teamwork and communication in surgical teams: implications for patient safety

TIP

FprhptappoAca

n

CTfAVo

RFfPiSCS

©P

eamwork and Communication in Surgical Teams:mplications for Patient Safetyeter Mills, PhD, MS, Julia Neily, RN, MS, MPH, Ed Dunn, MD, MPH, FACS

BACKGROUND: As part of a national program in the Department of Veterans Affairs to improve communicationwithin the health-care environment, the Medical Team Training questionnaire was developed toassess organizational culture, communication, teamwork, and awareness of human factorsengineering principles.

STUDY DESIGN: The Medical Team Training questionnaire was pilot tested with 300 health-care clinicians. Thefinal version of the Medical Team Training questionnaire was administered to an interdiscipli-nary group of 384 surgical staff members in 6 facilities as part of the Medical Team Trainingpilot project in the Department of Veterans Affairs.

RESULTS: The results revealed a pattern of discrepancies among physicians and nurses in which surgeonsperceive a stronger organizational culture of safety, better communication, and better teamworkthan either nurses or anesthesiologists do.

CONCLUSIONS: The Medical Team Training questionnaire was helpful in identifying hidden problems withcommunication before formal team training learning sessions, and it will be useful in focusingefforts to improve communication and teamwork in the operating room. (J Am Coll Surg 2008;
206:107–112. © 2008 by the American College of Surgeons)
ttCipimpbtTiist

MTTmmwcsiVh

ailures in coordination and communication among hos-ital clinicians have been associated with higher mortalityates in intensive care units,1 longer lengths of stay andigher nurse turnover in intensive care units,2 and greaterostoperative pain with lower functioning levels for pa-ients.3 Nursing reports of collaboration between nursesnd house staff were positively associated with improvedatient outcomes in medical intensive care units, but re-orts from the house staff were not associated with betterutcomes.4 Surgical teams at Department of Veteransffairs (VA) hospitals with low mortality rates communi-ated more effectively and more often than surgical teamsssociated with high mortality rates.5

Assessing the effectiveness of cooperation and commu-ication among surgical teams and ICU teams is critical to

ompeting Interests Declared: None.his article is the result of work supported with resources and the use of

acilities at the Veterans Affairs National Center for Patient Safety in Annrbor, MI, and the Veterans Affairs Medical Center, White River Junction,T. The views expressed in this article do not necessarily represent the viewsf the Department of Veterans Affairs or of the US government.

eceived May 2, 2007; Revised May 31, 2007; Accepted June 6, 2007.rom the Field Office of the Department of Veterans Affairs National Centeror Patient Safety, White River Junction, VT (Mills, Neily); Department ofsychiatry, Dartmouth Medical School, Hanover, NH (Mills); and the Med-

cal Team Training Program, Veterans Affairs National Center for Patientafety, Ann Arbor, MI (Dunn).orrespondence address: Peter Mills, PhD, MS, VAMC (11Q), 215 N Main

Nt, White River Junction, VT 05009.

1072008 by the American College of Surgeons

ublished by Elsevier Inc.

imely problem solving in the management of surgical pa-ients. The VA offers a program, developed by the Nationalenter for Patient Safety (NCPS) in 2003, entitled “Med-

cal Team Training” (MTT). This program is based on therinciples of crew resource management6 and is designed tomprove communication within the health-care environ-

ent. The primary objectives of the program are to im-rove the outcomes of patient care and staff job satisfactiony introducing crew resource management communica-ion principles for application in the clinical workplace.he questionnaire is administered before the MTT Learn-

ng Session to assess the safety climate and reveal differencesn perceptions between professional groups working in theame clinical units.This article summarizes development ofhe questionnaire and reports our initial findings.

ETHODShe MTT questionnaire was developed from the Teamraining questionnaire7 that had been used to evaluateedical quality improvement teams in facilitated improve-ent projects within the VA.7-14 The earlier questionnaireas modified to elicit more specific information related to

ommunication and teamwork between clinicians. This in-trument was pilot tested with 300 clinicians and admin-strators from the operating rooms (ORs) and ICUs in 6A medical centers (VAMCs) while the medical centersosted MTT Learning Sessions between September and
ovember 2003. After undergoing iterative changes, the
ISSN 1072-7515/08/$34.00doi:10.1016/j.jamcollsurg.2007.06.281

fbrLsaLdvba8p

SFMwWtocUvrcaq

tsi(cfsrcwttaopm

sCTbmSmsme

C

AQfaoc(apyAW

RTis

Frc

108 Mills et al Teamwork and Communication among Surgical Teams J Am Coll Surg

inal version of the MTT questionnaire was administeredefore the start of the Learning Sessions to 384 operatingoom staff and administrators in 6 VAMCs before theirearning Sessions from May through December 2005. Re-pondents were asked to rate the degree to which theygreed with statements in the questionnaire on a five-pointikert scale. Response options ranged from “1” (stronglyisagree) to “5” (strongly agree). The questionnaire wasoluntary and anonymous. The response rate (the num-er of completed questionnaires over the number ofttendees in the Learning Sessions) was 309/384, or0%. The percentages of professions represented are dis-layed in Figure 1.

