patient safety leadership walkrounds tm partners · pdf filepatient safety leadership...

423August 2005 Volume 31 Number 8

Reports about health care errors in the 1990s12

generated momentum for change that has per-manently altered the landscape of health care

delivery systems A clear goal has been to deploy tools toensure safer more reliable and simpler systems in healthcare For example computerized physician order entry3

makes information more usable pay-for-performanceplans reward better outcomes45 and work flow methodsimprove patients length of stay and smooth variability67

The implementation of such tools ultimately requiresactive participation by front-line care providers and theirwillingness to alter their patterns of work Yet this willnot happen unless health care and safety leaders promotethe belief in all providers that they can effectively partic-ipate in improving their work domain and that their inter-est and efforts are aligned with the overall interests oftheir organization and leaders89

Outside health care companies such as Alcoa andSouthwest Airlines have shown that leadership involve-ment in safety and reliability is essential to developappropriate employee attitudes and is a key ingredientto financial success1011 This focus on leadership stemsin part from operations management research thatshows that leaders do not address important problemsbecause they do not hear about them and that they underestimate front-line employee dissatisfactionwith service quality12ndash14 For example Tucker et al8

found that nurses tended to adapt to the inadequacies in

Patient Safety LeadershipWalkRoundsTM at Partners HealthCare Learning from Implementation

Leadership

Allan Frankel MDSarah Pratt Grillo MHA

Erin Graydon BakerCamilla Neppl HuberSusan Abookire MD

Marianne GrenhamPam Console

Mary OrsquoQuinnGeorge Thibault MDTejal K Gandhi MD

Background Brigham and Womenrsquos Hospital (BWH)began Patient Safety Leadership WalkRoundsTM inJanuary 2001 its experience along with that of threeother Partner Healthcare hospitals is reported

Collecting Data on WalkRounds Data were obtainedfrom interviews with patient safety personnelWalkRounds scribes and senior leaders

Findings A total of 233 one-hour WalkRounds dur-ing 28 months yielded 1433 commentsmdash30 related toequipment 13 to communications 7 to pharmacyand 6 to workforce Actions occurred quickly in smallhospitals Formal processes for managing larger issueswere necessary in large organizations Implementationfeasibility featured more prominently than severity indetermining actions

Discussion The study generated essential guidelinesfor successmdashfor example the supporting resourcesmust include the maintenance of effective informationdatabases that identify actions taken and the discus-sions during WalkRounds are influenced by who in lead-ership is participating their ability to quietly listen andwhether they have clinical or nonclinical backgrounds

Conclusions WalkRounds appears to be an effectivetool for engaging leadership identifying safety issuesand supporting a culture of safety

Article-at-a-Glance

Copyright 2005 Joint Commission on Accreditation of Healthcare Organizations


their environment rather than speak up about themThese nurses like physicians were no doubt trained todisplay (albeit never manage) endless capacity for vigi-lance and adaptability

Traditionally health care has had a culture of blamewhere people are punished for making errors Patientsafety efforts strive to create a culture of safety wherefront-line staff are comfortable speaking up about errorsand adverse events and leadership encourage this dia-logue Health carersquos need to better address this problemis underscored by the Joint Commission on Accreditationof Health Care Organizationsrsquo requirement that leader-ship participate in promoting safer care delivery15 and inthe emphasis on the development of a ldquoCulture of Safetyrdquoas articulated in the National Quality Forumrsquos SafePractices16ndash18 and The Leapfrogrsquos Group 4th Leap1920 and bythe United Kingdomrsquos National Patient Safety Agency2122

Patient Safety Leadership WalkRoundsTM 23 is a simplebut rigorous management tool designed to assist hospital leaders in implementing mechanisms for pro-moting safety learning about and hearing the concernsof front-line providers supporting appropriate accounta-bility concepts and allocating resources to areas ofgreatest risk Brigham and Womenrsquos Hospital (BWH)began WalkRounds in January 200123 which generatedinterest by other organizations in Partners HealthCare(Partners is composed of two acute care tertiary organizations four community hospitals two reha-bilitation hospitals and a psychiatric hospital) TheShaughnessy-Kaplan Rehabilitation Hospital beganimplementing WalkRounds in November 2001 followedby Newton-Wellesley Hospital (NWH) and SpauldingRehabilitation Hospital in January 2002 (Other Partnershospitals have since also started WalkRounds) We exam-ined these four hospitalsrsquo joint experience to address ina natural experiment the following questions Would hospital leadership be willing to openly dis-cuss operational failure safety and harm with front-lineproviders Would frank and open discussion occur in a publicsetting Could the information elicited be collected and aggre-gated in a useful manner Would the information collected affect actions orresource allocation

Implementing WalkRoundsAs the leadership and safetyquality personnel in the fourhospitalsmdashBWH Spaulding Rehabilitation Shaughnessy-Kaplan Rehabilitation and NWHmdashagreed to implementWalkRounds we [AF SPG] gave a two-hour presenta-tion to the hospital executives and members of the safetyand quality departments about theories promoting leader-ship involvement in safety and quality11 high reliability24

blame-free reporting25 and useful data categorization26 Aframework and timeline for implementation were sug-gested including a method for aggregating data Data col-lection began either on paper or in a spreadsheetprogram but BWH soon after beginning WalkRoundsdeveloped a computerized database to facilitate datamanipulation and to improve reporting of informationThe software was available for use by the other hospitals

Although WalkRounds inevitably is shaped by individ-ual personalities and organizational culture some basiccomponents as shown in Table 1 (page 425) wereencouraged After those components were suggestedleadership and safetyquality officers considered theprocess and resources available Some organizationschose to perform WalkRounds weekly others biweeklyor monthly Regardless of the scheduling WalkRoundswere occasionally cancelled secondary to executiveschedules or because the floors were too busy when theWalkRounds groups arrived (Table 2 page 426)

Mechanisms for feedback for reporting of commentsprovided during WalkRounds is described and depictedin Figure 1 (pages 427ndash428) At all four hospitals the per-sons charged with responsibility for patient safety par-ticipated in implementing WalkRounds The role of thescribe (Table 1) was assigned to persons with markedlydifferent positionsmdasha patient safety project analyst theadministrative assistant to a patient safety director thesenior administrative secretary for a quality departmentand a quality improvement coordinator

