perating Room Teamworkmong Physicians and Nurses:eamwork in the Eye of the Beholderartin A Makary, MD, MPH, J Bryan Sexton, PhD, Julie A Freischlag, MD, FACS,hristine G Holzmueller, BLA, E Anne Millman, MS, Lisa Rowen, RN, DNSc, Peter J Pronovost, MD, PhD
BACKGROUND: Teamwork is an important component of patient safety. In fact, communication errors are themost common cause of sentinel events and wrong-site operations in the US. Although efforts toimprove patient safety through improving teamwork are growing, there is no validated tool toscientifically measure teamwork in the surgical setting.
STUDY DESIGN: Operating room personnel in 60 hospitals were surveyed using the Safety Attitudes Question-naire. Surgeons, anesthesiologists, certified registered nurse anesthetists, and operating room nursesrated their own peers and each other using a 5-point Likert scale (1 very low, 5 very high).
RESULTS: Overall response rate was 77.1% (2,135 of 2,769). Ratings of teamwork differed substantiallyby operating room caregiver type, with the greatest differences in ratings shown by physicians:surgeons (F[4, 2058] 41.73, p 0.001), and anesthesiologists (F[4, 1990] 53.15, p 0.001).The percent of operating room caregivers rating the quality of collaboration and communication ashigh or very high was different by caregiver role and whether they were rating a peer or anothertype of caregiver: surgeons rated other surgeons high or very high 85% of the time, and nursesrated their collaboration with surgeons high or very high only 48% of the time.
CONCLUSIONS: Considerable discrepancies in perceptions of teamwork exist in the operating room, withphysicians rating the teamwork of others as good, but at the same time, nurses perceive team-work as mediocre. Given the importance of communication and collaboration in patient safety,health care organizations should measure teamwork using a scientifically valid method. TheSafety Attitudes Questionnaire can be used to measure teamwork, identify disconnects betweenor within disciplines, and evaluate interventions aimed at improving patient safety. ( J Am Coll
Surg 2006;202:746752. 2006 by the American College of Surgeons)
rrors in the operating room (OR) can have catastrophiconsequences for patients, families, caregivers and entirenstitutions. Retained sponges, wrong-site operations,
ompeting Interests Declared: None.upported by the Agency for Healthcare Research and Quality, grant num-ers 1UC1HS014246 and 1PO1HS1154401.resented at the American College of Surgeons 91st Annual Clinical Con-ress, San Francisco, CA, October 2005.
eceived November 15, 2005; Revised January 20, 2006; Accepted January0, 2006.rom the Departments of Surgery (Makary, Freischlag, Millman, Pronovost)nd Anesthesiology (Sexton, Holzmueller, Pronovost), Johns Hopkins Uni-ersity School of Medicine; Department of Health Policy and ManagementMakary, Millman, Pronovost), Johns Hopkins Bloomberg School of Publicealth; Department of Nursing (Rowen, Pronovost), John Hopkins Univer-
ity School of Nursing, and the Johns Hopkins Quality and Safety Researchroup (Makary, Sexton, Holzmueller, Millman, Pronovost), Johns Hopkinsedical Institutions, Baltimore, MD.orrespondence address: Martin A Makary, MD, MPH, Department of Sur-ery and Health Policy and Management, Johns Hopkins University Schoolf Medicine, Johns Hopkins Medical Institutions, 4940 Eastern Ave, Bldg
n-5, Baltimore, MD 21224. email: email@example.com
7462006 by the American College of Surgeons
ublished by Elsevier Inc.
ismatched organ transplants or blood transfusions cane the result of interpersonal dynamics, where commu-ication and collaboration breakdowns occur amongR team members.1-3 The Joint Commission on Ac-
reditation of Healthcare Organizations recently identi-ied breakdowns in communication as the leading rootause of wrong-site operations, and other sentinelvents.4 Teamwork is an integral component of a culturef good communication in the OR5 and, accordingly, isn important surrogate of patient safety. To this end, the999 Institute of Medicine report on medical error con-luded that hospitals need to promote effective teamunctioning as one of five principles for creating safeospital systems.6 The Joint Commission on Accredita-ion of Healthcare Organizations proposed that hospi-als measure culture beginning in 2007. A reliable andidely used measurement tool for the OR setting does
ot currently exist.
