by dividing the 1998population by the number of recordeddeaths from specific injuries that year.To determine the lifetime odds, the NSCtook the one-year odds and divided themby the life expectancy of a person bornin 1998, which is 76.7 years. The NSCalso published death and injury rates per100,000 for various occupations, activi-ties, and modes of travel.
For 1998, the NSC determined thatthe odds of dying from an injury (inten-tional or unintentional) during that yearwere 1 in 1,796. They also determinedthat the lifetime odds of dying from aninjury for a person born in 1998 were 1in 23. Looking only at accidental (unin-tentional) deaths in 1998, the oddsdecrease to 1 in 2,762 in that year and to1 in 36 lifetime. Figure 3–7 shows the1998 odds of death in the United Statesdue to an unintentional injury for vari-ous causes. Included in the table are theaverage one-year odds for a HEMScrewmember to die in a fatal crash. Asyou can see, the odds of suffering a fatali-ty in HEMS exceed that of motor-vehi-cle accidents and all other accidentaldeaths.
As previously discussed, one of theproblems with data that is normalized inthis fashion is that it does not take intoaccount the true amount of exposure dur-ing the year. It assumes all exposures areequal. A more accurate approach wouldbe to determine the exposure (time, etc.)that might result in a specified risk thatcan be compared to other activities.
Cause of Death(in rank order)
Deaths,1998
Death RatePer 100,000
All Causes, all ages 2,337,256 864.9Heart Disease 724,859 268.2Cancer 541,532 200.4
HEMS (range over a 22 year period: 0–699 per 100,000) 196
Stroke 158,448 58.6Chronic obstructive pulmonary diseases (COPD) 112,584 41.7AllAccidental Deaths 97,835 36.2
(Following is a select listing of accidental deaths)Transport Accidents 45,774 16.9
Motor-vehicle 43,501 16.1Air and space transport 692 0.3Water transport 692 0.3Railway 515 0.2Misc. 374 0.1
Falls 16,274 6.0Poisoning by solids, liquids, gases, vapors 10,255 3.8Drugs, medications, and biologicals 9,838 3.6Drowning 3,964 1.5Choking, Inhalation, ingestion of food or other object 3,515 1.3Fire and flames 3,255 1.2Complications/misadventures of surgery/medical care 3,228 1.2Natural and environmental factors 1,521 0.6Firearm, missile 866 0.3
Type of Accident or Manner of InjuryDeaths,
1998One-year
oddsLife-time
OddsHEMS Accidents (22-year average) 1,158Total deaths due to injuries 150,445 1,796 23All Accidental Deaths 97,835 2,762 36Transport Accidents 45,774 5,904 77
Motor-vehicle 43,501 6,212 81Railway 515 524,753 6,842Other road vehicle 235 1,149,991 14,993Water transport 692 390,532 5,092Air and space transport 692 390,532 5,092
Poisoning by solids and liquids 10,255 26,353 344Poisoning by gases and vapors 546 494,960 6,453Complications, misadventures o f surgical, medical care 3,228 83,720 1,092Falls 16,274 16,606 217Fire and flames 3,255 83,025 1,082Natural and environmental factors 1,521 177,678 2,317
Excessive heat 375 720,661 9,396Excessive cold 420 643,448 8,389Lightning 63 4,289,651 55,928Cataclysmic storms, and floods 204 1,324,745 17,272
Drowning, submersion 3,964 68,176 889Inhalation and ingestion of food 1,147 235,613 3,072Inhalation and ingestion of other object 2,368 114,125 1,488Mechanical suffocation 1,070 252,568 3,293Struck by falling object 723 373,787 4,873Machinery 1,018 265,470 3,461Adverse effects of drugs in therapeutic use 276 979,159 12,766Hanging, strangulation, and suffocation 5,726 47,197 615Firearms 17,424 15,510 202
Figure 3–7: Odds of Death Due to Unintentional Injury, Selected Causes, 1998(”1 in _____”) Adapted in part from: National Safety Council, http://nsc.org/lrs/statinfo/odds.htm
43Special Supplement
We have determined that in HEMSthere is an average one-year risk of deathof 1 in 1,158. The Book of Risks has iden-tified several activities that produce a 1-in-1,000 risk of death. In the 22-yearHEMS study period an estimated3,002,176 total hours have been flown.Adding together the estimated number ofcrewmembers each year yields a total of105,922. This corresponds to an averageexposure of 28.3 hours of flight time pro-ducing the estimated odds of 1-in-1,158.Adjusting the ratio to 1-in-1,000, we getan average exposure of 32.9 hours.
HEMS: The Risk to the Patient
There is some level of risk related toall aspects of healthcare—every proce-dure performed, every medication dis-pensed, and every patient transported.The possibility of adverse events or med-ical errors represents a significant risk toeach and every patient.
In Human Error, James Reason definesan error as “the failure of a plannedaction to be completed as intended (i.e.,error of execution) or the use of a wrongplan to achieve an aim (i.e., error ofplanning).” An accident represents anadverse outcome after an error.
The results of the “1991 HarvardMedical Practice Study” concluded that3.7% of New York hospitalizations result-ed in adverse events, of which 13.6% ledto death. A subsequent study publishedin 1999 found that in Utah andColorado, 2.9% of the hospital admis-sions experienced an adverse event, with6.9% of these resulting in a fatality.
According to the Institute of Medicinepublication, with over 33.6 millionadmissions to U.S. hospitals in 1997, theresults of these two studies would suggestthat between 44,000 and 98,000Americans die in hospitals each year as aresult of medical errors.
To put these figures in perspective, thiswould be roughly equivalent to the crashof a jumbo jet carrying 500 passengersevery 2–4 days. In addition, normalizingthe data yields a death rate between 131and 292 per 100,000 patients due tomedical errors.
It is noted that the total number ofdeaths and corresponding death rate issignificantly different than the statisticspresented for “complication ofsurgery/medical care” that are listed inFigure 3–6. The National Safety Councildeath rate is based upon the reportednumber of deaths in this category com-pared to the entire U.S. population. Theresearchers of the two cited studies basedtheir estimates upon a comprehensivereview of a sampling of medical recordsfor adverse events. The estimated num-ber of deaths was then calculated for thenumber of annual patients (1997) ratherthan the entire population. Accordingto the U.S. Department of Commerce,the population in the United States inJanuary 1997 was 266,490,000 people.Normalizing the estimated number ofdeaths from the New York andUtah/Colorado studies for the entire U.S.population yields a death rate per100,000 between 16.5 and 36.8 due tomedical errors. This is still significantlyhigher than the NSC rate of 1.2.
While air medical transport is not amedical treatment and aviation accidentswould not be considered a medical error,some could argue that these accidents
represent an adverse event in the health-care environment. In our 22-year study,we estimate that a total of 2,745,207patients have been flown by HEMS.Over this same time period, 21 patientshave lost their lives in HEMS accidents.This corresponds to a death rate of 0.76per 100,000 patients flown. This takesinto account only fatal injuries as a resultof helicopter accidents and does notaddress any medical errors or otheradverse events that could take place dur-ing transport. Based upon these figures,it would appear that there is a far greaterrisk to the patient of dying from anadverse event while hospitalized thanfrom an accident aboard a medical helicopter.
Occupational Risks: Deaths and Injuries in the Workplace
In 2000, there were 5,200 workplaceaccidental fatalities, while an additional3.9 million American workers suffereddisabling injuries on the job. The NSCreports an average death rate across allindustries at 3.8 per 100,000 workers,with mining and agriculture having thehighest rates. Figure 3–9 shows that ifHEMS “workers” were compared to thepublished NSC data, the average annualHEMS death rate is approximately ninetimes greater than the riskiest industriestracked by the NSC. However, it must bepointed out that this comparison is great-ly distorted when you consider the smallHEMS “population.” Even in 2001, withour largest estimated number ofcrewmembers, 2 fatalities resulted in adeath rate of 23 per 100,000.
The picture changes dramatically ifyou look at the rate of injuries, not just
Figure 3–8: Activities Producing a 1-in-1,000 Risk of Death Adapted in part from: Laudan, The Book of Risks,1994.
Activity Time/Effort Involved
Rock climbing 25 hours
HEMS transport 32.9 hours Skydiving 50 hours
Driving a motorcycle 55 hours
Skiing 340 hours
Flying on a 1,200 hours scheduled airline
Driving a car 52,000 miles
Industry Workers DeathsDeathRate1
DisablingInjuries
InjuryRate
All industries 136,402,000 5,200 3.8 3,900,000 2.86%. Avg) 196 0.16%
Agriculture 3,380,000 780 22.5 130,000 3.85%Mining, quarrying 520,000 110 21.2 20,000 3.85%Construction 8,949,000 1,220 13.6 470,000 5.25%Manufacturing 19,868,000 660 3.3 630,000 3.17%Transportation and public utilities 8,084,000 930 11.5 380,000 4.70%Trade 27,723,000 420 1.5 750,000 2.71%Services 47,611,000 630 1.3 940,000 1.97%Government 20,267,000 450 2.2 580,000 2.86%
HEMS (22 yr
Figure 3–9: U.S. Unintentional Work-Related Injuries and Deaths, 20001Per 100,000 workers. Adapted in part from: National Safety Council, Injury Facts 2001 Edition
44 AIR MEDICAL PHYSICIAN HANDBOOK
fatalities. The overall injury rate for allindustries is 2.86%, ranging from 1.97%to 5.25% of the workers. Over the 22-year period, the injury rate for HEMSranges from 0.0% to 0.61%, with an aver-age of 0.16%. This injury rate takes intoaccount only injuries that were sufferedin helicopter accidents. It does not takeinto account any other etiology of dis-abling injury (e.g., back injury, falls) thatcould afflict an air medical crewmemberwhile on duty.
A Pittsburgh study by Doyle et al.,however looked at occupational injuriesin air medical transport. Presented at the2002 Critical Care Transport medicineConference (CCTMC), this 3-year studyfound that the risk per flight of a crew-member sustaining a reportable injury isvery low. A total of 16,062 flights result-ed in only 86 injuries. However, Doylealso concluded that of the 140 flight per-sonnel, 59% had sustained a reportableinjury.
Although industries differ, the types ofinjuries that occur are common. Figure3–10 shows the various mechanisms ofinjury and the percent seen in all indus-tries. As this chart clearly shows, thereare inherent risks to all occupations thatcould result in injuries or death. In manyways the risks to health care providersmay be similar to other professions, butthere may also be greater risk in certainareas. For HEMS personnel, flying isindeed one of those risks. Health careworkers, including HEMS personnel, maybe exposed to other risks as well.
Violence in the workplace is an all toocommon problem, with homicide beingthe leading cause of occupational deathamong all workers in the United States.It is estimated that 1,000 deaths in theworkplace are due to assault each year.Unfortunately, health care workers arenot immune to work-related attacks.
According to a 1998 OSHAPublication, more assaults occur in thehealth care and social services industriesthan in any other. They cited Bureau ofLabor Statistics (BLS) data that showedhealth care and social service workershaving the highest incidence of injuriesdue to assault. According to one study byGoodman et al., in 1994, between 1980and 1990, 106 occupational violence-related deaths occurred among healthcare workers, including 27 pharmacists,
26 physicians, 18 registered nurses, 17nurses’ aides, and 18 health care workersin other occupational categories. A sepa-rate study using the National TraumaticOccupational Fatality database reportedthat there were 69 nurses killed at workbetween 1983 and 1989.
In a 1998 California study, they foundthat assault, hostage taking, rapes, rob-bery, and violent actions resulting indeath were reported in emergency rooms,mental health hospitals and clinics, andsocial service offices. A 1983 study byConn and Lion found that assaults bypatients in the hospital setting occurredin psychiatric units (41%), emergencyrooms (18%), medical units (13%), surgi-cal units (8%), and even pediatric units(7%). A 1988 study by Lavoie et al.,investigated 127 large, university-basedhospital emergency departments andreported that 43% had at least one physi-cal attack on a medical staff member permonth. Seven percent of the reportedacts of violence over a 5-year periodresulted in death.
Violence in the emergency department(ED) has seen an increase in recent years.ED personnel face a significant risk ofinjury from assaults by patients and areoften abused by relatives of patients orother persons associated with the patient.A study in the Journal of EmergencyNursing by Erickson and Williams-Evansfound that 82% of nurses surveyed report-ed they had been assaulted during theircareers—many going unreported. Stultz,in 1993 reported that nurses were themost frequent targets of assault and the
greatest percentage (25%) of assaultsoccurred in the ED. Of 51 homicidesinvolving health care workers, 23% werein the ED.
Unfortunately, air medical personnelare not immune to violent assaults. InJanuary 2002, two flight crewmembersheard a motor vehicle collision on thestreet adjacent to their base. Uponapproaching the vehicle to assist the driv-er, the two flight team members wereshot—one fatally.
Health care workers are exposed toother dangers as well, with the risk ofneedlestick injury and the transmission ofbloodborne pathogens is an industry-wideconcern. These injuries can result in seri-ous and potentially fatal infections fromthe hepatitis B virus (HBV), hepatitis Cvirus (HCV), or human immunodeficien-cy virus (HIV).
In the United States, there are morethan 8 million health care workers inhospitals and other health care settings.Estimates suggest that 600,000 to 800,000needlestick injuries occur annually, withabout half going unreported. A 1999study by the National Institute forOccupational Safety and Health esti-mates that in the average hospital, work-ers will incur approximately 30 needle-stick injuries per 100 beds per year.
In the United States, between 1985and 1999, there were 55 “documented”cases and 136 “possible” cases of occupa-tional HIV transmission to health careworkers. Combined data from more than20 worldwide studies of health care work-ers exposed to HIV-infected blood found
Rail vehicle1%
Self-inflicted4%
Fire, explosions3% Nonhighway accident
8%Aircraft Accident
4%
Highway Accident23%
Assaults11%
Struck byObject
9%Caught in
equipment7%
Exposure,electrical,caustic,etc.
8% Pedestrian6%
Watervehicle
2%
Falls14%
Figure 3–10: Fatal Occupational Injuries by Event, 2001Adapted from: Bureau of Labor Statistics, http://www.bls.gov/news.release/cfoi.t03.htm
45Special Supplement
an average transmission rate of 0.3% (21infections in 6,498 exposures).
HBV infections occur much more com-monly in health care workers. In 1995alone, an estimated 800 health care work-ers became infected with this virus. TheHBV transmission rate after a singleneedlestick exposure ranges from 6 to 30%if the health care worker is not immune toHBV. However, treated with hepatitis Bimmune globulin and initiating the hepa-titis B vaccine is more than 90% effectivein preventing HBV infection. About one-third to one-half of persons with acuteHBV infection develop symptoms of hep-atitis. Most acute infections resolve, but5% to 10% of patients develop chronicHBV infection that carries an estimated20% risk of dying from cirrhosis and 6%risk of dying from liver cancer.
Hepatitis C virus infection, the mostcommon chronic bloodborne infection inthe U.S., is prevalent in 1 to 2% of thegeneral population. Of the acute HCVinfections that have occurred annually, 2%to 4% have been in health care workersexposed to blood in the workplace.Unlike HBV, chronic infection develops in75 to 85% of patients, with active liverdisease developing in 70%. Of thepatients with active liver disease, 10% to20% develop cirrhosis, and 1% to 5%develop liver cancer. On the average, 1.8%of the health care workers with percuta-neous exposure to HCV become infected.
Another potentially dangerous andrisky endeavor for many occupations,especially health care workers, is rotatingshift work. Investigators have shownthat disruptions in circadian phases dueto rotating shift work are associated withdecreased performance, lapses of atten-tion, and increased reaction time. In1992, Gold et al., found that nurses onrotating schedules reported more “acci-dents” (including on-the-job errors, on-the-job personal injuries, and auto acci-dents due to sleepiness) than did nurseson other schedules.
Sleepiness has been blamed for approx-imately 200,000 to 400,000 motor-vehi-cle collisions per year in the UnitedStates. Considering that the NationalHighway Traffic Safety Administration(NHTSA) recorded 6,279,043 motor-vehicle accidents in 1999, sleepinesswould account for 3.2 to 6.4% of theaccidents that year. In comparison,
NHTSA estimates that alcohol wasinvolved in 7% of all crashes in 1999.
The practice of driving home after anight shift appears to be a significantoccupational risk for health care workers.Steele (1996) reported the results of asurvey of emergency medicine residentsthat found nearly 75% of the auto acci-dents and 80% of the near-crashesoccurred following a night shift. Thatsame year, Novak reported that approxi-mately 95% of night nurses working12–hour shifts reported having had anautomobile accident or near-miss acci-dent while driving home from nightwork.
Travel-Related Risks
How we choose to travel to work, dur-ing work, or for recreation predisposeseach of us to additional risks. Travel-related deaths account for nearly half ofall accidental deaths, with passenger carsand taxis having a significantly higherdeath rate than all other modes of travel(Figure 3-11).
There is a motor-vehicle death every13 minutes. However, mile for mile, it isfar riskier to walk or jog—if it involvescrossing an intersection—than to use anyform of motorized vehicle, includingmotorcycles. Pedestrian deaths, estimatedat nearly 6,000 each year, are more thandouble the rate of motorcycle deaths.
Figure 3-11 depicts the death rates per100,000,000 miles of travel for a three-year period for the major modes of trans-portation. From 1997-1999, scheduledairlines were the safest mode of trans-portation. Transit buses and school buseswere next, but had ten times the deathrate when compared to the airlines.Fatalities by automobile were about 22times greater per passenger-mile than bycity bus. In 22 years, HEMS has flownan estimated 3,002,176 hours, coveringan estimated 360,261,121 miles (using anaverage of 120 miles-per-hour).Normalizing our HEMS data for 137crewmember fatalities yields a death rateof 38.0 per 100,000,000 miles traveled.
How do you manage the risk of beinginvolved in an auto accident? Whileautomobile accidents kill a great manypeople, there are ways to manage the riskand reduce your chances of suffering afatal auto injury. In motor-vehicle colli-
sions, large cars are generally safer thansmall ones. You are roughly twice aslikely to die in a serious auto accident ifyou are in a small car rather than a largecar. Yet, for various reasons, many of uschoose to drive small cars. Driving atnight, mile for mile, is almost four timesmore likely to end in a fatal accidentthan driving during the daylight. Yetmost people still drive at night. The lap-shoulder seat belt reduces fatalities infront-end collisions by 42%, but it isreported that 32% of Americans neverwear seat belts. However, when you takeinto account car size and seat belt use,you are less likely to die in a large vehicleaccident wearing no seat belt than in asmall car wearing a seat belt.
Helmets are an important considera-tion when it comes to bicycles andmotorcycles. Helmets reduce the risk offatal injury to motorcyclists by 30%.Helmets can also reduce the risk of seri-ous head injury in bicycle accidents bymore than 70%.
Ground Ambulance Accidents
Emergency vehicle accidents are notunique to air medical transport. A fre-quent discussion among transport profes-sionals and other health care providersinvolves comparing the risk vs. benefit ofground vs. air transport. EMTALAguidelines require that the risks and ben-efits of all inter-facility transfers beexplained to the patient. This shouldinclude an explanation of the anticipatedrisks and benefits of the chosen mode oftravel—air or ground.
Unfortunately, the availability andaccuracy of statistics regarding groundambulance accidents is even worse thanthat of air medical transport. Similar toHEMS, no exposure data exists for
Mode of Transportation Death Rate Buses 0.04 Railroad 0.06 Scheduled Airlines 0.004 Passenger Cars and Taxis 0.87 HEMS 38.0
Figure 3–11: Transportation AccidentDeath Rates / 100,000,000 miles,1997 to 1999Adapted in part from: National Safety Council, Injury Facts 2001 Edition
46 AIR MEDICAL PHYSICIAN HANDBOOK
ground ambulances. There is no infor-mation nationwide as to the number ofmiles traveled, number of ambulances orthe number of patients transported. Thelack of this information makes it impossi-ble to make a meaningful comparisonbetween air and ground transport.However, some information is availableregarding ambulance accidents.
The National Safety Council publishesdata on crashes involving emergencyvehicles in the United States. The NSCannual tabulations are based upon theFatality Analysis Reporting System(FARS) and General Estimates System(GES) which are made available eachyear from the National Highway andSafety Administration. However, expertsin the field of ground EMS transportquestion the accuracy of these statistics.There is no central tracking system toidentify and capture ground ambulanceaccidents. Some accidents will be report-ed as “ground ambulance” accidents. Insome systems, however, where EMS is acomponent of the Fire Department, theaccident may be logged as havinginvolved a “fire” vehicle. Some accidentreports may list the vehicle as a “lighttruck.” Some states track ground ambu-lances aggressively. In New York Statealone, there are an average of 350 ambu-lance accidents each year, injuring anaverage of 2 people per day. However,many states do not have any system inplace to track accidents accurately.