Different combinations of administrative and clinicalleaders chose to participate at each hospital (Table 2)Leadership involvement seemed to be determined bypersonal interest and by who was initially asked Inthose hospitals where the chief executive officer showedinterest more involvement by a greater variety of per-sonnel was apparent No chief information officers orchief financial officers were asked to join initially and


425

ownership of the WalkRounds seemed to settle on theadministrative leadership especially but not exclusivelythose with clinical backgrounds Chief information offi-cers now participate regularly at BWH

Collecting Data on WalkRoundsData were obtained from structured interviews with

patient safety personnel who participated in the roundsinterviews with WalkRounds scribes and evaluation of thecollected information Unstructured interviews were usedto elicit how information was used in the organization toidentify the reporting structure within the administrationand to identify who took responsibility for feedback tofront-line employees and higher-level administrators

Concerns or events presented during WalkRoundswere divided into categories modified from Vincent etalrsquos incident analysis categories26 CommunicationEquipment Information Systems Laboratory PharmacyPatient Related Staff Related Department SpecificMiscellaneous for a total of 48 subcategories nestedwithin the categories For example Patient Related sub-categories were as followsmdashinfant transport monitor-ing patient flow patient issues and patient transport

Category modifications were made over time to makethe information more useful in identifying specificactions to perform or to allow the information to beaggregated so that reports would be more useful toadministrative heads and middle managers As statedBWH maintained its data in a computerized databaseMost other hospital information was on paper or in theprocess of being moved into a spreadsheet program orcomputerized database

A ldquocommentrdquo was defined as any concern or eventraised by an individual during the WalkRounds and bothconcerns and events were considered ldquooperational fail-uresrdquo3 The patient safety personnel in each hospital cat-egorized the comments (to allow the information to bemost effectively tailored for operational use) which oneof the authors [SPG] then recategorized into theresearch study database to ensure consistency

Actions performed at BWH were added to its data-base in real time Actions performed at other hospitals

August 2005 Volume 31 Number 8

An identified individual and probably the patientsafety manager or director should participate in allthe WalkRounds

Another person should perform the function ofscribe during the WalkRounds and document (a)location (b) who participated (c) the topics dis-cussed and (d) other factors that may be pertinentto context when reviewing comments

Administrative and clinical leaders should partici-pate in WalkRounds on a rotating basis

Middle management should be informed whenWalkRounds would occur in their areas and be provided with sample questions to help their staff prepare to discuss actual or poten-tial patient harm that results from system complexity

Concerns expressed by all individuals should beheard sympathetically but attempts should bemade to lead the conversations to correlate theconcerns with specific episodes of patient orprovider harm or potential harm

WalkRounds should be performed in all locationsthat affect clinical care including laboratoriesradiology pharmacy emergency departments andall patient care floors

WalkRounds should be scheduled up to one year inadvance choosing times not based on leadershipavailability but on the likelihood of front-lineprovider availability taking into consideration nursing shifts lulls in activity and when physiciansperform clinical rounds

All possible personnel from physicians to clean-ing staff should be included in the rounds andincluding patients in some rounds should be considered

The discussions should begin by explaining that therounds primarily seek insights about systems fail-ures that the purpose of the rounds is to act andto use the information elicited to identify where toallocate resources to improve safety quality andefficiency

At the end of the WalkRounds participants shouldbe asked as a way of disseminating the conceptsdiscussed to find two other persons with whomthey work and tell them about the WalkRounds andits purpose

Table 1 Basic Concepts for WalkRounds Hospitals

The list of categories and subcategories of comments in the comput-erized database can be obtained by e-mail request to Dr Frankel



CEO chief executive officer COO chief operating officer CMO chief medical officer CNO chief nursing officer CFO chief financial officer VP vice presi-dent CIO chief information officerdagger Calculations based on some WalkRounds visiting two or more units Dagger Institute for Helathcare Improvement Patient Safety Leadership WalkRounds Database httpwwwihiorgIHITopicsPatientSafetySafetyGeneralToolsPatientSafetyLeadershipWalkRoundsDatabasehtm (last accessed Jul 18 2005)

Table 2 Demographics of WalkRounds January 2001ndashMay 2003

Brigham andWomenrsquos Hospital

Newton-WellesleyHospital

SpauldingRehabilitation

Hospital

Shaughnessy-KaplanRehabilitation

HospitalWalkRounds initiated January 2001 January 2002 January 2002 November 2001Bed size 700 310 296 160

Frequency of rounds WeeklyInitially weekly butoccur almost biweeklyMoving to monthly

WeeklyInitially weeklyCurrently 2 times amonth

Method of scheduling 3 months in advanceSet time of 10 AMevery Wednesday

Cycle of allareasdepartmentsscheduled at once

One month in advance

Number of clinicalareasdepartments visited(all departments with clini-cally relevant activities)

65 35 20 10

Number of rounds conducted 73 23 60 45

Months between visits toeach unitdagger 12 15 5 3

Persons participating 372 70 150 135

Involvement of senior leadership

CEO COO CMO CNOCFO VPs CIO

Senior VP MedicalAffairs Senior VP ofPatient ServicesDirector of Quality andSafety

CEO VP Clinical Affairs

President VP of PatientCare Services VP ofRehab Services Director of ProgramDevelopment Presidentof Medical StaffMedical DirectorChiefof Patient Care ServicesChief of Rehab MedicineService Director ofQuality Management

Person(s) responsible forcoordinating and managingdata

Director of PatientSafety Manager ofPatient Safety ProjectAnalyst

Director of Quality andSafety Patient SafetyProject ManagerAdministrativeAssistant

Director of NursingDirector QualityManagement SeniorSecretary for QualityManagement

Director of QualityManagement QualityImprovementCoordinator

Method for managing dataComputerized data-baseDagger

Worksheet placed inbinder summary inspreadsheet program

Minutes typed aftereach round

Worksheet

Person(s) with bestoverview of WalkRoundsprocess

Patient SafetyManager

Director of QualitySafety Patient SafetyProject ManagerAdministrative Assistantfor QualitySafety

