Table 3 Relationships between teamwork and patient safety
Study | Topic | Primary topic | Sample & setting | Design & data collection methods | Assessment of variables | Analyses | Findings | Outcomes & effect sizes | Quality scored |
|---|---|---|---|---|---|---|---|---|---|
a) observational studies | |||||||||
Burtscher et al., 2010 [61] | Relationships between coordination activities and team performance under differing situational demands | yes | 19 anesthetists and 14 anesthesia nurses, 40 cases, teaching hospital, Switzerland | Video observation of anesthesia induction | Teamwork: observation system used for coding coordination activities & clinical work Patient safety: team performance (self-developed checklist) | Paired-sample t-test | 1) Compared to low-performing teams, high-performing teams increase task management during non-routine events 2) No changes in information management during non-routine events | 1) t(20) = −2.75, p < 0.05, []b 2) NS | 13.5 (15) |
Burtscher et al., 2011 [12] | Relationships between adaptive team coordination during non-routine events and clinical performance during anesthesia induction | yes | 15 anesthesia teams (1 resident, 1 nurse), teaching hospital, Switzerland | Video observation of simulated anesthesia induction | Teamwork: team coordination (structured observation)a Patient safety: decisions and execution latency (expert rating) | Pearson’s correlation | 1) Information management is a) negatively correlated with decision latency b) but not with execution latency 2) No correlations between task management and a) decision latency b) execution latency | 1a) r = −0.49, p = 0.003 1b) NS 2a) NS 2b) NS | 12.5 (15) |
Burtscher et al., 2011 [5] | Team mental model properties moderate link between monitoring behaviors and performance in anesthesia induction | yes | 31 teams (1 anesthesia resident, 1 anesthesia nurse), teaching hospital, Switzerland | Video observation of simulated anesthesia induction | Teamwork: Team mental model similarity and accuracy (concept mapping), monitoring behavior (structured observationa) Patient safety: adherence to anesthesia induction protocol (structured observationa) | Multiple hierarchical regression | 1) Teams with similar mental models perform well irrespective of team monitoring level; teams with dissimilar mental models only perform well when team monitoring is low 2) Team mental model similarity is only related to performance when team mental model accuracy is also high 3) Team performance is high when either team or system monitoring is high and the other is low 4) Mental model accuracy does not moderate relationship between systems monitoring and performance | 1) β = 0.36, p = 0.04, ΔR2 = 0.13, [ƒ2 = 0.21]b 2) β = 0.42, p = 0.02, ΔR2 = 0.17, [ƒ2 = 0.12]b 3) β = −0.36, p = 0.04, ΔR2 = 0.12, [ƒ2 = 0.28]b 4) NS | 14 (15) |
Catchpole et al., 2007 [64] | Relationships between non-technical skills and adverse events in the OR | yes | 42 operations (24 pediatric, 18 orthopedic), 2 hospitals, UK | Live & video observation | Teamwork: non-technical skills (NOTECHSa) Patient safety: Adverse events: minor problems, intraoperative performance, operating time | Multiple linear regression | Non-technical skills negatively predict 1) minor problems but not 2) intraoperative performance or 3) operating time | 1) B = −3.3, t = −2.2, p = 0.035, []b 2) NS 3) NS | 8 (15) |
Catchpole et al., 2008 [62] | Relationships between non-technical skills and errors in the OR | yes | 54 surgeons, anesthetists, and nurses, 48 operations (26 laparoscopic cholecystectomies, 22 carotid endarterectomies), 1 hospital, UK | Live observation of operation | Teamwork: NOTECHSa Patient safety: errors in surgical technique (observation clinical human reliability assessment technique), other procedural problems and errors (checklist), operating time | Multiple linear regression | 1a) Surgical leadership and management negatively predicts operating time, 1b) whereas anesthetic leadership and management in carotid endarterectomy positively predicts operating time 2a) nursing leadership and management negatively predict other procedural problems and errors 2b) whereas nursing leadership and management in carotid endarterectomy positively predicts operating time (2 predictors) 3a) surgical situation awareness negatively predicts errors in