747Vol. 202, No. 5, May 2006 Makary et al Teamwork in the Operating Room
Attitudes about teamwork are associated with erroreduction behaviors in aviation,7 with patient outcomesn intensive care units,8-10 and with nurse turnover in the
R.11 Good teamwork is associated with better job sat-sfaction,12 and less sick time taken from work.13 Dis-repant attitudes about teamwork have been suggesteds a considerable source of nurses dissatisfaction withheir profession14 that has led to the critical nursinghortage.15 They might be a root cause of errors in oper-tions, and surgeons are increasingly pressured to pre-ent negative outcomes.5 In the name of patient safety,here has been a plethora of new programs andrainingwith varying degrees of success. These initia-ives represent a stride in the right direction, but they areoid of reliable metrics to measure their effect on team-ork. We developed and validated a survey to measure
eamwork in the surgical setting. In this study, we usedhis tool to compare ratings of teamwork within andetween OR caregivers.
ETHODSur survey, the Safety Attitudes Questionnaire (SAQ)16
s a refinement of the Intensive Care Unit Managementttitudes Questionnaire.17,18 The latter was adapted
rom the Flight Management Attitudes Questionnaire19
nd its predecessor, the Cockpit Management Attitudesuestionnaire.20 These surveys are reliable, sensitive to
hange,21 and the elicited attitudes shown to predict per-ormance.7,22,23 There is a 25% overlap in item contentetween the SAQ and Flight Management Attitudesuestionnaire. We improved content validity and cre-
ted an OR version of the SAQ after reviewing the liter-ture on teamwork in the OR, conducting focus groups,nd asking OR caregivers to review the survey for con-ent relevance. Previous research suggested differences inerceptions of OR teamwork by OR caregiver type,17
nd, to this end, we focused on the ratings of teamworkhat OR caregivers give to one another. We used theommunication and collaboration section of the SAQ,here the respondent is asked to describe the quality of
Abbreviations and Acronyms
CRNA certified registered nurse anesthetistsOR operating roomSAQ Safety Attitudes Questionnaire
ommunication and collaboration you have experienced a
ith: eg, surgeons, anesthesiologists, surgical techni-ians, certified registered nurse anesthetists (CRNA),nd OR nurses (1 very low, 2 low, 3 adequate, high, 5 very high).The SAQ (Operating Room Version) was adminis-
ered to all OR caregivers in a Catholic health systemomprised of 60 hospitals in 16 states in July and Augustf 2004. No one was excluded and OR caregivers in-luded surgeons, anesthesiologists, surgical technicians,RNAs, and OR nurses. Random sampling was notsed because small sample sizes in caregiver positionsithin a hospital, instead, highly representative response
ates were sought from each institution. Surveys weredministered during preexisting departmental and staffeetings, with a pencil and return sealable envelope toaintain confidentiality. Individuals not captured in pre-
xisting meetings were hand-delivered a survey, pencil, andeturn envelope. All surveys were anonymous to thearegivers name but not to caregiver type or hospital.
tatistical analysissing ANOVA, we tested for differences in ratings of
ommunication and collaboration (previously calledeamwork ratings) that surgeons, anesthesiologists, surgi-al technicians, CRNAs, and OR nurses gave to eachther. In addition to the means used in ANOVA, we alsoresent the percent rating teamwork highly (high or veryigh) for each caregiver type. All statistical analyses wereerformed using SPSS version 12.0.
ESULTSf 2,769 questionnaires handed out in 60 hospitals
222 surgeons, 1,058 OR nurses, 564 surgical techni-ians, 170 anesthesiologists, and 121 CRNAs), a total of,135 surveys were returned, for an overall response ratef 77.1% (range across hospitals of 57% to 100%). ORurses (79%) had the highest response rate and CRNAsad the lowest (67%) (Table 1). Average respondent was3 years old with 10 years of experience at the currentospital. Surgeons (8.6% women) and anesthesiologists12.7% women) were predominantly men.
eamwork ratingseamwork ratings for each OR caregiver differed con-iderably by caregiver type, with the largest differences inerceptions of teamwork between physicians and non-hysicians. Table 2 shows the mean ratings of teamwork
nd ANOVA results. Physicians had the lowest overall
748 Makary et al Teamwork in the Operating Room J Am Coll Surg
atings of teamwork (3.68 of 5.00) and OR nursesscrub and circulating) were given the highest ratings ofeamwork (4.20 of 5.00). This, despite the fact that sur-eons and anesthesiologists rated teamwork within theirwn discipline the highest, their group received the low-st ratings overall. In addition, OR nurses, who wereiven the highest overall ratings of teamwork, ratedeamwork with surgeons as only 3.52 of 5.00, relative tohe higher ratings surgeons gave OR nurses (4.42 of.00).