Senior Secretary forQuality ManagementDirector of QualityManagement (inter-mittently vacant)

Director of QualityManagement



Mechanisms of Feedback and Reporting at the Four Partners HealthCare Hospitals

continued



Figure 1 Different mechanisms for feedback of information about comments expressed during WalkRounds and sub-

sequent actions are shown for Spaulding Rehabilitation Hospital Shaughnessy-Kaplan Rehabilitation Hospital

Brigham and Womenrsquos Hospital and Newton-Wellesley Hospital

Mechanisms of Feedback and Reporting at the Four Partners HealthCare Hospitals ((ccoonnttiinnuueedd))


429

were not identified or collected in real time becauseresources were not allocated to this effort Structuredinterviews were used to obtain or supplement databaseinformation regarding actions taken

Frequency of the WalkRounds and the total numberof locations per hospital affected the number of timeseach area was visited During 2frac12 years BWH the largesthospital conducted 73 rounds in 65 locations comparedwith Shaughnessy-Kaplan Rehabilitation Hospitalwhich in the course of two years conducted 45 roundsto 10 locations We considered the choice of frequencyan indicator of leadersrsquo and safety personnelrsquos interest inWalkRounds

FindingsNumber of CommentsThe BWH elicited on average 12 comments perWalkRounds whereas the other hospitals elicitedbetween 3 and 4 No single category stands out althoughspecific issues are predominant in some organizationssuch as equipment at both rehabilitation hospitals All the hospitals initially reported that the volume ofinformation collected was overwhelming The BWHrequested as a formal part of the WalkRounds that theparticipants in the rounds decide on the most important

issues The other hospitals prioritized actions accordingto severity and feasibility

Categories of CommentsThe categories with significant comments at all four

hospitals were equipment and communication-relatedissues followed by staff-related issues (three hospitals)especially work overload Information systems pharma-cy issues education and training policiesproceduresprotocols housekeeping and infection control appearedin more than 6 of the comments for at least one of thehospitals (Table 3 above)

Actions TakenThe BWH monitored actions and compiled more than

118 actions Information about actions taken at the threeother hospitals was collected for this article throughinterviews with the personnel who organizedWalkRounds NWH noted 12 actions during 18 monthsthat were specifically taken as a result of informationelicited during the WalkRounds and Shaughnessy-Kaplan noted 27 actions taken in 2 years as a result ofWalkRoundsndashobtained data

The actions included small local changes such as achange in a bathroomrsquos designation from house staff to


Table 3 Categories and Subcategories Containing gt 6 of the Comments Elicited

(N = Comments elicited)

Brigham andWomenrsquosHospital

(NN = 924)


(NN = 89)


Hospital

(NN = 221)

Shaughnessy-Kaplan Rehab

Hospital

(NN =199)CategoriesCommunication Related 115 (124) 17 (19) 28 (127) 20 (10)IncompleteInconsistent Documentation 8 (9)EquipmentSupplyFacility Related 206 (22) 24 (27) 76 (344) 120 (60)Equipment functionalitymaintenance 60 (6) 7 (8) 33 (15) 40 (20)Supply availabilityorganization 8 (9)Pharmacy Related 90 (97)Staff Related 9 (10) 22 (10) 13 (65)Work Overload 8 (9) 20 (9)Specific Items or DepartmentsEducationTraining 13 (6)PoliciesProceduresProtocols 8 (9)Housekeeping 14 (6)



patient use to major structural or resource allocationssuch as the hiring of a liaison to greet and guide patientsin an emergency room the building of new intensivecare unit doors and policy changes requiring hearbackor readback of telephone orders (Table 4 page 431ndash433)

Some comments voiced during WalkRounds were notaddressedmdashfor reasons of expense the infrequency of thecomment or low-risk nature of the problem or theimpracticality or infeasibility of a solution For examplethere usually was no simple resolution for staff-shortagecomplaintsmdasha comment was voiced that the hospitalneeded more unit assistants (but the hospital had beenunable to attract applicants for these positions)mdashor sig-nificant architectural changes In a few cases the issueswere felt to occur too infrequently to be worthy of effort

Personsmdashwhether patient safety personnel orscribesmdashparticipating regularly in the WalkRounds tend-ed to become skillful at identifying whom to turn to forspecific actions or projects At the BWH the patient safe-ty manager meets with the persons she believes can bestaddress a problem and responsibility for those actions isthen discussed during operations management meetingsand formalized Other hospitals had similar but less for-mal mechanisms to identify the locus of responsibilityfor an action although the scribes even in assistantpositions could identify after a few rounds where and towhom to turn

Mechanisms for Feedback About Comments and Actions

Mechanisms for feedback and reporting of informa-tion about comments expressed during WalkRounds andsubsequent actions differed across the four hospitals(Figure 1) Feedback to front-line staff about concernscomments and actions varied from frequent e-mails sentto individual providers to quarterly or six-month summa-tions of data sent to managers for dissemination to staff

Leadersrsquo Reactions to WalkRoundsThe WalkRoundsrsquo effect on leadership decision mak-

ing and education was assessed by unstructured inter-views The group uniformly found the time spent onWalkRounds worthwhile and mentioned actions thatwere not listed in the databases thereby identifying asource of information not tapped by the current data

collection at each hospital A sample of senior leadersrsquocomments is presented in Table 5 (page 434)