surgical techniques (3 predictors) 3b) whereas surgical situation awareness in carotid endarterectomy positively predicts operating time (3 predictors) 4) Teamwork dimensions a) leadership and management b) teamwork and cooperation c) problem solving and decision making d) situation awareness are not associated with patient safety dimensions e) errors in surgical technique f) other procedural problems and error g) operating time | 1a) β = −0.19, p = 0.023 1b) β = 0.81, p < 0.001, R2 = 0.717, [ƒ2 = 2.53]b,c 2a) β = −0.39, p = 0.012 2b) β = 0.41, p = 0.008, R2 = 0.69 [ƒ2 = 2.215]b,c 3a) β = −0.71, p < 0.001 3b) β = 1.97, p < 0.001, R2 = 0.19, [ƒ2 = 0.233]b,c 4ae-dg) 9 non-significant associations | 9 (15) |
Catchpole et al., 2008 [63] | Relationships between non-technical skills and safety threats, errors, and operative duration | yes | Physicians and nurses, 44 operations (24 pediatric, 20 orthopedic), 2 hospitals, UK | Live & video observation | Teamwork: NOTECHSa Patient safety: errors & threats (checklists and free observations) | Spearman’s rank correlation | 1) Positive correlation between non-technical skills and 1a) safety threats 1b) operative duration 1c) but not technical errors in pediatric surgery 2) No correlations between non-technical skills and 1a) safety threats 1b) operating time 1c) technical errors in orthopedic surgery | 1a) ρ = 0.58, p < 0.005 1b) ρ = 0.58, p < 0.005 1c) NS 2a) NS 2b) NS 2c) NS | 10 (15) |
Endacott et al., 2014 [81] | Relationships between leadership, teamwork and performance in medical emergencies | yes | 42 nurses, 15 teams, 1 hospital, Australia | Video observation of simulated emergency | Teamwork: Team Emergency Assessment Measure (TEAM) a Patient safety: performance of key treatment actions | Pearson correlation | Teamwork correlates positively with patient safety in the 1) respiratory distress and 2) hypovolemic shock but not in the 3) chest pain scenario | 1) r = .90, p < 0.001 2) r = .54, p < .05 3) NS | 11.5 (15) |
Kolbe et al., 2012 [65] | Relationships between speaking up and technical team performance/team interaction | no | 31 anesthesia teams (1 nurse, 1 resident), teaching hospital, Switzerland | Video observation of simulated anesthesia induction | Teamwork: Coding scheme for (non-)verbal team interactions Patient safety: technical team performance (adherence to checklist of standard anesthesia induction and target values) | Hierarchical linear regression | 1) Technical team performance is predicted by nurses’ levels of speaking up 2) but not by residents’ levels of speaking up | 1) β = 0.43, p = 0.017, R2 = 0.18, [f2 = 0.22]bc (2 predictors) 2) NS | 14 (15) |
Kuenzle et al., 2010 [67] | Relationship between shared leadership and anesthesia team performance under high and low task load | yes | 12 anesthesia teams (1 resident, 1 nurse), teaching hospital, Switzerland | Video observation of simulated anesthesia induction | Teamwork: Coding scheme for content-oriented and structuring leadership behavior Patient safety: performance (reaction time after non-routine event) | ANOVA | 1a) No differences in shared leadership behaviors of high-performing teams between nurses and residents 1b) during high- and low task load situations 2a) Residents show more leadership behaviors than nurses in low performing teams 2b) independent of task load | 1a) F(1, 20) = 0.00, p = 0.971, η2 = 0.000 1b) Interaction: NS 2a) F(1, 20) = 7.14, p = 0.015, η2 = 0.263 2b) Interaction: NS | 12.5 (15) |
Kuenzle et al., 2010 [66] | Relationship between shared leadership and anesthesia team performance under high and low task load | yes | 12 anesthesia teams (1 resident, 1 nurse), 1 hospital, Switzerland | Video observation of simulated anesthesia induction | Teamwork: structuring and content oriented leadership: structured observation Patient safety: performance (speed of correct management after non-routine event = high task load) | Spearman’s rank correlation Kruskal-Wallis-test | 1) Under high task load team performance and a) structuring and b) content-oriented leadership are not correlated 2) Under low task load, team performance and a) structuring, b) but not content-oriented leadership are negatively correlated 3) Interaction of leadership behavior and team experience is not associated with team performance | 1a) NS 1b) NS 2a) ρ = −0.56, p < 0.05 2b) NS 3) NS | 12 (15) |