In April 2002, Robert Davis reportedin USA Today that there might be anestimated 15,000 ambulance crashes ayear. This estimate is approximately threetimes the number of ground ambulanceaccidents that we find in the NSC annu-al publications. Other sources suggestfrom sample data that it is very reason-
able to estimate one fatal ambulancecrash each week, as many as ten seriousinjuries every day and as many as 10,000total injuries every year. Figure 3–12presents three years of NSC data forground ambulances, as well as otheremergency vehicles (police and fire).
USA Today reported that with “rela-tively few fatalities each year ( . . .out ofmillions of ambulance calls), federal offi-cials say there is no pattern that triggersany alarms,” and “. . . .there is not a hugesafety problem.”
There are obviously those who woulddisagree.
USA Today referenced a 1993 Houstonstudy that found that ambulances were13 times more likely to be involved in anaccident than other vehicles based uponthe number of accidents per miles driven.That study also determined that ambu-lances were five times more likely to beinvolved in an accident that resulted ininjuries.
Our analysis of the NSC data for1997–1999, found that ground ambu-lance had the highest percentage of fatalaccidents when comparing each categoryof emergency vehicle. Over the 3-yearperiod, 0.47% of all ground ambulanceaccidents resulted in a fatal injury. Firevehicles were a little better at 0.39% andpolice were 0.32%. In addition, if youconsider the number of total fatal injuriesper 1,000 emergency vehicle accidents,the ambulance remains the most lethal.Over the 3-year period, the ambulanceaveraged 5.2 fatal injuries per 1,000 acci-dents, compared to 4.8 deaths for firevehicles and 3.49 for police. The per-centage of accidents that resulted ininjuries is also the highest for ambulances(36%) compared to the fire (18%) andpolice (32%) vehicles.
From the 1997–1999 NSC data, infatal, multi-vehicle ambulance accidents,only 25% of the deaths were occupants ofthe ambulance. Less than 3% of thefatalities were the ambulance “driver”according to the NSC data. Over 22% ofthose killed were “emergency vehicle pas-sengers”. The NSC tables did not differ-entiate between the patients, medical per-sonnel, patient’s family or others whowere killed in the ambulance, making itimpossible to know the total number ofground ambulance personnel killed inaccidents. Of the remaining fatalities,two-thirds of those killed in these acci-dents were occupants of another vehicleand 8% of the fatalities were nonmotorists.
Recreational Risks
When it comes to sports and recre-ational activities, the National SafetyCouncil and other publications make lit-tle if any effort to compare or rank therelative risk of injury or death. There aretoo many variables and unknowns toconsider, including frequency and dura-tion of exposure, number of participantsand accurate numbers of injuries. Theonly injuries generally known are thosethat required emergency treatment.Most of the statistics come from emer-gency department logs.
The rising popularity of extreme sportswas documented in the 1999 Time maga-zine article.
More Americans than ever are injuringthemselves while pushing their personalrecreational limits. BASE jumping(jumping from Buildings, Antennas,Span/bridges and Earth/cliffs) has one ofthe sporting world’s highest fatality rates.In its 18-year history, 46 participants
have been killed.Currently, there are morethan a thousand jumpersin the U.S. and more get-ting into it every day. Thesport has never been morepopular.
While there has been asteady decline throughoutthe ’90s in the participa-tion in sports like baseball,touch football, and aero-bics, there has been rapidgrowth in adventure
Ambulance Fire Truck/Car Police Car97 98 99 Avg. 97 98 99 Avg. 97 98 99 Avg.
EV in fatal MV A 28 25 15 0.47% 20 17 17 0.39% 97 83 65 0.32% EV in injury MVA 1,465 2,306 1,473 36.4% 997 781 677 17.6% 8,305 8,210 8,645 32.4%Total number of MVAs 4,745 4,615 5,050 3,928 3,188 6,839 23,725 24,417 29,419
EV driver killed 0 2 0 2.7% 5 4 5 21.2% 19 17 17 19.6% EV passengers killed 8 7 2 22.7% 3 3 3 34.8% 6 3 4 24.4% OV occupants killed 21 18 11 66.7% 13 12 9 51.5% 54 66 40 59.0% Nonmotorist killed 2 2 2 8.0% 4 3 2 13.6% 24 10 11 16.6% Total Deaths 31 29 15 25 22 19 103 96 72
Total injuries 3,351 3,274 2,659 1,467 1,035 1,130 12,689 12,339 15,230
Figure 3–12: Emergency Vehicles Involved in Motor Vehicle Accidents, 1997–1999(EV=Emergency vehicle; OV=Other Vehicle) Adapted from: Injury Facts: 1999 Edition, 2000 Edition, and 2001 Edition. National Safety Council.
47Special Supplement
sports. Snowboarding has grown 113% infive years and now boasts nearly 5.5 mil-lion participants. Mountain biking,skateboarding, scuba diving, and othermore hazardous activities have seen moreand more people participate. In 1997 theU.S. Consumer Products SafetyCommission reported that 48,000 peoplewere treated in hospital emergency roomswith skateboarding-related injuries—33%more than the previous year. Visits tothe E.R. were also up for snowboardinginjuries (up 31%) and mountain climbing(up 20%). The US ParachuteAssociation reports that 10% of skydiverssuffer injuries requiring medical atten-tion. Annually, there are approximately30 deaths, or 1.2 deaths per 100,000jumps. In contrast, there have been only20 serious injuries reported from an esti-mated 1,000,000 bungee jumps between1988 to 1994. During the same timethere were 7 deaths, for a death rate of0.7 per 100,000 jumps.
Some participation and injury statisticswere available from the National SafetyCouncil and other sources. While theaccuracy of injury rates may not be precisefor the reasons previously mentioned, somecalculations are presented for comparison.
Any meaningful comparison betweenrecreational injury rates and work-relatedinjuries may be difficult. However, if youlook at the “all industry” occupationalinjury rate of 2.86%, it would suggest thatyou are more likely to get injured on thejob than participating in any of the abovesports except hockey. On the other hand,looking at the injury rate for HEMS per-sonnel of 0.16%, it would seem that youare less likely to get injured in a helicopteraccident than participating in these sports.
SECTION 4: SAFETYAND RISK MANAGE-MENT IN HEMS
No report on HEMS safety would becomplete without trying to identify waysto enhance program (and industry) safe-ty and reduce risk. While every air med-ical program has similar exposure charac-teristics, no two programs are exactlyalike. This section will look at four basicrequirements of a safety program, intro-duce some key aspects of risk manage-ment, and then highlight the benefits ofa multidisciplinary approach to improvesafety.
PRINCIPLES OF A SAFETYAND RISK MANAGEMENTPROGRAM
A safety and risk management pro-gram for an air medical program mustencompass two aspects. Most importantis doing everything possible to preventan accident from occurring. If this failsand an accident does happen, everythingmust be in place to mitigate the impactof the accident.
Being safe does not eliminate risk—itreduces it. There are four basic princi-ples that should guide the coordination,implementation, and evolution of a safe-ty and risk management program for airmedical operations. These principles areattitude, participation, education, andjudgment.
Attitude
Safety does not just happen, it is not aspecific event or a “thing”—it is an atti-tude. This perhaps is the most impor-tant component of the safety equationand may override all other aspects andvariables. Everyone must have the rightattitude about safety in order to partici-pate and survive in an air medical trans-port program. This, along with a com-mitment to safety, is essential and mustbe exhibited by every crewmember andevery manager. The attitudes shouldreflect the mission that safety must bethe program’s number one priority.
A number of obstacles may preventthe development of a sound safety pro-gram. An FAA Advisory Circular (AC)No. 60–22 defines attitude as “a personal
motivational predisposition to respondto persons, situations, or events in agiven manner that can, nevertheless, bechanged or modified through training.”Negative attitudes, however, are particu-larly difficult to overcome. Some peoplemay think that safety is not their respon-sibility and their actions are not likely toimpact the safety of flight or result in anaccident. In some cases, identifyingpotential problems may seem to be toothreatening to discuss and may simply beavoided. In other circumstances, denialoccurs as team members insist that thereis no safety problem.
Complacency is another serious prob-lem and in many ways may represent anegative attitude. Merriam-Webster’sCollegiate Dictionary defines compla-cency as “self-satisfaction accompaniedby unawareness of actual dangers or defi-ciencies.” Having learned things beforeor having done certain activities in thepast may result in overconfidence andeventually to errors in performance.Never having an accident or incidentdoes not assure continued safety if itresults in complacent attitudes. It resultsin smart people sometimes doing dumbthings. In 1901, Wilbur Wright wrote,“Carelessness and overconfidence areusually more dangerous than deliberatelyaccepted risks.” One hundred yearslater, this statement is still appropriate.
In 1997, the pilot of an AerospatialeAS-365N Dauphin transported two pas-sengers to a corporate ramp atIndianapolis International Airport.Upon shutting down the engines andapplying the rotor brake, one passengerexited and walked forward of the heli-copter and turned into the path of therotor system. The passenger was struckin the left temple by a main-rotor bladeand killed. In January 2001, a hospitalsecurity guard walked into the tail rotorof a Bell 206L as the helicopter waspreparing to depart the base hospitalhelipad. The security guard died of hisinjuries. It may be very difficult todetermine if complacency, carelessness,or overconfidence were contributing fac-tors in these accidents.
Participation
A safety program must be planned,instituted and practiced every day and
Recreational Activity Injury Rate Skydiving 10% Hockey 4.08% Rugby 2.4% Basketball 1.94% Football 1.66% Snowmobiling 1.34% Bicycle Riding 1.21% Baseball/Softball 1.07% Skateboarding 0.76% Roller/In-line Skating 0.41%
Figure 3–13 Adapted: National Safety Council,Injury Facts 1999 Edition
48 AIR MEDICAL PHYSICIAN HANDBOOK
on every flight. It is not enough toassume that hospital administration, pro-gram management, the pilots, or the avi-ation operator will be completely respon-sible for safety. The safety program mustbe multidisciplinary and responsive inorder to be successful. Safety is dynamic.Things change. Crewmembers (pilots,mechanics, medical and communica-tions) come and go, aircraft may change(primary or backup), weather conditionsvary, and a program may fly to hundredsof different locations. Every flight is dif-ferent. It takes teamwork, where individ-uals interact effectively and efficientlywith fellow crewmembers to maintain asafe aviation operation.
There is no such thing as a “free ride.”Whatever the role of a team member inthe program, each individual mustacknowledge the critical fact that safetyis their prime responsibility. It requiresactive participation, rather than passiveobservation. While medical personnelare not expected to be experts in avia-tion, each must be proficient with theirsafety-related responsibilities. At times,when delivery of medical care and safetymay seem to come into conflict, safetymust always take priority.
Each crewmember must recognizehis/her role to help identify, address, andhelp resolve (as appropriate) potentialsafety concerns. Every situation repre-sents an opportunity to learn and toimprove. A program should encourageevery member to identify opportunitiesfor improvement, either to enhance pro-gram safety or efficiency.
Education
Education is a key ingredient in identi-fying, understanding, and actively man-aging risk. For a pilot and air medicalcrew, being knowledgeable of the ele-ments that may increase or decrease therelated risk should lead to taking appro-priate steps to minimize unnecessaryexposure. The result would be to greatlyreduce the chance of being involved inan accident or incident.
There are many aspects of educationthat may be considered part of a safetyprogram geared to actively manage risk.Education and training obviously beginwith the pilots and mechanics and con-tinue with the medical crew and commu-
nication specialists. It must also includethe security and public safety personnelwho set up and/or secure landing zones orhelipads.
Another important aspect of educationis learning from past mistakes. In sepa-rate documents, both the Aircraft Ownersand Pilots Association (AOPA) AirSafety Foundation and the Flight SafetyFoundation support the belief that pilotscan learn valuable lessons from analysesof past accidents and incidents. Accidentdata often yield clues to safer operations,and can be applied at relatively little costand with no additional regulations. Byanalyzing mishaps, pilots, medical teams,and program administrators can learnabout potential risks and take proactivesteps to control them. The U.S. AirForce Guide to Mishap Investigation states,“the proper use of mishap experience isreducing mishap potential.”
It is important to realize that educationisn’t everything —it is merely the begin-ning. Being able to apply the knowledgeunder routine and emergency conditionsis expected of each and every crewmem-ber. Each crewmember must be able toperform his or her safety-related func-tions proficiently and independently. Itis an unnecessary distraction and risk forpilots and medical crewmembers to worryabout someone else who does not dohis/her job.
Judgment
FAA AC No. 60–22 defines judgmentas “the mental process of recognizing andanalyzing all pertinent information in aparticular situation, a rational evaluationof alternative actions in response to it,and a timely decision on which action totake.” In 1986, Arthur Negretteauthored an article entitled “SpatialOrientation: It Plays No Favorites.” Hereported that on the average, it takes ahelicopter pilot five seconds to recognizea hazard, determine the necessary correc-tive action, and respond.
Some people feel that good judgmentis an inherent characteristic and not onethat can be easily taught. Someone maybe given the tools through education andmonths or even years of experience, butmay still lack the ability to put it alltogether for optimal performance andoutcome.
The AC goes on to define the PoorJudgment (PJ) Chain as “a series of mis-takes that may lead to an accident orincident.” The AC states that there aretwo basic principles that are generallyassociated with the creation of a PJ chain:(1) one bad decision often leads to anoth-er; and (2) as a string of bad decisionsgrow, it reduces the number of subsequentalternatives for continued safe flight.
In the HEMS environment with itslimited resources (generally one pilot andtwo medical crewmembers), it is essentialfor each and every crewmember to be atthe top of their game, using sound judg-ment to recognize individual limitationsand changing conditions. Having alter-nate plans and knowing when to imple-ment them can impact both aviation andmedical safety. For example, in aviation,judgment is key when it comes to thedecision to initiate a flight or when toabort a flight. In medicine, it may beexemplified in your management of thepatient with a difficult airway. Whatdrugs, if any, should be considered? Whatis the backup plan if the intubation is notsuccessful? Having the knowledge andthe skill is essential. Sound judgment andproficiency are more likely to yield a posi-tive outcome. In HEMS this willenhance overall safety.
RISK MANAGEMENT
Risk management is a discipline fordealing with uncertainty; a science oflooking to the future through today’svision. It enables us to make a range ofinformed decisions about our environ-ment, health, safety, and our social andeconomic well being. It is about manag-ing resources wisely, protecting fromharm, and safeguarding assets. In HEMS,risk management should be directedtoward optimal flight safety as well as pro-viding the highest quality of medical care.
Effective risk management acknowl-edges and identifies threats, evaluatesand prioritizes the risks, considers theprobability a risk will materialize, andcontrols loss (preventing loss and reduc-ing the severity should a loss occur).Results must be evaluated and strategiesrevised as appropriate. With this infor-mation, a person or organization is ableto make an informed decision as to howthey will deal with various risks.
49Special Supplement
64%
44%
35%
25%
23%
23%
15%
9%
7%
7%
7%
6%
Scene Operations
Program Complacency
Mission-related Stress
5-minute Response Time
Pilot Fatigue
Other Threats*
Unforecasted/Lack of Weather Reporting
Inadequate Weather Minimums
Helicopter Mechanical Condition
Pressure to Accept Mission
Inadequate Pilot Training
12-Hour Duty Shift
Percentage of Reports
Identifying Risks
In this report, we have already identi-fied many of the issues that could repre-sent threats to the safety of flight. In thereports from the Air Medical AccidentAnalysis and the Helicopter AccidentAnalysis Team specific problems thathave led to accidents are identified.
Additional information is obtainedfrom the AirMed “2000 Aircrew Survey”that identified the situations that HEMSpilots felt posed the greatest threats (i.e.,greatest risk) to flight safety. Scene oper-ations (64%), program complacency(44%), and mission-related stress (35%)were the three most common concerns.Figure 4–1 shows the various risks thatwere identified.
A more comprehensive “EMS LinePilot Survey” was conducted byNEMSPA and HAI in 2000–2001, whichyielded a combined total of 304 respons-es. In general, there were considerablymore variables to choose from in this sur-vey, which often resulted in lower per-centages for each specific response.
The greatest risk to safety industry-wide was found to be management orcrew pressure (11%). Inexperience wasnext (9%), followed by LZ operations/hazards and weather reporting/forecastingat 8% each. At 7% were poor decision-making, night operations, and marginalVFR/inadvertent IFR conditions. Whenasked what risk factors contributed toaccidents in recent years, the most com-mon response was pushing weather mini-mums (14%), followed by complacency(11%), pilot complacency (10%), inex-perience (8%), and lack of IFR training(8%).
There were a number of questions inthe “EMS Line Pilot Survey” that furtherclarified the types of pressure the pilotsidentified, as seen in Figure 4–2.
Dealing with Risks
Techniques to Control Risk
There are six techniques to controlrisk. The first and most common optionis risk avoidance. If it is not worth assum-ing the risk, avoiding the exposure mayeliminate the risk. In air medical trans-port, an alternative would be groundtransport, which has its own inherentrisks. A program could also decide that
Figure 4–1: Greatest Threats to Flight Safety (n=148)“Other threats” included night operation, general aviation traffic and negotiating geography/mountainousterrain and others. Adapted from: Rau, 2000 Aircrew Survey, AirMed, Nov/Dec, 2000
Survey Variable Response Undue pressure from:
Program Management 32%Flight crews 14%Operator 13%Dispatch 13%
Pressure from:Crew/management to speed up response or lift-off times 55%Other pilots (indirectly) to launch/continue in marginal weather 13%Crew/management to launch/continue in marginal weather 8%
Pilots pressure themselves to:Launch/continue in marginal conditions 13%Fly when fatigued or ill 32%Speed up response or lift-off times 48%
Figure 4–2: Pilot-related Pressures (n = 304)Adapted from: EMS Line Pilot Survey, NEMSPA and HAI, 2001
scene accidents or flying at night posetoo great a threat and elect to avoidthese situations.
The second option is risk prevention.In this situation, the HEMS programwould take the necessary action to try toprevent a loss through comprehensivetraining (pilots, medical crews and com-municators), policies and procedures,and so on. A helicopter program maychoose to do scene flights only duringthe day or only land at pre-designatedlanding zones. A program may also con-clude that a twin-engine aircraft is less
likely to have a potentially catastrophicmalfunction (i.e., engine failure of a sin-gle-engine helicopter) or that two pilotswill reduce the likelihood of pilot error.All of these options strive to decrease thepossibility of an aircraft accident, but donot eliminate all possibility of that loss.
The third technique is loss reduction,which is an attempt to reduce the severi-ty of the loss. The decision to wear hel-mets or nomex are forms of loss reduc-tion in HEMS. Survival training andappropriate survival gear may also reducethe losses in an accident, but do nothing
50 AIR MEDICAL PHYSICIAN HANDBOOK
to prevent an accident from occuring.Segregation of risk is the next alterna-
tive. While the technique has similarcharacteristics to loss reduction, it alsohas very distinct features. A programmay choose to have separate helipads fortheir multi-helicopter program or for avisiting aircraft to use. If a helipad acci-dent were to occur, only one helicopterwould be involved. Some families usethis technique to control the risk of trav-el on commercial airlines. Rather thanall flying together on one airplane, thefamily will split up, taking two differentflights, reducing the risk of an accidenttaking an entire family.
The next alternative is risk transfer,when someone else assumes the risk. InHEMS, this could include not having aflight program or having another flightprogram undertake your transports (similarto risk avoidance). It may also be a factorin considering who is responsible for thevarious aspects of the Part 135 aviationoperation; who is responsible for the train-ing of the pilots and mechanics; or who isresponsible for the maintenance of the air-craft and compliance with all FederalAviation Regulations. It could be thehospital or flight program, or this could betransferred to a full-time aviation operator.Risk transfer can also be represented bythe various hull, liability, disability, andlife insurance policies that are in place toaddress various losses. Power-by-hour isyet another form of risk transfer.
Risk retention is the final option, wherea program or individual consciouslyassumes the risk, taking no specific stepsto reduce the risk or the severity of a loss.
HEMS-Specific Risks
Dozens of intervention strategies havebeen recommended in the Air MedicalAccident Analysis and the HAATreport. It is appropriate to highlight spe-cific HEMS risks and several considera-tions to deal with these risks.
Scene Operations. The concern overscene operations, as identified in Figure4–1, is long standing and justified—espe-cially at night. Potential landing zone(LZ) hazards (trees, wires, etc.) representa significant threat to safety. To improvesafety, 9% of the responses to the “EMSLine Pilot Survey” favored night visiongoggles.
