Table 1

 Studies measuring safety climate in health care

AuthorsInstrument details and survey sampleSafety climate factorsPsychometric analysesOutcome measuresResults
PPE, personal protective equipment (e.g. gloves, masks); UP, universal precautions; FA, factor analysis; ORMAQ, Operating Room Management Attitudes Questionnaire.
(1) DeJoy et al26See also DeJoy et al,37,38 Gershon et al,39 Guastello et al,40 McGovern et al41 for analyses based on parts of same data setSafety climate scale 35 items (part of a longer questionnaire)902 nurses, 322 physicians and 247 technicians (57% response rate) from 3 USA hospitalsSafety performance feedback Management commitment to safety Provision of PPE UP-related job hindrances Knowledge and information about UP Risk of infection Self-protective actions Work organisationExploratory factor analysis yielded 8 factors based on 23 items (some based on 3 items or less)α for the identified factors ranged from 0.39 (general work organisation) to 0.83 (management commitment)Regression analysesSelf-report scale on compliance; adherence to UPJob hindrance was strongest predictor of compliance in nurses and physicians. Safety performance feedback was a strong predictor of compliance in nurses but more modest predictor for physicians
(2) Gershon et al2646 item safety climate scale (part of longer questionnaire)1240 employees from a large urban USA medical research centre (60% response rate)Only employees with the highest risk for blood and body fluid exposure were selected for participationSenior management support Absence of job hindrances Cleanliness, orderliness Minimal conflict and good communication Safety-related feedback/training by supervisors PPE/engineering control equipment availabilityExploratory factor analysis yielded 6 factors based on 20 items (only 2 based on more than 3 items)α > 0.7 for all 6 factors Regression analysesSelf-report scale on compliance with universal safety precautions Exposure incident history“Cleanliness and orderliness”, “senior management support” and “absence of job hindrances” associated with compliance with safety practices Higher “senior management support” and “feedback/training” related to lower exposure to incidents
(3) Neal et al2716 items about safety climate (part of longer questionnaire)525 employees from an Australian hospital (56% response)Safety climate scale included items about:Management values Communication Training Safety systemsSafety climate defined by mean score from 16 items No factor analysis α = 0.93 for 16 item safety climate scale Structural equation modellingSelf-report of safety practices and procedural complianceSafety climate indirectly related to safety compliance
(4) Felknor et al28Safety climate 11 items based on Gershon et al.26 (part of longer quesionnaire)878 employees from 10 Costa Rican hospitals (96% response rate)Management commitment to safety Work area Unsafe work practices Reporting safety violationsNo FA57Regression analysisWork injuries Self-report compliance with safety practicesSafety climate inverse relationship with workplace injuries Positive relationship between safety climate and safety practices
(5) McCoy et al2921 item safety climate scale based on studies by Murphy et al37149 infection control practitioners from 149 USA hospitals (62% response rate)Management commitment Feedback Job demands Safety committee PPE availabilityExploratory FAα = 0.62–0.93 Logistic regression analysisPerceptions of adequacy of healthcare worker training to monitor co-workers’ adherence to standard precautions“Management commitment” and “feedback” positively related to training to observe co-workers’ standard precautions compliance“Job demands” inversely related to training to observe co-workers’ standard precautions compliance
(6) Vredenburgh3018 item scale based on Ostrom et al5762 risk managers from 62 USA hospitals (57% response rate)Rewards Training Management commitment Communication and feedback Selection ParticipationExploratory factor analysis; 6 factor solution did not correspond to the hypothesised dimensions Multiple regressionOccupational injuriesFactor 1 (reactive measures) and factor 2 (proactive measures) predicted injury rates
(7) Carrico3179 item questionnaire based on Offshore Safety Questionnaire6393 nurses in Delaware, USA (31% response)Communication Satisfaction safety Involvement Work pressure Safety attitudes Safety behavioursInternal reliability analysis of proposed safety climate dimensionsNoneLow mean scores indicated a somewhat poor safety climate for nurses
(8) Singer et al32Stanford/PSCI Culture Survey (82 items)6312 employees including attending physicians, senior executives and a 10% random sample of other hospital personnel at 15 USA hospitals (47% response rate)Organisation Department Production Reporting/seeking help Shame/self-awarenessExploratory FA yielded 5 factors based on 30 items (shame/self-awareness factor only based on 3 items)α not given for identified factorsNoneProblematic and neutral responses suggested “a lack of safety culture” in some hospitals
(9) Sorra and Nieva33Hospital Survey on Patient Safety (79 items)1437 staff at 21 USA hospitals (29% response rate)Supervisor/manager expectations and actions promoting patient safety Organisational learning Teamwork within units Communication openness Feedback/communication about error Response to error Staffing Hospital management support for patient safety Teamwork across hospital units Hospital handoffs and transitionsExploratory FA yielded 14 factors based on 66 items. Confirmatory FA yielded a 12 factor solution (2 factors that measured outcomes and 10 factors that measured safety climate) based on 42 items (6 factors only based on 3 items)α>0.7 for all factors except staffing (α = 0.63)CorrelationsSelf-report of:Number of events reported Overall patient safety grade Overall perceptions of safety Frequency of event reporting“Overall perceptions of safety” were correlated with “patient safety grade” and “hospital management support for patient safety”“Frequency of event reporting” was correlated with “feedback and communication about error”
ORMAQ studies47,62
(10) Itoh et al34Adapted Operating Team Resource Management Survey which included 57 items.66 doctors, 486 nurses and 43 pharmacists from 5 Japanese hospitals (91% response rate)Satisfaction with management Morale and motivation Communication Teamwork Power distance Own competence Recognition of stress Stress management ErrorNo factor structure emerged from FA (personal communication)α not given for proposed factors CorrelationRates of incident reporting for nurses in one hospital“Recognition of human error” and “power distance” were negatively correlated with rates of incident reporting
(11) Pronovost et al35Total length of questionnaire not given but it included 10 item safety climate scale 395 staff at a large USA teaching hospital (64% response rate)Supervisor and management commitment to safety Knowledge of how to report adverse events Understanding of systems as the cause of adverse eventsNo FAα not given for proposed factorsNoneParticipants perceived supervisors to have a greater commitment to safety than senior leaders
(12) Woods et al36Total length of questionnaire not given but it was adapted from the 60 item version of the ORMAQ 802 healthcare workers from an Australian Health Service Area (26% response rate)Organisational culture Communication Teamwork Assertiveness Performance shaping factors ErrorNo FAα not given for proposed factorsNoneNone related to outcomes