. 2023 Mar 29;23:300. doi: 10.1186/s12913-023-09332-8

Table 3.

The measurements of patient safety culture and adverse patient events, and their associations

First author -Year -Origin -Concepts PSC/ AE	PSC -Tool -Dimensions/items -Context -Participants PSC -n = -Response rate %	AE -Method -Number patients/records -Numbers and types of AEs assessed	Analysis -Methods -Study level -Number of units	Reported association Statistical values	Critical appraisal Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies [29]
Ausserhofer (2013) Swiss [34] Patient Safety Climate/ Patient Outcomes	SOS 1/9 (complete) –translated Surgery/medicine/ mixed Nurses n = 1,630 Response rate: 72%	Retrospective nurse-estimate Incidence of AEs last year – 7-point scale 1,630 nurses 7 AEs: Nurse-reported medication errors, pressure ulcers, patient falls, urinary tract infection, bloodstream infection, pneumonia (and patient satisfaction)	Bivariate and multivariate logistic regressions Unit and hospital level 132 units in 35 hospitals	No statistically significant associations between PSC scores and 6 selected AEs: pressure ulcers, urinary tract infection, bloodstream infection, medication errors, pneumonia, patient falls (nor any association between PSC and patient satisfaction) Increased scores of Rationing of nursing care were consistently associated with increased rates of ^bbloodstream infection, ^bmedication administration error, ^cpneumonia (and ^dpatient satisfaction) in multivariate analysis (Reduced nurse ratio increases AE rates) ^ap = 0.004 ^bp = 0.026 ^cp = 0.027 ^dp = 0.005	Fair
Bacon (2021) US [35] Organisational Safety Climate/ Mortality and Failure-to-rescue	Safety Climate Tool – revised (originally Zohar`s measure of safety climate) 0/33 Surgery Nurses, physicians, others n = 261 Response rate: NR	Chart review AHRQ Quality Indicators 10,823 patients 52,898 records 2 AEs: Failure-to-rescue (deep vein thrombosis/ pulmonary embolus, pneumonia, sepsis, cardiac arrest, hemorrhage) and in-hospital mortality	Multilevel models Bonferroni correlation Hospital level 2 hospitals	No statistically significant associations between PSC scores and rates of in-hospital mortality or failure-to-rescue Reported in p-values	Fair
Birkmeyer (2013) US [36] Safety Culture/ Complications	HSOPS + SAQ-OR + questions disruption 4/30 Surgery Surgeons, nurses/ operating room technicians, operating room administrators n = 184 Response rate: surgeons 95% nurses/ operating room technicians 82%, operating room administrators 68%	Chart review Standardised instrument 24,117 patients 10 types of AEs: Abdominal abscess, bowel obstruction, leak, bleeding, respiratory failure, renal failure, wound infection/ dehiscence, venous thromboembolism, myocardial infarction or cardiac arrest, death. The AE rate is the overall rate of the 10 AEs	Bivariate regression model Spearman`s Correlation (p) Hospital level 22 hospitals	Increased overall PSC scores from ^asurgeons and ^bnurses were associated with reduced AE rates. Increased scores of ^cHospital safety culture by nurses and increased scores of ^dOperating room safety by surgeons were associated with reduced AE rates ^ap < 0.001 bp < 0.011 ^cp < 0.002 ^dp < 0.045	Fair
Bosch (2011) Netherlands [32] Organisational Culture and Teamwork Climate/ Pressure Ulcers	Team Climate Inventory (TCI)- short + Competing Values Framework (CVF) (both translated) 0/14 + 5/20 Ward units in hospitals and nursing homes Physicians, nurses, nursing assistants n = 460 Response rate 41% for the hospital wards, and 39% for the nursing home wards	Prospective nurse-reporting and scoring 1,274 patients 1 AE: Nosocomial pressure ulcers	Bivariate and multilevel logistic regression (General Linear regression) Unit level 104 units	No statistically significant association between PSC scores, Team climate or Preventive quality management at ward level, and the prevalence of nosocomial pressure ulcers (^aIncreased scores of Institutional quality management were significantly correlated with increased scores of Preventive quality management at ward level.) ^ap < 0.001	Fair
Brown (2013) US [37] Safety Culture/ Adverse Patient Outcomes	HSOPS 12/42 (complete) + Global rating composite (4 items) NR Nurses NR Response rate: NR	Register National Database for Nursing Quality Improvement. The Collaborative Alliance for Nursing Outcomes (CALNOC). CALNOC indicator definitions 3 AEs: Health Acquired Pressure Ulcers, reported falls and falls with injury	Linear regression, Pearson product-moment correlations (r) Unit level 9 hospitals 37 units	^aIncreased scores of Teamwork within units were associated with reduced reported falls and ^bincreased scores of Management support were associated with increased rates of reported falls ^cIncreased Global rating composite was associated with reduced rates of Health-Acquired Pressure Ulcers (HAPU) ^dSkill mix, Staff turnover and Workload intensity are strongly corelated with PSC scores ^ap < 0.05 ^bp < 0.03 ^cp < 0.05 ^dp < 0.01–0.05	Poor
Brubakk (2019) Norway [38] Organisational Culture/ Mortality	SAQ (and organisational factors survey) 1/2 (+ 19/ 57) Emergency/ acute care Nurses, physicians, managers n = 8,800 Response rate: 72% 2010 77% in 2011 75% in 2012	Register The Norwegian Institute of Public Health 46,026 admitted patients 1 AE: Risk-adjusted 7-day mortality	Multivariable regression Unit and group level 20 hospitals 56 units	^aReduced scores of Patient safety climate and ^bLeadership were associated with increased 7-day mortality ^cIncreased scores of Workload perceived by nurses were associated with increased 7-day mortality rates. ^dIncreased Middle manager`s engagement levels were associated with reduced 7-day mortality rates ^ap < 0.003 ^bp < 0.045 ^cp < 0.028 ^dp < 0.037	Good
Camargo (2012) US/ 20 States [39] Safety Climate/ Adverse Events and Medical Error	Survey – constructed 9/50 Emergency/ acute care Nurses, physicians n = 3,562 Response rate: 66%	Chart review Standardised form: 18 questions 9,821 charts 3 AE-categories: Medical error, adverse event (preventable and non-preventable) and near miss (intercepted and non-intercepted) Scaled: Significant, serious, life threatening or fatal	Multivariable regression models Incident Rate Ratio (IRR) Unit level 62 units	No statistically significant association between PSC scores and preventable AE rates, nor was there an association between PSC scores and serious violations of treatment guidelines ^aIncreased PSC scores were significantly associated with increased “Intercepted near misses” ^aIRR 1.79 (1.06–3.03)	Fair
Davenport (2007) US [40] Organisational Safety Climate/ Outcomes	SAQ 6/30 (complete) Surgery Nurses, physicians, others n = 6,083 Response rate: 52%	Chart review NSQIP protocol 57,880 patients 2 AEs: Risk-adjusted surgical morbidity (patient having 1 or more out of 21 complications) and 30-day mortality	Multivariate logistic regression Spearman`s p correlation Hospital level 52 hospitals	No statistically significant association between PSC scores and rates of 30-day mortality or 30-day morbidity (21 postoperative complications) ^aIncreased scores of Communication/ collaboration with doctors were correlated with reduced rates of risk-adjusted morbidity Scores of Burnout was not correlated with AE rates ^ap < 0.01	Good
