Table 1

Main results of the reviewed studies

Author	Setting and subjects	Design method	Results and comments
Alvarez and Coiera52	Intensive care unit (ICU) teaching hospital; trauma centre and ICU 3 senior and 3 junior registrars, and 3 nurses	Observations Recorded conversations with lapel microphones	345 total conversation-initiating interruptions 492 total turn-taking interruptions
Blum and Lieu7	Children's hospital; teaching hospital 18 interns	Interns kept logs and recorded the pages they received	235 pages interrupted patient care activities 126 pages interrupted scheduled work rounds and educational conferences 88 pages interrupted personal care activities
Brixey et al15	Trauma section of the emergency department (ED) of a large teaching hospital 5 attending ED physicians and 8 nurses	Observations using a semistructured field note form on a tablet PC	Physicians experienced roughly 10 interruptions per hour Nurses experienced approximately 12 interruptions per hour Sources of interruptions: telephone, pager, other people, self After being interrupted, physicians and nurses resumed their primary task only after they performed 1–8 other tasks
Chisholm et al14	ED of urban teaching hospital; ED of suburban private teaching hospital; ED of rural county community hospital 30 emergency physicians (EPs)	Observations using standardised data collection form Conducted task analysis	Mean number of interruptions: 30.9 (9.7) Mean number of break-in-task: 20.7 (6.3)
Chisholm et al8	EDs of 5 non-teaching community hospitals; 22 primary care offices 22 EPs and 22 office-based primary care physicians (PCPs)	Observations Task analysis	EP's mean interruptions per hour (by source): total=9.7; care=5.6; patient=0.4; non-patient=2.3; telephone=1.4; break-in-task=5.4 PCPCs mean interruptions per hour (by source): total=3.9; care=1.9; patient=0.2; non-patient=1.4; telephone=0.6; break-in-task=1.8
Coiera et al53	ED of rural hospital; ED of urban tertiary teaching hospital 6 nurses and 6 doctors	Observations Recorded conversations with lapel microphones	Total interruptions: 393; nurses' interruptions: 185; doctors' interruptions: 208 Rate of interruptions for all subjects: 11.15 per hour; nurses' rate of interruptions: 11.2 per hour; doctors' rate of interruptions: 11.1 per hour
Coiera and Tombs54	Teaching hospital 8 physicians from general medicine and 2 nurses from the medical ward	Observations Recorded conversations with lapel microphones Interviews	Participants generated 43 pages and received 23 pages Participants generated 65 phone calls and received 31 phone calls
Dearden et al20	Inner city patient practice 1 general practitioner (GP) and 102 patients	Phase I: observations Phase 2: survey	Phase I:63 interrupted consultations (interruption rate=10.2 %) Sources and number of interruptions: phone=31; forms/prescription=24; other=8 Phase II: 65% of patients unaffected by the interruption 18% of patients had negative feelings about interruptions
Flynn et al21	Pharmacy in non-government, not-for-profit general medical–surgical hospital 14 pharmacists and 10 technicians	Tested visual acuity, hearing and distractibility Observations Video taped	2022 interruptions were detected, affecting 1143 prescription sets. Error rate of interrupted prescriptions sets=6.65% 2457 distractions were identified, affecting 1329 prescription sets Error rate of distracted prescriptions sets=6.55%
France et al11	Adult ED at Vanderbilt University Medical Center10 emergency medicine (EM) physicians, 5 post-graduate year-three (PGY-3) resident physicians and 5 post-graduate year-two (PGY-2) resident physicians	Observations using standardised data collection form on a handheld computer Task analysis NASA Task Load Index survey	2053 total tasks 333 interruptions were recorded. Of those, 93 were breaks-in-tasks and 210 were temporary interruptions EM physicians completed 102.4 tasks and were interrupted 16.0 times per 180 min observational period PGY-3 completed 108 tasks and were interrupted 17.6 times per 180 min observational period PGY-2 completed 97.8 tasks and were interrupted 11.0 times per 180 min observational period Categorisation of interruptions: face-to-face physician; face-to-face nurse; face-to-face other; lost chart, form or document; page; direct patient care; phone call; equipment malfunction; and other