Education plays an important role inscene safety. Educating pre-hospital per-sonnel on the selection and preparationof the LZ is key. Educating the medicalcrew and their active participation tolook for hazards on approach and landingis also essential. However, as the statisticsshowed, very few of the accidents thatoccurred on “scene” flights occurred whilelanding at the scene. More accidentsoccurred with the patient on board thanon any other leg of the scene missions.
Complacency. Program complacencyhas already been addressed. It is impor-tant for pilots, crewmembers, communi-cation specialists and administrators torecognize that complacency may be thegreatest danger in HEMS—the silentkiller. The complacent individual gener-ally exhibits a low level of awareness anddoes not recognize the need for action orinvolvement. This leads to mistakes.
Someone once was asked, “What is thedifference between ignorance and com-placency?” he responded, “I don’t knowand I don’t care.”
Pressure and Stress. Mission-relatedstress has also been addressed, in part,with regard to both self-imposed andexternally imposed pressure. In 1988,the NTSB report recognized both ofthese pressures as significant concerns.Thirteen years later, the “EMS Line PilotSurvey” identified management or crewpressure as the greatest risk to safetyindustry-wide. Have we made noprogress over the past decade?
Weather. It is of interest that two fac-tors that scored fairly low in Figure 4–1were pilot training (7%) and unforecast-ed/lack of weather reporting (15%). Inthe AirMed pilots’ survey, 85% of thepilots reported having made a forcedlanding once or twice in their HEMScareer for deteriorating weather condi-tions. A responding 32% reported atleast one forced or precautionary landingduring the past year due to weather. Thiswould seem to imply that the pilot’s abil-ity to identify and actively manage a riskdiminishes the perception of a particularthreat to safety. On the other hand, con-cerns that they may have little directcontrol over (i.e., scene LZs, programcomplacency, etc.) remain significant
threats. Weather, however, remains an all too
common factor in HEMS accidents.Despite thorough planning and adherenceto weather minimums, it is possible forany HEMS pilot to encounter unantici-pated weather and enter inadvertentinstrument meteorlogical conditions(IMC). The first line of defense for anypilot against inadvertent IMC is to takethe necessary steps to avoid the situation.Comprehensive weather planning is per-haps the most important step. Thisshould include access to updated weatherforecasts and a working knowledge ofweather charts and reports, as well asfamiliarity with local weather trends.Forecasts are not guarantees but forecastsfor marginal conditions are usually accu-rate. It is only a short step from marginalconditions to unflyable conditions. Amarginal forecast and wishful thinkingthat “the weather won’t be as bad as theysay” has no place in aviation, especially airmedical transport.
The pilot’s actions immediately afterencountering IMC will determine theoutcome of the flight. In many weatheraccidents the pilot waited just a little toolong to change his mind. In a significantnumber, deciding to divert only a fewminutes earlier would have kept theflight out of danger. Trained and profi-cient pilots, who have a plan of action inthe event of inadvertent IMC, are morelikely to experience a successful outcome.Another important defense against inad-vertent IMC is a willingness to land theaircraft. Such willingness runs counter tothe pressures HEMS pilots sometimes feelto complete their missions.
Although it is impossible to make adirect correlation, more experience inweather-related decision making shouldresult in a gradual reduction in some of theVFR into IMC accidents and incidents.
Night and Spatial Disorientation.Flying at night poses its own unique risksin aviation. This is especially true inHEMS. The FAA manual, AeronauticalDecision Making for Air AmbulanceHelicopter Pilots, notes that “even on theclearest night with VFR (visual flightrules) conditions, a pilot can come closeto IFR (instrument flight rules, i.e., inad-vertent IMC) operations if there is nomoon and/or no ground lights to estab-
51Special Supplement
lish a horizon reference.” In contrast,there could be an abundance of groundlights below and stars above that canseem to merge into a continuous sweep ofpinpoints that can deprive a pilot of anyhorizon reference. In both situations,with the loss of the visual reference (theground and/or horizon), the interpreta-tion of motion and position in relation tothe environment may be lost. An inex-perienced pilot who becomes disoriented,or who does not trust his/her instru-ments, may change direction, altitude,speed, etc., and if unable to compensateis likely to have an accident.
Another risk lurking in the night sky isthe unseen cloud. Clouds disappear easilyin the dark and a pilot can fly into onewithout seeing it coming. In all of thesenighttime situations, instrument trainingand proficiency may help mitigate thepotential risk.
FAA Advisory Circular 60–4A(February, 1983) addresses pilot’s spatialdisorientation. In this AC, it states: “Testsconducted with qualified instrument pilotsindicate that it can take as much as 35 sec-onds to establish full control by instru-ments after the loss of visual referencewith the surface.” While the tests wereperformed on fixed-wing aircraft, theresults may be more dramatic with helicop-ters since they require even more pilotintervention to maintain control.
Another group of fixed-wing pilotswere asked to identify their personalexperience with spatial disorientation.The most common sensory illusionsreported were:
• A sensation that one wing was lowalthough wings were level (60%)
• On leveling after banking, a tenden-cy to bank in opposite direction(45%)
• When in a turn, a feeling as if theywere straight and level (39%)
• Becoming confused in attempting tomix “contact” and instrument cues(34%)
• On recovery from a steep climbingturn, the feeling of turning in theopposite direction (29%)
This AC also points out that while vis-ibility may be above VFR minimum, thenatural horizon and surface referencesmay at times become obscured in low-
visibility conditions, on over-waterflights, and at night—especially insparsely populated areas. The AC con-cludes “You and only you have fullknowledge of your limitations. Knowthese limitations and be guided by them.”
Pilot Training and Experience. Thisleads directly into the issue of pilot train-ing and experience. In the AirMed sur-vey, pilot training scored very low as aperceived risk to flight safety.
In the HAI/NEMSPA survey, however,when asked “what can be done toimprove industry safety,” the most com-mon response was to improve the qualityand frequency of training (17%).
The Frazer articles and Hart lectureshave concluded that in the vast majorityof HEMS accidents, this did not appearto be a major factor. This report is notabout to address specific training require-ments for the aviation (pilots andmechanics) professionals. However, astudy by FlightSafety International (FSI)seems worthy of mention.
FlightSafety offers flight simulatortraining for both fixed-wing and rotor-wing pilots. Their full-motion simulatorscan reproduce various in-flight emergen-cies during various lighting and weatherconditions. FSI conducted a study todetermine the impact that simulatortraining had on the fixed-wing fatal acci-dent rate. FSI estimates that theytrained approximately 20% of the fixed-wing pilot population. Looking at fiveyears of accident data, they identified atotal of 471 accidents, or an average of94 per year. They postulated that if sim-ulator training made no difference, theFlightSafety trained pilots should haveaccounted for approximately 20% (92) ofthe total accidents. Instead, the FSI-trained pilots had only 3% (15) of thetotal accidents. The expected accidentfrequency was reduced by more than80%. FlightSafety concluded that: “Thebenefits of simulator training are obvious....and the safety record proves it.”
While simulator training may indeedbe beneficial, other considerations mayalso be factors. It is also possible that thecompanies that went to the effort andexpense of sending their pilots to simula-tor training are more safety-proactive inother areas as well.
Of interest, the Air Medical Accident
Analysis rated full-motion simulators ashighly effective, but low feasibility.Perhaps more interesting is the reponseto the “EMS Line Pilot Survey” whenasked “what can be done to improveindustry safety?” Of the 304 responses,only one selected “simulation.”
AIR MEDICAL RESOURCEMANAGEMENT
Air Medical Resource Management(AMRM) is not a unique concept. Basedupon more than 20 years of business andaviation models, AMRM is specificallydesigned for our industry. The goal is toprovide the methodology to make opti-mum use of the capabilities of the indi-viduals and aircraft systems to achievethe safest and most efficient completionof a flight.
In 1979, NASA suggested that busi-ness managerial concepts could beapplied in the cockpit to reduce the highnumber of “human-factors” accidentsoccurring with the airlines. Within 10years, Cockpit Resource Management—later expanded conceptually to CrewResource Management (CRM)—wasincluded in training worldwide at mostmajor airlines. The U.S. Air Force hadbegun full-scale CRM training of allcrews of multiperson aircraft. Today,there is sufficient evidence that CRMtraining and practice have improved avi-ation safety.
CRM is the effective management of allresources available to ensure that allgroup members are operating from acommon frame of reference and toward acommon goal of aviation safety. CRMprovides a framework for accomplishing agiven mission. In air medical transport,training programs have been and arebeing developed that teach CRM skillsand principles not only to pilots, but alsoto medical personnel, communicationspecialists, maintenance personnel andmanagement. In fact, at the 2000 AirMedical Safety Summit, the number onepriority identified by the industry leaders(aviation, medical and management) wasCRM and related training. At an all-day seminar prior to the Air MedicalTransport Conference in September2001, a new Air Medical ResourceManagement Train-the-Trainer course waspresented for the first time.
52 AIR MEDICAL PHYSICIAN HANDBOOK
The FAA’s Advisory Circular identifiescomponents of CRM to include team-work, communication skills, decisionmaking, workload management, situa-tional awareness, preparation and plan-ning, cockpit distractions, and stressmanagement.
Teamwork is key in AMRM and mustbe maximized to facilitate the transfer ofinformation. In HEMS, teamwork mustinclude everyone who could impact thesafety of flight, including the pilots, med-ical personnel, communication special-ists, maintenance personnel, air trafficcontrollers (ATC), and management.Much like a sports team, each personmust know his/her role, be an active par-ticipant and be able to execute his/herassignment when called upon. In sports,the goal is winning. In aviation, “win-ning” is a safe and efficient aviation oper-ation resulting in the safe return of theaircraft and crew at the conclusion ofeach and every flight.
Effective communication skills areessential in developing teamwork. Pilotscan increase the probability of a safeflight by overcoming barriers in commu-nication and learning to effectively seekand evaluate information. Communica-tion problems are often cited as a causalfactor in aircraft accidents and incidents.At a recent air medical conference, aleading safety expert in HEMS stated,“open and effective communicationmight have prevented perhaps as many as80% of EMS accidents.”
Aeronautical Decision Making (ADM)refers to a systematic mental process usedby pilots to consistently determine thebest course of action in response to agiven set of circumstances. ADMincludes risk assessment and stress man-agement. It also illustrates how personalattitudes can influence decision makingand how those attitudes can be modifiedto enhance safety in the cockpit. Gooddecision making skills are not necessarilyinherent and must be learned. A pilotmust seek and evaluate all relevant infor-mation, using all available resources,before making an important decision.These resources may include people(medical crewmembers, ATC, communi-cation specialists, other pilots) aircraftinstruments, documentation (flight man-uals, checklists, etc.), and sensations(vibrations, sights, smells, position).
Studies by a number of researchers haveshown that there is a strong correlationbetween errors in decision making andthe severity of accidents. While manyskill-related problems may result in minorinjuries and damage, faulty decision mak-ing processes often result in accidentswith serious injuries and fatalities.
Situational Awareness is the accurateperception and understanding of all thefactors and conditions going on aroundyou. In aviation, this deals with the fourfundamental risk elements that affectsafety before, during, and after the flight—the pilot, the aircraft, the environment,and the type of operation that compriseany given aviation situation. Thisrequires a pilot’s full attention. Here too,cockpit distractions must be kept to aminimum. The Aviation SafetyReporting System (ASRS) identified pilotdistraction as one of the most frequentcauses cited in the reported incidents.
Managing the workload is critical tothe single-pilot operation. Tasks must becarefully prioritized and the pilot mustavoid being distracted from his/her pri-mary duty of flying the aircraft. The sin-gle pilot can often benefit by utilizinghis/her resources and sharing tasks. Thiscould include having the medical crewhandle some of the communications withdispatch or EMS ground personnel, look-ing for hazards while landing or takingoff, requesting assistance or informationfrom ATC or from dispatch, and prudent-ly using automation such as an autopilot.
In the single-pilot HEMS operation,preflight planning and preparation is ofspecial importance. Generally, no oneelse is available to confirm radio frequen-cies and make radio calls, fix positionsand call out checklists. In a high-work-load or stressful situation, the pilot mustbe able to call upon his/her training anditems that may have been committed tomemory, such as frequencies and emer-gency procedures, that otherwise mightbe difficult to confirm in an emergency.In addition, as appropriate, the pilotshould utilize the other resources thatmay be available.
The effects of stress are often difficultto recognize and the inability to recog-nize this may be hazardous in aviation.Failure to manage stress often leads toeroded judgment, errors in decision mak-ing, decreased work performance, inat-
tention, degraded communication skills,preoccupation, and complacency. Apilot suffering from stress may forget orskip procedural steps, accept lower per-formance standards, and exhibit a ten-dency toward spatial disorientation andmisperceptions. These misperceptionsmay result in misreading maps, chartsand checklists, misjudgment of distanceand altitude, and loss of time perception.In a study of more than 700 naval avia-tors who had been involved in major air-craft mishaps over a four-year period, itwas discovered that those pilots whoexhibited the symptoms of inadequatestress coping were more likely to beinvolved in an aircraft mishap.
As you can see, the elements ofAMRM are intertwined and do not standalone. Together, they can be used toimprove work performance and the safetymargin in air medical transport. But crewresource management is not restricted tothe aircraft. It can also be an essentiallearning tool for people who worktogether in any environment, includingthe emergency department or traumaroom.
The Air Medical Accident Analysisrecommended CRM training as highlyeffective and moderately feasible. TheHAI/NEMSPA survey seemed to agree.Of the responses, 63% of the pilots sur-veyed found CRM effective in makingthe program safer and 41% found thatCRM made the program more efficient.Unfortunately, the pilot survey alsoshows that we have several hurdlesbefore us. A total of 38% of the respon-dents found CRM to be ineffectivebecause it “doesn’t work here” (3%), wasnot well presented (8%) or due to a lackof support from the program (6%), opera-tor (3%), flight crew (9%) and some ofthe pilots (10%). The lack of support bysome of the pilots is not surprising, asCRM proposes a way of doing things thatis contrary to the way many pilots havetrained and flown for years. Historically,pilot training leaves the pilot-in-com-mand (PIC) as the sole arbitrator as tohow to conduct a flight With AMRM,the PIC is still fully responsible but isencouraged to involve others who maybe in a position to contribute to the deci-sion–making process on behalf of safety.
53Special Supplement
SECTION 5: CONCLUSION
Every occupation has inherent risks.Medical professionals and transportationprofessionals are no different and areexposed to various risks every day.Transport Medicine combines these twofields into a unique medical specialtywhere the aviation and medical profes-sionals face uncommon challenges andrisks every day.
The risk of a helicopter accident isvery real in HEMS. Since 1972, it isestimated that HEMS has flown an esti-mated 3.0 million hours while transport-ing approximately 2.75 million patients.In 31 years (through September 2002)there have been 162 accidents involvingdedicated medical helicopters and fouraccidents involving dual-purpose heli-copters in the United States. In 67 fatalaccidents, 183 people have lost theirlives, including 144 crewmembers. Inthe early and mid-1980s, during theHEMS industry’s most rapid growth, weexperienced an alarming number of acci-dents. The early and mid-1990s showedimprovement, but 1998 to 2001 againshowed an increase in the number ofHEMS accidents across the nation.Despite this recent increase, however,the percentage of fatal accidents hasdeclined by more than a third comparedto the early 1980s. The fact remains,however, that since 1990, there has beenan average of 2.5 fatal accidents annually,taking the lives of 5 to 6 crewmemberseach year.
It must be pointed out that for all ofthe data that has been reviewed and ana-lyzed in this report, many of our numbersare very small. To draw conclusions overa five-year or 20-year period may besomewhat reasonable. However, many ofour calculations are greatly distorted dueto the small numbers for year-to-yearcomparisons. Due to a low occurrencerate, aircraft accidents are poor indicatorsof safety trends. In addition, there maybe limited first-hand information avail-able as to the real cause of an accident ifthe accident resulted in fatalities.
There is no typical HEMS accident.However, several observations are noted.A disproportionate number of HEMSaccidents occurred during night opera-tions, during the cruise phase of flight,
and on scene transports. Pilot error wasattributed as the direct or indirect causeof HEMS accidents approximately threetimes more often than mechanical fail-ure. Of the pilot errors, one-third wereweather related.
In 1988, the NTSB concluded thatpoor weather poses the greatest singlehazard to EMS helicopter operations.More than a decade later, deterioratingweather conditions continue to representa significant risk in HEMS. In general,the cause of the weather-related acci-dents does not appear to be a pilot’s dis-regard for established weather minimumsat takeoff. Instead, it is the pilot’sencounter with instrument meteorologi-cal conditions en route. In general,weather may not cause the accident, butit may increase the likelihood that anaccident will happen.
Weather is the second most commonfactor or cause of HEMS accidents. Ofthe weather-related HEMS accidents,over 85% occurred at night.Approximately 75% of all weather-relat-ed HEMS accidents resulted in fatalities.The correlation between weather-relatedaccidents and cruise flight is very strong.Degrading weather conditions can signif-icantly compromise a pilot’s ability to seeand avoid obstacles—especially while atcruise speed.
Pilot fatigue and total hours of flighttime do not appear to be significant fac-tors in HEMS accidents. Looking atHEMS incidents, however, suggests thatIFR rating and currency may be veryhelpful, if not invaluable, to overcome asituation and avoid an accident. In addi-tion, communication problems, timepressures, and distractions are frequentlyidentified as contributing risk factors inHEMS incidents.
The magnitude of injuries and aircraftdamage are significant considerations inHEMS accidents. HEMS accidents aremore likely to result in fatalities or seri-ous injuries than other helicopter acci-dents. While pre- and post-impact firesoccur in only a small percentage ofHEMS accidents, nearly half of all theaccidents result in the destruction of thehelicopter. No conclusions, however,can be made regarding single- vs. twin-engine aircraft.
Our HEMS accident rates and fatalityrates are based upon estimated exposure
data. Data for the past fourteen years hasbeen determined through several indus-try-wide surveys and various calculations.It is possible that our survey results haveunderestimated the number of HEMSprograms and dedicated helicopters byten percent or more. Therefore, our pro-posed exposure data may also be underes-timated. As a result, our calculated acci-dent and fatality rates could be overstat-ed by an estimated ten percent. This dif-ference, however, does not impact theoverall trends identified in HEMS acci-dents nor our comparison with other avi-ation operations.
In the early and mid-1980s, the acci-dent rate for HEMS was dramaticallyhigher than all other aviation operations.Since 1987, however, we have seen a sig-nificant decrease in the HEMS accidentrate to approximately one-third of whatwe experienced in the early to mid-1980s. The HEMS accident rate hasremained consistently below the accidentrates for both general aviation and allhelicopter operations since the late1980s. The fatality rate has also seen sig-nificant improvement since the late ’80s.Despite a recent increase, the fatal acci-dent rate is reduced by approximately75% compared to the early 1980s.
Finally, comparing the HEMS risks toother occupations is very difficult due tothe relatively small size of the “popula-tion” at risk. Looking strictly at thenumbers, HEMS appears to have a signif-icantly higher death rate than otheroccupations or causes of accidentaldeath. Only heart disease and cancerhave a higher fatality rate when com-pared to the 22-year average fatality ratefor HEMS.
HEMS accidents are not caused by asingle event, but by a chain of events. Inmost accidents, numerous risk factors canbe identified. Acting on any of theserisks and breaking the chain at any pointmay prevent an accident from occurring.The United States Aircraft InsuranceGroup (USAIG) has concluded thatcomplacency is a factor in over 50% ofall helicopter accidents. No one canafford to take a passive role in HEMS.The safety of flight requires the right atti-tude and active participation. Everypilot, mechanic, medical crewmember,communication specialist, and adminis-trator must be fully knowledgeable of
54 AIR MEDICAL PHYSICIAN HANDBOOK
their role and responsibilities. Each mustbe committed to a safe operation and toongoing risk management. To fly safe, aprogram must fly smart. Nothing takesthe place of comprehensive training, pro-ficiency, and sound judgment. Animportant training component should beAir Medical Resource Management. Thegoal of AMRM is to improve crew com-munications and interactions by address-ing teamwork, communication skills,decision making, workload management,situational awareness, preparation andplanning, cockpit distractions, and stressmanagement. The focus of the teammust be on doing “what is right” ratherthan on “who is right.”
The risks in HEMS cannot be underes-timated. In addition, the cumulativeeffects of multiple risk factors must be
considered when making decisions oneach and every transport. Risk manage-ment is a major component of the deci-sion-making process. It relies on situa-tional awareness, problem recognition,and exercising good judgment to reducerisks associated with each flight.