Fan (2016) US/ Minnesota [41] Safety Culture/ Surgical Outcomes	HSOPS 12/42 (complete) Surgery NR n = 1,926 Response rate: 43%	Prospective reporting as defined by the National Healthcare Safety Network NR 1 AE: Postoperative colon surgery SSI/ number of operations performed	Bivariate and multivariate linear regression Pearson product-moment correlations (r) Unit level 7 units/hospitals	Increased scores of following PSC dimensions were associated with reduced rates of Colon SSI: ^aTeamwork across units ^bTeamwork within units ^cOrganisational learning ^dFeedback and communication about error ^eCommunication openness* ^fOverall perception of safety ^gManagement support for patient safety ^hSupervisor/manger expectations and actions promoting safety ⁱNon-punitive response to error* ^jFrequency of events reported ^kHandoffs and transitions when adjusting for ASA and surgical volume ^a r = -0.96, [-0.76, -0.99] ^b r = -0.88, [-0.38, -0.98] ^c r = -0.95, [-0.71, -0.99] ^d r = -0.92, [-0.56, -0.99] ^e r = -0.85, [-0.26, -0.98] ^f r = -0.90; [-0.45, -0.99] ^g r = -0.90, [-0.44, -0.98] ^h r = -0.85, [-0.25, -0.98] ⁱ r = -0.78, [-0.07, -0.97] non-significant when adjusting for ASA and surgical volume ^j r = -0.76, [-0.01, -0.96] ^kp < 0.05*	Fair
Garrouste-Orgeas (2015) France [42] Safety Culture/ Medical Error	SAQ-ICU 6/63 (complete) ICU Nurses, physicians, others n = 1,534 Response rate: 77.2%	Prospective reporting/ observation 8 h /5 days /2 weeks combined to chart review 4 AEs: Error administration anticoagulant medication, error prescribing anticoagulant medication, error administration insulin, accidental removal of a central venous catheter, accidental extubating	Multivariate hierarchical model Unit level 31 units	Limited statistically significant association between PSC scores and rates of medical errors/ patient days^a Increased scores of ^bDepression symptoms, ^cICU organisation (40% off work previous day), ^dStaff-specific safety training programme and ^ePatient level/ workload were associated with increased AE rates Increased scores of Burnout were not statistically significantly correlated with increased AE rates ^ap = 0.04–0.87 ^bp = 0.01 ^cp = 0.01 ^dp = 0.001–0.02 ^ep = < 0.0001- 0.03	Good
Han (2020) South Korea [43] Patient Safety Culture/ Adverse Events	HSOPS 12/42 (complete) NR NR n = 212 Response rate: 86%	Retrospective nurse-estimate Incidence of AEs last year – 7-point scale 212 nurses 5 AEs: Falls, medication errors, pressure ulcers, health-associated infections (surgical site, urinary tract, central-line associated bloodstream infections and ventilator-associated pneumonia) and physical restrain ≥ 8 h (Combined into a binominal variable: “never/happened”)	Bivariate regression and multiple logistic regression. Odds Ratio (OR) Hospital level 2 hospitals	Increased scores of the following PSC dimensions were associated with reduced AE rates: ^aSupervisor/ manger expectations and actions promoting safety and 4 AEs ^bCommunication openness and 4 AEs ^cManagement support for patient safety and 3 AEs ^dTeamwork across units and 3 AEs ^eTeamwork within units and 2 AEs ^fFeedback and communication about error and 2 AEs ^gNonpunitive response to error and 1 AE ^hHandoffs and transitions and 1 AE ⁱOrganisational learning – continuous improvement and 1 AE Increased scores for Patient safety competencies were associated with reduced AE rates Odds Ratio (OR): ^a OR 0.33–0.39 ^b OR 0.25–0.51 ^c OR 0.22–0.55 ^d OR 0.29–0.47 ^e OR 0.23–0.51 ^f OR 0.43–0.52 ^g OR 2.08 ^h OR 2.02 ⁱ OR 0.053 ^j OR 0.024–0.049	Fair
Haynes (2011) US [44] Safety Climate/ Postoperative Morbidity and Mortality	SAQ-OR NR/6 Surgery Nurses, physicians, others n = 281 pre intervention n = 257 post intervention Response rate: 97.7%	Chart reviews and communication with clinical teams Charts were reviewed at discharge or in 30 days 19 AEs: Acute renal failure, bleeding requiring ≥ 4 units of red cell transfusion within 72 h after surgery, cardiac arrest requiring cardiopulmonary resuscitation, coma for ≥ 24 h, deep venous thrombosis, myocardial infarction, unplanned intubation, ventilator use for ≥ 48 h, pneumonia, pulmonary embolism, stroke, major wound disruption, surgical site infection, sepsis, septic shock, systemic inflammatory response syndrome, unplanned return to the OR, vascular graft failure and death	Correlation analysis Spearman`s correlation (p) Hospital level 8 hospitals	Increased PSC scores were associated with reduction in postoperative complication rate^a including 18 AE rates and mortality (The measurement was related to an intervention.) ^a r = 0.7143 p < 0.0381	Fair
Hofmann (2006) US [45] Safety Climate/ Medication Error and Patient Outcomes	Zohar`s measure of safety climate – revised and Error Orientation Scale 3/9 + 3/13 Surgery/ medicine Nurses n = 1,127 Response rate: NR	Chart review Coordinators collected the AE frequency over 3 months 2 AEs: Medication Errors and Urinary Tract Infections	Bivariate regression and multiple logistic regression Unit level 42 hospitals 81 units	A) Increasing overall PSC scores significantly predict reduced rates of ^amedication errors and ^burinary tract infections B) Regression with^abshow that^a was significantly moderated by Patient complexity A)^a -1.51 p < 0.05 ^b -1,27 p < 0.05 B) ^ab -7.85 p < 0.05	Fair
Huang (2010) US [46] Safety Culture/ Outcomes	SAQ-ICU 6/60 (complete) ICU Nurses, physicians, others n = 2,103 Response rate: 47.9%	Register PICCM clinical national database 65,978 patients 2 AEs: Hospital mortality and LOS	Linear regression model and multivariate logistic regression Unit level 30 units	^aIncreased scores of Perceptions of management were associated with reduced mortality rates ^bIncreased scores of Safety climate were associated with reduced LOS ^ap = 0.005 ^bp = 0.003	Fair
Hwang (2011) Korea [47] Safety Climate/ Medical Errors	SAQ – translated 17 items from 4 dimensions + 4 items added due to Korean context 2/21 NR Nurses n = 1,923 Response rate: 89.7%	Nurse-estimate experienced errors in retrospective questionnaire n = 277 nurses AEs last year: Yes/ No Frequency of AEs last year	Multiple logistic regression Hospital level 33 hospitals	Nurses with better scores of ^aworkgroup and ^borganisation-level Safety climate were associated with reduced error rates Odds Ratio: ^a(OR = 0.73) p < 0.001 ^b(OR = 0.69) p < 0.001	Poor
Kakemam (2021) Iran [48] Patient Safety Culture/ Adverse Events	HSOPS- Persian 12/42 (complete) Emergency/ acute care, ICU, surgery, medicine, NR Nurses n = 2,995 Response rate: 51.1%	Retrospective nurse-estimate Incidence of AEs last year, 7-point scale 2,995 nurses 6 AEs: Pressure ulcer, patient falls, adverse drug events, surgical wound infection, complaints from patients or their family, infusion or transfusion reaction	Bivariate and multiple logistic regression models Hospital level 32 hospitals	^aIncreased scores of nine PSC dimensions were significantly associated with a reduced perception of AE rates in at least two out of six AEs ^ap < 0.001	Fair
Kline (2008) Canada [49] Patient Safety Culture/ Adverse Events	Patient Safety Culture 2005 Survey – Database 1/5 NR NR (reported as nursing leaders in primary study) n = 298 (408/417 in primary study) Response rate: 83%/ 72% (reported in primary study)	Register Regional Incident Reports in forms by any health staff 5,070 incident reports/ 3,093 non-incident reports Severity range 1–4 and contributing factors/ categorised as 7 incident types: Care and treatment, injury or death, falls, medication discrepancy, medication incident, test or results, vaccine	Hierarchical linear regression – multilevel Unit level 3 hospitals 40 units	^aResource intensity predicts incident severity level ^bPSC predicts “adverse event severity” over “case resource intensity” ^ap < 0.001 ^bp < 0.05 R² = 0.093	Fair