Friedman et al12	ED of quaternary care teaching 11 EPs	Observations using standardised data collection form	400 interruptions were detected (4.4 interruptions per hour) 378 interruptions were categorised by source: nurses=53.7%; physicians=31.8%; family member=5.8% 87.5% of interruptions resulted in little movement 9.75% of interruptions resulted in more than 3 m movement
Harvey et al9	2 teaching hospitals 10 interns and nursing staff of 3 representing medical units	Interns kept logs and recorded the pages they received Nursing staff kept logs and recorded pages they made to interns	Interns' recorded reasons for pages: prescribing of medication (45%), patient assessment (24%), reporting of laboratory results (17%), starting intravenous line or venipuncture (8%), death pronouncement (1%), resuscitation (1%), wrong number (1%), not recorded (4%) 19% of pages interrupted direct patient care
Healey, Primus et al33	Operating theatre in teaching hospital 4 consultant urologists and their surgical teams	Observations	Mean interruption duration/case: 5.66 min Mean interruption duration/case duration: 13.05% Source of interruptions: conversation=198; phone=130; bleeper=26; equipment=58; procedure=36; environment=163; monitor=3
Healey, Sevdalis et al34	Operating theatre in teaching hospital Surgical team of anaesthetists, surgeons, nurses and their assistants	Observations	Mean interference per case: 50.14 min No of distractions/interruption experienced: surgeons=276; nurses=213; anaesthetist=116
Hedberg and Larsson10	Medical ward; geriatric rehabilitation ward; primary healthcare unit 6 nurses (2 from each ward/unit)	Observations	85 interruptions detected Source and number of interruptions: patient=21; family=7; assistant nurse=23; nurse=8; physician=8; secretary=7; noise=11 Nurses' activities and number of interruptions: direct patient care=53; indirect patient care=27; other=5 29% of interruptions occurred during medication administration 14% of interruptions occurred during documentation
Laxmisan et al55	Adult ED within large tertiary, teaching hospital Staff in ED	Observations Interviews	Attending physicians experienced an interruption every 9 min Residents experienced an interruption every 14 min
Pape16	Medical–surgical nursing unit in acute care hospital Nurses: control group (n=8); focused protocol (n=8); Medsafe (n=8)	Quasi-experimental three-group design of the medication administration process	Control group experienced 484 distractions Focused protocol group experienced 180 distractions (focused vs control: p<0.001) Medsafe protocol group experienced 64 distractions (Medsafe vs control: p<0.001; Medsafe vs focused: p<0.014)
Pape et al17	5 nursing units in large hospital 78 staff nurses	Observations before and after intervention Questionnaire	Mean distraction score before intervention: 42 Mean distraction score after intervention: 31 Pre/post results significantly different: p<0.001
Paxton et al22	General practices In 1990: 34 nurses; In 1991: 33 nurses; both years: 85 GPs and 1,930 patients	Observations Patient and GP questionnaires	1990: nurses observed a total of 3081 interruptions 1991: nurses observed a total of 1,729 interruptions GPs observed a total of 4,030 interruptions 4% of patients interrupted, felt it was an intrusion
Peleg et al18	Urban primary clinic 2 adult family doctors	Doctors recorded interruptions experienced before and after intervention	Preintervention: 528 interruptions in 379 consultations Postintervention: 402 interruptions in 355 consultations Pre/post decrease of 19% of interruptions (p = 0.21)
Potter et al1	General acute medicine nursing unit Dyad of 1 registered nurse (RN) and 1 patient care technician	Observations	RN experienced 43 different interruptions classified as: delays in staring, direct disruptions, or indirect disruptions
Potter et al2	Large tertiary medical centre in the Midwest 7 RNs	Observations by human factors engineer (HFE) and nurse researcher (NR) Task analysis	Interruptions observed by HFE: 261 (mean=5.9 per h) Interruptions observed by NR: 151 (mean=3.4 per h) 47% of interruptions occurred when nurses were performing interventions; 22% occurred during medication preparation; Interruptions averaged 7% of nurses work time