In the Forum section of theSeptember/October 2001 Air MedicalJournal, Ed MacDonald, the President ofNEMSPA wrote about “the next acci-dent”. This article should be requiredreading for everyone in HEMS. Itaddresses the “it can’t happen here” men-tality and suggests why it can happen toyou.
Tragically, there will be a next accidentand more of our colleagues will lose theirlives. Maybe not today, maybe not thismonth, or even this year. Hopefully, not
for a long, long time. No one expects itand everyone assumes the next accidentwill involve someone else’s program.
No one can eliminate the risks relatedto HEMS. Some may choose to avoidthe risk. Programs may close and indi-viduals may decide to pursue a differentoccupation. Others may choose to influ-ence risk. Organizational culture caninfluence risk as much as any/all otherfactors. The cultures, in this case, repre-sent the collective beliefs that shapebehavior toward safety and a safe HEMSprogram. Working together, every mem-ber of a flight program must play a role toactively manage risk and to avoid takingunnecessary risk. Your safety demands it.Lives depend upon it.
Do whatever it takes. Don’t be thenext accident!
Acknowledgements
Nearly two years ago, the UCAN Safety Committee undertook a simple task to review a fewarticles pertaining to air medical accidents. Our Section Chief of Emergency Medicine, Dr.James Walter had challenged us for answers regarding the magnitude of risk in HEMS. Did weas a Safety Committee, flight program, or industry have enough information to fully appreciatethe safety concerns and risks encountered every day? We decided to meet his challenge headon. Within a short time, this safety report took on a life of its own and our investigation andresearch were underway.
There are a great many people who participated and helped complete this project. My thanksto the entire UCAN Safety Committee for their dedication to this study and for enduring theendless revisions, additions, and various deadlines. I am also grateful to the entire UCAN crewand the Emergency Medicine faculty who encouraged us as we worked through the variousphases of the project. Numerous individuals, operators and aircraft manufacturers shared data-base information that made our calculations possible. Without the endorsement of the AMPABoard of Directors and Pat Petersen, this report would never have been published in this formatand in its entirety. The immediate response and support from Madeleine Byers (CJ Systems)and Sandy Kinkade (Bell Helicopter) further strengthened our resolve to pursue this publicationand to obtain the necessary funding to make this document available to the entire air medicalcommunity.
My thanks to each and every association, corporation and company who providing us withthe financial means to publish and distribute this document to our colleagues: Foundation forAir Medical Research, Bell Helicopter, CJ Systems, American Eurocopter, Fitch and Associates,AAMS, NFPA, NEMSPA, Air Methods, Sikorsky, Golden Hour Data Systems, ASTNA,Petroleum Helicopters, Metro Aviation, Turbomeca Engine Corporation, Heli-Dyne, AgustaAerospace Corporation, IAACCT, Keystone Helicopter; Innovative Engineering and NAACS.
And finally, to Jane, Maddy and Jake. Thank you for putting up with the long hours neededto complete this project. Now, more than ever, I appreciate and understand the importance ofputting “safety above all”—for myself, for my crew, for my colleagues. For my family.
—Ira Blumen
55Special Supplement
Att
ach
men
t 1
:H
EM
S A
ccid
en
t Su
mm
ary
, 1
99
8-2
00
2
Date
Ai
rcra
ft M
issi
on
Phas
e Da
y /
Airc
raft
Tim
e N
Num
ber
Prog
ram
Ci
ty, S
tate
Ty
pePh
ase
of
flig
ht
Pt
Inju
ries
Ni
ght
Wea
ther
Fire
St
atus
Op
erat
or
Com
men
ts /
Desc
riptio
n /
NTSB
Pro
babl
e Ca
use(
s)
1/11
/98
222U
TAI
RMED
Salt
Lake
City
,ON
SPO
BCR
U1
4FNi
ght
Snow
,PI
FDe
stro
yed
Air
Stor
m fr
ont m
ovin
g in
, win
d gu
sts
exce
edin
g 35
mph
.@
225
0
N222
UH
UT
wind
>35
mph
Met
hods
Fata
lities
:3 c
rewm
embe
rs a
nd 1
pat
ient.
NTSB
Fin
al: P
ilot E
rror.
3/23
/98
205A
-1Lo
s An
geles
Los
Ange
les,
ONS
POB
TKF
1 4F
, 2S
Day
Clea
r, ca
lmUn
k De
stro
yed
Publ
icEn
rout
e to
MVA
and
dur
ing
trans
port
to L
A Ch
ildre
n’s H
ospi
tal,
@ 0
740
N9
0230
City
Fir
Air O
psCA
wind
s 20
pilo
t rep
orte
d in
fligh
t em
erge
ncy.
Fata
lities
: 3 c
rew,
1 pa
tient
.Un
it 80
60
mile
vis.
NT
SB P
relim
inar
y: S
epar
atio
n of
tail
roto
r blad
es a
nd 9
0-de
gree
gear
box
durin
g cr
uise
flig
ht
5/24
/98
206
L-3
Air E
vac
EMS
Wes
t Plai
ns,
INT
PGT
TKF
0 3S
Da
y Cl
ear
No
Dest
roye
dAi
r Eva
cAi
rcra
ft lo
st p
ower
dur
ing
take
off
at ~
60
ft alt
itude
and
low
air
@ 1
235
N2
7AE
MO
EMS
spee
d. H
ard
landi
ng in
to p
arkin
g lo
t, m
ain ro
tor b
lade
stru
ck
light
pol
e an
d air
craf
t rol
led o
nto
its ri
ght s
ide.
NT
SB F
inal
: Im
prop
er m
aint
enan
ce.E
ngin
e’s a
cces
sory
gea
rbox
wa
s im
prop
erly
asse
mbl
ed, r
educ
ing
oil f
low
to th
e tu
rbin
e
shaf
ting,
lead
ing
to a
tota
l los
s of
eng
ine
powe
r.
6/5/
9835
0BA
Valle
yHa
rling
en,
ONS
PGT
CRU
0 3F
Ni
ght
5 m
iles,
PIF
Dest
roye
dTe
x-Ai
rEn
rout
e to
a ru
ral s
cene
, the
heli
copt
er c
rash
ed 1
9 m
iles
past
@ 0
549
N9
11VA
Ai
r Car
e
TX
brok
en c
eilin
g
th
e ac
ciden
t site
. Im
pact
ed tr
ees
and
terra
in. V
isibi
lity
seve
rely
at 1
,400
ft;
rest
ricte
d by
thick
sm
oke
from
fire
s in
Mex
ico.
Grnd
fog
NTSB
Fin
al: P
ilot e
rror.
Cont
inue
d fli
ght i
nto
adve
rse
weat
her
co
nditi
ons
resu
lting
in a
loss
of c
ontro
l due
to s
patia
l diso
rient
atio
n.
7/29
/98
222B
SkyL
ife o
fFr
esno
, CA
ONS
PGT
LNG
0 3N
Ni
ght
Not a
fact
or
No
Sign
ifica
nt
Roge
rsW
hile
landi
ng a
t sce
ne o
f acc
iden
t, en
coun
tere
d sig
nific
ant
@ 2
248
N911
RACe
ntra
l CA
Helic
opte
r bl
owin
g di
rt an
d du
st ~
5 ft
from
gro
und,
resu
lting
in lo
ss o
fgr
ound
refe
renc
e. H
elico
pter
rolle
d to
its
left,
landi
ng o
n its
top.
Crew
was
ext
ricat
ed w
ith a
ssist
ance
, but
was
uni
njur
ed.
NTSB
Fin
al: P
ilot e
rror.
8/20
/98
222
SPIn
tens
ive A
irSi
oux
Falls
, IN
T PG
T CR
U 0
3F
Nigh
t VM
CPI
F De
stro
yed
RMH
Enro
ute
to p
ick u
p a
patie
nt. R
outin
e ra
dio
cont
acts
up
to 5
min
@ 2
114
N30S
VSD
10 m
iles;
out.
At ~
2125
, com
m c
ente
r rec
eived
a c
all fr
om S
penc
er H
ospi
tal
8,50
0 ft
info
rmin
g th
em th
at th
e he
licop
terh
ad n
ot a
rrive
d.
over
cast
;NT
SB F
inal
: Mec
hani
cal f
ailu
re.I
n-fli
ght b
reak
-up
trave
ling
at
wind
s 12
kno
ts,
abou
t 130
kno
ts a
t 960
ft a
bove
the
grou
nd. A
fatig
ue c
rack
of t
he
to 2
6 gu
st
sw
ashp
late
oute
r rin
g pi
n of
the
main
roto
r ass
embl
y re
sulte
d in
th
e se
para
tion
of th
e pi
n, a
nd u
ltim
ately
the
in-fl
ight
bre
akup
.
8/28
/98
BK-1
17To
peka
Tope
ka, K
S M
AT
LNG
0 3N
Da
y No
t a fa
ctor
No
Si
gnifi
cant
St. L
ouis
Retu
rnin
g fro
m 2
day
s of
main
tena
nce
at a
n alt
itude
of 3
00-4
00 ft
, @
105
3 N2
30H
Life
star
Helic
opte
rspi
lot h
eard
a “l
oud
bang
” fol
lowe
d by
the
aircr
aft r
otat
ing
to th
erig
ht.
Pilo
t was
abl
e to
touc
h do
wn u
prig
ht th
en th
e air
craf
t rol
led
onto
its
left s
ide.
NOT
COU
NTED
IN T
OTAL
S.
NTSB
Fin
al: P
ilot e
rror.
Pilo
t fail
ed to
ass
ure
the
engi
ne c
owlin
g wa
s se
cure
d, re
sulti
ng in
the
cowl
ing
sepa
ratin
g fro
m th
e he
licop
ter a
nd d
amag
ing
the
main
roto
r and
tail
roto
r.
11/2
9/98
MD-
900
Life
Fligh
tBo
ise, I
D ON
S PO
B TK
F 1
4N
Dusk
Over
-Cas
tNo
No W
ind-
Idah
oDe
parti
ng fr
om M
VA in
a re
mot
e ca
nyon
, airc
raft
stuc
k an
d@
175
6 N9
77LF
to D
ark
10m
i vis
scre
en a
nd
Helic
opte
rsse
vere
d un
mar
ked
powe
r lin
es 1
50 ft
abo
ve th
e gr
ound
. Pilo
tm
ain ro
tor
then
det
erm
ined
that
the
helic
opte
r was
con
trolla
ble
and
disp
layed
blad
esno
unu
sual
fligh
t cha
ract
erist
ics, a
nd c
hose
to p
roce
ed to
his
dest
inat
ion.
Airc
raft
com
plet
ed m
issio
n an
d lan
ded
unev
entfu
lly a
t ho
spita
l with
pat
ient a
nd c
rew.
Pos
t-flig
ht e
xam
inat
ion
reve
aled
craz
ing
of th
e wi
ndsc
reen
and
dam
age
to 4
of t
he 5
main
roto
r bl
ades
requ
iring
majo
r rep
air/re
plac
emen
t. NT
SB F
inal
: Pilo
t erro
r.
56 AIR MEDICAL PHYSICIAN HANDBOOK
Inst
ruct
or p
ilot (
PIC)
had
bee
n do
ing
a 3-
hr c
heck
ride
with
new
pilo
t, wi
th la
st m
aneu
ver “
hydr
aulic
s of
f”. P
IC th
en in
itiat
ed a
nor
-m
al ta
keof
f (wi
th h
ydra
ulics
‘ON’
) and
dur
ing
the
take
off,
the
heli-
copt
er ro
lled
over
to th
e lef
t. P
IC re
porte
d it
felt
like
a co
mpl
ete
hydr
aulic
s fa
ilure
. Ro
tor b
lades
stru
ck th
e gr
ound
, kno
ckin
g ta
ilne
arly
off,
front
of t
he a
ircra
ft wa
s de
stro
yed.
NT
SB F
inal
: Und
eter
min
ed.
Fligh
t abo
rted
twice
due
to w
eath
er e
nrou
te to
refe
rring
hos
pita
l.Pi
lot a
ttem
pted
a p
reca
utio
nary
land
ing
in d
ecre
ased
visi
bilit
y.Ai
rcra
ft st
ruck
tree
and
hou
se 1
.5 m
iles
from
hos
pita
l. NT
SB F
inal
:Pi
lot e
rror.
Inad
verte
nt fl
ight
into
IMC,
sub
sequ
ent s
patia
l diso
rien-
tatio
n an
d lo
ss o
f con
trol;
also
inac
cura
te w
eath
er fo
reca
st.
2-he
licop
ter s
cene
. Dur
ing
appr
oach
for l
andi
ng a
ppro
ach,
the
pilo
tno
ticed
pow
er li
nes
runn
ing
para
llel t
o th
e ro
ad b
ut d
id n
ot c
onsid
-er
them
to b
e a
haza
rd. O
n ta
ke-o
ff, th
e he
licop
ter h
it th
e po
wer
lines
and
the
pilo
t lan
ded
the
helic
opte
r in
an a
djac
ent f
ield.
Unde
rsid
es o
f the
main
roto
r blad
es w
ere
dam
aged
and
2 ta
il ro
tor
blad
es w
ere
dest
roye
d. N
TSB
Fina
l: Pi
lot e
rror.
Failu
re to
main
tain
clear
ance
from
pow
er li
nes.
Sun
glar
e wa
s a
fact
or.
Helic
opte
r cra
shed
as
they
wer
e re
turn
ing
to a
rem
ote
base
afte
rde
liver
ing
a pa
tient
to V
alley
Hos
pita
l Med
ical C
ente
r. N
TSB
Fina
l:Pi
lot e
rror.
Cont
inue
d VF
R fli
ght i
n de
terio
ratin
g IF
R co
nditi
ons
resu
lting
in s
patia
l diso
rient
atio
n an
d su
bseq
uent
loss
of c
ontro
l.
Whi
le lif
ting
off a
nd h
over
ing,
tail
roto
r stru
ck h
anga
r. Ta
il ro
tor
and
skid
s da
mag
ed. D
amag
e to
bui
ldin
g. N
TSB
Fina
l: Pi
lot e
rror.
Failu
re to
main
tain
visu
al se
para
tion
with
bui
ldin
g.
On a
ppro
ach
to s
cene
in ru
ral a
rea,
har
d lan
ding
occ
urre
d in
tofie
ld.
Dam
age
to s
kids,
tail
boom
and
nos
e co
wlin
g.NT
SB F
inal
: Pilo
t erro
r.M
isjud
ged
flare
dur
ing
landi
ng.
IFR
fligh
t, 2
pilo
t airc
raft.
Relo
catio
n fli
ght f
rom
rem
ote
airpo
rt(1
,381
ft M
SL) b
ase
back
to h
ospi
tal.
Afte
r tak
e-of
f, at
abo
ut 1
,600
feet
, the
co-
pilo
t (SI
C) w
ho w
as fl
ying,
beg
an a
des
cend
ing
left t
urn
and
subs
eque
ntly
cras
hed
into
risin
g te
rrain
on
a tre
e-co
vere
dslo
pe a
t ~ 1
,000
ft e
levat
ion.
NTS
B Fi
nal:
Pilo
t erro
r.Fa
ilure
of t
hePI
C to
ade
quat
ely s
uper
vise
the
SIC,
and
main
tain
a p
ositi
ve c
limb.
Helic
opte
r was
on
appr
oach
to a
n in
term
ediat
e re
fueli
ng s
ite d
urin
gan
inte
rhos
pita
l tra
nsfe
r. A
witn
ess
saw
piec
es o
f the
main
roto
rsy
stem
sep
arat
e fro
m th
e he
licop
ter b
efor
e th
e cr
ash.
NT
SB F
inal
: Mec
hani
cal f
ailu
re.
Corro
sion
of th
e te
nsio
n-to
rsio
nst
rap
resu
lting
in fa
tigue
cra
ckin
g an
d su
bseq
uent
sep
arat
ion
of th
est
rap
and
main
roto
r blad
e fro
m th
e he
licop
ter.
A ba
ck-u
p he
licop
ter w
as in
ser
vice.
Upo
n ta
keof
f, re
achi
ng th
eed
ge o
f a ro
ofto
p ho
spita
l heli
pad,
the
aircr
aft d
id a
n ab
rupt
and
violet
yaw
to th
e lef
t, th
e no
se tu
cked
dow
nwar
d an
d th
e air
craf
tst
arte
d lo
osin
g cle
aran
ce. P
ilot l
ande
d th
e air
craf
t on
the
pave
dst
reet
belo
w, b
ut th
e ta
ilboo
m h
it a
brick
wall
. NT
SB F
inal
: Pilo
t erro
r.In
adeq
uate
pre
fligh
t plan
ning
/ pr
epar
atio
n —
aux
iliary
pow
er u
nit c
ord
was
atta
ched
to th
e he
licop
ter d
urin
g th
e he
licop
ter’s
take
off a
ttem
pt.
Date
Ai
rcra
ft M
issi
on
Phas
e Da
y /
Airc
raft
Tim
e N
Num
ber
Prog
ram
Ci
ty, S
tate
Ty
pePh
ase
of
flig
ht
Pt
Inju
ries
Ni
ght
Wea
ther
Fire
St
atus
Op
erat
or
Com
men
ts /
Desc
riptio
n /
NTSB
Pro
babl
e Ca
use(
s)
12/1
3/98
AS 3
50-B
ASh
anno
nSa
n An
toni
o,
MAT
HO
V 0
2N
Clea
r Su
bsta
ntial
So
uthw
est
@ 1
745
N9
11M
VM
ediva
c On
eTX
Helic
opte
rs
2/12
/99
AS 3
55Li
feFli
ght
Toled
o, O
H IN
T PR
F LN
G 0
3S
Nigh
t IM
CNo
De
stro
yed
CJI
@ 1
720
N355
MF
– Sc
atte
red
snow
squ
alls.
2/13
/99
BK-1
17 B
-1He
rman
nHo
usto
n, T
XON
S PO
B CL
B 2
5N
Day
Clea
r No
Su
bsta
ntial
Ow
n@
164
5
N220
HLi
fe F
light
Part
135
4/3/
99BO
-105
Fligh
t for
Las
Vega
s, N
V IN
T PR
F CR
U 0
3F
Nigh
t IM
CPI
FDe
stro
yed
Met
ro@
235
0
N105
HH
Life
Snow
ing
Aviat
ion
4/11
/99
BK-1
17Ba
y Fli
ght
St. P
eter
sbur
g,ON
S PG
T HO
V 0
3N
Day
Clea
r No
Si
gnifi
cant
RM
H@
164
5
N163
BK
FL
5/15
/99
222U
TLi
felin
e Ro
ckfo
rd, I
L ON
S PG
T AP
P 0
3N
Nigh
t 40
00 ft
No
Subs
tant
ial
Air
@ 2
122
N781
SAce
iling,
M
etho
ds
4mi v
is.
