Development and validation of a patient safety culture questionnaire in acute geriatric units.

BACKGROUND

Older patients (≥65 years) are exposed to more harm resulting from adverse events in hospitals than younger patients. Theoretical considerations and empirical findings suggest that safety culture is the key to improving the quality of health care.

OBJECTIVE

To describe the development of a German-language instrument for assessing patient safety culture (PSC) and its reliability and validity; to verify criterion validity by means of a cross-sectional analysis of the impact of PSC on clinical quality that compares acute geriatric units with a sample from intensive care, surgery and trauma surgery departments, and to report variations in the PSC profile between these groups.

METHODOLOGY

Using a review of existing safety culture surveys, multidimensional scaling procedures and expert interviews, we tested the content and convergent validity of a 158-item questionnaire completed by 508 physicians and nurses from 31 acute geriatric units and 7 comparison departments. Criterion validity was verified by various regression models with a self-reported measure of adverse events. Differences in PSC profiles were analyzed using a one-factorial ANOVA and regression models.

RESULTS

We identified 7 constructs of PSC and demonstrated substantial convergent and criterion validity. In the acute geriatric units, higher levels of 'management commitment to patient safety' and lower levels of 'error fatalism' were associated with a reduced incidence of medical errors. In the comparison group, only the variable 'active learning from mistakes' was relevant for safety performance. Our results also indicate that acute geriatric units display higher standards than the comparison group in all the aspects of patient safety examined.

CONCLUSION

It is possible to measure salient features of PSC using a valid and reliable survey. Some aspects of PSC are more closely related to safety events than others. In acute geriatric units, patient safety appears to be influenced mainly by management's determination of how things are done whereas improvement of the system itself in a more incremental manner is required in the other high-risk ward types.