cale developmentactor analysis was conducted on the final version of theTT questionnaire, using Principal Component Analysis

ith Varimax rotation software (SPSS version 12.1 forindows; SPSS, Inc). Factor analysis is a statistical method

hat builds “factors” of intercorrelated questions. The the-ry behind it is that questions that ask about the sameoncept should be similar, so are correlated to each other.sing factor analysis to statistically identify factors pro-

ides evidence of the validity of the concepts the factorsepresent, because the factors have been independently dis-overed. Four factor subscales were identified. Further itemnalysis resulted in the removal of four questions from theuestionnaire because of relatively low factor scores.

The final subscales, questions, and alpha scores (a statis-ic that measures the degree to which each subscale is mea-uring a similar principle) for the respondents are displayedn Table 1. The first subscale is “organizational culture”alpha � 0.786) and represents the perception that thelinician is supported by the organization and feels com-ortable interacting with others on the clinical team. Theecond subscale, “communication,” (alpha � 0.819) rep-esents the perception that clinicians make efforts to ex-hange critical information. The third subscale, “team-ork,” (alpha � 0.858) represents the perceptions that

here is mutual respect, appreciation, and collaboration onhe clinical team. The fourth subscale, “human factorswareness,” (alpha � 0.842) represents an understandingf basic human factors principles, such as the tendency foreople to make mistakes and the effects of fatigue on hu-

Abbreviations and Acronyms

MTT � Medical Team TrainingOR � operating roomVAMC � Veterans Affairs Medical Center

an performance. a

The questionnaire was further validated in a separatetudy conducted at the Michael E DeBakey VA Medicalenter in Houston, TX. After enrolling in the Medicaleam Training program and initiating briefings and de-riefings in the OR, the Houston facility tracked improve-ents in communication using the MTT questionnaire.

ignificant improvements in communication were docu-ented by repeat application of the MTT questionnaire to

urgical staff. These improvements were also associated withore timely administration of prophylactic antibiotics and

nhanced deep vein thrombophlebitis prophylaxis.15

The Institutional Review Board (IRB) at Dartmouthollege approved this project (CPHS #16923).

nalysisuestionnaire responses for operating room personnel

rom three major categories—surgeons, anesthesiologistsnd certified registered nurse anesthetists (CRNAs), andperating room nurses—were aggregated for the six medi-al centers and compared using analysis of varianceANOVA) analysis. The subscale scores were comparedcross the three clinician groups using Bonferroni post hocrobes. Multivariate analysis of variance (MANOVA) anal-sis was conducted to detect differences between facilities.ll analyses were conducted using SPSS version 12.1 forindows (SPSS, Inc).

ESULTSable 2 displays demographic data on the VAMCs involved

n this study. The final analysis was conducted on 233 ORtaff members self-identified as either nurses (n � 139),

igure 1. Percentage of professions among identified operatingoom staff (n � 309). The category “physicians” represents physi-ians who are neither surgeons nor anesthesiologists.

nesthesiology providers (anesthesiologists or CRNAs;

nradtp

pvla

v

T

V

123456

V

TI

O

C

T

H

MO

109Vol. 206, No. 1, January 2008 Mills et al Teamwork and Communication among Surgical Teams

� 49), or surgeons (n � 45, Fig. 2). Table 3 displays theesults of the ANOVA analysis for each of the 4 subscalescross the 3 professional groups. There was a significantifference among professions on perceptions of organiza-ional culture, communication, and teamwork. Post hocrobes (Table 4) revealed a general pattern that surgeons’

able 2. Demographic Data

AMCSurgical cases/year

(fiscal year 2005 data), nMedical school

affiliation

4,051 Yes1,771 Yes5,365 Yes4,422 Yes2,952 Yes2,241 Yes

able 1. Items and Coefficient Alpha for Medical Team Traitems

rganizational cultureMedical Team Training for staff meets one of our organization’sOur health-care facility has a nonpunitive method of investigatinI like my job.Our team views problems from a systems perspective rather thanOur team leadership is open to feedback and input from all teamHealth-care personnel should always monitor each other for signsI am comfortable intervening in a procedure if I have concerns ab