DiscussionGuidelines for SuccessThe power of the WalkRounds is visible in this study inthe change in leadership perceptions Yet there are aseries of essential guidelines for success as follows The supporting resources must include the mainte-nance of effective information databases that identify inreal time actions taken This validates the WalkRoundsand aids in timely feedback to front-line personnel Theresources necessary at the BWH for example include25 of a patient safety managerrsquos position and approxi-mately frac12-day per week from a research assistant or sen-ior secretarial position Although it is likely that manytopics were discussed in WalkRounds at all four hospitalsthe BWH documented almost three times the number asdid the others There are numerous possibilities for thisdiscrepancy Anecdotally the WalkRounds conversationsappeared equally engaging and tended to last from 40 to60 minutes However early on BWH fortified the mecha-nisms for collecting and managing data by developing arobust database and assigning a research assistant to thetasks of scribe and data input The other hospitals built inthese processes more slowly and tended to use paper or aspreadsheet program to maintain their data Scheduling and timing of the rounds is a routine sec-retarial process but simple tricks can markedly affect theroundsrsquo productivity For example informing the unit aday or two before may help elicit more comments duringthe hour and scheduling should be primarily based noton a senior leaderrsquos availability but on the greatest accessto all providers and the ebb and flow of clinical intensityin the locations visited Scheduling WalkRounds at 5 AMin one of the hospitals allowed senior leaders to interactwith the night-shift personnel without major disruptionto clinical activities The night rounds tended to visit twounits in an hour to maximize exposure and because thoseshifts tended to be less fully staffed The most effective use of the time occurs if there is a choreographed set of steps performed on the infor-mation elicited The WalkRounds discussions must be carefully monitored and documented the contributingfactors relating to each comment must be identified and



continued

Table 4 Actions Taken in Various Departments in the Four Hospitals

Department Responsible Actions Taken (Comments)

Cardiology Stress TestingDevote one hour a day to inpatient testsModify low-risk chest pain protocol

Clinical Labs

During off-shifts supervisors will alter priorities if RN requests assistance with draw Blood Bank is working with NursingPharmacy on including bar-coding technology forinfusionsPatient identification process reinforced with phlebotomy staff and is now an annualcompetency for nursing staff and reviewed with house staff

Chief Medical Officer

Communication Department created an algorithm to act as a back-up system for pag-ing system outagesPartnersrsquo Web-based on-call system in place to enable consistent communicationregarding physician coverageWeb-based attending coverage has increased communication between MDsRNsregarding coverageSecurity will attend codes to help enforce crowd control Roles and responsibilitiesreviewed with code leaderThe new ADT system was trialed to assist in the reduction of wait time for meds byimproving the admitting process from PACU LampD or NICU to CWNtower floors

Dana Farber Cancer Institute

To improve the communication between the physicianrsquos office and the floor with directadmits from Dana Farber the patients should be admitted to the infusion room to bereceived by RN with a reportResident training reviewing guidelines regarding threshold for calling for helpAdmitting alerted to the need to not admit patients with the same name to the same podIS looking into putting flags on patients with similar spelling and sounding namesPain service relinquished PCA service to all services to improve response time Pain service will keep epiduralsRequest for ldquodo not recalculaterdquo message to appear for in-house patientrsquos orders

DialysisChange house staff bathroom to patient bathroomInstall bedpan washerPurchase fax machine

Emergency Department Liaison to greet and guide patients at check in

Education

RNs (3) attended Posey in-service on restraints and fallsIn-service on how to tie and untie restraintsAnnual nursing competencies will include hot packcold pack competenciesInstructions for reporting equipment in need of repair reviewed with staffSafety flip chart revised to include a section with instructions on how to handle equip-ment in need of repair

Engineering Maintenance

Call light pull cords replaced with more durable materialIntercom system on CWN8 replacedAdapting all beds for compatible call system plug-ins for 11C 4B and 16Carpet replaced

Equipment RepairRehab Departments pulse oximeters (s) repaired and returned to serviceUAs assigned to clean IV pole wheels and to identify any preventive maintenance issues



continued

Table 4 Actions Taken in Various Departments in the Four Hospitals (continued)


Equipment Acquisition

Footstools ordered to elevate patientrsquos legs when sitting in chairTwo Arjo lifts purchasedTwo Advantax alarm beds on 4th floor addedNew O2 ldquoErdquo cylinder rack purchased for the second floorO2 tubing transitioned to color tubingAdditional privacy curtains provided to Marblehead Outpatient siteAdditional commodes purchased including heavy duty for bariatric patientsAdditional bed alarms purchasedAdditional BP cuffs purchasedSix new stretchers purchasedAdditional Geri Chairs ordered

Facilities

4th floor has become the model for efficient use of space for future renovationsICUs are slated for redesign which will include installation of privacy doorsAdditional family waiting space has been added for ICUsAutomatic door installed in 8CReplace carpets and buckled floorsTubes replaced for pneumatic system

Facility Improvement

Wooden hand rails were checked and tightened throughout facilityVersaframes (safety handrails that are attached to toilets) were checked and tightenedin all bathroomsTelephone jack moved from external wall into Patient Family Lounge on the secondfloorHardware installed on windows in Middleton Outpatient site to prevent openinggreater than 6 inchesNew signage for clean and dirty precaution gown binsWindow shutters replaced by pull shades

Gastrointestinal Department Label equipment with instructions for disposalHouse Officers Install computer in on-call room

Materials Management

Determine appropriate par levels (standard numbers) for DNR forms on the floorsMattress-provided bedrail extenders to assist in preventing patient falls from beds withhigh air mattresseslow side railsBiohazard kits are now stocked on every pod and on-site for easy availability Equipment repair program in progress that will begin with an inventory of all existingequipment include a tracking and repair processIncreased supply of commode bucketsImproved materials management support through increased standardization andimprovements in customer service

Nursing

Alaris infusion pumps implementedNew monitors for neurology patients installedNursing policies updated to incorporate ldquoread backrdquo policy for all verbal and telephoneorders


433

all information must be placed into a database that allowscomments to be associated with actions Use of and inputinto the computerized database at the BWH was key tothe success of its rounds allowing reports to be generat-ed tying together comments and actions identifying lociof responsibility and facilitating effective feedback tofront-line providers and up the administrative ladderThose hospitals using paper and a spreadsheet programnoted that documentation was a time-consuming processthat was periodically put aside for issues that appearedmore urgent This undermined the power of the rounds WalkRounds can be performed easily in publicly openareas such as nursing stations and patient care hallwaysVisibility of the rounds was perceived as useful in pro-moting leadershiprsquos investment in safety and hospital

administrationrsquos interest in identifying problems andaddressing them Concern about confidentiality and thetype of sensitive topics that might be discussed (such asepisodes of patient harm) was initially voiced by all thehospitals organizations but turned out to be a nonissue The discussions during WalkRounds are influencedby who in leadership is participating their ability to qui-etly listen and whether they have clinical or nonclinicalbackgrounds Patient safety personnel influence the con-versations by how effectively they cite human factorsand systems theory All these factors affect front-lineworkersrsquo willingness to speak up The types of com-ments elicited at each hospital differed with a prepon-derance of the comments in the two rehabilitationhospitals centering on equipment issues It is possible


