Powell-Cope GM (James A. Haley VAMC)
Luther S (James A. Haley VAMC)
Spehar A (James A. Haley VAMC)
Quigley P (James A. Haley VAMC)
Rosenberg D (James A. Haley VAMC)
Palacios P (James A Haley VAMC)
McKnight S (National Center for Patient Safety)
Nelson A (James A Haley VAMC)
Culture of safety is a critical feature of high reliability organizations. In 2000 the VA National Center for Patient Safety administered a 112-item questionnaire nationwide to measure employees' perceptions of safety culture. While the survey provided important data, employees criticized the length of the questionnaire, and researchers identified redundancies. The objective of our study was to develop a shorter reliable and valid version of the questionnaire.
A cross sectional survey design was used. After reviewing several instruments, the research team modified the NCPS questionnaire and added new items. Dependent variables were overall rating of the facility’s safety culture and an error reporting scale. 3232 surveys were mailed to randomly selected employees in VISN 8. 1750 usable anonymous surveys were returned (54% response rate). Exploratory factor analysis was used to determine factor structure, Cronbach’s alpha and item analysis were used to evaluate internal consistency, and linear regression was conducted to evaluate construct validity.
Forty-one percent of the sample was direct care providers; 11% were administrators. Factor analysis, employing principal factor extraction followed by Promax rotation, was conducted on 66 survey items. A five-factor solution was retained which accounted for approximately 85% of the variance in the initial extraction. Oblique factor solution was interpreted. Factor loadings were use to identify a reduced set of items that reliably measured the factors. The reduced set included 40 items with six and nine items retained for each factor. Alpha’s were .85 or higher for four of the five factors. Regression analysis found that scale scores for all five factors were positively associated with rating of hospital safety, while scores for four of the five factors were associated with error reporting behaviors (p < .05). Factor 1, Organizational Communication, was found to have the strongest association in both models.
We successfully developed a reliable and valid 40-item survey to measure employees’ perceptions of safety culture.
This brief survey can be used by health care organizations to measure employees’ perceptions and to track perceptions over time. The questionnaire can be used to satisfy JCAHO requirements related to measurement of safety culture.