ommunicationOur team routinely briefs procedures before starting them.Our team routinely debriefs procedures after completing them.Our team has a specific way of ensuring that all members understWorkload and task distribution are clearly communicated in ourDuring surgical and diagnostic procedures, everyone on the team

eamworkEveryone on our team is comfortable giving feedback to other teaMorale on our team is high.Our team members understand each other’s strengths and weaknOur team members have mutual respect for each other.Our team has a successful method for resolving conflicts betweenI know the first and last names of all members of my surgical teamThe surgeon and anesthesiologist maintain open channels of comOur team has a shared vision of how to improve.uman factors awareness (scale reversed)Competent clinicians never make mistakes.Less experienced staff should not question the decisions of more eNurses should not question decisions made by attending physiciaFatigue does not affect my ability to perform my work tasks effecWhen I am interrupted, my patient’s safety is not affected.Stress and distractions do not affect my ability to perform my wo

edical Team Training scale: overall alpha � 0.881.R, operating room.

AMC, Department of Veterans Affairs Medical Center.

erceptions differ from those of nurses and anesthesia pro-iders, and nurses and anesthesiologists tend to have simi-ar perceptions of teamwork, communication effectiveness,nd organizational culture.

MANOVA analysis of the 6 individual VA facilities re-ealed a significant interaction effect between facility and

horized beds, nCardiovascular surgeryand university affiliate

Neurosurgeryprograms

237 No No59 No No

237 Yes Yes585 Yes Yes191 Yes Yes119 No No

Questionnaire (n � 283)Alpha

0.786rategic goals.erse events.

meone’s fault.”bers.ress and fatigue.hat is occurring.

0.819

ll important communications.environment.are of what is happening.

0.858embers.

members.ing the conduct of procedures in the OR.cation during procedures in our OR.

0.842

enced staff.

.

ks effectively.

Aut

ning

key stg adv

as “somemof stout w

and aworkis aw

m m

esses.

teamdur

muni

xperins.tively

rk tas

ptposroa

DCuqgtqpt

cOTpbdt

mzgpgprdvfsacctr“cArmaiop

TTS

nO

C

T

H

C


rofession (F � 1.98, p � 0.001). This finding indicateshat five of the six operating rooms repeated the same basicattern in which the perceptions of nurses and anesthesi-logists were relatively similar and both differentiated fromurgeons. One facility (no. 4) had a very different pattern ofesponses. In facility 4, the surgeons were much more awaref problems with organizational support, situationalwareness, and communication.

ISCUSSIONommunication failures in the operating room are notncommon and can jeopardize patient safety.16,17 Conse-uently, assessment of communication effectiveness in sur-ical teams will enable timely problem solving and poten-ial interventions in the provision of surgical care. Ouruestionnaire instrument is a useful means for identifying aattern of differences among surgical teams in their percep-ions of organizational culture, teamwork, and communi-

able 3. Means, Standard Deviations, and ANOVA Results fraining Questionnaireubscale Nurses Surgeons

139 45rganizational cultureMean 3.21 3.64SD 0.68 0.89

ommunicationMean 2.87 3.41SD 0.78 0.86

eamworkMean 2.91 3.41SD 0.76 0.84uman factors awarenessMean 3.58 3.43SD 0.92 0.76

Figure 2. Breakdown of professional groups analyzed (n � 233).

RNA, certified registered nurse anesthetist.

ation effectiveness. We are able to differentiate betweenR staff members with different communication styles.he most common pattern is that nurses and anesthesiaroviders tend to perceive their environment similarly, andoth differ significantly from surgeons’ perceptions. In ad-ition, surgeons perceive the environment more positivelyhan the other two groups do.

More specifically, our findings revealed that surgeons areore likely to report their perceptions of a strong organi-

ational culture for patient safety. For example, this cate-ory includes the item, “I am comfortable intervening in arocedure if I have concerns about what is occuring.” Sur-eons may believe this to be the case; nurses and anesthesiaroviders are less likely to report the same. Surgeons alsoate teamwork and communication during their proce-ures more favorably relative to nurses and anesthesia pro-iders in those procedures. An example would be the dif-erential responses to the following question: “Duringurgical and diagnostic procedures, everyone on the team isware of what is happening.” This disparity between surgi-al team members signals a problem if the surgeon per-eives that other team members are well informed whenhey are reporting the opposite. Surgeons also report higheratings on teamwork effectiveness with responses such as,morale on our team is high,” and “everyone on our team isomfortable giving feedback to other team members.”gain, nurses and anesthesia providers are less likely to

eport their agreement with the surgeons. If surgical teamembers have disparate perceptions about how well they

re communicating or collaborating with each other, hows it possible for them to be collaborating optimally withther members of the surgical team for the care of theiratients?