RN registered nurse MD physician ADT admission discharge transfer PACU postanesthetic care unit LampD labor and delivery NICU neonatal intensivecare unit CWN Center for Women and Newborns IS information systems PCA patient-controlled analgesia UA unit assistant IV intravenous BP bloodpressure ICU intensive care unit DNR do-not-resuscitate ID identification MRI magnetic resonance imaging C-spine cervical spine PA physician assis-tant FTE full-time equivalent CT computerized tomography



Obstetrics-Gynecology andLabor and Delivery

Plans for bar-coding will address the issue of new ID tags for babies because the cur-rent bands are not durableInfant security reviewed with parents to increase awareness regarding not leaving thefloor with babyDeveloped admissions criteria for newborns

Pediatrics

Tape over all numbers on microwave except ldquo30 secondsrdquo to prevent someone fromoverheating a hot pack Sign posted on microwave to be used for heating food ldquoFood OnlyrdquoReorganize IV solutions

Pharmacy

To improve communication regarding delays in meds pharmacists are to call if therewill be a delay in responding to text pagersNICU obtained a table top SuremedProposed in FY04 that Pharmacy mix all IVsCoumadin Protocol Sheet reinstitutedClarification that daily dosing medication dispensing time can be flexiblePolicy for splitting medication was reviewed

Radiology

Changes to MRI schedule to improve wait time for stat MRIC-spine protocol reviewed to identify failure modes Dynamic scheduling has improvedstandardization for intake information in attempt to increase communication betweenprovidersAlgorithm developed for nurses to call during patient emergencies in recoverymdashPAs andfellows to coverFTE added in Radiology CT for 3rd shiftPurchased additional MRI compatible pumps



that these issues arose because unlike acute care hospi-tals rehabilitation hospitals tend to have solid multidis-ciplinary teamwork structures designed for longerinpatient stays and must make do with less reimburse-ment for patients who are often no less acutely ill thanthose in acute care hospitals A concern of patient safety personnel was that they would be assigned to address the problems they

uncovered whereas the appropriate process would be touse the patient safety personnel to identify problems andthen hand off to others the responsibility for actionFormalized associations must be made to delineate howthe patient safety personnel will interact with middle man-agers and executives in identifying who should be respon-sible for actions The BWH after a year of WalkRoundsinvited directors and middle managers to participate asthe individuals most likely to shepherd actions to fruitionThe vice president of materials management now attendssome rounds However each hospital discovered that toolarge a group hindered open and rich discussion and iden-tified an optimal number of individuals to participate TheBWH found that ideally no more than three to four indi-viduals should visit the area designated for theWalkRounds although they generally had four to five inthe group The size of the group in total would vary byhow many individuals from the floor participated Surveys of participants revealed that four out of fiveco-workers later discussed the rounds with their peersHowever participants at the two hospitals with a Web-based incident reporting system did not believe that therounds increased event reporting The surveys were alsouseful in eliciting ideas for improving the rounds withthe most common suggestions being to include more anddifferent types of staff and to have the rounds occurmore frequently Categorization of data in two of the hospitals startedwith Vincentrsquos criteria but these were modified usinggrounded theory27 (that is building categories from thedata rather than creating the categories and then assign-ing the data) when categories were lacking For exam-ple issues related to computerized physician order entrywere common in one organization and this required aspecific category Supply unavailability was subdividedon the basis of where the supplies originated therebymaking it easier to identify who should be responsiblefor addressing an identified problem Categorization of data by severity scoring (frequency timesharm or likelihood of harm) is done in most of the hospi-tals but is difficult to use in resource allocation It is help-ful however in identifying trends Completion of most ofthe actions has been based on ease of implementationrather than risk of harm We note this with particularinterest because of its possible implications On the one

CEO ldquoThe WalkRounds reminds me to pay attentionto the day-to-day issues that confront staff andthis awareness is in my head when making biggerdecisions For example the prioritization and speedof resources For example we bought the OR equip-ment necessary for operating on very largepatients but the WalkRounds helped to speed upthe purchase of equipment for managing largepatients on the floors This amounted to about a$30000 expenditurerdquo

CEO ldquoItrsquos been helpful in getting me out to hearfrom the staff I hear about issues from the exec-utives and lsquohigher upsrsquo but the rounds help to clarifymy perceptions and to alleviate misperceptions alsoto talk to staff about these concerns Itrsquos fascinatingand helpful to hear the front-line perspective Otheractions wersquove taken that we wouldnrsquot have been asquick to act on include the development of a liaisonposition in the emergency department The writingof the job description was affected by theWalkRounds and because of my insights I was ableto discuss directly with staff how the position shouldbe used and to articulate to the staff what their per-ceptions about the position should be Wersquore alsoworking to reconfigure the intake area and intakeprocess in the emergency department Hearing fromthe staff during the WalkRounds about the difficul-ties there helped to push that along fasterrdquo

VP of Patient Care Services ldquoIn regards to person-al education and insights I enjoy doing theWalkRounds personally as I see it as a way to con-nect with the staff during their routine activitiesinstead of just at staff meetings or informal lunch-es I feel the lsquoblameless culturersquo engendered ishealthy for any organization and truly fosters alearning environmentrdquo