Lee (2018) Canada [50] Organisational Safety Culture/ Adverse Events	HSOPS NS/7 NR Nurses n = 1,053 Response rate: NR	Retrospecitve nurse-estimate Incidence of AEs last year, 7-point scale 1,053 nurses 3 AEs: Medication error, patient falls with injury, urinary tract infection (+ quality of care)	Multilevel ordinal logistic and linear regression Pearson and Spearman correlation Hospital level 63 hospitals	^aIncreased Overall organisational safety culture was associated with reduced rates of reported medication errors, falls with injury and urinary tract infections ^aIncreased scores of Overall organisational safety culture increased the quality of care ^ap < 0.05	Fair
Mardon (2010) US [51] Patient Safety Culture/ Adverse Events	HSOPS 12/42 (complete) NR NR n = 56,480 Response rate: 51%	Register HCUP at AHRQ – Patient safety indicators (PSI) NR 8 AEs: Complications of anesthesia, death in low mortality diagnostic related groups, failure to rescue, foreign body left in during procedure, transfusion reaction, birth trauma – injury to neonate, obstetric trauma vaginal delivery with or without instrument or with cesarean delivery. Composite score	Bivariate correlations and multivariate logistic regression Hospital level 179 hospitals	Increased score on following PSC dimensions were moderately associated with reduced PSI composite score: ^aFrequency events reported ^bHandoffs and transitions ^cManagement support for patient safety ^dOrganisational learning – continuous improvement ^eStaffing ^fTeamwork across units ^gTeamwork within units ^hOverall perceptions of patient safety ⁱ Supervisor/manager expectations and actions ^j Patient safety grade ^kHSOPS composite average The PSC dimensions Communication openness, Feedback and communication about error and Number of events reported were not significantly correlated to PSI composite score Bivariate: ^{1−3a,b,c,e,f,h,k}p < 0.001 ^d,g,ip < 0.01 ^jp < 0.05	Fair
McLinton (2019) Australia [52] Physical Safety Climate and Psychosocial Safety Climate/ Patient Incidents	Psychosocial safety climate – 12 4/12 (complete) NR Nurses, physicians, managers, others n = 436/ 60 teams (groups of individuals with an identifiable leader) Response rate: NR	Institutional incident safety system data NR Average number of incidents/ patient Any events causing harm or “near miss” accident (i.e. medication errors and falls)	Multilevel correlation and hierarchical linear model Pearson correlation Individual and group level 1 hospital	Increased ^aPsychosocial safety climate composite scores and ^bBurnout were significantly associated with reduced rates of patient incidents Increased Psychosocial safety climate composite scores were significantly associated with reduced scores of ^cBurnout and increased scores of ^dEngagement Increased Physical safety climate composite scores were significantly associated with increased scores of ^eBurnout Increased scores in Psychosocial safety climate and Physical safety climate were significantly associated with reduced scores of ^fEmotional demand, ^fBullying, ^fSkill discretion ^ap < .0.001 ^bp < .0.01 ^c < 0.01- ind. level < 0.01- team level ^d < 0.001- ind. level < 0.05- team level ^e < 0.05- ind. level < 0.001- team level ^fp < .0.01–0.001	Fair
Najjar (2015) Palestine [53] Patient Safety Culture/ Adverse Events	HSOPS – Arabic 12/42 (complete) Surgery/ medicine/ obstetrics Nurses, physicians, others n = 316 Response rate: 74%	Chart review Global Trigger Tool (GTT) 640 Records 54 Triggers	Bivariate regression Spearman rho correlation Unit level 2 hospitals 8 units	Increased scores in 8/15 PSC dimensions were significant associated with reduced AE rates: ^aAggregate safety culture ^bHospital management support ^cNon-punitive response to error ^dOpen communication/ feedback received on error ^eTeamwork within units ^fSupervisor expectations and action promoting patient safety ^gOrganisational learning ^hPatient safety grade ^a,d p < .0.001 ^b,ep < 0.002 ^cp < 0.020 ^fp < 0.003 ^gp < 0.011 ^hp < 0.018	Good
Odell (2019) US [54] Hospital Safety Culture/ Surgical Outcomes	SAQ – modified + engagement surgeons 8/57 (complete) Surgery Nurses, physicians, others n = 871 Response rate: 47%	Register American College of Surgeons (ACS) NSQIP database NR 4 AEs: Risk-adjusted morbidity, mortality, DSM and unplanned readmission rates Morbidity measure captures cardiac arrest requiring resuscitation, myocardial infarction, ventilator dependence > 48 h, pneumonia, progressive renal insufficiency, acute renal failure, sepsis or septic shock, deep incisional, organ space, superficial surgical site infection, stroke/ CVA, unplanned intubation, urinary tract infection, dehiscence DSM includes complications in the morbidity measure except for ventilator dependence, superficial SSI, stroke/ CVA, and additionally includes venous thromboembolism	Linear regression and hierarchical logistic regression ^ap < 0.007 ^b p < 0.004 ^c p < 0.23 ^d p < 0.52 Hospital level 49 hospitals	Increased PSC composite scores were associated with reduced rates of ^apostoperative morbidity and^b DSM No significant association between PSC and ^cthe risk of ^dmortality or readmission	Fair
Olds (2017) US [55] Hospital Safety Climate/ Mortality	HSOPS + (named Multi-state Nursing Care and Patient safety study survey) NR/7 Emergency/ acute care Nurses n = 27,009 Response rate: 39% (non-responders assessed) (97)	Register Discharge records 852,974 patients 1 AE: In-hospital mortality	Bivariate correlation and multivariate logistic regression Hospital level 600 hospitals	^aIncreased PSC composite score was correlated with reduced mortality ^bPerception of safety climate is not predictive of patient mortality beyond the Effect of nurse environments ^ap < 0.001 ^bp < 0.316	Good
Profit (2020) US [56] Safety Culture/ Quality of Care	SAQ 6/30 (complete) NICU Nurses, physicians, others n = 2,073 Response rate: 62.9%	Register CPQCC clinical data NR 9 AEs: Antenatal corticosteroids, hypothermia, pneumothorax, healthcare-associated infection, chronic lung disease, retinopathy screen, discharge on any human milk, growth velocity, mortality	Correlation tests Pearson r correlation ^ap < 0.01 ^bp < 0.05 Unit level 44 units	^aIncreased scores of Teamwork climate and ^bSafety climate were correlated with a reduction in 1/9 of the metrics, healthcare associated infections (HAI)	Good
Quach (2021) US [31] Safety Climate/ Adverse Events	CESARS (ORCA`s organisational culture) 7/28 (+ 6/23) (complete) Outpatient/ homes Nurses, physicians, others n = 1,397 (first survey)/ n = 1,645 (second survey) Response rate: 26.4% and 27.7%	Register FY2017-FY2018 Minimum Data Set VHA 4 AEs: New/ worsened pressure ulcers, falls, major injuries from falls, catheter use	Bivariate logistic regression Group level 56 CLCs	Increased scores of Supervisor`s commitment to safety were associated with ^areduced rates of falls (clinicians) and ^breduced rates of catheter use (nurses) Increased scores of Environmental safety were associated with ^creduced rates of pressure ulcers (clinicians), ^dreduced rates of major injuries from falls (nurses), and ^ereduced rates of catheter use (nursing assistants) ^fIncreased scores of Global ratings were associated with higher level of catheter use for nurses and nursing assistants ^a,ep < 0.05 ^b,c,d,fp < 0.01	Fair
Rosen (2010) US [57] Hospital Safety Climate/ Safety Outcomes	PSCHO 11/42 NR Physicians, managers, others n = 9,309 Response rate: 50%	Chart review PSI software discharge records 13 AEs: Complications of anesthesia, decubitus ulcer, failure to rescue, iatrogenic pneumothorax, infection due to medical care, postoperative (po) fracture, po hemorrhage or hematoma, po physiologic and metabolic derangement, po respiratory failure, po pulmonary embolism or deep vein thrombosis, po sepsis, po wound dehiscence, accidental puncture or laceration	Linear regression Hospital and group level 30 hospitals	No statistically significant association between PSC overall scores and rates of PSIs or PSI composite rates Increased scores of individual dimensions were correlated with reduced rates of specific PSIs: Fear of blame and punishment for making mistakes with ^adecubitus ulcer and ^bpostoperative pulmonary embolism or deep vein thrombosis Perception of lower psychological safety with ^cfailure to rescue Overall emphasis on safety with ^ddecubitus ulcer and ^eiatrogenic pneumothorax ^acdep < 0.05 ^bp < 0.01	Fair