Lubbert et al., 2009 [68] | Relationship between team organization and treatment errors | yes | 378 video registrations of patients treated in the emergency room, 1 hospital, The Netherlands | Video observation | Teamwork & patient safety: Self-developed checklist measuring adherence to advanced trauma life support (ATLS) guidelines | t-test | 1) Errors in team organization dimension evident leadership are associated with more deviations from treatment protocol, whereas 2) errors in team organization dimension effective leadership are not | 1) p = 0.01 (no other indicators reported) 2) NS | 6 (15) |
Manser et al., 2009 [11] | Relationships between different coordination patterns and team performance | yes | 46 anesthesia residents, 23 teams, USA | Video observation of simulated anesthesia emergency | Teamwork: self-developed coding scheme for coordination Patient safety: clinical performance (adherence to malignant hyperthermia treatment guidelines) | Hierarchical regression analysis | 1) Time spent on coordination dimensions a) task management b) but not information management c) or coordination via work environment negatively predicts performance 2) Time spent on task management categories a) task distribution b) but not planning c) clarification d) initiating action e) or assistance negatively predicts performance 3) Time spent on information management categories a) situation assessment b) but not information transfer c) decision making d) or feedback/acknowledgement negatively predicts performance | 1a) β = −0.47, p < 0.01, ΔR2 = 0.243, [f2 = 0.32]b 1b) NS 1c) NS 2a) β = −0.54, p < 0.01, ΔR2 = 0.340, [f2 = 0.52]b 2b) NS 2c) NS 2d) NS 2e) NS 3a) β = −0.57, p < 0.05, ΔR2 = 0.227, [f2 = 0.29]b 3b) NS 3c) NS 3d) NS | 11.5 (15) |
McCulloch et al., 2009 [6] | Relationships between non-technical skills and technical errors | yes | 54 surgeons, anesthetists and nurses, 48 observations before and 55 observations after training, teaching hospital, UK | Uncontrolled pre-post-training Live observations of operations | Teamwork: NOTECHSa Patient safety: technical errors (Observation Clinical Human Reliability Assessment, OCHRA)a | Spearman’s rank correlation | 1) Negative correlation between a) overall non-technical skills and technical errors b) especially for surgical sub-team 2) Negative correlation between a) situational awareness and technical errors b) especially for surgical sub-team | 1a) ρ = −0.215, p = 0.024 1b) ρ = −0.236, p = 0.013 2a) ρ = −0.300, p = 0.001 2b) ρ = −0.436, p < 0.0001 | 11.5 (18) |
Mishra et al., 2008 [69] | Relationships between non-technical skills and technical errors | yes | 26 observations (nurses, surgeons, anesthetists), teaching hospital, UK | Live observation of operation | Teamwork: NOTECHSa Patient safety: OCHRAa | Spearman’s rank correlation | 1) No correlation between technical errors and a) leadership & management b) teamwork & cooperation c) problem-solving and decision-making in the d) overall team, or e) surgeon f) anesthetists g) and nurses subgroup 2) Negative correlation between situation awareness and technical errors for a) overall team b) surgeon subgroup c) but not anesthetists d) and nurses subgroup | 1ad) NS 1ae) NS 1af) NS 1ag) NS 1bd) NS 1be) NS 1bf) NS 1bg) NS 1 cd) NS 1ce) NS 1cf) NS 1cg) NS 2a) ρ = −0.505, p = 0.009 2b) ρ = −0.718, p = 0.001 2c) NS 2d) NS | 10 (15) |
Ottestad et al., 2007 [70] | Development and psychometric testing of tool to measure resuscitative skills and to compare interns and teams regarding ideal management of septic shock | no | 23 observations (ICU residents), USA | Video observation of emergency simulation | Teamwork: NOTECHSa Patient safety: Adherence to Surviving Sepsis Campaign Guidelines | Pearson’s correlation | Positive correlation between non-technical skills and team sepsis management | r = 0.4, p = 0.05 | 7.5 (15) |