CEO chief executive officer OR operating room VP vice president

Table 5 Sample of Senior Leadersrsquo Reactionsto WalkRounds


435

hand this might indicate that middle managersrsquo goal is toact judiciously and fix problems most amenable tochange The other more ominous possibility is that mid-dle managers donrsquot feel they are adequately positioned toact or the climate around them is appropriately respon-sive to risk and therefore stop trying to make changeseven when they know risk is elevated28 Although true forthose problems documented in the databases this wasless true when evaluating the influence of theWalkRounds on executive behavior overall as noted bysenior leadersrsquo comments provided in this article At the BWH initially only negative comments elicitedduring the WalkRounds were placed into the databaseOver time positive comments have been included tooAs a result reports to the participating senior leadershave a more balanced perspective of operations and mayfacilitate further support of actions perceived as usefulby staff These positive comments are now also trackedso that changes in their number may be monitored Before WalkRounds is conducted in a specific area ofthe organization review of previous WalkRoundsrsquo com-ments from that floor or unit should be performed Onarrival of the WalkRounds group the comments should bediscussed including any actions taken or planned toaddress these previously identified issues Questions aboutthe effect of those actions and plans should be elicited Finally tying the steps together in introducingWalkRounds requires modest resources and ensureseffective use of leadership time (Table 6 right) Applyingresources to this process helped promote an environ-ment preoccupied with safety and supportive of greateroperational transparency and effectiveness

Limitations and Next StepsAre the resources and leadership time spent to sup-

port the WalkRound process appropriate if the majorityof the changes that resulted are minor Indeed many ofthe WalkRounds actions were perceived to be small anda few even inconsequential There are three answers tothis question

First we did not prospectively evaluate whether theWalkRounds experience influenced leadership in makingmajor budget allocations or operational decisionsHowever leadersrsquo comments in our debriefs with themdid indicate that WalkRounds influenced decision making


Week 1 Introductory session to Leadership and Safety and

Quality Personnel Introductory session to middle management Teach patient safety personnel how to use comput-

erized database to collect and manage information Perform a ldquopilotrdquo WalkRounds to test concept fol-

lowed by discussion about data collected andplacement of information into database

Week 2 Identify a central and appropriately authorized

committee (one run by the chief operating officeror equivalent) to whom WalkRounds data will bediscussed and actions assigned WalkRounds shouldbe a standing agenda item for this committee

Identify how patient safety personnel will learn aboutand track actions (eg participating on committeeand debriefs with leadership on a regular basis)

Week 3 Send out hospitalwide notice of plans to begin

WalkRounds Ask for floors to volunteer to be the first Identify and develop with assistance from Marketing

if feasible feedback mechanisms Perform pilots offeedback and reporting

Develop feedback process for immediately afterrounds to those who participated about the con-cerns discussed that day

Develop feedback plan to specific locations andindividuals about actions takenmdashthis could be daysor even months later

Develop monthly or periodic report for the opera-tions committee

Develop a report to the Board of Trustees (or itsquality subcommittee)

Week 4 Identify leadership to participate in WalkRounds Schedule WalkRounds for 6 months to one year Write and sign performance agreements for those

participating Leaders agree to participate not canceland perform X number of WalkRounds per yearPatient safety personnel agree to manage data andfeedback in a timely fashion Operations committeeagrees to complete action items in X months withgoal to improve cycle times by Y percent in one year

Table 6 Introductory Steps to ImplementingEffective WalkRounds



about some major expenses that were not noted in thehospital databases primarily because the keepers of thedatabase had no mechanism to elicit that informationfrom the senior executives Further research is requiredand would likely entail direct observation of WalkRoundsand then executive-level and budget-related meetings toformally associate the discussions

Second there is reasonable evidence to suggest thathighly effective industries achieve success becausefront-line workers participate in decision making open-ly voice concerns and are able to modify their environ-ments even to address the small operational failuresthat lead to inefficiency or error29ndash31 The conundrum isthat addressing small operational failures often requiresan interdisciplinary effort that is beyond the capacity ofan individual or unit but is feasible through facilitationby leadership89 Resolving many small issues may beprecisely the goal of WalkRounds However aside fromthe sometimes-substantial benefit achieved by manysmall interventions an alliance formed between thefront line and leadership in the course of WalkRoundsmay ultimately make it easier to implement major

improvementsmdashand improvements that are more likelyto be successful Such major improvements would notnecessarily be evident in the WalkRounds databasebecause it is replete with the concerns of persons whoview predominantly their own domains and unlike theview we hope exists in the executive suite who donrsquothave an overview of the whole system

Third aside from the actions taken there is early evi-dence that WalkRounds quickly positively affects nurs-ing perceptions of their work environments acomponent likely to influence nursing retention ratesData collection and actions taken will ultimately consti-tute only one measure of WalkRoundsrsquo total effect Itsstrongest effect may lie in its improvement of the abilityto hold onto nurses in a climate of dwindling nursingcapacity and increasing health care demand3233

The research remains to be done and will requiresome years to produce to answer the question ofwhether addressing many small processes is a key com-ponent of safe care However some shining examples ofsuccess based on a constant drumbeat of attention todetail are evident outside health care112934 Proving thevalue of the WalkRounds based on outcome data is

going to be a daunting task and not one we were able todo in this study We focussed more on implementingWalkRounds but continue our studies to assess theeffect of these rounds

Are there other tools to achieve this goal PossiblyMultidisciplinary root cause analyses as developed in theVeterans hospitals35 brings together leadership and front-line workers to address problems and behaviorally basedteamwork training in various models is sweeping intohealth care and promotes leadership-team memberalliances and the voicing of concerns36 Each will have itsbenefits All implemented together this toolkit couldreshape how we deliver organized health care

ConclusionsWalkRounds appears to affect resource allocation influ-ences actions taken to improve safety quality and effi-ciency and appears to be an effective tool for engagingleadership identifying safety issues and supporting aculture of safety If supported by teamwork training andappropriate accountability policies WalkRounds willhelp lead to the culture of safety

The authors thank Nancy Kruger Gary Gottlieb Andy Whittemore TonySciola Maurice Greenbaum Jack Skowronski Richard ZaniewskiLeslie Selbovitz Paulette DiAngi Elaine Bridge and Michael Jellinekfor their outstanding leadership and willingness to give time and feed-back to both implementing and researching WalkRounds