Shahian (2018) US [58] Hospital Safety Culture/ Mortality	HSOPS 12/42 (complete) Emergency/ acute care NR n = 257 hospital- surveys n = 834 average/ hospital Response rate: 54% (5–100%)	Register MEDPAR 1,609 patients 19,357 discharges 1 AE: Risk-adjusted mortality	Multivariate hierarchical logistic regression Hospital level 171 hospitals	No statistically significant association was found between PSC scores and rates of 30-day mortality Reported as OR	Fair
Singer (2009) US [59] Hospital Safety Climate/ Safety Performance	PSCHO 8/38 Emergency/ acute care Nurses, physicians, others n = 18,223 Response rate: 52%	Register MEDPAR- PSI 12 AEs: Complications of anesthesia, decubitus ulcer, iatrogenic pneumothorax, infection due to medical care, postoperative (po) hip fracture, po hemorrhage or hematoma, po physiologic and metabolic derangement, po respiratory failure, po pulmonary embolism or deep vein thrombosis, po sepsis, po wound dehiscence, accidental puncture or laceration	Multilevel logistic regression Hospital level 91 hospitals	Increased scores in the PSC dimension ^aFear of shame/ blame and ^bOverall PSC were associated with reduced PSI composite rates ^abp < 0.05	Fair
Smits (2012) Netherlands [60] Patient Safety Culture/ Unintended Events	HSOPS – Dutch version named COMPaZ 11/40 Surgery/ medicine/ Emergency and acute care Nurses, physicians, managers, others n = 542 Response rate: 56%	Prospective reporting Staff wrote reports of all unintended events 1,885 Events 8 Classifications: Materials and equipment, diagnosis and treatment, medication, protocols and regulations, incorrect data and substitutions, collaboration with resident physicians and consultants, collaboration with other departments and other	Multilevel logistic regression Unit level 20 hospitals 28 units	Increased scores in 3/11 PSC dimensions, ^aNonpunitive response to error, ^bHospital management support and ^cWillingness to report were significantly associated with reduced rates of unintended events (medication, materials/ equipment and collaboration with resident physicians/consultants) ^a,bp < 0.01–0.05 ^cp < 0.001–0.01	Fair
Steyrer (2013) Austria [61] Safety Climate/ Medical Error	VSCQ 4/40 ICU Nurses, physicians n = 734 Response rate: 41.4% (nurses) and 35.2% (physicians)	Prospective reporting Form to record predefined medical errors 48 h 378 patients 7 categories AEs: Administration of medication, unplanned dislodgement of airways, arterial lines, central venous catheters, urinary catheters, enteral nutrition probes, or drains. Error rate: rate of ratio affected by errors in an ICU/ total number of patients	OLS regression (Ordinary Least Squares) Unit level 57 units	Increased scores of following PSC dimensions were significantly associated with reduced AE composite rates ^aIncreased scores of Workloads increases the error composite rate, and ^b increased scores of Safety climate reduced the AE composite rate PSC scores were more associated with reduced AE composite rate than safety tools ^ap < 0.01 ^bp < 0.05
Tawfik (2019) US/ California [62] Safety Climate and Strength/ Outcomes	SAQ 0/7 NICU Nurses, physicians, others n = 2,073 Response rate 62.9%	Register CPQCC clinical data 6,682 patients 4 AEs: LOS, infections, chronic lung disease and mortality	Logistic linear regression Unit level 44 units	Increased scores of Safety climate strength (the consistency of responses) were significantly associated with reduced ^aLOS Safety strength and Safety climate predicted LOS more than Safety climate separately, and increased scores of Safety strength were associated with lower odds of infection, but not other secondary outcomes ^ap < 0.001	Good
Thomas-Hawkins (2015) US [33] Patient Safety Culture/ Adverse Events	HSOPS – modified 2/5 Outpatient/ homes Nurses n = 422 Response rate 52%	Retrospective nurse-estimate Incidence of AEs last year – series of survey items, 7-point scale 422 Nurses 13 AEs: Medication error, complaints from patient/family, vascular access infection, vascular assess infiltration, hospital admission, skipped dialysis, shortened dialysis, dialysis hypertension, falls without injuries, falls with injury, bleeding from vascular access, emergency room use, vascular access thrombosis	Logistic regression Unit level From 47 states	Increased scores of Poor to failing patient safety grade were significantly associated with reduced rates of ^amedication error, ^bcomplaints from patient/ family, ^cvascular access infection, ^dhospital admission and ^eskipped dialysis, ^ffalls without injuries, ^gbleeding from vascular access and ^hemergency room use Increased scores of Patient handoffs and transitions were significantly associated with reduced rates of ⁱvascular access thrombosis, ^jcomplaints from patient/ family, ^kskipped dialysis, ^l shortened dialysis, ^memergency room use, ⁿbleeding from vascular access, ^o vascular access infection, ^pmedication error and ^qvascular access infiltration Increased scores of overall PSC were significantly associated with lower odds of frequent rates of medication errors by nurses, patient hospitalisation, vascular access infection, and patient complaints ^a,b,i,j,k,lp < 0.001 ^c,d,e,m,n,op < 0.01 ^f,g,h,p,qp < 0.05	Poor
Valentin (2013) Austria [63] Safety Climate/ Medical Error	VSCQ 5/53 (complete) ICU Nurses, physicians n = 2,563 Response rate: 41.5% (nurses) and 35.2% (physicians)	Prospective reporting Form to record predefined AEs 48 h 795 patients 2 AEs: Medication errors and dislodgement errors	Multivariate logistic regression Unit level 57 units	^aIncreased scores of Safety climate overall were significantly associated with reduced AE rates ^bIncreased scores of Workloads at patient level were statistically significantly associated with increased AE rates ^abp < 0.01	Fair
Wang (2014) China [64] Patient Safety Culture/ Adverse Events	HSOPS 12/ 42 (complete) Surgery, medicine, Emergency/ acute care, Intensive Care Unit Nurses n = 463 Response rate: 72.3%	Retrospective nurse-estimate Incidence of AEs last year, 7-point scale 463 nurses 7 AEs: Pressure ulcers, prolonged physical restraint, complaints from patient/family, medicine errors, infusion or transfusion reaction, patient falls, surgical wound infection	Bivariate and Multivariate logistic regression Unit and hospital level 7 hospitals 28 units	Increased scores of the following PSC dimensions were significantly associated with reduced rates of specified AEs: Organisational learning – continuous improvement with ^apressure ulcers, ^bprolonged physical restraint and ^ccomplaints from patient/family Frequency of event reporting with ^dmedicine errors and ^epressure ulcers Feedback and communication about error with ^fpressure ulcer and ^ginfusion or transfusion reaction Hospital Management support for patient safety with ^hmedicine error and ⁱinfusion or transfusion reaction Supervision expectations and actions promoting safety with ^jcomplaints from patient/ family Non-punitive response to error with ^kpressure ulcers Handoffs and Transitions with ^linfusion or transfusion reaction ^ap = 0.002 ^bp = 0.019 ^cp = 0.013 ^dp = 0.021 ^ep = 0.006 ^fp = 0.037 ^gp = 0.041 ^hp = 0.006 ⁱp = 0.027 ^jp = 0.029 ^kp = 0.045 ^lp = 0.034	Fair