Schmutz et al., 2015 [79] | Relationships between coordination, task type and performance in medical emergencies | yes | 277 nurses, resident and senior physicians, 68 teams, 7 hospitals, Germany | Video observation of simulated pediatric emergency | Teamwork: Coordination behaviors via CoMeT–E (Coordination System for Medical Teams - Emergency) observation toola Patient safety: Clinical performance via key treatment steps checklist | Hierarchical linear regression | 1a) Coordination behavior closed-loop communication is positively associated with clinical performance, whereas 1b) coordination behaviors task distribution and 1c) providing information without request are not. 2a) Task type moderates relationship 1a) in that it is stronger in rule-based compared to knowledge-based tasks 2b) Task type did not moderate relationship 1b) 2c) Task type did not moderate relationship 1c) | 1a) β = .25, p < .05 1b) NS 1c) NS 2a) β = −.52, p < .01 2b) NS 2c) NS | 14 (15) |
Schraagen et al., 2011 [85] | Relationships between non-routine events, teamwork and patient outcomes | yes | 1 pediatric cardiac surgery team, 40 operations, The Netherlands | Cross-sectional self-report questionnaire, live observation of operations, record review | Teamwork: observation tool derived from NOTECHSa, ANTSa, NOTSSa, and OTASa Patient safety: 30-day postsurgical complications, operating time | Pearson’s correlation, ANOVA | 1) Positive correlation between non-technical skills and a) operating time b) but not postsurgical complications 2) Explicit coordination of anesthetists is associated with higher levels of postsurgical complications | 1a) r = 0.45, p < 0.05 1b) NS 2) Muncomplicated = 12.88, Mminor complications = 21.55, []b Mmajor complications = 16.40, F(2,36) = 4.78, p < 0.01, []b | 10 (16) |
Schraagen et al., 2011 [86] | Relationships between non-routine events, teamwork and patient outcomes | yes | 1 pediatric cardiac surgery team, 40 operations, The Netherlands | Cross-sectional self-report questionnaire, live observation of operations, record review | Teamwork: NOTECHSa Patient safety: 30-day postsurgical complications | Pearson’s correlation, ANOVA | Teamwork and cooperation is associated with higher levels of postsurgical complications | Muncomplicated = 3.19, Mminor complications = 3.44, Mmajor morbidity = 3.28, F(2,36) = 3.85, p < 0.05, η2 = 0.18 | 8.5 (16) |
Siassakos et al., 2010 [80] | Relationships between individual team members’ knowledge, skills, and attitudes and team performance | no | 19 teams (physicians and midwives), 6 maternity units, UK | Video observation of obstetric emergency simulation, self-report questionnaire | Teamwork: SAQ subscale team climatea Patient safety: team performance (magnesium administration) | Kendall’s rank correlation | No correlation between teamwork climate and performance | NS | 8 (16) |
Siassakos et al., 2011 [72] | Relationships between teamwork skills and behaviors and team performance in emergency situations | yes | 47 teams (2 physicians and 4 midwives each), 6 maternity units, UK | Video observation | Teamwork: Team analytical toola Patient safety: performing key actions | Kendall’s rank correlation | 1) Positive correlation between speed of magnesium administration and a) skills b) behavior c) and overall teamwork 2) Negative correlation between time needed to put patient in recovery position and a) skills b) behavior c) but not overall teamwork 3) Negative correlation between time needed to administer oxygen and a) skills b) behavior c) and overall teamwork 4) Negative correlation between time needed to sample blood and a) skills b) behavior c) and overall teamwork | 1a) τ = 0.54, p < 0.001 1b) τ = 0.41, p = 0.001 1c) τ = 0.51, p < 0.001 2a) τ = −0.29, p = 0.012 2b) τ = −0.25, p = 0.026 2c) NS 3a) τ = −0.39, p < 0.001 3b) τ = −0.28, p = 0.014 3c) τ = −0.41, p < 0.001 4a) τ = −0.35, p = 0.002 4b) τ = −0.35, p = 0.002 4c) τ = −0.35, p < 0.002 | 8.5 (15) |
Siassakos et al., 2011 [71] | Relationships between teamwork and clinical efficiency in emergency situations | yes | 114 physicians and nurses, 19 teams, 6 maternity units, UK | Video observation | Teamwork: self-developed observation system Patient safety: performing key action (speed of magnesium administration) | Kendall’s rank correlation | 1) Positive correlation between closed-loop communication and clinical efficiency 2) Positive correlation between unambiguous communication and clinical efficiency 3) No correlations between clinical efficiency and a) SBAR communication style b) team coordination c) situational awareness d) leadership style e) supportive language f) task support by senior clinician | 1) τ = 0.46, p = 0.022 2) τ = 0.53, p = 0.004 3a) NS 3b) NS 3c) NS 3d) NS 3e) NS 3f) NS | 8 (15) |