Rhoades et al23	Primary care outpatient clinics of teaching hospital 22 residents and their patients	Observations using standardised collection form Questionnaire	Residents interrupted patients an average of 2 times/visit Residents using computer interrupted 66% of the visits Knock on door interrupted 15% of the visits Beepers interrupted 8% of the visits Residents leaving the room interrupted 33% of the visits Patients who felt they had not spoken enough experienced significantly more interruptions
Sevdalis et al24	UK Hospital 16 surgeons, 26 nurses and 20 anaesthetists/operating departmental practitioners (OPDs)	Disruption in surgery index survey	Estimated frequency of interruptions experienced (out of 100%): surgeons=25%; nurses=42%; anaesthetist/OPDs=37% Participants estimated that their colleagues experience more interruptions than they do (p<0.05) Participants estimated that more disruptions contribute to errors for their colleagues than for themselves (p<0.05) Patient-related disruptions were judged more serious to contribute to errors over other disruptions (p<0.05) Goal of surgical procedure was judged to be obstructed by individual (p<0.01), operating room environment (p<0.05) and communication disruptions (p<0.05)
Sevdalis et al56	Operating suites of a large teaching hospital Surgical teams	Observations Rated level of distractions	Focused on case-irrelevant communication (CIC) as distractions 167 CIC events were recorded (3.48 per operation) 26.95% CIC events were classified as irrelevant comment/query by team staff Surgeons were most likely initiators (35.8%) and recipients (61.73%) of CIC events (p<0.05) CIC events related to equipment/provisions were considered most distracting
Shvartzman and Antonovsky57	4 family practice units 4 physicians	Observations	136 interruptions were recorded (mean=1.36 interruptions per consultation) Sources of interruptions (and number out of 117): nurse=51; student=41; physician=10; patient=7; maintenance worker=5; clerical worker=3 Reason for interruption (and number out of 136): sign prescription=33; obtain patient file=29; consultation=21; telephone=19; appointment diary entry=3; other=31
Spencer et al58	ED of metropolitan teaching hospital 4 RNs and 4 medical officers	Observations Recorded conversations with lapel microphones	One-third of events were considered interruptions (mean rate of 15 interruptions per person per hour)
Tucker19	Nursing units in 9 hospitals 26 nurses	Observations Interviews	194 operational failures were observed (average of 1 every 74 min) 6% of failures cost hospital average of US$0.82 11% of failures cost hospital average of US$414
Tucker and Spear41	6 US hospitals Phase I: 11 nurses; Phase II: 6 nurses; Phase III: 136 nurses	Phase I: observations Phase II: interviews Phase III: survey	Observed 8.4 operational failures per 8 h shift Observed a total of 955 interruptions; 45 were caused by operational failures; 910 were caused by patient care considerations
Westbrook et al13	Teaching hospital in Sydney 19 doctors: 7 registrars; 5 residents and 7 interns	Observations by clinically experienced RNs Recorded data on a multidimensional work task classification system on a PDA	All doctors were interrupted 2.9 times per hour Registrars were interrupted 2.9 times per hour Residents were interrupted 2.5 times per hour Interns were interrupted 3.3 times per hour 74% of interrupted tasks were resumed within the observation period (1 h)
Wiegmann et al25	Operating rooms of one medical institution Surgical teams	Observations	341 disruption events were recorded (8.1 per hour; 11.0 per surgical case) 52% of the disruption events were categorised as teamwork/communication; 17% extraneous interruptions; 12% supervisory/training-related; 11% equipment/technology; and 8% resource-based Rate of errors increased linearly with increases in the rate of disruptions (r=0.47, p<0.05)
Wolf et al59	Urban, acute care hospital 7 RNs	Observations by HFE and NR	Observed an average of 3.4 interruptions per hour RNs who prepared medication following protocol were interrupted 1.3 times per hour RNs who prepared medication violating protocol were interrupted 0.6 times per hour
Zheng et al60	Surgery suites of a tertiary care hospital Surgical teams	Video taped	114 disruption events were recorded per hour 11% of the disruption events caused delays in the surgical workflow Conversations were the largest portion of the disruption events (71 events per hour)