J

Allan Frankel MD is Director of Patient SafetyPartners Healthcare Boston Sarah Pratt Grillo MHAis Project Manager Health Research and EducationalTrust Chicago Erin Graydon Baker is Patient SafetyManager Brigham and Womenrsquos Hospital (BWH) Bostonand Camilla Neppl Huber is Patient Safety ProjectAnalyst Susan Abookire MD formerly Director ofQuality and Safety Newton-Wellesley Hospital (NWH) Newton Massachusetts is Chair Quality andPatient Safety Mount Auburn Hospital CambridgeMassachusetts Marianne Grenham is Director of Quality and Risk Management Spaulding RehabilitationHospital Boston and Pam Console is AdministrativeAssistant Quality Management Mary OrsquoQuinn isDirector of Quality Management Shaughnessy-KaplanRehabilitation Hospital Salem Massachusetts GeorgeThibault MD is Vice President of Clinical AffairsPartners Healthcare Tejal K Gandhi MD is Director ofPatient Safety BWH Please address reprint requests toAllan Frankel MD AfrankelPartnersorg



1 Brennan TA et al Incidence of adverse events and negligence inhospitalized patients Results of the Harvard Medical Practice Study lN Engl J Med 324370ndash376 Feb 7 19912 Institute of Medicine To Err Is Human Building a Safer Health

System Washington DC National Academy Press 20003 Kuperman GJ Gibson RF Computer physician order entryBenefits costs and issues Ann Intern Med 13931ndash39 Jul 1 20034 Milstein A et al Improving the safety of health care The Leapfrog Initiative Eff Clin Pract 3313ndash316 NovDec 2000 httpwwwacponlineorgjournalsecpnovdec00milsteinhtm (lastaccessed Jun 16 2005)5 Milstein A Adler NE Out of sight out of mind Why doesnrsquot wide-spread clinical quality failure command our attention Health Aff

(Millwood) 22119ndash127 MarndashApr 20036 McManus ML Variability in surgical caseload and access to inten-sive care services Anesthesiology 981491ndash1496 Jun 20037 Litvak E et al Emergency department diversion Causes and solu-tions Acad Emerg Med 81108ndash1110 Nov 20018 Tucker AL Edmondson AC Spear SJ When problem solving pre-vents organizational learning Journal of Organizational Change

Movement 15(2)122ndash137 20029 Tucker AL Impact of operational failures on hospital nurses andtheir patients Journal of Operations Management 22151ndash169 Apr200410 Arndt M How OrsquoNeill Got Alcoa Shining Business Week Feb 52001 p 39 11 Freiberg K Freiberg J Nuts Southwest Airlinesrsquo Crazy Recipe for

Business and Personal Success Austin TX Bard Press 199612 Auty S Long G Tribal warfare and gaps affecting internal servicequality International Journal of Service Industry Management

10(1)7ndash22 199913 Peters LH OrsquoConnor EJ Eulberg JR Situational constraintsSources consequence and future considerations In Rowland KFerris G (eds) Research in Personnel and Human Resources

Management vol 3 Greenwich CT JAI Press 1985 pp 79ndash11414 Nelson R Winter S Organizational Capabilities and Behavior

An Evolutionary Theory of Economic Change Cambridge MAHarvard University Press 1982 15 Joint Commission on Accreditation of Healthcare OrganizationsComprehensive Accreditation Manual for Hospitals The Official

Handbook Oakbrook Terrace IL Joint Commission Resources 2004 16 National Quality Forum (NQF) The National Quality Forum Safe

Practices for Better Healthcare A Consensus Report WashingtonDC NQF 200317 NQF A National Framework for Healthcare Quality Measurement

and Reporting Washington DC NQF 200218 McGlynn EA et al Establishing national goals for quality improve-ment Med Care 41(1 suppl)I16ndashI29 Jan 2003

19 The Leapfrog Group Leapfrog Survey Summary (results as ofOctober 31 2003) httpwwwleapfroggrouporgReadoutpdf (accessedFeb 14 2004)20 Birkmeyer JD et al Leapfrog Safety Standards Potential Benefits

of Universal Adoption Washington DC The Leapfrog Group 200021 Ranger CA Bothwell S Making sure the right patient gets the rightcare Qual Saf Health Care 13329 Oct 2004 22 National Health System National Patient Safety Agency SevenSteps to Patient SafetymdashStep 1 http81144177110sitemediadocuments492_Final20Step201pdf (last accessed Apr 29 2004)23 Frankel A et al Patient Safety Leadership WalkRoundsTM Jt

Comm J Qual Safety 2916ndash26 Jan 200324 Weick K Managing the Unexpected Assuring High Performance

in an Age of Complexity San Francisco Jossey-Bass 200125 Billings CE Some hopes and concerns regarding medical event-reporting systems Lessons from the NASA safety reporting systemArch Pathol Lab Med 122214ndash215 Mar 199826 Vincent C Taylor-Adams S Stanhope N 1998 Framework for ana-lyzing risk and safety in clinical medicine BMJ 3161154ndash1157 Apr 11199827 Kuhn T The Structure of Scientific Revolutions ChicagoUniversity of Chicago Press 197028 MacDuffie JP The road to root cause Shop-Floor problem-solvingat three auto assembly plants Management Science 43(4)479ndash502199729 Edmondson AC Bohmer RM Pisano GP Disrupted routinesTeam learning and new technology implementation in hospitalsAdministrative Science Quarterly 46(4)685ndash716 200130 Roberts K New challenges in organizational research High relia-bility organizations Industrial Crisis Quarterly 3(3)111ndash125 1989 31 Roberts K Some characteristics of one type of high-reliabilityorganizations Organization Science 1(2)160ndash167 199032 Thomas EJ et al The effect of executive walk rounds on nursesafety climate attitudes A randomized trial of clinical units BMC

Health Serv Res 528 Apr 11 2005 33 Thomas EJ et al Correction The effect of executive walk roundson nurse safety climate attitudes A randomized trial of clinical unitsBMC Health Serv Res 546 Jun 10 2005 34 Grabowski M Roberts K Risk mitigation in large-scale systemsLessons from high-reliability organizations California Management