Thomas et al., 2006 [73] | Relationship between teamwork and quality of care | yes | 118 teams (physicians, nurses, respiratory therapists), resuscitation room, teaching hospital, USA | Video observation of neonatal resuscitation | Teamwork: Frequency of different teamwork behaviors Patient safety: Neonatal Resuscitation Program (NRP) Guidelines | Spearman’s rank correlation | 1) Negative correlation between team communication and a) overall quality of resuscitation, b) non-compliance with all NRP steps, and c) non-compliance during preparation and initial steps 2) Negative correlation between team management and a) noncompliance with all NRP steps, and b) noncompliance during preparation and initial steps but not c) overall quality of resuscitation, 3) Negative correlation between team leadership and a) overall quality of resuscitation, but not with b) noncompliance with all NRP steps, and with c) non-compliance during preparation and initial steps | 1a) ρ = −0.236, p = 0.007 1b) ρ = −0.214, p = 0.014 1c) ρ = −0.230, p = 0.008 2a) ρ = −0.201, p = 0.021 2b) ρ = −0.252, p = 0.003 2c) NS 3a) ρ = −0.288, p < 0.001 3b) NS 3c) NS | 9.5 (15) |
Tschan et al., 2006 [74] | Relationships between directive leadership, structuring inquiry and performance regarding different phases | yes | 109 clinicians (nurses, residents, senior physicians), 21 teams, ICU, university hospital, Switzerland | Video observation and transcription of emergency simulation | Teamwork: directive leadership and structuring inquiry Patient safety: clinical performance (key actions, hands-on time) | Pearson’s correlation | 1) Phase 1 (nurses only): positive correlation between performance and a) directive leadership and b) structuring inquiry 2) Phase 2 (residents and nurses): positive correlation between performance and a) resident directive leadership during first 30 s, no correlation between performance and b) resident directive leadership per second c) resident structuring inquiry per second d) resident structuring inquiry during first 30 s 3) Phase 3 (nurses, residents, senior physicians): positive correlation between performance and a) senior physician structuring inquiry, no correlation between performance and b) resident structuring inquiry c) senior physician d) resident directive leadership | 1a) r = 0.445, p < 0.05 1b) r = 0.216, p < 0.05 2a) r = 0.522, p < 0.05 2b) NS 2c) NS 2d) NS 3a) r = 0.428, p < 0.01 3b) NS 3c) NS 3d) NS | 11.5 (15) |
Tschan et al., 2009 [75] | Relationships between team communication and perceptual biases of individuals and accuracy of diagnosis | yes | 53 physicians, 20 teams, university hospital, Switzerland | Video observation of hand-over simulation | Teamwork: coding of communication and behavior Patient safety: diagnostic performance | ANOVA | 1) Groups considering more diagnostic information are not more likely to find the correct diagnosis 2) Groups showing a) more explicit reasoning b) more talking to the room are more likely to find the correct diagnosis | 1) NS 2a) F(2, 15) = 5.750, p = 0.014 2b) χ2 = 8.598, df = 2, p = 0.007 | 11 (15) |
Westli et al., 2010 [76] | Relationship between teamwork skills/shared mental models and clinical performance | yes | 27 trauma teams, Norway | Video observation of emergency simulations | Teamwork: ANTSa and Anti-Air Teamwork Observation Measure (ATOM) Patient safety: Team global medical management, key actions of trauma management | Pearson’s correlation | 1) Negative correlation between supporting behavior and performing key actions 2) Negative correlation between poor coordination and medical management 3) Positive correlation between information exchange and medical management 4) Negative correlation between poor situational awareness and performing key actions 5) Positive correlation between providing information and medical management 6a-u) 21 non-significant correlations between teamwork and patient safety variables | 1) r = −0.37, p < 0.05 2) r = −0.36, p < 0.05 3) r = 0.34, p < 0.05 4) r = −0.40, p < 0.05 5) r = 0.51, p < 0.01 6a-u) NS | 10.5 (15) |
Wiegmann et al., 2007 [77] | Relationship between (teamwork-related) surgical flow disruptions and surgical error | yes | 31 cardiac operations, 1 hospital, USA | Live observation of operation | Teamwork: teamwork-related surgical flow disruptions Patient safety: surgical errors | Multiple regression | Teamwork-related surgical flow disruptions positively predict surgical errors | β = 0.692, p < 0.001, adj. R2 = 0.553, [f2 = 1.24]bc (5 predictors altogether) | 11 (15) |
