Self-report instruments have been widely used to better understand variations in patient safety climate between physicians and nurses. Research is needed to determine whether differences in patient safety climate reflect true differences in the underlying concepts. This is known as measurement equivalence, which is a prerequisite for meaningful group comparisons. This study aims to examine the degree of measurement equivalence of the responses to a patient safety climate survey of Chinese hospitals and to demonstrate how the measurement equivalence method can be applied to self-report climate surveys for patient safety research.
Using data from the Chinese Hospital Survey of Patient Safety Climate from six Chinese hospitals in 2011, we constructed two groups: physicians and nurses (346 per group). We used multiple-group confirmatory factor analyses to examine progressively more stringent restrictions for measurement equivalence.
We identified weak factorial equivalence across the two groups. Strong factorial equivalence was found for Organizational Learning, Unit Management Support for Safety, Adequacy of Safety Arrangements, Institutional Commitment to Safety, Error Reporting and Teamwork. Strong factorial equivalence, however, was not found for Safety System, Communication and Peer Support and Staffing. Nevertheless, further analyses suggested that nonequivalence did not meaningfully affect the conclusions regarding physician–nurse differences in patient safety climate.
Our results provide evidence of at least partial equivalence of the survey responses between nurses and physicians, supporting mean comparisons of its constructs between the two groups. The measurement equivalence approach is essential to ensure that conclusions about group differences are valid.