6/14
/99
S-76
AUn
ivers
ityLe
xingt
on K
T OT
H N/
A CR
U 0
4F
Nigh
t IM
C Fo
g;No
De
stro
yed
PHI
@ 2
208
N274
3Eof
Ken
tuck
y<
1/4
mile
visib
ility;
Win
ds c
alm
7/17
/99
BK-1
17He
rman
nHo
usto
n, T
X IN
T PG
T CR
U 0
3F
Day
Rain
in
No
Dest
roye
d Ow
n@
123
1
N110
HH
Life
Fligh
t th
e ar
eaPa
rt 13
5
8/10
/99
206L
Life
Beat
Air
Cape
Giar
deau
,IN
T PG
T TK
F 0
3N
Day
N/A
No
Subs
tant
ial
St. L
ouis
@ 1
138
N810
FM
edica
l M
O
Helic
opte
r
57Special Supplement
Date
Ai
rcra
ft M
issi
on
Phas
e Da
y /
Airc
raft
Tim
e N
Num
ber
Prog
ram
Ci
ty, S
tate
Ty
pePh
ase
of
flig
ht
Pt
Inju
ries
Ni
ght
Wea
ther
Fire
St
atus
Op
erat
or
Com
men
ts /
Desc
riptio
n /
NTSB
Pro
babl
e Ca
use(
s)
9/10
/99
BO-1
05Fir
st F
light
Melb
ourn
e, F
L ON
S PG
T AP
P 0
3S
Nigh
t Fo
ggy
No
Sign
ifica
nt
Met
ro@
031
4
N911
HR
Aviat
ion
11/1
7/99
206L
–1M
ercy
Flig
ht/
Grea
t Fall
s, M
T ON
S PO
B TK
F 1
4N
Day
Win
d gu
sts
No
Su
bsta
ntial
Om
ni-
@ 1
350
N519
EH
Med
fligh
t to
15
knot
s fli
ght
02/2
6/00
412
Life
Star
Kn
oxvil
le, T
N ON
S PG
T M
AN
0 3N
Ni
ght
Clea
r No
Si
gnifi
cant
Ow
n Pa
rt@
020
0
N411
UT
Nigh
t13
5
3/10
/00
BO-1
05Li
fest
ar
Amar
illo, T
X ON
S PO
B CR
U 1
4F
Nigh
t Fo
g PI
F De
stro
yed
Tem
sco
@ 0
605
N3
35T
4/14
/00
222
Life
link
III
St. P
aul,
MN
OTH
PRF
CRU
0 2N
Da
y Cl
ear
No
Sign
ifica
nt
Air
@ 1
610
N2
25LL
M
etho
d
4/25
/00
BK11
7Ba
yflig
ht
St. P
eter
sbur
g,
INT
PRF
CRU
0 3F
Da
y No
t a F
acto
r No
De
stro
yed
RMH
@ 1
215
N4
28M
B
FL
5/6/
00BK
117
Unive
rsity
Ci
ncin
nati,
OH
FUL
N/A
LNG
0 1N
Ni
ght
VMC
No
Subs
tant
ial
PHI
@ 2
335
N911
NCHo
spita
l
Appr
oach
ing
a sc
ene
LZ, t
he h
elico
pter
beg
an to
des
cend
rapi
dly
from
~ 3
00 ft
. Pilo
t app
lied
colle
ctive
con
trol a
nd e
ngin
e po
wer,
but t
he h
elico
pter
con
tinue
d to
des
cend
, col
lidin
g wi
th tr
ees
and
then
rolle
d on
to it
s rig
ht s
ide
in s
wam
py te
rrain
. NT
SB F
inal
: Pilo
t erro
r.Fa
ilure
to re
cogn
ize e
ntry
into
set
tling
with
pow
er d
urin
g ap
proa
ch to
land
and
failu
re to
take
rem
edial
actio
n to
esc
ape
from
set
tling
with
pow
er.
Helic
opte
r res
pond
ed to
a s
ki re
sort,
with
the
LZ
in a
n op
en a
rea
near
ski
lift t
ower
s. O
n ta
ke-o
ff, w
ith tr
ees
dire
ctly
in fr
ont,
the
pilo
t dec
ided
to tu
rn th
e he
licop
ter t
o th
e lef
t, ho
ver t
o an
ope
nar
ea a
nd d
epar
t dow
nslo
pe. .
. .”
Afte
r the
heli
copt
er m
oved
left
20-3
0 ft,
the
pilo
t felt
the
tail
of th
e he
licop
ter “
rota
te a
brup
tlylef
t.” P
ilot t
ried
to m
ainta
in c
ontro
l and
retu
rn to
the
LZ, b
ut th
eta
il ro
tor s
truck
a li
ft to
wer.
Helic
opte
r lan
ded
hard
. NT
SB F
inal
: Pilo
t erro
r.Cl
eara
nce
from
an
objec
t was
not
m
ainta
ined
. Gus
ting
wind
con
ditio
ns w
as a
fact
or.
Airc
raft
had
arriv
ed o
n sc
ene
and
was
repo
sitio
ning
in th
e LZ
due
to p
rese
nce
of a
ste
ep s
lope
. Th
e ta
il ro
tor s
truck
a s
mall
tree
while
man
euve
ring.
Afte
r con
tact
, the
TR
and
TR g
earb
ox s
epar
at-
ed fr
om th
e air
craf
t. F
lying
deb
ris fr
om s
epar
atin
g co
mpo
nent
sca
used
furth
er d
amag
e to
airc
raft
fuse
lage.
NTSB
Fin
al: P
ilot e
rror.
Failu
re to
main
tain
visu
al lo
okou
t res
ultin
g in
col
lisio
n wi
th tr
ee.
Resp
onde
d to
a s
cene
repo
rtedl
y clo
se to
the
TX/O
K st
ate
line.
Fog
repo
rted
form
ing
while
the
aircr
aft w
as o
n sc
ene.
The
pilo
tan
d cr
ew li
fted
with
a p
atien
t on
boar
d at
~06
05.
No ra
dio
com
-m
unica
tion
was
esta
blish
ed a
fter l
ift-o
ff. D
ue to
fog
in th
e ar
ea,
wrec
kage
was
not
fou
nd u
ntil
~11
00 h
rs.
NTSB
Fin
al:
Pilo
t erro
r. Fa
ilure
to m
ainta
in c
ontro
l of a
ircra
ft as
a re
sult
of c
ontin
ued
fligh
t int
o kn
own
adve
rse
weat
her.
Fac
tors
inclu
de d
ark
nigh
t con
ditio
ns, f
og,
low
ceilin
g, a
nd p
ilot’s
lack
of
tota
l ins
trum
ent f
light
tim
e.
Durin
g cr
uise
, pilo
t los
t of c
ontro
l of a
ircra
ft an
d lan
ded
on a
two
stor
y bu
ildin
g. M
ajor d
amag
e to
skid
s. N
o in
jurie
s.NT
SB F
inal
: Mec
hani
cal.
Pylo
n m
ount
ed s
uppo
rt as
sem
bly
sepa
rate
d fro
m tr
ansm
issio
n ca
se d
ue to
fatig
ue fa
ilure
of t
heth
read
ed s
tuds
and
dow
el pi
ns, r
esul
ting
in fa
ilure
of t
he fl
ight
cont
rol s
yste
m; a
lso, i
nade
quat
e m
ainte
nanc
e pr
oced
ures
by
com
pany
main
tena
nce
pers
onne
l.
Crew
had
dro
pped
off
a pa
tient
at B
ayfro
nt M
edica
l Cen
ter.
Depa
rted
for b
ase
(8 m
in fl
ight
), fly
ing
a ne
w ro
ute
in re
spon
se to
noise
com
plain
ts fr
om n
eighb
ors
along
the
prev
ious
ly di
rect
rout
e.3-
4 m
in in
to fl
ight
, col
lided
with
the
radi
o tra
nsm
issio
n to
wer g
uywi
re a
nd th
e st
eel t
ower
480
feet
abo
ve th
e gr
ound
. NT
SB F
inal
: Pilo
t erro
r.Fa
ilure
to m
ainta
in c
leara
nce
with
towe
rre
sulti
ng in
col
lisio
n.
Afte
r cro
ssin
g th
e ed
ge o
f the
LZ
and
almos
t in
a ho
ver,
he h
eard
alo
ud n
oise
or b
ang
from
the
rear
of t
he h
elico
pter
. Sim
ulta
neou
sly,
the
left r
udde
r ped
al pu
shed
rear
ward
, and
the
nose
sta
rted
tom
ove
to th
e rig
ht. P
ilot m
ade
a ha
rd la
ndin
g.
NTSB
Fin
al: P
ilot e
rror.
Misj
udgm
ent o
f clo
sure
rate
resu
lting
inco
llisio
n wi
th b
uild
ing.
Fac
tors
invo
lved:
tailw
ind
and
stuc
k wi
ndso
ck.
58 AIR MEDICAL PHYSICIAN HANDBOOK
Date
Ai
rcra
ft M
issi
on
Phas
e Da
y /
Airc
raft
Tim
e N
Num
ber
Prog
ram
Ci
ty, S
tate
Ty
pePh
ase
of
flig
ht
Pt
Inju
ries
Ni
ght
Wea
ther
Fire
St
atus
Op
erat
or
Com
men
ts /
Desc
riptio
n /
NTSB
Pro
babl
e Ca
use(
s)
7/16
/00
BK 1
17Li
fe S
tar
Allen
, TX
ONS
PGT
LNG
0 3N
Ni
ght
Not a
Fac
tor
No
Dest
roye
d Om
ni-fl
ight
@ 0
140
N312
LS(T
exas
)
7/24
/00
A St
ar
Geor
gia
Bapt
ist
Atlan
ta, G
A IN
T PR
F CR
U 0
3F
Nigh
t Cl
ear
No
Dest
roye
d Cr
itica
l @
023
0 N9
11AM
Li
fe F
light
Care
Med
-flig
ht
7/28
/00
222
UTLi
fe L
ink
III
Min
neap
olis,
UN
K N/
A TK
F 0
1N
Day
Clea
r No
Su
bsta
ntial
Ai
r Met
hods
@ 1
140
N224
LLM
N
10/1
6/00
AS35
5Du
ke
Durh
am, N
C IN
T
CRU
0 1F
Ni
ght
Clea
r UN
KDe
stro
yed
CJI
@ 2
355
N3
55DU
Life
Flig
ht
10/1
4/00
206
LCl
assic
Ja
cob
Lake
, ON
S PO
B TK
F 1
4M
Day
Not a
fact
or
No
Dest
roye
d Ow
n@
122
7 N2
233F
Life
guar
d III
AZPa
rt 13
5
11/1
3/00
BO 1
05Fli
ght
Pahr
ump,
NV
ONS
PGT
LNG
0 3N
Ni
ght
Not a
fact
or
No
Subs
tant
ial
Met
ro
@ 2
048
91
1VH
fo
r Life
Aviat
ion
12/1
8/00
365
N1No
ne
Wes
t Miff
lin,
MAT
0
3M
Day
Not a
fact
or
No
Dest
roye
d CJ
I@
153
0
N89S
M
PA
12/2
2/00
206
Criti
cal A
ir
Wilc
ox, A
Z IN
T PR
F LN
G 0
3M
Nigh
t No
t a fa
ctor
No
Su
bsta
ntial
Cr
itica
l @
033
1
N288
JB
Med
icine
Air M
ed
01/2
2/01
206
L-1
Air E
vac
Quin
cy, I
L IN
T PG
T TK
F 0
1F
Nigh
t Cl
ear
No
Min
or
Air E
vac
@ 0
005
N6
1AE
on g
roun
dEM
S In
c.2N
Whi
le m
aneu
verin
g air
craf
t at a
n LZ
of a
sce
ne re
spon
se, t
he ta
ilro
tor s
truck
a tr
ee.
NTSB
Fin
al: P
ilot e
rror.
Failu
re to
main
tain
obs
tacle
clea
ranc
ewh
ile h
over
ing.
Airc
raft
was
retu
rnin
g fro
m a
call
in S
ylves
ter,
GA.
Radi
o co
ntac
twa
s lo
st a
t 023
0. T
here
was
no
May
day
or d
istre
ss s
igna
l. Th
eair
craf
t was
foun
d in
a w
oode
d ar
ea.
NTSB
Fin
al: P
ilot e
rror.
Pilo
t exp
erien
ced
spat
ial d
isorie
ntat
ion
resu
lting
in lo
ss o
f con
trol o
f airc
raft.
Fac
tors
invo
lved:
dar
kni
ght l
ight
con
ditio
ns.
On li
ftoff
from
the
helip
ad, t
he ta
il ro
tor h
it th
e pa
d lig
ht.
The
pilo
t lan
ded
and
shut
dow
n th
e he
licop
ter.
NTSB
Fin
al: P
ilot e
rror.
Inad
equa
te p
refli
ght,
impr
oper
ve
rtica
l tak
eoff,
and
not
obt
ainin
g cle
aran
ce fr
om h
elipa
d lig
ht.
Fact
ors
invo
lved:
tailw
ind
take
off,
helip
ad li
ght.
Enro
ute
to h
ospi
tal,
the
main
roto
r gea
rbox
(MGB
) oil
pres
sure
warn
ing
light
illu
min
ated
. Afte
r lan
ding
the
crew
wen
t by
grou
nd.
Mec
hani
c be
lieve
d th
e oi
l pre
ssur
e sw
itch
had
faile
d an
d di
scon
-ne
cted
the
wire
. The
pilo
t did
a ru
n-up
and
hov
er.
The
AC th
ento
ok o
ff an
d wa
s re
porte
d do
wn 1
6 m
inut
es la
ter.
NTSB
Fin
al: M
echa
nica
l.Fa
ilure
to c
ompl
y wi
th m
anuf
actu
rer’s
inst
ruct
ions
for c
orre
ctin
g illu
min
ated
main
roto
r gea
rbox
oil
pres
sure
war
ning
ligh
t, re
sulti
ng in
failu
re o
f the
main
roto
r gea
r-bo
x du
e to
oil
star
vatio
n, lo
ss o
f main
roto
r RPM
, and
an
unco
ntro
lled
desc
ent.
On ta
keof
f fro
m s
cene
, hit
trees
. Po
ssib
le lo
ss o
f tail
roto
r ef
fect
ivene
ss.
NTSB
Fin
al: P
ilot e
rror.
In-fl
ight
loss
of c
ontro
l dur
ing
lift-o
ffdu
e to
impr
oper
plan
ning
and
dec
ision
s. F
acto
rs in
volve
d: h
igh-
dens
ity a
ltitu
de, h
elico
pter
weig
ht, a
nd la
ck o
f sui
tabl
e ta
ke-o
ffar
ea.
Resp
ondi
ng to
sce
ne.
2 fe
et o
ff th
e gr
ound
read
y to
land
whe
n a
vehi
cle e
nter
ed it
s LZ
. Th
e pi
lot w
as a
bout
to a
scen
d wh
en h
eno
ticed
a s
et o
f pow
er li
nes
in h
is pa
th.
He p
ower
ed u
p to
turn
arou
nd w
hen
his
skid
hit
the
grou
nd a
nd th
e he
licop
ter l
ande
d on
its s
ide.
Pilo
t was
abl
e to
clim
b ou
t, cr
ew h
ad to
be
“cut
out
”.NT
SB F
act f
indi
ng d
ocum
ent.
Tail
roto
r con
trol l
oss
durin
g po
st-5
00 h
r main
tena
nce
oper
atio
nch
eck.
Sev
eral
atte
mpt
s we
re m
ade
to la
nd.
On th
e las
t atte
mpt
the
pilo
t los
t con
trol a
nd e
xecu
ted
a co
ntro
lled
cras
h lan
ding
.NO
T CO
UNTE
D IN
TOT
ALS.
Pilo
t exp
erien
ced
sudd
en o
nset
of i
llnes
s du
e to
app
aren
t foo
dpo
isoni
ng.
A ve
ry h
ard
landi
ng w
as m
ade
at W
ilcox
Airp
ort.
NTSB
Fin
al: P
ilot i
ncap
acita
tion.
Due
to n
ause
a. P
ilot c
ollap
sed
onto
cyc
lic c
ausin
g in
adve
rtent
main
roto
r con
tact
with
gro
und.
Hosp
ital s
ecur
ity o
ffice
r walk
ed u
nder
neat
h th
e ta
il bo
om w
hile
the
AC w
as ru
nnin
g. T
he o
ffice
r was
stru
ck in
the
head
by
the
tail
roto
r res
ultin
g in
a fa
tal w
ound
. NT
SB F
inal
: Gro
und
crew
erro
r.Se
curit
y gu
ard
faile
d to
main
-ta
in c
leara
nce
from
tail
roto
r.
59Special Supplement
Date
Ai
rcra
ft M
issi
on
Phas
e Da
y /
Airc
raft
Tim
e N
Num
ber
Prog
ram
Ci
ty, S
tate
Ty
pePh
ase
of
flig
ht
Pt
Inju
ries
Ni
ght
Wea
ther
Fire
St
atus
Op
erat
or
Com
men
ts /
Desc
riptio
n /
NTSB
Pro
babl
e Ca
use(
s)
02/2
8/01
412
St.M
ary’s
Gran
d
MAT
0
1F
Day
Not a
fact
or
No
Dest
roye
d PH
I@
102
4 N4
12SM
Air L
ifeJu
nctio
n,
CO
03/2
3/01
206L
1EM
S Ai
r Se
neca
Fall
s,
UNK
UNK
CRU
0 2S
Da
y No
t a fa
ctor
No
Su
bsta
ntial
EM
S Ai
r@
152
0 N2
138Y
Serv
ices
NYSe
rvice
sof
NY
of N
Y
04/0
6/01
222U
TW
yom
ing
Alco
va, W
Y ON
S PG
T LN
G 0
3N
Day
Not a
fact
or
No
Subs
tant
ial
CJ S
yste
ms
@ 1
715
N222
LFLi
fe F
light
04/2
3/01
206L
3Cr
itica
l Air
Phoe
nix,
IN
T
CRU
3N
Da
y Cl
ear
No
Subs
tant
ial
Criti
cal A
ir@
143
0 N2
15M
Med
icine
AZM
edici
ne
05/0
5/01
BO-1
05C
Mer
cy
Med
ford
, IN
T PO
B CR
U1
4N
Day
Not a
fact
or
INF
Subs
tant
ial
Own
135
@ 1
700
N105
RH
Fligh
tOR
06/0
3/01
369D
Hana
pepe
, ON
S PG
T CR
U O
4N
Day
Not a
fact
or
No
Subs
tant
ial
Smok
ey
@ 1
620
N110
9VHI
Mou
ntain
Helic
opte
rs
07/2
0/01
BK11
7No
rth T
exas
Ad
diso
n,
INT
PGT
CRU
0 2S
Nigh
t Cl
ear
N0
Subs
tant
ial
Omni
-@
160
3N3
13LS
Li
feSt
arTX
1Mfli
ght
07/2
1/01
S-76
Los
Ange
les,
1M
Nigh
t No
t a fa
ctor
No
Subs
tant
ial
Helin
et
@ 0
049
N769
BBCA
Aviat
ion
08/1
8/01
355F
1Ca
re F
light
Re
no, N
V ON
S PO
B TK
F 1
4N
Day
Clea
r No
Su
bsta
ntial
RM
H@
142
5 N5
3LH
09/2
2/01
AS35
0En
loe
Ch
ico, C
A ON
S PG
T LN
G 0
1FNi
ght
Clea
r No
De
stro
yed
Enlo
e@
200
6 N9
11NT
Med
ical
1SM
edica
lCe
nter
1NCe
nter
10/0
7/01
AS35
5F1
Sout
hwes
tRo
sebu
d,
ONS
PGT
CRU
0 2M
Nigh
t Cl
ear
No
Dest
roye
d So
uthw
est
@ 2
250
N911
BBHe
licop
ters
,TX
1NHe
licop
ters
,In
c.
Inc.
10/1
9/01
350B
2M
ed A
irRo
swell
, PR
N/
A DE
C 0
2F
Day
Not a
fact
or
UNK
Dest
roye
d M
edica
l Air
@ 1
458
N111
DTNM
2STr
ansp
ort
Depa
rted
the
hosp
ital h
elipa
d fo
r pos
t main
tena
nce
test
flig
ht th
atre
quire
d au
toro
tatio
n. Im
pact
ed th
e gr
ound
app
roxim
ately
24
min
utes
late
r, 11
mile
s so
uth
of G
rand
Jun
ctio
n.
NTSB
Fin
al: P
ilot E
rror.
Pilo
t fail
ed to
main
tain
roto
r spe
ed d
ur-
ing
an in
tent
iona
l aut
orot
atio
n, re
sulte
d in
a lo
ss o
f con
trol.
Pilo
t hea
rd b
angs
and
exp
erien
ced
yawi
ng a
nd p
ower
loss
. He
dete
rmin
ed th
at ta
il ro
tor t
hrus
t was
lost
. Upo
n pr
ecau
tiona
rylan
ding
the
aircr
aft r
olled
. NT
SB F
inal
: Mec
hani
cal.
The
loss
of a
bol
t in
a Th
omas
cou
plin
gon
the
tail
roto
r driv
e sh
aft,
for u
ndet
erm
ined
reas
ons,
dur
ing
clim
b, w
hile
oper
atin
g ov
er u
nsui
tabl
e te
rrain
.
In h
over
, pilo
t atte
mpt
ed to
reor
ient a
ircra
ft wh
en th
e ta
il ro
tor
stru
ck a
55-
gallo
n ba
rrel.
Airc
raft
lande
d wi
th d
amag
e.NT
SB F
inal
: Pilo
t Erro
r.
Airc
raft
expe
rienc
ed to
tal p
ower
loss
and
pilo
t com
plet
ed a
n au
toro
tatio
n lan
ding
. NT
SB p
relim
inar
y.
Powe
r los
s an
d fir
e in
#1
engi
ne.
Emer
genc
y lan
ding
at a
irpor
tan
d fir
e ex
tingu
ished
by
airpo
rt pe
rson
nel.
NTSB
pre
limin
ary.
Not d
edica
ted
EMS.
Enr
oute
with
fire
fight
ers
to p
ick u
p a
“med
ical e
mer
genc
y” a
nd e
xper
ience
d en
gine
failu
re.
Auto
rota
ted
into
tree
s.