Williams et al., 2010 [78] | Relationships between teamwork behaviors and resuscitation errors | yes | 12 resuscitation teams, NICU, teaching hospital, USA | Video observation of resuscitation | Teamwork: frequency of different teamwork behaviors Patient safety: Neonatal Resuscitation Program (NRP) Guidelines | Spearman’s rank correlation, generalized linear mixed model (GLM) | 1) Negative correlation between vigilance and NRP errors 2) No correlation between workload management and NRP errors 3) NRP errors are associated with a) more assertions before the error b) less teaching after the error 4) No associations between NRP errors and a) information sharing before error b) information sharing after error c) inquiry before error d) inquiry after error e) assertion after error f) teaching before error | 1) ρ = −0.62, p = 0.031 2) NS 3a) OR = 1.44, p = 0.008, 95 % CI 1.10 – 1.89 3b) OR = 0.59, p = 0.028, 95 % CI 0.37 – 0.94 4a) NS 4b) NS 4c) NS 4d) NS 4e) NS 4f) NS | 10 (15) |
b) survey studies | |||||||||
Brewer, 2006 [87] | Relationships between culture, team characteristics/processes and patient safety/hospital financial performance | yes | 430 nurses, physicians and other medical care providers, 16 surgical units, 4 acute care hospitals, USA | Cross-sectional self-report questionnaire, record review | Teamwork: Positive team processes: Relational Coordination Scalea Negative team processes scale Patient safety: patient falls (incident reporting system), length of stay (hospital records) | Pearson’s correlation | 1) Positive intra-team processes correlate positively with a) length of stay b) but not with patient falls 2) No correlation between negative team processes and a) length of stay b) patient falls | 1a) r = 0.59, p < 0.05 1b) NS 2a) NS 2b) NS | 10 (16) |
Chan et al., 2011 [88] | Validity of a team-based tool to assess success of a team-based intervention to reduce central line associated blood stream infections (CLABSI) | no | 46 ICUs, 35 hospitals, USA | Secondary analyses of longitudinal RCT, self-report questionnaire, record review | Teamwork: Team check-up tool (TCT) Patient safety: Central line associated bloodstream infections (CLABSI) | Cox regression | No association between teamwork and duration to reach zero CLABSI’s after intervention | NS | 10 (19) |
Chang & Mark, 2009 [89] | Antecedents (teamwork, nurse & patient factors) of severe and non-severe medication errors | yes | 1 671 nurses, 279 units, 146 hospitals, USA | Longitudinal self-report questionnaire, record review | Teamwork: Relational Coordination Scalea Patient safety: medication errors (hospital incident reports) | Generalized estimating equations (GEE) | Relational coordination predicts neither 1) severe nor 2) non-severe medication errors | 1) NS 2) NS | 9 (16) |
Edmondson, 2004 [10] | Relationship between team/organizational characteristics, team leadership and medication errors | yes | 159 nurses, physicians and pharmacists, 8 hospitals, USA | Cross-sectional self-report questionnaire, record review | Teamwork: Team/organizational characteristics and team leadership (self-developed questionnaire) Patient safety: medication error (hospital incident reports & self-reported) | Spearman’s rank correlation | Positive correlation between 1) nurse manager coaching 2) nurse manager direction setting and 3) unit relationship quality and a) detected and b) intercepted medication errors but not with c) non-preventable drug complications | 1a) ρ = 0.74, p < 0.03 1b) ρ = 0.71, p < 0.03 1c) NS 2a) ρ = 0.74, p < 0.03 2b) ρ = 0.83, p < 0.03 2c) NS 3a) ρ = 0.74, p < 0.03 3b) ρ = 0.76, p < 0.03 3c) NS | 11 (16) |
Fasolino et al., 2012 [90] | Relationships between nurse characteristics, nurse practice environment, team member effectiveness and medication error | yes | 163 nurses, 11 surgical units, 1 hospital, USA | Cross-sectional self-report questionnaire, record review | Teamwork: team member effectiveness survey Patient safety: medication errors (hospital incident reports) | Spearman’s rank correlation | Team member effectiveness is positively correlated with medication error | ρ = 0.19, p < 0.01 | 12 (16) |