ofessional Groups on Four Subscales of the Medical Team

nesthesiologists and CRNAs F score p Value

49

3.23 6.24 0.0020.62

2.72 10.38 0.000050.72

2.72 10.16 0.000060.64

3.72 1.31 0.2730.82

or Pr

A

ptattT

sshnrgoarmclttwe

vqFvpfiqtsfbieicnqi

yt

TS

O

C

T

H

*

111Vol. 206, No. 1, January 2008 Mills et al Teamwork and Communication among Surgical Teams

Facility 4 did not follow the pattern found in other hos-itals in our study. We did not formally assess the charac-eristics of the facilities or the staff. But based on our inter-ctions with the staff in this facility, we became aware thathis group had already been exposed to team training andhat some of the senior leaders had experience in aviation.his may account for the difference in their responses.These results are consistent with findings from other

tudies. Awad and colleagues15 reported their baseline re-ults from an earlier version of this questionnaire, revealingigher ratings from surgeons than anesthesiologists andurses. Thomas and associates18 reported teamwork surveyesults with critical care nurses and physicians working to-ether who viewed the success of their teamwork with eachther differently. Again, physicians reported more favor-ble perceptions of teamwork. DeFontes and Surbida19 alsoeported that physicians viewed teamwork with nursesore positively than nurses viewed teamwork with physi-

ians. Finally, Makary and coauthors20 reported from aarge survey study of 60 hospitals that surgical staff rateeamwork within their own profession more highly thaneamwork in other domains. Surgeons rate overall team-ork more highly than nurses working in the same OR

able 4. Bonferroni Post Hoc Probes of ANOVA Analysis (nubscale Professional group (1)

rganizational culture Nursing SurAn

Surgeons NuAn

Anesthesiologists and CRNAs NuSur

ommunication Nursing SurAn

Surgeons NuAn


eamwork Nursing SurAn

Surgeons NuAn


uman factors awareness Nursing SurAn

Surgeons NuAn


The mean difference is significant at the 0.05 level.

nvironment. c

Another important component of this study was to pro-ide further evidence of the construct validity of the MTTuestionnaire. This evidence was revealed in several ways.irst, the MTT questionnaire was based on a previouslyalidated questionnaire, developed for general medical im-rovement teams, that was already capable of reliably dif-erentiating between high- and low-performing medicalmprovement teams.7 Second, the four subscales of theuestionnaire were independently identified using the sta-istical method of factor analysis, suggesting that each sub-cale measures a unique concept that is somewhat differentrom the other subscales. This idea is further corroboratedy the relatively high alpha scores for each of the subtests,ndicating the high intercorrelation of the questions withinach subscale. Third, positive changes in the questionnairen another study15 were associated with positive clinicalhanges and staff reports of improved communication. Fi-ally, the patterns of communication identified by the MTTuestionnaire were very similar to communication patternsdentified in other studies18-20 with similar types of clinicians.

There are several limitations to this study. Our data anal-sis was limited to six VA medical centers, and most ques-ionnaire respondents were nurses. In addition, this was a

3)ssional group (2) Mean difference (1–2) Significance

s �0.42406* 0.002iologists and CRNAs �0.01942 1.000

0.42406* 0.002iologists and CRNAs 0.40463* 0.020

0.01942 1.000s �0.40463* 0.020s �0.53659* 0.000iologists and CRNAs 0.14849 0.769


�0.14849 0.769s �0.68508* 0.000s �0.49347* 0.001iologists and CRNAs 0.17326 0.508


�0.17326 0.508s �0.66673* 0.000s 0.14746 0.973iologists and CRNAs �0.14296 0.972

�0.14746 0.973iologists and CRNAs �0.29042 0.322

0.14296 0.972s 0.29042 0.322

� 23Profe

geonesthesrsingesthesrsinggeongeonesthesrsingesthesrsinggeongeonesthesrsingesthesrsinggeongeonesthesrsingesthesrsinggeon

ross-sectional study with neither followup information on

plmoh

ptScvmpaetnpabs

ASAADC

R

1

1

1

1

1

1

1

1

1

1

2


atient outcomes nor observed staff behaviors. Neverthe-ess, these results indicated that the MTT questionnaire