NTSB
pre
limin
ary.
NOT
INCL
UDED
IN T
OTAL
S.
Whi
le en
rout
e an
eng
ine
faile
d. W
hile
turn
ing
towa
rds
the
airpo
rtth
e 2n
d. E
ngin
e fa
iled.
The
AC
cras
hed.
NT
SB F
inal
: Pilo
t Erro
r.Fa
ilure
to fo
llow
pref
light
che
cklis
t and
turn
on
fuel
pum
ps re
sulti
ng in
fuel
star
vatio
n an
d po
wer l
oss.
Had
trans
porte
d an
org
an h
arve
st te
am to
hos
pita
l. H
elico
pter
rolle
d on
to it
s rig
ht s
ide
while
sta
ndin
g un
man
ned
with
bot
hen
gine
s op
erat
ing
and
roto
rs tu
rnin
g.
NTSB
Fac
t. NO
T IN
CLUD
ED IN
TOTA
LS.
Pilo
t los
t visu
al re
fere
nce
upon
take
off f
rom
acc
iden
t sce
ne
(bro
wn o
ut).
Atte
mpt
ed to
land
and
airc
raft
rolle
d ov
er.
NTSB
Fin
al: P
ilot E
rror.
Selec
tion
of u
nsui
tabl
e lan
ding
zone
.
Appr
oach
ing
scen
e LZ
with
flig
ht n
urse
usin
g ni
ght v
ision
go
ggles
. Enc
ount
ered
bro
wn o
ut a
nd a
borte
d lan
ding
. Im
pact
ed tr
ees
and
cras
hed.
NT
SB p
relim
inar
y.
Diffi
culty
find
ing
scen
e an
d cr
ew in
form
ed p
t. to
be
take
n by
grou
nd tr
ansp
ort.
Bot
h en
gine
s lo
st p
ower
and
pilo
t ini
tiate
dau
toro
tatio
n re
sulti
ng in
a h
ard
landi
ng a
nd s
lide.
Airc
raft
sepa
rate
d fro
m s
kids
and
cam
e to
rest
on
its s
ide.
Fue
l sys
tem
foun
d to
hav
e a
tota
l of 2
qua
rts o
f fue
l.
NTSB
pre
limin
ary.
LZ tr
ainin
g. 2
pol
ice o
ffice
rs k
illed,
pilo
t and
ano
ther
pol
ice o
ffice
rse
rious
. NT
SB p
relim
inar
y.
60 AIR MEDICAL PHYSICIAN HANDBOOK
Date
Ai
rcra
ft M
issi
on
Phas
e Da
y /
Airc
raft
Tim
e N
Num
ber
Prog
ram
Ci
ty, S
tate
Ty
pePh
ase
of
flig
ht
Pt
Inju
ries
Ni
ght
Wea
ther
Fire
St
atus
Op
erat
or
Com
men
ts /
Desc
riptio
n /
NTSB
Pro
babl
e Ca
use(
s)
11/0
9/01
A119
IHC
Life
Flig
ht
Ogde
n, U
T IN
T N/
A AP
P 0
2M
Day
Clea
r No
Si
gnifi
cant
Ow
n @
144
5N1
19RX
Part
135
11/1
4/01
BO-1
05Ba
nnoc
kPo
cate
llo.
MAT
N/
A CR
U 0
1S
Nigh
t No
t a fa
ctor
PI
F De
stro
yed
Own
@ 0
033
N93L
FLi
fe F
light
IDPa
rt 13
5
01/1
8/02
BK 1
17Un
ivers
ityCl
evela
nd,
OTH
PGT
TKF
0 2F
Nigh
t No
t a fa
ctor
UN
KDe
stro
yed
CJ@
002
5 N6
26M
EHo
spita
lsOH
1SSy
stem
sM
edEv
ac
01/2
0/02
109
Airli
ftSe
attle
, WA
ONS
PGT
TKO
0 1S
Da
y Li
ght R
ainNo
Si
gnifi
cant
CJ
@ 0
750
N55N
WNo
rthwe
st30
00 ft
. Sy
stem
sce
iling
03/2
1/02
AS-3
50B
Mou
ntain
Susa
nville
, OT
H PR
F CR
U 0
1FDa
y No
t a fa
ctor
No
Subs
tant
ial
Mou
ntain
@ 1
335
N118
4HLi
fe F
light
CA2S
Li
fe F
light
6/21
/02
AS-3
50-B
2Li
feNe
t No
rfolk,
NE
INT
PGT
CLB
0 3F
Da
y VF
R No
De
stro
yed
RMH
@ 1
207
N852
HW
09/0
7/02
222U
TM
ercy
Air
Nipt
on, C
A ON
S PG
T CR
U 0
3F
N VF
R PI
F De
stro
yed
Mer
cy A
ir@
042
8N4
17M
ASe
rvice
s
09/0
9/02
206
Care
fligh
t Si
oux
Falls
, SD
INT
POB
UNK
1 4F
N
UNK
UNK
Dest
roye
d Om
ni-
@ 2
152
N400
SLfli
ght
Upon
app
roac
h, ro
tor R
PM d
ecay
ed a
nd d
id n
ot re
cove
r. Ha
rdlan
ding
and
skid
s to
rn o
ff an
d air
craf
t rol
led o
nto
side.
Fligh
t was
for a
sur
vey
of n
ew h
elipa
d.NT
SB F
inal
: Mec
hani
cal.
Impr
oper
rigg
ing
of th
e ro
tary
var
i-ab
le di
ffere
ntial
tran
sfor
mer
by
the
man
ufac
ture
r, re
sulti
ng in
inco
rrect
fuel
sche
dulin
g to
the
fuel
cont
rol u
nit.
Pilo
t lan
ded
as a
pre
caut
ion
seco
ndar
y to
fuel
trans
fer p
ump
prob
lem, M
edica
l cre
w an
d pa
tient
wen
t by
grou
nd.
Mec
hani
car
rived
5 h
ours
late
r to
repa
ir air
craf
t. P
ilot t
ook
off a
nd w
asre
turn
ing
to b
ase
and
expe
rienc
ed d
ecre
ased
visi
bilit
y an
dcr
ashe
d. T
he m
echa
nic
saw
“a b
right
glo
w” a
nd fo
llowe
d th
efli
ght p
ath
of th
e air
craf
t. T
he m
echa
nic
extin
guish
ed th
e fir
e,di
scon
nect
ed th
e ba
ttery
, and
call
ed fo
r help
.NT
SB F
inal
: Pilo
t erro
r.Fa
ilure
to m
ainta
in a
dequ
ate
clear
ance
of te
rrain
dur
ing
initi
al cli
mb.
AC li
fting
from
roof
top
helip
ad w
hen
it st
ruck
sid
e of
hos
pita
lan
d cr
ashe
d to
gro
und.
NTSB
pre
limin
ary.
Powe
r los
s sh
ortly
afte
r tak
eoff
nece
ssita
ting
emer
genc
y lan
ding
from
alti
tude
of a
ppro
x. 3
00 ft
. In
itial
retu
rn fl
ight
abo
rted
due
to w
eath
er.
Atte
mpt
ed to
retu
rn to
bas
e th
e fo
llowi
ng d
ay w
hen
cras
h oc
curre
d.NT
SB p
relim
inar
y.
AC re
turn
ing
to b
ase
and
collid
ed w
ith th
e su
rface
of l
ake.
NTSB
pre
limin
ary.
Pilo
t inf
orm
ed d
ispat
cher
of a
“ped
al bi
ndin
g in
the
right
ped
al.”
At 5
0 to
100
feet
AGL
airc
raft
“sta
rted
spin
ning
and
des
cend
ing
until
the
nose
dro
pped
and
the
helic
opte
r im
pact
ed th
e te
rrain
.”NT
SB p
relim
inar
y.
En ro
ute
to a
ccid
ent s
cene
. W
itnes
ses
repo
rted
aircr
aft w
as
“flyin
g lo
w an
d ve
ry fa
st” a
nd im
pact
ed th
e gr
ound
nos
e-lo
w.Po
st-im
pact
fire
occ
urre
d. N
o di
stre
ss c
all o
r pro
blem
s re
porte
d.NT
SB p
relim
inar
y.
Airc
raft
was
enro
ute
to S
ioux
Fall
s, S
D. L
ast p
ositi
on re
port
was
at 2
158.
Airc
raft
was
foun
d by
farm
ers
in a
soy
bean
field
at
appr
oxim
ately
093
0 (9
/10/
02),
not f
ar fr
om la
st p
ositi
on re
port.
Ther
e wa
s no
dist
ress
call
. NT
SB p
relim
inar
y.
Mis
sion
Typ
e M
issi
on P
hase
Ph
ase
of F
light
In
jurie
s Fi
re
FUL
Refu
eling
PRPu
blic
Relat
ions
PGT
Goin
g to
get
the
patie
ntAP
PAp
proa
chDE
CDe
scen
tM
ANM
aneu
verin
gN
None
INF
In F
light
INT
Inte
rhos
pita
lTN
GTr
ainin
gPO
BPa
tient
on
Boar
d CL
BCl
imb
HOV
Hove
rTK
FTa
keof
fM
Min
orNo
No F
ireM
AT
Main
tena
nce
OTH
Othe
rPR
FRe
turn
ing
from
pat
ient
CRU
Crui
se
LNG
Land
ing
TXG
Taxin
gS
Serio
usPI
FPo
st-Im
pact
Fire
ONS
Onsc
ene
FFa
tal
UNK
Unkn
own
Sour
ces:
LeR
oy J
acks
on,
Air
Met
hods
; C
onni
e Sc
hnei
der,
Fitc
h an
d A
ssoc
iate
s; C
ON
CE
RN
Net
wor
k; N
TSB
Avi
atio
n A
ccid
ent S
ynop
ses,
<ht
tp:/
/ww
w.n
tsb.
gov>
;an
d Pr
elim
inar
y H
elic
opte
r A
ccid
ent R
epor
ts,
<ht
tp;/
/saf
ecop
ter.
arc.
nasa
.gov
>.
Key:
61Special Supplement
1. “1997 Population Profile of the United States.” U.S. Department ofCommerce Economics and Statistics Administration, Bureau of theCensus. Current Population Reports, Special Studies P23-194.<http://www.census.gov/prod/3/98pubs/p23-194.pdf>
2. “1999 Annual Report File and 2000 Early Assessment Files.” NationalCenter for Statistics and Analysis, U.S. Department ofTransportation, National Highway Traffic Safety Administration.<http://www.nhtsa.dot.gov/people/ncsa/pdf/2Kassess_rev3.pdf>
3. “1999 Nall Report, Accident Trends and Factors for 1998.” Air SafetyFoundation, Aircraft Owners and Pilots Association (AOPA),Frederick, Maryland.<http://www.aopa.org/asf/publications/99nall.html>
4. Abrahamson, D. “Risk Takers and Thrill Seekers.” InternetInvestment Education.<http://www.netinvestedu.com/columns/insurance.html>
5. “Aeronautical Decision Making.” FAA Advisory Circular AC No: 60-22, December 13, 1991. <http://ntl.bts.gov/DOCS/Ac60-22.html>
6. “Air Medical Accident Analysis: Final Report.” Air Medical ServiceAccident Analysis Team. Richard M. Wright, Jr., Chair. HelicopterAssociation International, Alexandria, VA, September, 2001.
7. “Accident Synopsis.” National Transportation Safety Board.<http://www.ntsb.gov/NTSB/Month.asp>
8. Aircraft Owners and Pilots Association - AOPA Online.<http://www.aopa.org>
9. “Annual Transport Statistics.” Hospital Aviation. 8 (3): 1989.
10. “Annual Transport Statistics.” The Journal of Air Medical Transport.9 (3): 1990.
11. “Annual Transport Statistics.” The Journal of Air Medical Transport.10 (3): 1991.
12. “Aviation Accident Statistics.” National Transportation SafetyBoard. <http://www.ntsb.gov/aviation/Stats.htm>
13. “Aviation Safety Summary FY 2001.” United States Department ofAgriculture, Forest Service<http://www.aviation.fs.fed.us/library/fy01avsumm.pdf>
14. Aviation Safety Reporting System (ASRS).<http://asrs.arc.nasa.gov>
15. Asimos, A., et al. “Tuberculosis Exposure Risk in EmergencyMedicine Residents.” Academic Emergency Medicine, pp. 1044-1049:1999.
16. Avala, J. Crash Risk for Fire Fighting Airplanes. NBC Nightly News,August 27, 2002.
17. Aviation Safety Reporting System. “ASRS Rotorcraft Incident Study,Data Summary.” Battelle, ASRS, Mountain View, CA: 1999.
18. Bottini, E., et al. “Incidence and Nature of the Most Common RugbyInjuries Sustained in Argentina.” British Journal of Sports Medicine,34 (2): 2000.
19. Brennan, TA., Leape LL, Laird, NM., et al. “Incidence of adverseevents and negligence in hospitalized patients: Results of theHarvard Medical Practice Study I.” N Engl J Med. 324:370-376,1991.
20. Collett, H. M. “Aeromedical Accident Trends.” Hospital Aviation, 7(2): 1988.
21. Collett, H. M. “Air Medical Helicopter Transport.” Hospital Aviation,7 (7): 1988.
22. Collett, H. M. “Accident Trends for Air Medical Helicopters.”Hospital Aviation, 8 (2): 1989.
23. Collett, H. M. “Mid-Year Report.” Hospital Aviation, 8 (7): 1989.
24. Collett, H. M. “1989 Accident Review.” The Journal of Air MedicalTransport, 10 (2): 1991.
25. Collett, H. M. “Air Medical Accident Rates.” The Journal of AirMedical Transport. 10 (2): 1991.
26. Collett, H. M. “Mid-Year Report.” The Journal of Air MedicalTransport, 10 (7): 1991.
27. Collett, H. and Mikat, A. “Two Engines Versus One.” AirMed, 3 (2):1997.
28. Commission on Accreditation of Medical Transport Systems.Accreditation Standards, 5th Edition. Anderson SC, January, 2002.
29. CONCERN Network. The Air & Surface Transport NursesAssociation (formerly the National Flight Nurses Association).<http://www.astna.org/concern.html>
30. Conn, L. M., & Lion, J.R. “Assaults in a University Hospital.”Philadelphia, PA: W. B. Saunders & Co., 1983.
31. Connell, L. J. and Patten, M. “Emergency 911 - EMS HelicopterOperations.” ASRS Directline. No. 6: 1993.<http://asrs.arc.nasa.gov/directline_issues/dl6_ems.htm>
32. Connell, L. J. and Reynard, W. D. NASA-Ames Research CenterMoffett Field, California, U.S. “Incident Reports Highlight Hazardsin EMS Helicopter Operations.” Helicopter Safety, Flight SafetyFoundation, Alexandria, Virginia, 21 (4): 1995.
33. Connell, L. J. and Reynard, W. D. “Emergency Medical ServicesHelicopter Incidents Reported to the Aviation Safety ReportingSystem.” NASA-Ames Research Center Moffett Field, California,U.S. April, 1993.
34. Davis, R. “Speeding to the Rescue Can Have Deadly Results.” USATODAY, April 4, 2002.<http://www.usatoday.com/news/nation/2002/03/21/usat-ambulance(acov).htm>
35. Dodd, R. “Factors Related to Occupant Crash Survival in EmergencyMedical Service Helicopters.” Unpublished doctoral dissertation,Johns Hopkins University, Baltimore, Maryland, U.S. 1992.
36. “Directory of Air Medical Programs.” AirMed, 6 (3): 2000.
37. Doyle TJ, Delbridge TR, Cole JS, Nicholas DH. “An Analysis ofOccupational Injuries in Air Medical Transport.” STAT MedEvacand the University of Pittsburgh, Pittsburgh, PA. Presented at the2002 Critical Care Transport Medicine Conference, Las Vegas,April, 2002.
38. “Drowsy Driving and Automobile Crashes.” National Center onSleep Disorders Research (NCSDR) and the National HighwayTraffic Safety Administration (NHTSA) Expert Panel on DriverFatigue and Sleepiness.<http://www.nhtsa.dot.gov/people/injury/drowsy_driving1/drowsy.html>
39. Elling, R. “Dispelling Myths on Ambulance Accidents.” JEMS, July,1989.<http://www.medicalpriority.com/articles/dispellingmyths1.htm>
40. Erickson, L., & Williams-Evans, S. (2000). Attitudes of emergencynurses regarding patient assaults. Journal of Emergency Nursing, 26(3): 2000.
BIBLIOGRAPHY
62 AIR MEDICAL PHYSICIAN HANDBOOK
41. FAR Part 1 — Definitions and Abbreviations. United StatesDepartment of Transportation, Title 14 of the Code of FederalRegulations (CFR), Federal Register, February 18, 1998.<http://www.faa.gov/avr/AFS/FARS/far-1.txt>
42. FAR Part 61 — Certification: Pilots, Flight Instructors, and GroundInstructors. United States Department of Transportation, Title 14 ofthe Code of Federal Regulations (CFR), Federal Register, October 10,2000. <http://www.faa.gov/avr/AFS/FARS/far-61.txt>
43. FAR Part 91 - “General Operating and Flight Rules.” United StatesDepartment of Transportation, Title 14 of the Code of FederalRegulations (CFR), Federal Register, Sept. 29, 2000.<http://www.faa.gov/avr/AFS/FARS/far-91.txt>
44. FAR Part 121 - “Operating Requirements: Domestic, Flag, andSupplemental Operations.” United States Department ofTransportation, Title 14 of the Code of Federal Regulations (CFR),Federal Register, October 10, 2000.<http://www.faa.gov/avr/AFS/FARS/far-121.txt>
45. FAR Part 135 - “Operating Requirements: Commuter and On-Demand Operations and Rules Governing Persons on Board SuchAircraft.” United States Department of Transportation, Title 14 ofthe Code of Federal Regulations (CFR), Federal Register, Sept. 29,2000. <http://www.faa.gov/avr/AFS/FARS/far-135.txt>
46. “Fatal occupational injuries by event or exposure, 1996-2001,” U.S.Department of Labor, Bureau of Labor Statistics.<http://www.bls.gov/news.release/cfoi.t03.htm>
47. Federal Aviation Administration. <http://www.faa.gov>
48. Flight Safety Foundation. <http://www.flightsafety.org>
49. FlightWeb For Air Medical Professionals. <http://www.flightweb.com>
50. Frazer, R. “Air Medical Accidents: A 20-Year Search forInformation.” AirMed, 5(5) September/October 1999.
51. Frazer, R. S. “Weather Accidents and the Air Medical Industry.”AirMed, 6(3) May/June 2000.
52. Frazer, R. S. “Air Medical Accidents Involving Collisions withObjects”, AirMed, 20(3) May/June 2001.
53. Frazer, R. S. “Air Medical Accidents Attributed to Maintenance”,AirMed, 21(3) May/June 2002.
54. Gold D., et al.: “Rotating shift work, sleep, and accidents related tosleepiness in hospital nurses.” Am J Public Health. 82(7):1011-4; 1992.
55. Goodman, R.; Jenkins, L; and Mercy, J. “Workplace-RelatedHomicide Among Health Care Workers in the United States, 1980through 1990.” JAMA 272(21): 1686-1688; 1994.
56. Greenfeld, K. T. “Life on the Edge”. Time Magazine, September 6,1999.
57. Gronczewski, C. A. “Axiom: Finished a Night Shift? Always Take aBrief Nap Before Driving.” Emergency Medicine News, 23(1) February,2001.
58. “Guidelines for Security and Safety of Health Care and CommunityService Workers,” State of California OSHA, March 10, 1998.<<http://www.dir.ca.gov/DOSH/dosh_publications/hcworker.html>
59. “Guidelines for Preventing Workplace Violence for Health Care andSocial Service Workers,” U.S. Department of Labor, OccupationalSafety and Health Administration, OSHA 3148, 1998.<<http://www.osha-slc.gov/SLTC/workplaceviolence /guideline.html>
60. Harris, F.D., Iseler, L. and Kasper, E. “U.S. Civil RotorcraftAccidents, 1963 through 1997.” NASA/TM-2000-209597,USAAMCOM-TR-00-A-006<http://safecopter.arc.nasa.gov/Pages/AccidentUpdates/NASA%20TP%20209597.pdf>
61. Harris, J. S. “Crew Resource Management Applies To Single-pilotFlight Operations.” Helicopter Safety, Flight Safety Foundation,Alexandria, VA, 21(5) September/October, 1995.
62. Harris, J. S. “Reports Show Pilot Error as the Major Cause ofHelicopter Accidents in U.S. On-demand Operations.” HelicopterSafety, Flight Safety Foundation, Alexandria, VA, 25(5)September/October, 1999.