First author
-Year
-Origin
-Concepts PSC/ AE

PSC
-Tool
-Dimensions/items
-Context
-Participants PSC
-n =
-Response rate %

AE
-Method
-Number patients/records
-Numbers and types of AEs assessed

Analysis
-Methods
-Study level
-Number of units

Reported association
Statistical values

Critical appraisal
Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies [29]

Ausserhofer (2013) Swiss [34]

Patient Safety Climate/

Patient Outcomes

SOS

1/9 (complete) –translated

Surgery/medicine/ mixed

Nurses

n = 1,630

Response rate: 72%

Retrospective nurse-estimate

Incidence of AEs last year –

7-point scale

1,630 nurses

7 AEs: Nurse-reported medication errors, pressure ulcers, patient falls, urinary tract infection, bloodstream infection, pneumonia (and patient satisfaction)

Bivariate and multivariate logistic regressions

Unit and hospital level

132 units in 35 hospitals

No statistically significant associations between PSC scores and 6 selected AEs: pressure ulcers, urinary tract infection, bloodstream infection, medication errors, pneumonia, patient falls (nor any association between PSC and patient satisfaction)

Increased scores of Rationing of nursing care were consistently associated with increased rates of ^bbloodstream infection, ^bmedication administration error, ^cpneumonia (and ^dpatient satisfaction) in multivariate analysis

(Reduced nurse ratio increases AE rates)

^ap = 0.004

^bp = 0.026

^cp = 0.027

^dp = 0.005

Fair

Bacon (2021)

US [35]

Organisational Safety Climate/

Mortality and Failure-to-rescue

Safety Climate Tool – revised (originally Zohar`s measure of safety climate)

0/33

Surgery

Nurses, physicians, others

n = 261

Response rate: NR

Chart review

AHRQ Quality Indicators

10,823 patients

52,898 records

2 AEs: Failure-to-rescue (deep vein thrombosis/ pulmonary embolus, pneumonia, sepsis, cardiac arrest, hemorrhage) and in-hospital mortality

Multilevel models

Bonferroni correlation

Hospital level

2 hospitals

No statistically significant associations between PSC scores and rates of in-hospital mortality or failure-to-rescue

Reported in p-values

Fair

Birkmeyer (2013) US [36]

Safety Culture/

Complications

HSOPS + SAQ-OR + questions disruption

4/30

Surgery

Surgeons, nurses/ operating room technicians, operating room administrators

n = 184

Response rate: surgeons 95% nurses/ operating room technicians 82%, operating room administrators 68%

Chart review

Standardised instrument

24,117 patients

10 types of AEs: Abdominal abscess, bowel obstruction, leak, bleeding, respiratory failure, renal failure, wound infection/ dehiscence, venous thromboembolism, myocardial infarction or cardiac arrest, death. The AE rate is the overall rate of the 10 AEs

Bivariate regression model

Spearman`s

Correlation (p)

Hospital level

22 hospitals

Increased overall PSC scores from ^asurgeons and ^bnurses were associated with reduced AE rates. Increased scores of ^cHospital safety culture by nurses and increased scores of ^dOperating room safety by surgeons were associated with reduced AE rates

^ap < 0.001 bp < 0.011

^cp < 0.002

^dp < 0.045

Fair

Bosch (2011)

Netherlands [32]

Organisational Culture and Teamwork Climate/

Pressure Ulcers

Team Climate Inventory (TCI)- short +

Competing Values Framework (CVF) (both translated)

0/14 + 5/20

Ward units in hospitals and nursing homes

Physicians, nurses, nursing assistants

n = 460

Response rate 41% for the hospital wards, and 39% for the nursing home wards

Prospective nurse-reporting and scoring

1,274 patients

1 AE: Nosocomial pressure ulcers

Bivariate and multilevel logistic regression

(General Linear regression)

Unit level

104 units

No statistically significant association between PSC scores, Team climate or Preventive quality management at ward level, and the prevalence of nosocomial pressure ulcers

(^aIncreased scores of Institutional quality management were significantly correlated with increased scores of Preventive quality management at ward level.)

^ap < 0.001

Fair

Brown (2013)

US [37]

Safety Culture/

Adverse Patient Outcomes

HSOPS

12/42 (complete) +

Global rating composite (4 items)

Nurses

Response rate: NR

National Database for Nursing Quality Improvement. The Collaborative Alliance for Nursing Outcomes (CALNOC). CALNOC indicator definitions

3 AEs: Health Acquired Pressure Ulcers, reported falls and falls with injury

Linear regression, Pearson product-moment correlations (r)

Unit level

9 hospitals

37 units

^aIncreased scores of Teamwork within units were associated with reduced reported falls and ^bincreased scores of Management support were associated with increased rates of reported falls

^cIncreased Global rating composite was associated with reduced rates of Health-Acquired Pressure Ulcers (HAPU)

^dSkill mix, Staff turnover and Workload intensity are strongly corelated with PSC scores

^ap < 0.05

^bp < 0.03

^cp < 0.05

^dp < 0.01–0.05

Poor

Brubakk (2019) Norway [38]

Organisational Culture/

Mortality

SAQ (and organisational factors survey)

1/2 (+ 19/ 57)

Emergency/ acute care

Nurses, physicians, managers

n = 8,800

Response rate: 72% 2010

77% in 2011

75% in 2012

The Norwegian Institute of Public Health

46,026 admitted patients

1 AE: Risk-adjusted 7-day mortality

Multivariable regression

Unit and group level

20 hospitals

56 units

^aReduced scores of Patient safety climate and ^bLeadership were associated with increased 7-day mortality

^cIncreased scores of Workload perceived by nurses were associated with increased 7-day mortality rates. ^dIncreased Middle manager`s engagement levels were associated with reduced 7-day mortality rates

^ap < 0.003

^bp < 0.045

^cp < 0.028

^dp < 0.037

Good

Camargo (2012) US/ 20 States [39]

Safety Climate/

Adverse Events and Medical Error

Survey – constructed

9/50

Emergency/ acute care

Nurses, physicians

n = 3,562

Response rate: 66%

Chart review

Standardised form: 18 questions

9,821 charts

3 AE-categories: Medical error, adverse event (preventable and non-preventable) and near miss (intercepted and non-intercepted) Scaled: Significant, serious, life threatening or fatal

Multivariable regression models

Incident Rate Ratio (IRR)

Unit level

62 units

No statistically significant association between PSC scores and preventable AE rates, nor was there an association between PSC scores and serious violations of treatment guidelines

^aIncreased PSC scores were significantly associated with increased “Intercepted near misses”

^aIRR 1.79 (1.06–3.03)

Fair

Davenport (2007) US [40]

Organisational Safety Climate/

Outcomes

SAQ

6/30 (complete)

Surgery

Nurses, physicians, others

n = 6,083

Response rate: 52%

Chart review

NSQIP protocol

57,880 patients

2 AEs: Risk-adjusted surgical morbidity (patient having 1 or more out of 21 complications) and 30-day mortality

Multivariate logistic regression

Spearman`s p correlation

Hospital level

52 hospitals

No statistically significant association between PSC scores and rates of 30-day mortality or 30-day morbidity (21 postoperative complications)

^aIncreased scores of Communication/ collaboration with doctors were correlated with reduced rates of risk-adjusted morbidity

Scores of Burnout was not correlated with AE rates

^ap < 0.01

Good

Fan (2016)

US/ Minnesota [41]

Safety Culture/

Surgical Outcomes

HSOPS

12/42 (complete)

Surgery

n = 1,926

Response rate: 43%

Prospective reporting as defined by the

National Healthcare Safety Network

1 AE: Postoperative colon surgery SSI/ number of operations performed

Bivariate and multivariate linear regression

Pearson product-moment correlations (r)

Unit level

7 units/hospitals

Increased scores of following PSC dimensions were associated with reduced rates of Colon SSI:

^aTeamwork across units

^bTeamwork within units

^cOrganisational learning

^dFeedback and communication about error

^eCommunication openness*

^fOverall perception of safety

^gManagement support for patient safety

^hSupervisor/manger expectations and actions promoting safety

ⁱNon-punitive response to error*

^jFrequency of events reported

^kHandoffs and transitions when adjusting for ASA and surgical volume

^a r = -0.96, [-0.76, -0.99]

^b r = -0.88, [-0.38, -0.98]

^c r = -0.95, [-0.71, -0.99]

^d r = -0.92, [-0.56, -0.99]

^e r = -0.85, [-0.26, -0.98]

^f r = -0.90; [-0.45, -0.99]

^g r = -0.90, [-0.44, -0.98]

^h r = -0.85, [-0.25, -0.98]

ⁱ r = -0.78, [-0.07, -0.97]

*non-significant when adjusting for ASA and surgical volume

^j r = -0.76, [-0.01, -0.96]

^kp < 0.05

Fair

Garrouste-Orgeas (2015) France [42]

Safety Culture/

Medical Error

SAQ-ICU

6/63 (complete)