Review 39152ndash162 Summer 1997 35 Khuri SF Daley J The National Surgical Risk Program In BarbourGL (ed) Redefining a Public Health System How the Veterans

Health Administration Improved Quality Measurement SanFrancisco Jossey-BassScribner 1996 pp 156ndash16636 Leonard M Frankel A Simmonds T Achieving Safe and Reliable

Healthcare Strategies and Solutions Chicago Health AdministrationPress 2004

References



their environment rather than speak up about themThese nurses like physicians were no doubt trained todisplay (albeit never manage) endless capacity for vigi-lance and adaptability

Traditionally health care has had a culture of blamewhere people are punished for making errors Patientsafety efforts strive to create a culture of safety wherefront-line staff are comfortable speaking up about errorsand adverse events and leadership encourage this dia-logue Health carersquos need to better address this problemis underscored by the Joint Commission on Accreditationof Health Care Organizationsrsquo requirement that leader-ship participate in promoting safer care delivery15 and inthe emphasis on the development of a ldquoCulture of Safetyrdquoas articulated in the National Quality Forumrsquos SafePractices16ndash18 and The Leapfrogrsquos Group 4th Leap1920 and bythe United Kingdomrsquos National Patient Safety Agency2122

Patient Safety Leadership WalkRoundsTM 23 is a simplebut rigorous management tool designed to assist hospital leaders in implementing mechanisms for pro-moting safety learning about and hearing the concernsof front-line providers supporting appropriate accounta-bility concepts and allocating resources to areas ofgreatest risk Brigham and Womenrsquos Hospital (BWH)began WalkRounds in January 200123 which generatedinterest by other organizations in Partners HealthCare(Partners is composed of two acute care tertiary organizations four community hospitals two reha-bilitation hospitals and a psychiatric hospital) TheShaughnessy-Kaplan Rehabilitation Hospital beganimplementing WalkRounds in November 2001 followedby Newton-Wellesley Hospital (NWH) and SpauldingRehabilitation Hospital in January 2002 (Other Partnershospitals have since also started WalkRounds) We exam-ined these four hospitalsrsquo joint experience to address ina natural experiment the following questions Would hospital leadership be willing to openly dis-cuss operational failure safety and harm with front-lineproviders Would frank and open discussion occur in a publicsetting Could the information elicited be collected and aggre-gated in a useful manner Would the information collected affect actions orresource allocation

Implementing WalkRoundsAs the leadership and safetyquality personnel in the fourhospitalsmdashBWH Spaulding Rehabilitation Shaughnessy-Kaplan Rehabilitation and NWHmdashagreed to implementWalkRounds we [AF SPG] gave a two-hour presenta-tion to the hospital executives and members of the safetyand quality departments about theories promoting leader-ship involvement in safety and quality11 high reliability24

blame-free reporting25 and useful data categorization26 Aframework and timeline for implementation were sug-gested including a method for aggregating data Data col-lection began either on paper or in a spreadsheetprogram but BWH soon after beginning WalkRoundsdeveloped a computerized database to facilitate datamanipulation and to improve reporting of informationThe software was available for use by the other hospitals

Although WalkRounds inevitably is shaped by individ-ual personalities and organizational culture some basiccomponents as shown in Table 1 (page 425) wereencouraged After those components were suggestedleadership and safetyquality officers considered theprocess and resources available Some organizationschose to perform WalkRounds weekly others biweeklyor monthly Regardless of the scheduling WalkRoundswere occasionally cancelled secondary to executiveschedules or because the floors were too busy when theWalkRounds groups arrived (Table 2 page 426)

Mechanisms for feedback for reporting of commentsprovided during WalkRounds is described and depictedin Figure 1 (pages 427ndash428) At all four hospitals the per-sons charged with responsibility for patient safety par-ticipated in implementing WalkRounds The role of thescribe (Table 1) was assigned to persons with markedlydifferent positionsmdasha patient safety project analyst theadministrative assistant to a patient safety director thesenior administrative secretary for a quality departmentand a quality improvement coordinator

Different combinations of administrative and clinicalleaders chose to participate at each hospital (Table 2)Leadership involvement seemed to be determined bypersonal interest and by who was initially asked Inthose hospitals where the chief executive officer showedinterest more involvement by a greater variety of per-sonnel was apparent No chief information officers orchief financial officers were asked to join initially and


425




























Hospital









65 35 20 10

















Worksheet









continued








429














(NN = 924)


(NN = 89)


Hospital

(NN = 221)


Hospital















continued




Clinical Labs








Education







continued





Facilities







Nursing



433









Pediatrics


Pharmacy


Radiology












435


































J























References


425




























Hospital









65 35 20 10

















Worksheet









continued








429














(NN = 924)


(NN = 89)


Hospital

(NN = 221)


Hospital















continued




Clinical Labs








Education







continued





Facilities







Nursing



433









Pediatrics


Pharmacy


Radiology












435


































J























References








Hospital









65 35 20 10

















Worksheet









continued








429














(NN = 924)


(NN = 89)


Hospital

(NN = 221)


Hospital















continued




Clinical Labs








Education







continued





Facilities







Nursing



433









Pediatrics


Pharmacy


Radiology












435


































J























References




continued








429














(NN = 924)


(NN = 89)


Hospital

(NN = 221)


Hospital















continued




Clinical Labs








Education







continued





Facilities







Nursing



433









Pediatrics


Pharmacy


Radiology












435


































J























References








429














(NN = 924)


(NN = 89)


Hospital

(NN = 221)


Hospital















continued




Clinical Labs








Education







continued





Facilities







Nursing



433









Pediatrics


Pharmacy


Radiology












435


































J























References














continued




Clinical Labs








Education







continued





Facilities







Nursing



433









Pediatrics


Pharmacy


Radiology












435


































J























References



continued





Facilities







Nursing



433









Pediatrics


Pharmacy


Radiology












435


































J























References


433









Pediatrics


Pharmacy


Radiology












435


































J























References











435


































J























References












J























References






















References


patient safety leadership walkrounds tm partners · pdf filepatient safety leadership...

Documents