63. Harris, J. S. “Data Show Same U.S. Fatal-accident Rate for Single-turbine and Twin-turbine Helicopters.” Helicopter Safety, Flight SafetyFoundation, Alexandria, VA, 25(1) January/February, 1999.
64. Harris, J. S. “Every Helicopter Pilot Must Be Prepared forInadvertent Entry into Instrument Meteorological Conditions.”Helicopter Safety, Flight Safety Foundation, Alexandria, VA, 22(2)March/April, 1996.
65. Harris, J. S. “For Helicopter Pilots, Managing Stress is part of FlyingSafely.” Helicopter Safety, Flight Safety Foundation, Alexandria, VA,21(1) January/February, 1995.
66. Harris, J. S. “U.S. Hospital-based EMS Helicopter Accident RateDeclines Over the Most Recent Seven-year Period.” Helicopter Safety,Flight Safety Foundation, Alexandria, VA, 20(4) July/August, 1994.
67. Hart, S. G., “EMS Accident and Incident Trends,” NASA-AmesResearch Center, Moffett Field, CA. Presented at the AMTCConference, Salt Lake City, UT, 2000.
68. Hart, S. G., “Civil Medevac Accidents,” NASA-Ames ResearchCenter, Moffett Field, CA. Presented at the Proceeding of the 11thInternational Symposium on Aviation Psychology, Columbus, OH.March, 2001. In Press, Ohio State University, 2001.
69. “Helicopter Accident Analysis Team . Final Report, July 1998.”NASA-Ames Research Center, Moffett Field, CA., 1998.
70. “Helicopter Accident Database and Statistics.” HelicopterAssociation International, Alexandria, VA.<http://www.rotor.com/safety/stat.htm>
71. “Helicopter Accident Statistics Year to Date: 2000.” NASA CivilHelicopter Safety. <http://safecopter.arc.nasa.gov/pages/accidentupdates/accstatsD.html>
72. Helicopter Association International. <http://www.rotor.com>
73. “Injury Facts.” 1999 Edition; 2000 Edition; and 2001 Edition.National Safety Council. Itasca, IL, 1999, 2000, 2001.
74. Jackson, L. Accident/Incident Database, 1978-2000. (unpublished)
75. Koenig, Robert L. “FAA Identifies CRM-related Issues and TrainingNeeds in Flight-inspection Missions.” Human Factors & AviationMedicine, Flight Safety Foundation, Alexandria, VA, 44(1)January/February, 1997.
76. Landsberg, B. “ The Sterile Cockpit.” Air Safety Foundation, AOPA.AOPA Pilot Magazine, June, 2000. <http://www.aopa.org/asf/asfarticles/sp0006.html>
77. Landsberg, B. “The ‘98 Nall Report.” Air Safety Foundation, AOPA.AOPA Pilot Magazine, December, 1998.<http://www.aopa.org/asf/asfarticles/sp9812.html>
78. Laudan, L. “The Book of Risks: Fascinating Facts About ChancesWe Take Every Day.” John Wiley and Sons, Inc., NY. 1994.
79. Lavoie, F., Carter, G. L., Denzel, D. F., & Berg, R. L. “EmergencyDepartment Violence in United States Teaching Hospitals.” Annals ofEmergency Medicine, 17(11), 1227-1233; 1988.
80. MacDonald, E. “The Next Accident” Air Medical Journal. 20(5),September-October, 2001.
81. Mayfield, T. “1993 Annual Transport Statistics and TransportCharges Survey.” Air Medical Journal. 12(10) October, 1993.
63Special Supplement
82. Mayfield, T. “1994 Annual Transport Statistics and Transport FeesSurvey.” Air Medical Journal. 13(4) April, 1994.
83. NASA’s Civil Helicopter Safety. <http://safecopter.arc.nasa.gov>
84. National Highway Traffic Safety Administration (NHTSA).<http://www.nhtsa.gov>
85. National Safety Council. <http://www.nsc.org>
86. National Transportation Safety Board, Aviation Accident Statistics.<http://www.ntsb.gov/aviation/Stats.htm>
87. “NEMSPA/HAI Joint EMS Line Pilot Survey 2001.” National EMSPilots Association, 2001.<http://www.nemspa.org/EMS%20Line%20Pilot%20Survey%202001.html>
88. Novak R, Auvil-Novak S: “Focus group evaluation of night nurseshiftwork difficulties and coping strategies.” ChronobiologyInternational 1996 13(6):457-463.
89. “Pilot’s Spatial Disorientation.” FAA Advisory Circular AC No: 60-4, February 9, 1983. <http://www.faa.gov/avr/afs/acs/60-4a.pdf>
90. “Preliminary Helicopter Accident Reports,” NASA Civil HelicopterSafety.<http://safecopter.arc.nasa.gov/Pages/AccidentUpdates/AccSum00.html>
91. “Preventing Needlestick Injuries in Healthcare Settings.” NationalInstitute for Occupational Safety and Health (NIOSH), Cincinnati,Ohio. Publication No. 2000-108, November, 1999. <<http://www.cdc.gov/niosh/2000-108.html>
92. Rau, W. “The 1999 Annual Transport Statistics and Fees Survey.”AirMed, 5(4) July/August, 1999.
93. Rau, W. “2000 Annual Transport Statistics and Fees Survey.” AirMed,5(4) July/August, 2000.
94. Rau, W. “2000 Aircrew Survey.” AirMed, 6(6) November/December,2000.
95. Rhee, K.J., Holmes, E.M., Moecke, H., Thomas, F.: “A Comparison ofEmergency Medical Helicopter Accident Rates in the United Statesand the Federal Republic of Germany” Aviation, Space &Environmental Medicine, 61(8) August 1990.
96. “Safety Facts: An Examination of Simulator-based Training.”FlightSafety International, West Palm Beach, Florida, 1999.
97. “Safety Reports. Aviation Safety Data Accessibility Study Index:Analysis and Interpretation of Safety Data,” January 20, 1997Prepared for the Office of System Safety, Federal AviationAdministration.<http://nasdac.faa.gov/safety_info_study/analysis.htm>
98. “Safety Statistics and Review,” Safety Pilot, AOPA Online, Air SafetyFoundation, AOPA. <http://www.aopa.org/asf/subjectidx.html#safety>
99. Safety Summit - Meeting Minutes, Dallas, Texas, April, 2000<http://www.aams.org/safetysummitmin000407.html>
100. Schneider, C. 1998-2000 Accidents and Incidents Summary.(unpublished)
101. Shapiro, M., et al. “The Perils of Bungee Jumping.” The Journal ofEmergency Medicine, 13(5) 1995.
102. Steele MT, Ma OJ, et al. “The Occupational Risk of Motor VehicleCollisions for Emergency Medicine Residents.” Acad Emerg Med;6(10), 1050-1053; 1999.
103. Stultz, MS. “Crime in hospitals 1986-1991: The Latest IAHSSSurveys.” Journal of Healthcare Protection Management, 9(2): 1993.
104. Thomas, EJ, Studdert DM, Newhouse JP., et al. “Costs of MedicalInjuries in Utah and Colorado.” Inquiry. 36:255-264, 1999.
105. Title 14 of the Code of Federal Regulations (CFR). United StatesDepartment of Transportation, Federal Aviation Regulations.<http://www.faa.gov/avr/AFS/FARS/far_idx.htm>
106. “To Err is Human: Building a Safer Health System.” Kohn LT,Corrigan JM, and Donaldson MS, (Ed); Committee on Quality ofHealth Care in America, Institute of Medicine, National AcademyPress, 2000. <http://www.nap.edu/books/0309068371/html/>
107. “Traffic Safety Facts 1999,” U.S. Department of TransportationNational Highway Traffic Safety Administration. DOT HS 809 086.<http://www.nhtsa.dot.gov/people/ncsa/pdf/Alcohol99.pdf>
108. “U.S. Civil Helicopter Safety Statistics: Summary Report (1970 -2001).” Helicopter Association International, Alexandria, Virginia.<http://www.rotor.com/safety/stat70.01.xls>
109. U.S. National Transportation Safety Board. “Safety Study:Commercial Emergency Medical Service Helicopter Operations.Report No. NTSB/SS-88/01.” 1988.
110. U.S. National Transportation Safety Board: “Notification andReporting of Aircraft Accidents or Incidents and Overdue Aircraft,and Preservation of Aircraft Wreckage, Mail, Cargo, and Records.”Code of Federal Regulations, Cite: 49CFR830.2. Revised October 1,2000. <http://frwebgate.access.gpo.gov/cgi-bin/getcfr.cgi?TITLE=49&PART=830&SECTION=2&YEAR=2000&TYPE=TEXT>
111. Veillette, Patrick R. “Human Error Cited as Major Cause of U.S.Commercial EMS Helicopter Accidents.” Flight Safety Digest, FlightSafety Foundation, Alexandria, Virginia, April/May, 2001.
112. “Weather-Related Accidents: How Cockpit Resource Managementand Aeronautical Decision Making Could Make a Difference.”FlightSafety International.
113. “What Are the Odds of Dying.” NSC Statistics. National SafetyCouncil, 2001 <http://www.nsc.org/lrs/statinfo/odds.htm>
64 AIR MEDICAL PHYSICIAN HANDBOOK
Dr. Blumen kindly asked me if I wouldcomment on this penetrating and in-depth report. As someone who has neverbeen shy about offering an opinion aboutEMS safety, I welcomed this opportunityto contribute as one who has loved thisavocation for many years. First, I wouldlike to commend Ira and the UCANSafety Committee for this thorough andcomprehensive labor of love. Much timeand energy has gone into this effort and Ihope it may change the way we look atEMS aviation safety evermore.
“Consensus is that the current EMShelicopter safety record is appalling.Unfortunately, that is the only consensusamong the various participants. There isnot even agreement about what consti-tutes an EMS accident. In 1986, forexample, one can cite from reputablesources that there were 16 EMS accidents(Hospital Aviation Magazine), or 22(American Medical News), or 28 as report-ed by CBS television’s Sixty Minutes orHelicopter News, or 14 or 17, as reportedby the FAA in two separate forums.Similarly, the number of EMS helicopterfatalities has been variously reported as13, 15, 19, 21, and 25.”1 This is from anarticle by Ira Rimson, written in theNational EMS Pilots Association’s AirNet in 1987. Reporting methodologycontinues as a problem for statisticalpurists even today. We do, however, con-tinue to see the same type of accidentswith the same persistent causes as yearspast. Whether or not statistics indicate atrend or a community-wide problem doesnot override the fact that one (preventa-ble) accident is too many.
Although this report attempts to col-late all available information concerningEMS Aviation Safety or occasional lackthereof, quantifying EMS aviation opera-tions and accidents has been imprecise,to say the least. Although the numbersand underlying causes have been a chal-lenge to decipher due to inconsistentmethods for classifying EMS accidents,incidents, forced, precautionary, and hardlandings, we still can draw valid andhelpful information as to the risksinvolved. The statistics at a macro level
are important—but management of therisks happens at a local program and per-sonal level. It is important not to get toowrapped around global statistics and getmore concerned about managing safety atthe level we can truly affect. As A. E.Housman states, “Statistics in the handsof an engineer are like a lamppost to adrunk—they’re used more for supportthan illumination.” I would recommendthat you take all of these studies and boilthem down to something usable at yourbase. A local risk assessment, based onthe hard, cold lessons of others, may helpyou look more honestly at yourself. Ifyou derive nothing worthwhile from thisentire study or believe that it only hap-pens to someone else, you may wish tolook into the mirror for the problem.
Probably the first civilian EMS acci-dent in this country occurred in May1978 in an Alouette and its cause wasreported as mechanical and engine-relat-ed. That was not, however, where anearly underlying cause for EMS accidentswas born. If we only look into accidentsfrom the civilian sector of EMS we willmiss some valuable clues to some verybasic risks and causes of EMS accidents.Critical patients were being flown longbefore the first hospital-based programbegan in Colorado in 1972. As early as1970, military air ambulance organiza-tions returning from Viet Nam werebeginning the Military Assistance toSafety and Traffic program (MAST).Helicopters had been extensively used totransport the wounded, sick, and injuredin the Korean and Viet Nam wars and asfar back as Burma.
The MAST program utilized militaryaeromedical assets and personnel to pro-vide much needed assistance to the civil-ian sector. Military air ambulances beganto provide emergency transportation ofsick and injured from the streets andhighways in several states. By the lateseventies, more than two dozen locationsthroughout the nation were served bymilitary emergency aviation services. Theentire EMS arena in the United Stateswas beginning its transition into thesophisticated and responsive systems we
see today. The MAST program was acritical link in the system’s genesis as wellas a model for EMS aviation in the civil-ian sector. The MAST program also suf-fered from some of the similar stressorsthat have carried over to today’s EMSaviation cultures. Pilots who risked theirlives evacuating their fellow soldiers fromthe mountains and rice paddies of VietNam were now being asked to do thesame for their fellow citizens on thefarms, highways, and mountains of theUnited States. This ingrained noble andvery human drive, which perhaps couldbe termed “the rescuer ethic,” still com-pels much of what we do and feel today.
Having served in this avocation asboth a military and civilian EMS pilot, itis a pressure that one feels every time thetones go off. It is why many of us do thiswork. It is the adrenalin jolt that drivesus to do great things. It drives us intotaking risks. It occasionally pushes us totake foolish risks and go beyond our bestjudgment. Sometimes that risk-takingpushes us beyond recovery. There is atime for heroes, but whether or not thatis a risk we can really accept needs to beaddressed directly with each member ofan air medical team, management, andthe communities we serve.
This underlying ethic is one commonto the dispatchers, pilots, nurses, para-medics, physicians, program managers,policemen, firemen, and almost everyoneinvolved in the emergency and publicsafety field. To deny that the “rescuer orhero mentality” exists is to somehowdeny who we are or who we want to be.It is also something that rarely appears onan EMS aircraft accident report. It is dif-ficult to quantify and a bigger challengeto manage. It is the “elephant in theroad” that no one readily admits to. It isa very real characteristic that drives us todo what we do—yet is seldom addressedas a core value or accident cause.
Every time I read an accident reportwhere the pilot and crew perish in ablinding snowstorm, foggy meadow, driv-ing thunderstorm, or pitch-black hillsideI always ask, “What were they doingthere?” “Why didn’t they turn back?”
RISK MANAGEMENT, BENEFITS AND COST
EDITORIAL COMMENTS
65Special Supplement–Editorial Comments
“Why did they go at all?” Often as not,the team is on their way home or trans-ferring a relatively stable patient whomight as easily been transferred byground ambulance or waited a few hoursin their hospital room. Combine the heromentality with get-homeitis and we have adeadly mix. Look at the statisticsthroughout this report and you will seesome of the catastrophic results of pilotsand crews unnecessarily flying into dete-riorating weather conditions or pushingwell beyond their personal limits. Mosthelicopter operators train their pilots tohave a corporate, professional mentality.That is a good start. We cannot denythat there are many pressures on an EMSpilot that bring the “rescuer mentality”to the surface. How we as an EMS com-munity deal with this intense drive tohelp others is a topic that the EMS avia-tion community must openly and regular-ly discuss. A written policy is just eye-wash and falls far short of effective riskmanagement.
If we look at NTSB accident reportsand accept them at face value, we will belimited in truly understanding the corevalues that really created the accidents.In those reports we will find that “thepilot flew into a hill, hit wires or build-ings, or disregarded regulations or compa-ny policies and procedures.” Those arereal causal elements—but simply do notgo deep enough. We must understandthe organizational culture and personalvalues that drove a pilot into a situationthat was needless and preventable.
Human factors remain as the numberone cause of EMS accidents. “A study of87 accidents form 1987 through 2000found that human error was the primarycausal factor in 76 percent. The greatestconcentration of human error occurredin the en-route phase of flight and ofteninvolved faulty in-flight planning anddecision making or inadequate evalua-tion of weather information.”2
In his 2001 study, Pat Veillette foundthat “forty-one of the 87 accidents(47%), including 26 of the 32 fatal acci-dents (81%) occurred during the en-route phase of flight. Of the en-routeaccidents, 68% resulted from humanerror.”3 “Twenty-six percent of the acci-dents—and 53 percent of the fatal acci-dents—occurred in low-visibility orinstrument meteorological conditions.”4
In the EMS Line Pilot survey conduct-ed by NEMSPA and HAI in 2000-2001,pilots who flew Emergency MedicalService helicopters reported that man-agement or crew pressure was most oftenthe greatest risk to safety as a whole inthe industry. It was closely followed bynight operations, inexperience, weatherreporting or forecasting, and poor deci-sion making. When asked about factorsthat have contributed to the rise in EMSaccidents, pilots responded with “pushingweather minimums, complacency, lack ofIFR training, and inexperience.”5
EMS pilots are routinely called upon tolaunch on a moment’s notice, day ornight, 24/7, to unprepared landing zoneswith marginal weather reporting for theirdestination(s) and routes. This substan-tially raises the risk over that of the aver-age commercial pilot who is able to flyfrom approved airport to approved air-port with adequate planning time andofficial weather reporting. Throw in apalatable sense of urgency and the stageis set.
The Air Medical Accident Analysisreport, conducted by the subcommitteeresulting from the April 2000 AirMedical Summit, performed a thoroughstudy of 20 air medical accidents thatoccurred from 1993 through 2000. Asbackground, this report stated: “Between1987 and 1997, there were on averagefour air medical helicopter accidents peryear for the industry. By 1997, the acci-dent rate for AMS (Air Medical Service)operations had been reduced to 1.97accidents per 100,000 flight hours from ahigh of 17.08 in 1987. In 1998, however,the number of accidents rose to a nineyear high of seven, but more alarming,was the rise of fatalities to fourteen, thehighest number since the peak year of1986. In 1999, the number of accidentsrose even higher to ten, the highest alsosince the peak year of 1986. Fatalitieswere down to ten but still higher thanthe average of six.”6
We have discussed statistics and safetyrisks throughout Dr. Blumen’s document.I would like to move our focus to thesolutions. Some of these are technologi-cal and institutional. I would submit,however, that the real answers are instrong personal and organizational safetycultures enforced by proactive and awaremanagement. Levelheaded professionals
must replace risk takers and adrenalinjunkies. Pilots and crewmembers whoplace safety values below that of personalthrill seeking or a mistaken sense of hero-ics must change their spots or find newprofessions. Managers, hospital adminis-trators, accountants, and program direc-tors must insure that their pilots andflight teams have proper tools and facili-ties to do their jobs safely.
Technology will provide stronger, moredependable, and ergonomically friendlyequipment, both in the aircraft that wefly and in the gadgets associated withflight following, air traffic control, terrainavoidance, GPS, night vision equipment,avionics, flight instrumentation, andcontrols. Many are readily availabletoday. Many flight programs today areusing 30-year-old technology, underpow-ered, or marginally safe aircraft, andexpecting their pilot and flight teams tomake due. Some programs continue toask their pilots to fly multiple and dissim-ilar airframes on a routine basis. Someprograms utilize a spare aircraft that isdissimilar or inadequate for the mission.There is a managerial blind eye to therisks that those cost-saving measures cre-ate. Often budget constraints and apolitically driven decision processexclude the pilot effectively from the air-craft selection process. Occasionally, theRFP process creates a situation wherecosts take priority over safety. Medicalpersonnel often have the final word inselecting the aircraft and often do sobased on medical needs with tokenregard to the most important tool in theprocess—the aircraft. There are manyaircraft in use today that are missingwhat I would consider critical elementsfor an optimum EMS helicopter.
These critical aircraft requirements, inmy opinion, are one single type aircraftwith an adequate margin of power andperformance to do any of the missions aprogram requires in its area of operations.This should account for weather, terrain,and all environmental factors. The air-craft should have adequate avionics,lighting, and safety features. It shouldhave sufficient space, efficient medicalconfiguration, and ergonomics to safelyand efficiently treat the type and numberof patients to be flown. A single type ofhelicopter model with similar ergonomicsand systems in the program aircraft
66 AIR MEDICAL PHYSICIAN HANDBOOK
reduces risks and maintenance complica-tions as well. Dissimilar aircraft createanother obvious risk ignored by manyprograms and operators. Perhaps there isa sound risk management reason why aSouthwest, United, or Delta pilotremains solely on one airframe, type, andmodel. Aside from the obvious training,standardization, and maintenance advan-tages, there are very valid and oftenignored risk management reasons to keeppilots in only one aircraft type andmodel. If you’ve ever rented a car differ-ent than the one you drive at home andsearched for the parking brake release orwindshield wiper button in a dark park-ing lot or rainy freeway, you shouldunderstand the term “negative habittransfer.” One question is whether or notthinly stretched community, hospital, orprogram director’s budgets are willing andable to afford that in the future—or evenin the present.