ICU

Nurses, physicians, others

n = 1,534

Response rate: 77.2%

Prospective reporting/ observation

8 h /5 days /2 weeks combined to chart review

4 AEs: Error administration anticoagulant medication, error prescribing anticoagulant medication, error administration insulin, accidental removal of a central venous catheter, accidental extubating

Multivariate hierarchical model

Unit level

31 units

Limited statistically significant association between PSC scores and rates of medical errors/ patient days^a

Increased scores of ^bDepression symptoms, ^cICU organisation (40% off work previous day), ^dStaff-specific safety training programme and ^ePatient level/ workload were associated with increased AE rates

Increased scores of Burnout were not statistically significantly correlated with increased AE rates

^ap = 0.04–0.87

^bp = 0.01

^cp = 0.01

^dp = 0.001–0.02

^ep = < 0.0001- 0.03

Good

Han (2020)

South Korea [43]

Patient Safety Culture/

Adverse Events

HSOPS

12/42 (complete)

n = 212

Response rate: 86%

Retrospective nurse-estimate

Incidence of AEs last year –

7-point scale

212 nurses

5 AEs: Falls, medication errors, pressure ulcers, health-associated infections (surgical site, urinary tract, central-line associated bloodstream infections and ventilator-associated pneumonia) and physical restrain ≥ 8 h

(Combined into a binominal variable: “never/happened”)

Bivariate regression and multiple logistic regression. Odds Ratio (OR)

Hospital level

2 hospitals

Increased scores of the following PSC dimensions were associated with reduced AE rates:

^aSupervisor/ manger expectations and actions promoting safety and 4 AEs

^bCommunication openness and 4 AEs

^cManagement support for patient safety and 3 AEs

^dTeamwork across units and 3 AEs

^eTeamwork within units and 2 AEs

^fFeedback and communication about error and 2 AEs

^gNonpunitive response to error and 1 AE

^hHandoffs and transitions and 1 AE

ⁱOrganisational learning – continuous improvement and 1 AE

Increased scores for Patient safety competencies were associated with reduced AE rates

Odds Ratio (OR):

^a OR 0.33–0.39

^b OR 0.25–0.51

^c OR 0.22–0.55

^d OR 0.29–0.47

^e OR 0.23–0.51

^f OR 0.43–0.52

^g OR 2.08

^h OR 2.02

ⁱ OR 0.053

^j OR 0.024–0.049

Fair

Haynes

(2011) US [44]

Safety Climate/

Postoperative Morbidity and Mortality

SAQ-OR

NR/6

Surgery

Nurses, physicians, others

n = 281

pre intervention

n = 257

post intervention

Response rate: 97.7%

Chart reviews and communication with clinical teams

Charts were reviewed at discharge or in 30 days

19 AEs: Acute renal failure, bleeding requiring ≥ 4 units of red cell transfusion within 72 h after surgery, cardiac arrest requiring cardiopulmonary resuscitation, coma for ≥ 24 h, deep venous thrombosis, myocardial infarction, unplanned intubation, ventilator use for ≥ 48 h, pneumonia, pulmonary embolism, stroke, major wound disruption, surgical site infection, sepsis, septic shock, systemic inflammatory response syndrome, unplanned return to the OR, vascular graft failure and death

Correlation analysis

Spearman`s correlation (p)

Hospital level

8 hospitals

Increased PSC scores were associated with reduction in postoperative complication rate^a including 18 AE rates and mortality

(The measurement was related to an intervention.)

^a r = 0.7143

p < 0.0381

Fair

Hofmann (2006)

US [45]

Safety Climate/

Medication Error and Patient Outcomes

Zohar`s measure of safety climate – revised and Error Orientation Scale

3/9 + 3/13

Surgery/ medicine

Nurses

n = 1,127

Response rate: NR

Chart review

Coordinators collected the AE frequency over 3 months

2 AEs: Medication Errors and Urinary Tract Infections

Bivariate regression and multiple logistic regression

Unit level

42 hospitals

81 units

A) Increasing overall PSC scores significantly predict reduced rates of ^amedication errors and ^burinary tract infections

B) Regression with^abshow that^a was significantly moderated by Patient complexity

A)^a -1.51 p < 0.05

^b -1,27 p < 0.05

B) ^ab -7.85 p < 0.05

Fair

Huang (2010)

US [46]

Safety Culture/

Outcomes

SAQ-ICU

6/60 (complete)

ICU

Nurses, physicians, others

n = 2,103

Response rate: 47.9%

PICCM clinical national database

65,978 patients

2 AEs: Hospital mortality and LOS

Linear regression model and multivariate logistic regression

Unit level

30 units

^aIncreased scores of Perceptions of management were associated with reduced mortality rates

^bIncreased scores of Safety climate were associated with reduced LOS

^ap = 0.005

^bp = 0.003

Fair

Hwang (2011)

Korea [47]

Safety Climate/

Medical Errors

SAQ – translated 17 items from 4 dimensions + 4 items added due to Korean context

2/21

Nurses

n = 1,923

Response rate: 89.7%

Nurse-estimate experienced errors in retrospective questionnaire

n = 277 nurses

AEs last year: Yes/ No

Frequency of AEs last year

Multiple logistic regression

Hospital level

33 hospitals

Nurses with better scores of ^aworkgroup and ^borganisation-level Safety climate were associated with reduced error rates

Odds Ratio:

^a(OR = 0.73) p < 0.001

^b(OR = 0.69) p < 0.001

Poor

Kakemam (2021)

Iran [48]

Patient Safety Culture/

Adverse Events

HSOPS- Persian

12/42 (complete)

Emergency/ acute care, ICU, surgery, medicine, NR

Nurses

n = 2,995

Response rate: 51.1%

Retrospective nurse-estimate

Incidence of AEs last year, 7-point scale

2,995 nurses

6 AEs: Pressure ulcer, patient falls, adverse drug events, surgical wound infection, complaints from patients or their family, infusion or transfusion reaction

Bivariate and multiple logistic regression models

Hospital level

32 hospitals

^aIncreased scores of nine PSC dimensions were significantly associated with a reduced perception of AE rates in at least two out of six AEs

^ap < 0.001

Fair

Kline (2008) Canada [49]

Patient Safety Culture/

Adverse Events

Patient Safety Culture 2005 Survey – Database

1/5

NR (reported as nursing leaders in primary study)

n = 298 (408/417 in primary study)

Response rate:

83%/ 72% (reported in primary study)

Regional Incident Reports in forms by any health staff

5,070 incident reports/

3,093 non-incident reports

Severity range 1–4 and contributing factors/ categorised as 7 incident types: Care and treatment, injury or death, falls, medication discrepancy, medication incident, test or results, vaccine

Hierarchical linear regression – multilevel

Unit level

3 hospitals

40 units

^aResource intensity predicts incident severity level

^bPSC predicts “adverse event severity” over “case resource intensity”

^ap < 0.001

^bp < 0.05

R² = 0.093

Fair

Lee (2018) Canada [50]

Organisational Safety Culture/

Adverse Events

HSOPS

NS/7

Nurses

n = 1,053

Response rate: NR

Retrospecitve nurse-estimate

Incidence of AEs last year,

7-point scale

1,053 nurses

3 AEs: Medication error, patient falls with injury, urinary tract infection (+ quality of care)

Multilevel ordinal logistic and linear regression

Pearson and Spearman correlation

Hospital level

63 hospitals

^aIncreased Overall organisational safety culture was associated with reduced rates of reported medication errors, falls with injury and urinary tract infections

^aIncreased scores of Overall organisational safety culture increased the quality of care

^ap < 0.05

Fair

Mardon (2010)

US [51]

Patient Safety Culture/

Adverse Events

HSOPS

12/42 (complete)

n = 56,480

Response rate: 51%

HCUP at AHRQ – Patient safety indicators (PSI)

8 AEs: Complications of anesthesia, death in low mortality diagnostic related groups, failure to rescue, foreign body left in during procedure, transfusion reaction, birth trauma – injury to neonate, obstetric trauma vaginal delivery with or without instrument or with cesarean delivery. Composite score

Bivariate correlations and multivariate logistic regression

Hospital level

179 hospitals

Increased score on following PSC dimensions were moderately associated with reduced PSI composite score:

^aFrequency events reported

^bHandoffs and transitions

^cManagement support for patient safety

^dOrganisational learning – continuous improvement

^eStaffing

^fTeamwork across units

^gTeamwork within units

^hOverall perceptions of patient safety

ⁱ Supervisor/manager expectations and actions

^j Patient safety grade

^kHSOPS composite average

The PSC dimensions Communication openness, Feedback and communication about error and Number of events reported were not significantly correlated to PSI composite score

Bivariate:

^{1−3a,b,c,e,f,h,k}p < 0.001

^d,g,ip < 0.01

^jp < 0.05

Fair

McLinton (2019)

Australia [52]

Physical Safety Climate and Psychosocial Safety Climate/

Patient Incidents

Psychosocial safety climate – 12 4/12 (complete)

Nurses, physicians, managers, others

n = 436/ 60 teams (groups of individuals with an identifiable leader)

Response rate: NR

Institutional incident safety system data

Average number of incidents/ patient

Any events causing harm or “near miss” accident (i.e. medication errors and falls)

Multilevel correlation and hierarchical linear model

Pearson correlation

Individual and group level

1 hospital

Increased ^aPsychosocial safety climate composite scores and ^bBurnout were significantly associated with reduced rates of patient incidents

Increased Psychosocial safety climate composite scores were significantly associated with reduced scores of ^cBurnout and increased scores of ^dEngagement

Increased Physical safety climate composite scores were significantly associated with increased scores of ^eBurnout

Increased scores in Psychosocial safety climate and Physical safety climate were significantly associated with reduced scores of ^fEmotional demand, ^fBullying, ^fSkill discretion

^ap < .0.001

^bp < .0.01

^c < 0.01- ind. level

< 0.01- team level

^d < 0.001- ind. level

< 0.05- team level

^e < 0.05- ind. level

< 0.001- team level

^fp < .0.01–0.001

Fair

Najjar (2015) Palestine [53]

Patient Safety Culture/

Adverse Events

HSOPS – Arabic

12/42 (complete)

Surgery/ medicine/ obstetrics

Nurses, physicians, others

n = 316

Response rate: 74%

Chart review

Global Trigger Tool (GTT)

640 Records

54 Triggers

Bivariate regression

Spearman rho correlation

Unit level

2 hospitals

8 units

Increased scores in 8/15 PSC dimensions were significant associated with reduced AE rates:

^aAggregate safety culture

^bHospital management support

^cNon-punitive response to error

^dOpen communication/ feedback received on error

^eTeamwork within units

^fSupervisor expectations and action promoting patient safety

^gOrganisational learning

^hPatient safety grade

^a,d p < .0.001

^b,ep < 0.002

^cp < 0.020

^fp < 0.003

^gp < 0.011

^hp < 0.018

Good

Odell (2019)

US [54]

Hospital Safety Culture/

Surgical Outcomes

SAQ – modified + engagement surgeons

8/57 (complete)

Surgery

Nurses, physicians, others

n = 871

Response rate: 47%

American College of Surgeons (ACS) NSQIP database

4 AEs: Risk-adjusted morbidity, mortality, DSM and unplanned readmission rates

Morbidity measure captures cardiac arrest requiring resuscitation, myocardial infarction, ventilator dependence > 48 h, pneumonia, progressive renal insufficiency, acute renal failure, sepsis or septic shock, deep incisional, organ space, superficial surgical site infection, stroke/ CVA, unplanned intubation, urinary tract infection, dehiscence

DSM includes complications in the morbidity measure except for ventilator dependence, superficial SSI, stroke/ CVA, and additionally includes venous thromboembolism

Linear regression and hierarchical logistic regression

^ap < 0.007

^b p < 0.004

^c p < 0.23

^d p < 0.52

Hospital level

49 hospitals

Increased PSC composite scores were associated with reduced rates of ^apostoperative morbidity and^b DSM

No significant association between PSC and ^cthe risk of ^dmortality or readmission

Fair

Olds (2017)

US [55]

Hospital Safety Climate/

Mortality

HSOPS + (named Multi-state Nursing Care and Patient safety study survey)

NR/7

Emergency/ acute care

Nurses

n = 27,009

Response rate: 39% (non-responders assessed) (97)

Discharge records

852,974 patients

1 AE: In-hospital mortality

Bivariate correlation and multivariate logistic regression

Hospital level

600 hospitals

^aIncreased PSC composite score was correlated with reduced mortality

^bPerception of safety climate is not predictive of patient mortality beyond the Effect of nurse environments

^ap < 0.001

^bp < 0.316

Good

Profit (2020)

US [56]

Safety Culture/

Quality of Care

SAQ

6/30 (complete)

NICU

Nurses, physicians, others

n = 2,073

Response rate: 62.9%

CPQCC clinical data

9 AEs: Antenatal corticosteroids, hypothermia, pneumothorax, healthcare-associated infection, chronic lung disease, retinopathy screen, discharge on any human milk, growth velocity, mortality

Correlation tests

Pearson r correlation

^ap < 0.01

^bp < 0.05

Unit level

44 units

^aIncreased scores of Teamwork climate and ^bSafety climate were correlated with a reduction in 1/9 of the metrics, healthcare associated infections (HAI)

Good

Quach (2021)

US [31]

Safety Climate/

Adverse Events

CESARS

(ORCA`s organisational culture)

7/28 (+ 6/23)

(complete)

Outpatient/ homes

Nurses, physicians, others

n = 1,397 (first survey)/

n = 1,645 (second survey)

Response rate: 26.4% and 27.7%

FY2017-FY2018 Minimum Data Set VHA

4 AEs: New/ worsened pressure ulcers, falls, major injuries from falls, catheter use

Bivariate logistic regression

Group level

56 CLCs

Increased scores of Supervisor`s commitment to safety were associated with ^areduced rates of falls (clinicians) and ^breduced rates of catheter use (nurses)

Increased scores of Environmental safety were associated with ^creduced rates of pressure ulcers (clinicians), ^dreduced rates of major injuries from falls (nurses), and ^ereduced rates of catheter use (nursing assistants)

^fIncreased scores of Global ratings were associated with higher level of catheter use for nurses and nursing assistants

^a,ep < 0.05

^b,c,d,fp < 0.01

Fair

Rosen (2010)

US [57]

Hospital Safety Climate/

Safety Outcomes

PSCHO

11/42

Physicians, managers, others

n = 9,309

Response rate: 50%

Chart review

PSI software discharge records

13 AEs: Complications of anesthesia, decubitus ulcer, failure to rescue, iatrogenic pneumothorax, infection due to medical care, postoperative (po) fracture, po hemorrhage or hematoma, po physiologic and metabolic derangement, po respiratory failure, po pulmonary embolism or deep vein thrombosis, po sepsis, po wound dehiscence, accidental puncture or laceration

Linear regression

Hospital and group level

30 hospitals

No statistically significant association between PSC overall scores and rates of PSIs or PSI composite rates

Increased scores of individual dimensions were correlated with reduced rates of specific PSIs:

Fear of blame and punishment for making mistakes with ^adecubitus ulcer and ^bpostoperative pulmonary embolism or deep vein thrombosis

Perception of lower psychological safety with ^cfailure to rescue

Overall emphasis on safety with ^ddecubitus ulcer and ^eiatrogenic pneumothorax

^acdep < 0.05

^bp < 0.01

Fair

Shahian (2018)

US [58]

Hospital Safety Culture/

Mortality

HSOPS

12/42 (complete)

Emergency/ acute care

n = 257 hospital- surveys

n = 834 average/ hospital

Response rate: 54% (5–100%)

MEDPAR

1,609 patients

19,357 discharges

1 AE: Risk-adjusted mortality

Multivariate hierarchical logistic regression

Hospital level

171 hospitals

No statistically significant association was found between PSC scores and rates of 30-day mortality

Reported as OR

Fair

Singer (2009)

US [59]

