Do task and item difficulty affect overestimation of one’s hand hygiene compliance? A cross-sectional survey of physicians and nurses in surgical clinics of six hospitals in Germany

Abstract

Abstract Background One barrier to hand hygiene compliance is overestimation of one’s own performance. Overconfidence research shows that overestimation tends to be higher for difficult tasks, which suggests that the magnitude of overestimation also depends on how it is assessed. Thus, we tested the hypothesis that overestimation was stronger for hand hygiene indications with low compliance (i.e., high difficulty), and the hypothesis that self-reported overall compliance based on a single item is higher than based on “5 Moments of Hand Hygiene” (WHO-5) items, since the single item implies an aggregation across indications. Methods In the WACH trial (German Clinical Trials Register [DRKS] ID: DRKS00015502), a questionnaire survey was conducted among physicians and nurses in nine surgical clinics (general/visceral surgery or orthopedics/trauma surgery) of six German hospitals. Self-reported compliance was assessed both by a single item and the WHO-5-items using percentage scales. These were compared with each other and with direct observations. Relative frequencies of the WHO-5 indications used to calculate the WHO-5-based self-reported overall compliance rate were estimated by a systematized review of the literature (see appendix). In analysis, t-tests, Chi2-tests and multiple linear regressions were used. Results Ninety-three physicians (response rate: 28.4%) and 225 nurses (30.4%) participated. Significant compliance differences between physicians and nurses were found for direct observations and were in favor of nurses, while no such differences were found for self-reports. Across the WHO-5, overestimation showed inverse correlations with observed compliance (physicians: r = −0.88, p = 0.049; nurses: r = −0.81, p = 0.093). Support for the hypothesis that the self-reported overall compliance based on one item is higher than that based on WHO-5 items was found for physicians (M = 87.2 vs. 84.1%, p = 0.041; nurses: 84.4 vs. 85.5%, p = 0.296). Exploratory analyses showed that this effect was confined to orthopedic/trauma surgeons (89.9 vs. 81.7%, p = 0.006). Conclusion Among physicians, results indicate stronger hand hygiene overestimation for low-compliance indications, and when measurements are based on a single item versus the five WHO-5 items. For practice, results contribute to infection prevention and control’s understanding of overestimation as a psychological mechanism that is relevant to professional hand hygiene.