Teamwork and safety attitudes among staff in critical care units and the relationship to patient mortality
© BioMed Central Ltd. 2004
Published: 15 March 2004
In 2002, West and colleagues reported a direct and strong association between the quality of human resource practices and patient mortality in NHS hospitals – higher levels of staff teamworking, training, development and appraisals were associated with lower patient mortality. The objectives of this study were to elicit teamwork and safety attitudes among staff in critical care units and to investigate whether these attitudes were related to case-mix-adjusted outcomes for patients.
The Safety Attitudes Questionnaire (SAQ) is a valid, reliable questionnaire that elicits caregiver attitudes for six factors: teamwork climate, safety climate, job satisfaction, working conditions, stress recognition, and perceptions of management. Staff received a sealed envelope containing a copy of the SAQ, a cover letter, a pencil, and a FREEPOST (direct to ICNARC) return envelope. Absolute confidentiality was maintained on questionnaires with the sole identifiers being Site Code and job category. Permission to link staff attitudes data to patient case mix and outcome data was sought from the Director. The relationship between patients' mortality and staff attitudes was investigated in a hierarchical logistic regression model after adjustment for case mix.
A total of 106 units participated in the survey, approximately 43% nationally. The mean number of staff was 69.8 (range 21–142). The staff response rate was 67.5%, ranging from 30.0% to 95.5% across the units. Of 4859 respondents, positive attitudes were highest for safety climate (40.4%), teamwork climate (35.0%), job satisfaction (31.9%) and working conditions (22.3%), and lower for stress recognition and perceptions of management, 12.1% and 9.7%, respectively. There was significant variation for all six factors across units. Overall 14,064 (19.8%) admissions died in the unit and 21,254 (29.9%) admissions died before ultimate discharge from hospital. Ultimate hospital mortality varied between 16.7% and 54.9% across units. Neither hospital nor unit mortality were associated with the percentage of staff positive within any domains of the SAQ either before or after adjustment for case mix.
Contrary to the results reported by West and colleagues for NHS hospitals and by Wheelan and colleagues for 17 US intensive care units, despite a wide variation in staff attitudes, negative attitudes were not associated with a poorer, risk-adjusted patient outcome in critical care.