Hoffer Gittell et al., 2000 [9] | Relationship between relational coordination and quality of care/length of stay | yes | 338 physicians, nurses, and other medical care providers, 9 hospitals, USA | Cross-sectional self-report questionnaire, record review | Teamwork: Relational Coordination Scalea Patient safety: Length of stay | Hierarchical linear regression | Relational coordination is associated with decreased length of stay | B = −53.77, p < 0.001, []b | 13 (16) |
Hwang & Ahn, 2015 [83] | Relationships between teamwork and error reporting | yes | 576 nurses, 2 acute care hospital, South Korea | Cross-sectional self-report questionnaire | Teamwork: Teamwork perceptions questionnaire (TPQ)a Patient safety: occurrence of and reporting medical errors | Logistic regression | Teamwork dimensions 1) team structure, 2) team leadership, 3) situation monitoring, 4) mutual support, and 5) communication are positively associated with error reporting No information on relationship between teamwork and occurrence of medical errors | 1) OR = 0.92, 95 % CI 0.50 –1.692) OR = 1.13, 95 % CI 0.78 –1.623) OR = 0.96, 95 % CI 0.52 – 1.78 4) OR = 1.23, 95 % CI 0.66 – 2.30) 5) OR = 1.82, 95 % CI 1.05 - 3.14 | 12.5 (16) |
Kalisch & Lee, 2010 [60] | Relationship between teamwork and missed nursing care | yes | 2216 nurses, 40 acute care units, 4 hospitals, USA | Cross-sectional self-report questionnaire | Nursing Teamwork Surveya MISSCARE Surveya | Pearson’s correlation Multiple linear regression | 1) Negative correlation between missed nursing care and a) trust b) team orientation c) backup behavior d) shared mental model e) team leadership 2) After controlling for various covariates, overall teamwork scores negatively predict missed nursing care | 1a) r = −0.31, p < 0.01 1b) r = −0.28, p < 0.01 1c) r = −0.31, p < 0.01 1d) r = −0.32, p < 0.01 1e) r = −0.29, p < 0.01 2) B = −0.254, p < 0.001, ΔR2 = 10.9, [f2 = 0.124]b | 12.5 (16) |
Leroy et al., 2012 [8] | Mediation and moderation relationships between leader behavioral integrity for safety, team psychological safety, team priority of safety, and treatment errors | yes | 580 nurses and head nurses, 4 hospitals, Belgium | Longitudinal self-report questionnaire | Teamwork: Team Psychological Safety Scalea Patient Safety: head nurses’ reports of treatment errors | Path analysis | 1) Good overall model fit 2) Within path model, team psychological safety at time 1 positively predicts treatment errors at time 2 | 1) χ2 = 6.72, p = 0.03, SRMR = 0.07, RMSEA = 0.02, CFI = 0.98 2) β = 0.28, p = 0.02 | 14 (16) |
Manojlovich et al., 2007 [82] | Relationships between perceived work environments, nurse-physician communication and patient outcomes | yes | 462 nurses, 25 ICUs, 8 hospitals, USA | Cross-sectional self-report questionnaire | Teamwork: parts of ICU Nurse-Physician Questionnairea Patient safety: nurse-reported adverse events (medication errors, ventilator-associated pneumonia, catheter-associated sepsis) | Random intercept multilevel models | Nurse-physician communication negatively predicts 1) ventilator-associated pneumonia 2) catheter-associated sepsis and 3) medication errors | 1) B = −0.045, p < 0.05, R2 = 0.09, [f2 = 0.1]b,c 2) B = −0.049, p < 0.05, R2 = 0.14, [f2 = 0.16]b,c 3) B = −0.047, p < 0.01, R2 = 0.11, [f2 = 0.12]b,c | |
Manojlovich et al., 2009 [91] | Relationship between nurse-physician communication and patient outcomes | yes | 462 nurses, 25 ICUs, 8 hospitals, USA | Cross-sectional self-report questionnaire, record review | Teamwork: ICU Nurse-Physician Questionnairea Patient safety: adverse outcomes ventilator-associated pneumonia, bloodstream infections, and pressure ulcers | Pearson’s correlation | No correlation between nurse-physician communication subscales 1) timeliness 2) accuracy 3) openness and 4) understanding and patient safety indicators a) ventilator-associated pneumonia b) bloodstream infections and c) pressure ulcers | 1a-4c) 12 non-significant associations | 11 (16) |
Ogbolu et al., 2015 [84] | Relationships between nurse work environment and patient safety | no | 222 nurses, Nigeria | Cross-sectional self-report questionnaire | Teamwork: nurse-physician relationsa Patient safety: Patient safety: one item from AHRQa | Generalized linear mixed modeling | Relationship between nurse-physician relations and patient safety not reported (only relationship between aggregate NWI scale and patient safety) | - | 10 (16) |