Each OR caregiver rated teamwork with their ownolleagues highly within their peer group at their hospi-al. Surgeons rated teamwork among surgeons highly,ith 85.2% describing the teamwork with surgeons as
high or very high (Fig. 1). Similarly, anesthesiologistsated teamwork among anesthesiologists very highly andRNAs rated CRNAs very well (scores were 95.8 and2.7, respectively). In fact, surgeons perceived thatveryone in the OR is doing a good job in terms ofeamwork (Fig. 2). Figures 3A, 3B, and 3C display theontrast between surgeons and nurses, surgeons and an-sthesiologists, and anesthesiologists and nurses, respec-ively, and Figures 4A and 4B demonstrate interpositionifferences in teamwork among all members of the OR.uch differences underscore the disconnect in teamworknd the methodological barrier in aggregating measuresf teamwork in surgery.
able 1. Characteristics of Respondents Surveyed and Res
urgeon 73 222/305 48.3 9.9nesthesiologist 77 170/220 45.8 9.3RNA 67 121/181 44.6 10.R nurse 79 1,058/1,335 43.3 10.otal 77 2,135/2,769 42.6 11.
Values are mean SD.RNA, certified registered nurse anesthetist; OR, operating room.
able 2. ANOVA Results for Teamwork Ratings by and of Ea
atings of df F p Value Surg
urgeons 4, 2058 41.73 0.001 4nesthesiologists 4, 1990 53.15 0.001 4RNAs 4, 1571 37.36 0.001 4R nurses 4, 2061 12.93 0.001 4
urgical technicians 4, 2044 6.17 0.001 4
1 very low; 5 very high.Scrub and circulating.
RNAs, certified registered nurse anesthetists; df, degrees of freedom; OR, operati
ISCUSSIONubstantial discrepancies in perceptions of teamwork ex-st in the OR, with physicians rating the teamwork ofthers as good, and at the same time, nurses perceiveeamwork as poor. These findings mirror similar resultsf discrepant attitudes about collaboration betweenhysicians and nurses in intensive care units.18
Based on our findings, surgeons and anesthesiologistsppear more satisfied with physiciannurse collabora-ion than nurses. Nurses did not reciprocate the highatings of teamwork given by physicians. This mightave been a result of fundamental and long-standingifferences between nurses and physicians, includingtatus, authority, gender, training, and patient-care re-ponsibilities. It might also be a result of different ideasf what constitutes effective teamwork. Discussionsith respondents during survey feedback presentationsighlighted that nurses often describe good collabora-ion as having their input respected, and physicians of-en describe good collaboration as having nurses whonticipate their needs and follow instructions. Histori-ally, there are differences between the expectations thathysicians and nurses bring to a communication en-ounter. Nurses are trained to communicate more holis-ically, using the story of the patient, and physiciansre trained to communicate succinctly using the head-
e Rates by Operating Room Caregiver Position
Women Experience inposition (y)*
hospital (y)*n %
8.6 19 17.4 9.41 12.3 9.2012.7 21 15.8 8.18 10.6 8.6050.0 63 14.7 12.32 9.5 9.3589.0 942 13.9 10.04 10.7 8.6968.5 1,462 13.7 10.47 10.0 9.08
perating Room Provider TypeMean ratings* of teamwork by
OverallAnesthesiologists CRNAs OR nurses
4.03 3.72 3.52 3.684.80 4.25 3.85 3.964.58 4.67 3.94 4.044.31 4.10 4.25 4.204.17 3.95 4.07 4.10
749Vol. 202, No. 5, May 2006 Makary et al Teamwork in the Operating Room
ines.24 Differences in communication expectations andechniques might have roots in medical and nursing ed-cational cultures.
pproachabilityood teamwork-related behaviors can lead to better pa-
ient outcomes.25 One of the best-studied laboratories ofhis science has been the aviation industry. Research inommercial aviation has demonstrated important tiesetween teamwork and performance.26 The link be-ween teamwork and safety was most obvious after planerash investigations exposed cockpit crew members re-uctance to question a captains performance as a rootause of aviation accidents. Surveys to assess culture inhe cockpit and predict performance were subsequentlyeveloped.27 We applied the aviation model to a culturalssessment of teamwork in medicine and found similarntimidation or lack of approachability barriers.17
urse hesitancy to express concernshe willingness of personnel to speak up about aatient-safety concern is an important part of safety inhe operating room. The traditional hierarchy of surgeryas often discouraged speaking up to a surgeon, andurses can be hesitant to confront a surgeon on issues ofatient care because they might have less training or
OR Nurses rateOR Nurses
igure 1. Teamwork as viewed within peer groups by operating roomaregiver role. Anesthes., anesthesiologist; CRNA, certified registeredurse anesthetists; OR, operating room.
xperience in dealing with a patients medical condition. a
n addition, there might be social barriers involving race,ender, and socioeconomic status. A nurses perceptionf...