The other question is whether or notoperators or programs are really willing totake all of the steps necessary to aggres-sively manage risks. Hospital and corpo-rate CFOs who approve budgets must notbe lured into false economies that elevaterisks. Insurance companies and govern-ment agencies are now, or will be, exert-ing their influence on unsafe, ineffectiveor redundant programs. The MBA men-talities who believe profits and costs arethe only measures of good business mustadd safety as an equal partner to theirthought process. If we want to make adifference in our day-to-day risk manage-ment, we must take a hard look at how“that’s the way we’ve always done it”affects us today.
The Air Medical Service SafetySummit’s Air Medical Accident AnalysisFinal Report7 concluded its study of 20recent air medical accidents with the fol-lowing interventions that rated high inboth effectiveness and feasibility. Theywere:
• Enhance the training for night flyingoperations
• Enhance the training for mountainflying operations
• Equip aircraft with TerrainAvoidance Warning Systems(TAWS)
• Equip aircraft with Radar Altimeters• Provide aircraft with mission-essen-
tial equipment
• Improve the content of weatherbriefings
The top six in the high effectivenessand moderate feasibility were:
• Conduct/enhance annual IFR profi-ciency checks
• Conduct/enhance training toimprove understanding of weatherbriefings
• Enhance overall training in recur-rent, professional knowledge, etc.
• Conduct/enhance training inAeronautical Decision Making(ADM)
• Establish integrated and structuredPilot Training Programs
• Conduct/enhance mission-orientedtraining
This report was distributed to the AirMedical Services subcommittee of HAI,the Air Medical Safety AdvisoryCommittee (AMSAC), and theAAMS/CORE Safety Committee fortheir review and suggestions. A majorkey to the training issue is that it shouldbe mission oriented. If the pilot isexpected to find and land in an LZ in themountains on a pitch-black foggy nightor land in a dusty or snowy LZ, regularand recurring training should meet thatrequirement. Unfortunately, many opera-tors routinely train on safe and sterilerunways or helipads. If we are to trulylower our risks, we must train in the sameenvironment and mission conditions wewill encounter. Instrument trainingshould involve real inadvertent IMC sit-uations under real mission profiles. In thereal EMS world, this means that trainingdollars must come from vendors and pro-grams alike to better manage risks. Wehave some real solutions. We must havethe will as well.
When the EMS line pilots were askedfor their suggestions to improve safety inthe NEMSPA line pilot survey, the topvote getter was, “Increase quality and fre-quency of training.” This was followingclosely by “Improve pilots’ salaries andbenefits,” and “Night Vision Goggles.”We need to stop ignoring the hard, coldfact that our pilots cannot see like bats inthe dark. We have been pretendingsomehow for years that once you are anEMS pilot, you become magicallyendowed with built-in sonar and night
vision skills. Pilots and crews must learnto say “no” when asked to perform out-side of their limits. The right equipmentincludes things like night-vision devices,night suns, and skid lighting. If youdon’t have adequate lighting or NVGassistance, don’t use your rotor blades ascurb feelers as you plough through a darknight.
Another ominous finding in theNEMSPA/HAI survey was that over 25%of EMS pilots either had not receivedany crew resource management trainingor they felt it was ineffective or not wellpresented8. When asked about the effec-tiveness of their training or preparationfor their present position, significantnumbers (>10%) of pilots responded thatthe following areas were weak: Flightcrew dynamics or interaction, CrewResource Management, Aircraft systems,and mission planning.
In response to the need for improvingand standardizing Crew ResourceManagement, the Air Medical SafetyAdvisory Committee (AMSAC) pushedthe development of the Air MedicalResource Management (AMRM) pro-gram. Through the efforts of MichelleNorth, an exportable AMRM packagewas produced and “Train-the-Trainer”sessions initiated. It is the intention ofthe AMSAC that it become available toall programs and that they continuallytrain in this invaluable resource.
We should all take heed of one partic-ular result of the EMS pilot survey. Afteryears of warnings to the EMS aviation com-munity about the need for pilots to makeflight go-no go decisions independent of pres-sure, approximately 20% of pilots surveyedresponded that “occasionally, some flightcrews do pressure a pilot to launch or contin-ue a mission.” Pilots also responded insignificant numbers that they had beenpressured to take flights by managementand that local competition created somepressure to fly. Pilots were also pressuredto speed up launch times creating oppor-tunities to miss critical tasks. Most pro-gram managers as well as dispatchers,pilots, and medical crewmembers mustcontrol their perceived need to hurry upor pressure pilots. It is clear that some donot. Sometimes our most seasoned medicor nurse is also the most adamant aboutthe “need for speed.” Sometimes it is thecrewmember who has watched too many
67Special Supplement––Editorial Comments
“911” TV shows or has a “Rambo” self-image. These human factors are clearlymanageable and these risks avoidablewith effective leadership and personaldiscipline.
The Air Medical Safety AdvisoryCommittee (http://www.amsac.org/) “wasenvisioned to be an operator drivenforum dedicated to the sharing and devel-opment of safety information and initia-tives for the AMS industry.” It is dedicat-ed to seeking solutions to some of theEMS aviation community’s tough issues asthey relate to competition, flight and dutytime, fatigue countermeasures, standard-ized criterion for Air Medical Servicesincident/accident reporting, and otherEMS safety issues. The organization con-tinues today with participation from oper-ators, programs, NASA, FAA, insurance,and air medical professional organizations.Ask your operator if they participate.Unfortunately, a few operators have notthought it worth their time and effort.
The bottom line is that there are nonew accident causes. If we want to fixwhat’s broken at a human level we needonly to look at Pat Veillette’s chart atFigure 1-21 of this report and see whattype of human factors are causing acci-dents. The top three are “risk taking,pre-flight planning, and in-flight decisionmaking.” We have met the enemy and itis us.
In conclusion, NO pilot should everhave to make critical flight decisionsunder the thumbscrew of peer pressure,competition, job security, or self-inflictedsense of urgency. If you want a quickbarometer of your program’s safety cul-ture ask a few key questions such as:
When was the last time someone fromthe hospital, vendor, or corporate man-agement attended a program safety meet-
ing, rode along at 2:00 AM, or simply satdown in the crew lounge and chattedwith the crews about the things that real-ly matter?
When was the last time you did“hands-on” extrication, survival, or crashdrills under realistic mission-orientedprofiles?
What percentage of your annual andrequired training is devoted to the biggestand riskiest tool in your medical kit—theaircraft—or scene safety—or helipad safe-ty—or weather—or Air MedicalResource Management?
Do your aviation maintenance techni-cians have adequate time, facilities,resources, and full support to do theircritical work?
What happens when everybody knowsthat one of the pilots is a cowboy or isconstantly pushing the envelope?
Who picks your monitors, defibrilla-tors, IV kits, or traction splints? Whopicks your aircraft?
What happens if a crewmember, flightcommunicator, manager, or doctor pres-sures a pilot or crew to fly?
How often do you train the folks whoset up your LZs?
What happens if someone says “no” or“let’s go back”?
Can you ask these questions in yourorganization?
When was the last time you did? Didanything change?
This is hardly a comprehensive list.Each of us should continually and hon-estly assess our safety attitudes, values,and culture. We don’t need any moreheroes and monuments to the tragic endof noble intentions. We need enlight-ened managers who will be fiercely inde-pendent and effective when it comes tosafety issues. We need as much priority
on risk management as we have on pub-lic relations, charting, medical training,nursing or paramedic or piloting skills.
How do administrators select their pro-gram directors or lead pilots? What skillsets does one need to be a manager?Must a program director be a nurse? Areleadership and management skills moreimportant than medical, nursing, avia-tion, or technical skills for a manager?How much emphasis is there that leadersthoroughly know and use risk manage-ment principles? Do present-day EMSprogram management courses place ahigh priority on safety and risk manage-ment? How long will we continue to dothings the way “we’ve always donethem”?
We need corporations, programs, andoperators who always put safety firstthrough action not words. We needmanufacturers who will produce aircraftthat can do the job. There must be ade-quate, appropriate, and well-maintainedequipment that can do the mission. Weneed pilots who will participate in theprocess and have the courage to speakout for safety issues and hold theirground. We need safety cultures thatsupport those who say “no” for safety’ssake. We need the will and the courageto change. “There are three kinds of peo-ple: Those who make things happen,those who watch things happen, andthose who ask, “What happened?”–(Casey Stengel)
Ed MacDonaldCo-Chair, AAMS/CORE Safety CommitteeImmediate Past President, National EMS Pilots AssociationSafety Representative, NEMSPA
References:1 Rimson, Ira J., PE. “The EMS Helicopter Accident Syndrome.” Air Net (National Association of EMS Pilots Newsletter), November 1987, Volume 2, Number 12. P.1.2 Veillette, Patrick R., PHD. Flight Safety Digest, Volume 20, Number 4-5, Flight Safety Foundation, April-May 2001. P.1.3 Veillette, Patrick R., PHD. Flight Safety Digest, Volume 20, Number 4-5, Flight Safety Foundation, April-May 2001. P.2.4 Veillette, Patrick R., PHD. Flight Safety Digest, Volume 20, Number 4-5, Flight Safety Foundation, April-May 2001. P.2.5 Joint Helicopter Association International/National EMS Pilots Association Line Pilot Survey, 2001<http://www.nemspa.org/EMS%20Line%20Pilot%20Survey%202001.htm>6 “Air Medical Accident Analysis Final Report” Air Medical Service Safety Summit Sub-committee, Dick Wright, Chairperson, Helicopter AssociationInternational. September 20. 2001.7 “Air Medical Accident Analysis Final Report” Air Medical Service Safety Summit Sub-committee, Dick Wright, Chairperson, Helicopter AssociationInternational. September 20. 2001.8 Joint Helicopter Association International/National EMS Pilots Association Line Pilot Survey, 2001<http://www.nemspa.org/EMS%20Line%20Pilot%20Survey%202001.htm>
68 AIR MEDICAL PHYSICIAN HANDBOOK
References:1 Thomas SH, et. al. Trauma HEMS Medical Transport: Annotated Review of Selected Outcomes-Related Literature, Prehospital Emer Care 2002;6:359-712 Thomas SH, et. al. Nontrauma Helicopter Emergency Medical Services Transport: Annotated Review of Selected Outcomes-Related Literature, Prehospital Emer Care2002;6:242-2553 Gearheart PA, Wuerz R, Localie AR, Cost-effectiveness analysis of helicopter EMS for trauma patients. Ann Emerg Med. 1997;30:500-5064 Teng TO, Five Hundred Life-Saving Interventions and Their Cost Effectiveness. Society for Risk Analysis. 1995;Vol. 15:3
CAN WE HAVE BENEFIT WITHOUT RISK?
We are pleased to contribute to defininga context for the extremely importantwork being presented by Dr. Ira Blumenand his colleagues at the University ofChicago Aeromedical Network (UCAN).
The passion and commitment of practi-tioners in air medicine is legendary.Following from the successive militaryexperience in Korea and then Vietnam,helicopter evacuation of the criticallyinjured, often in the most dangerous andtrying of circumstances, took on nearmythic status. The ethos of “finding away” came home and into the civilianworld of air medicine. Thirty years afterthe first helicopter program began opera-tion at St. Anthony’s, the debates aboutvalue—the interface of benefit, risk, andcost—continues.
In recent months, Thomas and col-leagues in Boston have reviewed and pub-lished annotations of the best empiricalstudies for the use of air medical interven-tion for critically ill and injured patients.1, 2
In addition, recent studies comparing costof intervention3, 4, increased mortality afterprogram closure5, and a cost-benefit analy-sis of air medicine6, 7 increasingly supportthe evidence base for measurable benefitsacross a wide range of disease and injuryprocesses. Marrying the unique technolo-gy of aviation with critical care medicinehas not only improved care for the critical-ly ill and injured, but improved access andequity within the healthcare system. Butat what cost and is the cost worthwhile?In this case, it is not only the cost of carebut also the human cost in lost or impairedlives through aviation accidents.
While the benefits for individualpatients and the wider population aredemonstrable, it is equally important tounderstand safety. The issues of aviationmisadventure coupled with the Institute ofMedicine Report8, noting alarming rates ofmedical misadventure leading to preventa-ble death, must give both patients andproviders great pause in assessing safety.
Safety throughout medicine is of great
concern to individual patients, the public,and providers of care. After a single acci-dent recorded in 1996, the air medicalcommunity over the past five years hasseen an upsurge in the number of aviationaccidents and incidents leading to deathand serious injury. How safe is this enter-prise and do the benefits outweigh therisks? The earliest test of medicine—“first, do no harm”—must be answered.
The short answer is that it is difficult toanswer these questions. We have longassumed that benefits outweigh risks whileeach of us wonders and worries aboutexperiencing an accident firsthand.While the number of accidents hasincreased, it is impossible to understand ifthe actual rate is increasing. To measurerates one must have both a numerator anda denominator—in this case the numberof accidents measured against exposure—the number of flights and the number offlight hours. Sadly, and frustratingly, ithas been nearly impossible to measure and“how safe” remains virtually unknowable.
Competitive pressures between pro-grams, Part 135 Operators, vendors, thelack of central data repository, and thecosts of gathering and analyzing data haveall played a part in the creation of a con-textual black hole as regards the safety ofair medicine. Operators, the FAA,NASA, air medical providers, and insur-ance underwriters have become increas-ingly frustrated with the current lack ofdata. While there have been a number ofinitiatives in the past two years—the cre-ation of the Air Medical Safety AdvisoryCouncil (AMSAC) the ASRS programfrom NASA, the Root Cause Study GroupReport, and the accident database fromHAI—the overall understanding of riskand safety remains limited. Why is thisimportant? Simply that the absence ofgood data and analysis is corrosive onmany fronts from poor regulation—rulesthat do not fix problems, to escalatinginsurance premiums, media alarm, andmost worryingly, to increasing distrust on
the part of the public. The publication of this paper changes
the discussion on all fronts. Until thisreport there has been no real effort to col-lect or examine the underlying data totruly have any understanding of overallsafety and the risks of air medical inter-vention. Understanding risk is essentialfor both providers and patients. There arerisks throughout medicine and in allambulance transport, whether by groundor air. The questions each of us mustanswer in the delivery of any medicalintervention and therapy are:
Do we understand the risk and have wetaken every step to minimize the risk?, and
Do the benefits outweigh the risks?This paper by Blumen, et al., is a huge
step forward in answering these questions.While the gathered research is still limitedby the lack of a central repository, thegathered and compared data from manyfronts allows a reasonable set of assump-tions to measure risk and safety in air med-icine. Stated another way, the real ques-tion is: Can we eliminate the medical riskin any given therapy and at what cost tobenefit? The answer is no—without atleast some risk we would not have benefits.
Most importantly the final sections ofthe report look at the risk to providers andpatients. The news is sobering toproviders while good for patients.Without question the issue of risk is tiedto exposure. This is a message we musttake home. Managing risk—identifica-tion, avoidance, reduction, and manage-ment are key strategies that each of usmust employ ever day. Every provider andparticipant in air medicine should readand re-read this report, take it to heart,and then change your practice.
Thomas Judge, CCT-PExecutive Director LifeFlight of MaineChair, AAMS Safety CommitteeMember, AMSACAAMS Region IV Director
69Special Supplement––Editorial Comments
5 Mann NC, et. al. Injury Mortality Following the Loss of Air Medical Support from Rural Interhospital Transport, Acad Emer Med 2002;9:694-6986 Powell DG, et.al. The Impact of a Helicopter Emergency Medical Services Program on Morbidity and Mortality, AMJ, 1197;16:27 Oppe S, De Charro, FT, The effect of medical care by a helicopter trauma team on the probability of survival and the quality of life of hospitalised victims.Accident Analysis and Prevention, 2001; 33129-1388 “To Err is Human: Building a Safer Health System.” Kohn LT, Corrigan JM, and Donaldson MS, (Ed); Committee on Quality of Health Care in America,Institute of Medicine, National Academy Press, 2000.
ANGELS OF MERCY OR ANGELS OF DEATH
In the late 1980s a television news-magazine referred to the air medicalindustry as the “Angels of Mercy orAngels of Death” in reaction to the highaccident and fatality rate. This ratereached an all-time high of 13.14 acci-dents per 100,000 hours flown. In con-trast, during this same time frame com-mercial airliners were experiencing a rateof .002/100,000 hours. As this safetyreport has shown, the accident rates andfatal accident rates per 100,000 flighthours are down dramatically from whatwe experienced in the mid-’80s.Unfortunately, from 1998 through 2001,we have had more accidents than in anyfour-year period in HEMS history, includ-ing the “death star” years. In addition,the fatal accident rates are the highestthey have been since the early 1990s.
What has changed and what has not?We are flying more sophisticated equip-ment. Oddly enough, during a recentHelicopter Association InternationalExposition Safety Symposium, a group ofhelicopter manufacturers conducted apanel discussion on the high accidentrate. It was their premise that they[manufacturers] have re-designed, re-engineered, re-structured, automated, andimproved the basic flying machines, yetwe are crashing at the same, if not higherrate than was experienced prior to allthese improvements.
We added more pilots to air medicalprograms to reduce exposure and fatiguewhile on duty as fatigue was considered tobe a major factor in the accident chain ofevents. We increased weather minimumsin the hopes that standards would encour-age better decision making. We formedan accreditation group to promote com-petition in achieving excellence and pro-fessionalism. Associations were formed toprovide a forum and infrastructure toattack safety issues head on. The industrywas aggressive and came together with
safety as their coat of arms.The fruits of this labor seemed to pro-
vide quite a harvest as 1990 came to aclose with no fatalities. Had we nippedthe beast in the bud? Unfortunately, wereturned to a smattering of accidents inthe early and mid-1990s. The number ofaccidents began to escalate in 1998 andcontinued through 2001 when we had 13accidents.
The industry took a deep breath andsaid, “where do we go from here?” Were-grouped, met in mass, identified seveninitiatives to break the chain of accidentsand attempted to provide an action planwith which to proceed. The FAA wasanxious for our industry to come up withan in-house solution. But did we?
Unfortunately, unilaterally there hasbeen little change in the way we do busi-ness. An Air Medical ResourceManagement course has been developed.Fielding and implementation is slow asfinancial support for safety education andtraining is not uniformly endorsedthroughout the industry. The AirMedical Safety Advisory Council wasformed in the hopes that Part 135 ven-dors could provide some insight and solu-tions to industry safety trends and shareinformation to aid in the prevention ofrepetitive safety infractions.
As in all organizational structures,progress is impeded by the very large geo-graphical nature of our business. There isan underlying sense of “breath holding”until the end of the day in hopes thatanother significant event [accident] hasn’t occurred. And then a new daybegins, as do our hopes. We still hear ofrepercussions for “whistle blowers” onsafety issues. Individuals within organiza-tions are afraid to come forward withsafety of flight issues for fear of losingtheir jobs. Aberrant behaviors are some-times rewarded rather than punished.Between operators there is little
exchange of information for fear of dis-closing proprietary issues. Successes aswell as failures are not shared. And inthese tough financial times in the health-care industry, competition is the dragonin disguise for faulty decision making,cutting safety corners, eliminating safetyinfrastructures, training and education,and general apathy toward developing asafety culture.
But all is not lost. There are multiplethings your organization can do to get onthe safety bandwagon and bring the acci-dent rate to zero. This must, however,start at the top with management buy-inthat safety is the only imperative toexceed mission accomplishment.
This report may raise your awarenesswith regard to safety and some uniquerisk assessments. But safety must be anintegral part of everything your programthinks, says, and does. Safety has tobecome an attitude and a way of life.You don’t “get safe” when you come towork. It must permeate the mind-set ofyour organization. Resources must becommitted to safety training and educa-tion. Formal safety standards must be set,must be trained, and must be adhered to.Safe behavior should be rewarded andunsafe behavior should have serious con-sequences. Open communication mustbe encouraged. And above all, yourorganization must develop a safety cul-ture that promotes the motto, “if it’s notworth doing safely, it’s just not worthdoing!”
No, all is not lost, but without aggres-sive, proactive, and committed attentionto individual organizational safety infra-structures, the accident rate will notchange, and we may be re-crownedAngels of Mercy or Angels of Death.
Michelle North, Ph.D.PresidentThe Wisdom Well
Funding to support this research and publication was provided in part by:
Grant support to fund the publication of this report has been received in part from:
AeroMed SoftwareManagement Solutions for Critical Care TransportInnovative Engineering