Hospital Safety Climate/

Safety Performance

PSCHO

8/38

Emergency/ acute care

Nurses, physicians,

others

n = 18,223

Response rate: 52%

MEDPAR- PSI

12 AEs: Complications of anesthesia, decubitus ulcer, iatrogenic pneumothorax, infection due to medical care, postoperative (po) hip fracture, po hemorrhage or hematoma, po physiologic and metabolic derangement, po respiratory failure, po pulmonary embolism or deep vein thrombosis, po sepsis, po wound dehiscence, accidental puncture or laceration

Multilevel logistic regression

Hospital level

91 hospitals

Increased scores in the PSC dimension ^aFear of shame/ blame and ^bOverall PSC were associated with reduced PSI composite rates

^abp < 0.05

Fair

Smits (2012) Netherlands [60]

Patient Safety Culture/

Unintended Events

HSOPS – Dutch version named COMPaZ

11/40

Surgery/ medicine/ Emergency and acute care

Nurses, physicians, managers, others

n = 542

Response rate: 56%

Prospective reporting

Staff wrote reports of all unintended events

1,885 Events

8 Classifications: Materials and equipment, diagnosis and treatment, medication, protocols and regulations, incorrect data and substitutions, collaboration with resident physicians and consultants, collaboration with other departments and other

Multilevel logistic regression

Unit level

20 hospitals

28 units

Increased scores in 3/11 PSC dimensions, ^aNonpunitive response to error, ^bHospital management support and ^cWillingness to report were significantly associated with reduced rates of unintended events (medication, materials/ equipment and collaboration with resident physicians/consultants)

^a,bp < 0.01–0.05

^cp < 0.001–0.01

Fair

Steyrer (2013) Austria [61]

Safety Climate/

Medical Error

VSCQ

4/40

ICU

Nurses, physicians

n = 734

Response rate: 41.4% (nurses) and 35.2% (physicians)

Prospective reporting

Form to record predefined medical errors 48 h

378 patients

7 categories AEs: Administration of medication, unplanned dislodgement of airways, arterial lines, central venous catheters, urinary catheters, enteral nutrition probes, or drains. Error rate: rate of ratio affected by errors in an ICU/ total number of patients

OLS regression (Ordinary Least Squares)

Unit level

57 units

Increased scores of following PSC dimensions were significantly associated with reduced AE composite rates

^aIncreased scores of Workloads increases the error composite rate, and ^b increased scores of Safety climate reduced the AE composite rate

PSC scores were more associated with reduced AE composite rate than safety tools

^ap < 0.01

^bp < 0.05

Tawfik (2019)

US/ California [62]

Safety Climate and Strength/

Outcomes

SAQ

0/7

NICU

Nurses, physicians, others

n = 2,073

Response rate 62.9%

CPQCC clinical data

6,682 patients

4 AEs: LOS, infections, chronic lung disease and mortality

Logistic linear regression

Unit level

44 units

Increased scores of Safety climate strength (the consistency of responses) were significantly associated with reduced ^aLOS

Safety strength and Safety climate predicted LOS more than Safety climate separately, and increased scores of Safety strength were associated with lower odds of infection, but not other secondary outcomes

^ap < 0.001

Good

Thomas-Hawkins (2015) US [33]

Patient Safety Culture/

Adverse Events

HSOPS – modified

2/5

Outpatient/ homes

Nurses

n = 422

Response rate 52%

Retrospective nurse-estimate

Incidence of AEs last year – series of survey items, 7-point scale

422 Nurses

13 AEs: Medication error, complaints from patient/family, vascular access infection, vascular assess infiltration, hospital admission, skipped dialysis, shortened dialysis, dialysis hypertension, falls without injuries, falls with injury, bleeding from vascular access, emergency room use, vascular access thrombosis

Logistic regression

Unit level

From 47 states

Increased scores of Poor to failing patient safety grade were significantly associated with reduced rates of ^amedication error, ^bcomplaints from patient/ family, ^cvascular access infection, ^dhospital admission and ^eskipped dialysis, ^ffalls without injuries, ^gbleeding from vascular access and ^hemergency room use

Increased scores of Patient handoffs and transitions were significantly associated with reduced rates of ⁱvascular access thrombosis, ^jcomplaints from patient/ family, ^kskipped dialysis, ^l shortened dialysis, ^memergency room use, ⁿbleeding from vascular access, ^o vascular access infection, ^pmedication error and ^qvascular access infiltration

Increased scores of overall PSC were significantly associated with lower odds of frequent rates of medication errors by nurses, patient hospitalisation, vascular access infection, and patient complaints

^a,b,i,j,k,lp < 0.001

^c,d,e,m,n,op < 0.01

^f,g,h,p,qp < 0.05

Poor

Valentin (2013) Austria [63]

Safety Climate/

Medical Error

VSCQ

5/53 (complete)

ICU

Nurses, physicians

n = 2,563

Response rate: 41.5% (nurses) and 35.2% (physicians)

Prospective reporting

Form to record predefined AEs 48 h

795 patients

2 AEs: Medication errors and dislodgement errors

Multivariate logistic regression

Unit level

57 units

^aIncreased scores of Safety climate overall were significantly associated with reduced AE rates

^bIncreased scores of Workloads at patient level were statistically significantly associated with increased AE rates

^abp < 0.01

Fair

Wang (2014)

China [64]

Patient Safety Culture/

Adverse Events

HSOPS

12/ 42 (complete)

Surgery, medicine,

Emergency/ acute care, Intensive Care

Unit

Nurses

n = 463

Response rate: 72.3%

Retrospective nurse-estimate

Incidence of AEs last year, 7-point scale

463 nurses

7 AEs: Pressure ulcers, prolonged physical restraint, complaints from patient/family,

medicine errors, infusion or transfusion reaction, patient falls, surgical wound infection

Bivariate and Multivariate logistic regression

Unit and hospital level

7 hospitals

28 units

Increased scores of the following PSC dimensions were significantly associated with reduced rates of specified AEs:

Organisational learning – continuous improvement with ^apressure ulcers, ^bprolonged physical restraint and ^ccomplaints from patient/family

Frequency of event reporting with ^dmedicine errors and ^epressure ulcers

Feedback and communication about error with ^fpressure ulcer and ^ginfusion or transfusion reaction

Hospital Management support for patient safety with ^hmedicine error and ⁱinfusion or transfusion reaction

Supervision expectations and actions promoting safety with ^jcomplaints from patient/ family

Non-punitive response to error with ^kpressure ulcers

Handoffs and Transitions with ^linfusion or transfusion reaction

^ap = 0.002

^bp = 0.019

^cp = 0.013

^dp = 0.021

^ep = 0.006

^fp = 0.037

^gp = 0.041

^hp = 0.006

ⁱp = 0.027

^jp = 0.029

^kp = 0.045

^lp = 0.034

Fair

Abbreviations: PSC Patient Safety Culture, AE Adverse Events, SOS Safety Organizing Scale, US United States, NR Not Reported, AHRQ Agency of Healthcare Research and Quality, HSOPS Hospital Survey of Patient Safety Culture, SAQ Safety Attitude Questionnaire, OR Operating Room, TCI Team Climate Inventory, CVF Competing Values Framework, CALNOC Collaborative Alliance for Nursing Outcomes, HAPU Hospital-Acquired Pressure Ulcers, IRR Incidence Rate Ratio, NSQIP National Surgical Quality Improvement Program, SSI Surgical Site Infection, ASAAmerican Society of Anesthesiologists, ICU Intensive Care Unit, OR Odds Ratio, PICCM Project IMPACT Critical Care Medicine, LOS Length of Stay, HCUP Healthcare Cost and Utilization Project, PSI Patient Safety Indicators, GTT Global Trigger Tool, ACS American College of Surgeons, DSM Death or Serious Morbidity, CVA Cerebrovascular Accident, NICU Neonatal Intensive Care Units, CPQCC California Perinatal Quality Care Collaborative, HAI Healthcare Associated Infections, CECARS Community Living Center Employee Survey of Attitudes about Resident Safety, ORCA Organizational Readiness to Change Assessment, VHA Veterans Health Administration, CLC Community Living Centers, PSCHO Patient Safety Climate in Healthcare Organizations, MEDPAR Medicare Provider Analysis and Review File, VSCQ Vienna Safety Climate Questionnaire, OLS Ordinary Least Squares