ay be useful in identifying differences in the perceptionf teamwork and communication effectiveness betweenealth-care professionals in the surgical environment.In our Medical Team Training program, we conduct

reparation and planning conference calls with changeeams from each participating facility before the Learningession. These calls are focused on identifying communi-ation problems among clinicians. Our questionnaire pro-ides a helpful tool for uncovering these problems that staffay not have otherwise recognized. In addition, specific

atterns of suboptimal communication can be identifiednd targeted for intervention in the program. These differ-nces in perception related to teamwork and communica-ion highlight the need for improvement in team cohesive-ess and a shared mental model. One of the goals of ourrogram is to facilitate an increased awareness of teamworknd communication challenges among clinical team mem-ers and to introduce tools for improving collaboration forafe patient care.

uthor Contributionstudy conception and design: Mills, Neily, Dunncquisition of data: Mills, Neily, Dunnnalysis and interpretation of data: Millsrafting of manuscript: Mills, Neily, Dunnritical revision: Mills, Neily, Dunn

EFERENCES

1. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. An eval-uation of outcome from intensive care in major medical centers.Ann Int Med 1986;104:410–418.

2. Shortell SM, Zimmerman JE, Rousseau DM, et al.The perform-ance of intensive care units: does good management make adifference? Med Care 1994;32:508–525.

3. Gittell JH, Fairfield KM, Bierbaum B, et al. Impact of relationalcoordination on quality of care, postoperative pain and func-tioning, and length of stay. Med Care 2000;38:807–819.

4. Baggs JG, Ryan SA, Phelps CE, et al. The association betweeninterdisciplinary collaboration and patient outcomes in a med-
ical intensive care unit. Heart Lung 1992;21:18–24.
5. Young GJ, Charns MP, Daley J, et al. Best practices for managingsurgical services: the role of coordination. Health Care ManageRev 1997;22:72–81.

6. Musson DM, Helmreich RL. Team training and resource man-agement in health care: current issues and future directions.Harvard Health Policy Rev 2004;5:25–35.

7. Mills PD, Weeks WB. Characteristics of successful quality im-provement teams: lessons from five collaborative projects in theVHA. Joint Commission Journal on Quality and Safety 2004;30:152–162.

8. Weeks WB, Mills PD, Dittus RS, et al. Using an improvementmodel to reduce adverse drug events in VA facilities. Joint Com-mission Journal on Quality Improvement 2001;27:243–254.

9. Mills PD, Weeks WB, Surott-Kimberly BC. A multi-hospitalsafety improvement effort and the dissemination of new knowl-edge. Joint Commission Journal on Quality and Safety 2003;29:124–133.

0. Mills PD, Waldron J, Quigley PA, et al. Reducing falls andfall-related injuries in the VA system. J Healthcare Safety Quar-terly 2003;v1:25–33.

1. Weeks WB, Mills PD, Waldron J, et al. A model for improvingthe quality and timeliness of compensation and pension examsin VA facilities. J Healthcare Management 2003;48:252–261.

2. Volicer L, Mills PD, Hurley AC, Warden V. Home care forpatients with dementia. Fed Practitioner 2004;21:13–28.

3. Neily J, Howard K, Quigley P, Mills PD. One-year follow-upafter a collaborative breakthrough series on reducing falls andfall-related injuries. Joint Commission Journal on Quality andSafety 2005;31:275–285.

4. Mills PD, Neily J, Mims E, et al. improving the bar code med-ication administration system in VHA. Am J Health-SystemPharm 2006;63:1442–1447.

5. Awad SS, Fagan SP, Bellows C, et al. Bridging the communica-tion gap in the operating room with medical team training. Am JSurg 2005;190:770–774.

6. Lingard L, Espin S, Whyte S, et al. Communication failures inthe operating room: an observational classification of recurrenttypes and effects. Quality Safety Health Care 2004;13:330–334.

7. Gawande AA, Zinner MJ, Studdert DM, Brennan TA. Analysisof error reported by surgeons at three teaching hospitals. Surgery2003;133:614–621.

8. Thomas EJ, Sexton JB, Helmreich RL. Discrepant attitudesabout teamwork among critical care nurses and physicians. CritCare Med 2003;31:956–959.

9. DeFontes J, Surbida S. Preoperative safety briefing project. Per-manente J 2004;8:21–27.

0. Makary MA, Sexton JB, Freischlag JA, et al. Operating roomteamwork among physicians and nurses: teamwork in the eye of
the beholder. J Am Coll Surg 2006;202:746–752.

teamwork and communication in surgical teams: implications for patient safety

Documents