Taylor et al., 2012 [92] | Relationships between safety climate, teamwork and patient adverse events | no | Nurses in 29 units, 1 hospital, USA | Cross-sectional & longitudinal self-report questionnaire, record review | Teamwork: SAQ subscale team climatea Patient safety: patient falls & injuries, deep vein thrombosis and pulmonary embolism records | Multilevel logistic regression | Positive team climate is associated with 1) fewer decubitus ulcers, but not 2) less patient falls & injuries or 3) pulmonary embolisms and deep vein thrombosis one year later | 1) OR = 0.56, 95 % CI 0.30 - 0.82, p < 0.01 2) NS 3) NS | 13.5 (16) |
Vogus et al., 2007 [93] | Moderation of relationship between team safety organizing behaviors and medication errors by trust in manager and existence of care pathways | yes | 1033 nurses & 78 nurse managers, 78 units, 10 acute-care hospitals, USA | Cross-sectional self-report questionnaire, record review | Teamwork: Safety Organizing Scale (SOS)a Trust in manager: 2 items Care pathways: 1 item Patient safety: medication errors (number of errors reported to unit’s incident reporting system up to 6 months after survey data collection) | Multilevel Poisson regression | 1) Safety organizing negatively predicts medication errors 2) Trust in manager has no impact on reporting of medication errors when level of safety organizing is high. When safety organizing is low and trust in manager is high, more errors are reported 3) Use of care pathways has no impact on reporting of medication errors when safety organizing is low. When safety organizing is high and care pathways are extensively used, fewer errors are reported | 1) β = −0.29, p < 0.01, 95 % CI −0.57 to −0.01 2) β = −0.68, p < 0.001, 95 % CI −1.03 to −0.32 3) β = −0.82, p < 0.001, 95 % CI −1.31 to −0.33 | 13 (16) |
Wheelan et al., 2003 [94] | Relationship between teamwork and patient mortality | yes | 349 healthcare providers, 17 ICUs, 9 hospitals, USA | Cross-sectional self-report questionnaire, record review | Teamwork: Group Development Questionnairea Patient safety: Standardized mortality rates | Pearson’s correlation | Level of group development correlates negatively with mortality rates | r = −0.66, p = 0.004 | 12 (16) |
Yun et al., 2005 [95] | Moderation of relationship between contingent leadership and team effectiveness by severity of patient trauma and team experience | yes | 91 members of trauma resuscitation teams, 1 hospital, USA | Cross-sectional self-report questionnaire, scenario method | Teamwork & patient safety: Team Effectiveness Scalea, Team leadership, severity of trauma and team experience manipulated across scenarios | General linear model (GLM) | 1) Interaction of leadership/severity of injury: Team effectiveness dimension quality health care is high when patient was not severely injured/leadership is empowering or patient was severely injured/leadership was directive 2) Interaction of leadership/team experience: quality health care is highest when leadership is empowering, independent of team experience 3) 3-way-interaction: quality health care is highest when team is experienced and leadership is empowering, independent of patient condition. When team is inexperienced, quality health care is highest when leadership is empowering and patient is not severely injured, or when leadership is directive and patient is severely injured | 1) Severely injured patient: Mdirective leaders = 3.06, 95 % CI 2.83 – 3.27, Mempowering leaders = 2.72, 95 % CI 2.50 – 2.95. Non-severely injured patient: Mempowering leaders = 3.91, 95 % CI 3.69 – 4.13, Mdirective leaders = 2.16, 95 % CI 1.94 – 2.38, F = 119.48, p < 0.01, η2 = 0.26. 2) Experienced team: Mempowering leadership = 3.65, 95 % CI 3.42 - 3.82, Mdirective leadership = 2.48, 95 % CI 2.25 - 2.70. Inexperienced team: Mempowering leadership = 2.99, 95%CI 2.76 - 3.21, Mdirective leadership = 2.74, 95 % CI 2.51 - 2.96, F = 23.19, p < 0.01, η2 = 0.06. 3) Inexperienced team/severely injured patient: Mdirective leadership = 3.19, 95 % CI 2.89 - 3.49, Mempowering leadership = 2.13, 95 % CI 1.82 - 2.44. Inexperienced team/non-severely injured patient: Mempowering leadership = 3.85, 95 % CI 3.57 - 4.12, Mdirective leadership = 2.28, 95 % CI 2.00 - 2.56, F = 7.31, p < 0.01, η2 = 0.04. | 14.5 (16) |