Michael Schiffinger1, Markus Latzke, Johannes Steyrer. 1. Michael Schiffinger, PD, is Senior Scientist, Interdisciplinary Institute for Management and Organisational Behaviour and Center for Empirical Research Methods, Vienna University of Economics and Business, Austria. E-mail: michael.schiffinger@wu.ac.at. Markus Latzke, Dr, is Assistant Professor, Interdisciplinary Institute for Management and Organisational Behaviour, Vienna University of Economics and Business, Austria. Johannes Steyrer, PhD, is Tenured Professor, Interdisciplinary Institute for Management and Organisational Behaviour, Vienna University of Economics and Business, Austria.
Abstract
BACKGROUND: Safety climate (SC) and more recently patient engagement (PE) have been identified as potential determinants of patient safety, but conceptual and empirical studies combining both are lacking. PURPOSES: On the basis of extant theories and concepts in safety research, this study investigates the effect of PE in conjunction with SC on perceived error occurrence (pEO) in hospitals, controlling for various staff-, patient-, and hospital-related variables as well as the amount of stress and (lack of) organizational support experienced by staff. Besides the main effects of PE and SC on error occurrence, their interaction is examined, too. METHODOLOGY/APPROACH: In 66 hospital units, 4,345 patients assessed the degree of PE, and 811 staff assessed SC and pEO. PE was measured with a new instrument, capturing its core elements according to a recent literature review: Information Provision (both active and passive) and Activation and Collaboration. SC and pEO were measured with validated German-language questionnaires. Besides standard regression and correlational analyses, partial least squares analysis was employed to model the main and interaction effects of PE and SC on pEO, also controlling for stress and (lack of) support perceived by staff, various staff and patient attributes, and potential single-source bias. FINDINGS: Both PE and SC are associated with lower pEO, to a similar extent. The joint effect of these predictors suggests a substitution rather than mutually reinforcing interaction. Accounting for control variables and/or potential single-source bias slightly attenuates some effects without altering the results. PRACTICE IMPLICATIONS: Ignoring PE potentially amounts to forgoing a potential source of additional safety. On the other hand, despite the abovementioned substitution effect and conjectures of SC being inert, PE should not be considered as a replacement for SC.
BACKGROUND: Safety climate (SC) and more recently patient engagement (PE) have been identified as potential determinants of patient safety, but conceptual and empirical studies combining both are lacking. PURPOSES: On the basis of extant theories and concepts in safety research, this study investigates the effect of PE in conjunction with SC on perceived error occurrence (pEO) in hospitals, controlling for various staff-, patient-, and hospital-related variables as well as the amount of stress and (lack of) organizational support experienced by staff. Besides the main effects of PE and SC on error occurrence, their interaction is examined, too. METHODOLOGY/APPROACH: In 66 hospital units, 4,345 patients assessed the degree of PE, and 811 staff assessed SC and pEO. PE was measured with a new instrument, capturing its core elements according to a recent literature review: Information Provision (both active and passive) and Activation and Collaboration. SC and pEO were measured with validated German-language questionnaires. Besides standard regression and correlational analyses, partial least squares analysis was employed to model the main and interaction effects of PE and SC on pEO, also controlling for stress and (lack of) support perceived by staff, various staff and patient attributes, and potential single-source bias. FINDINGS: Both PE and SC are associated with lower pEO, to a similar extent. The joint effect of these predictors suggests a substitution rather than mutually reinforcing interaction. Accounting for control variables and/or potential single-source bias slightly attenuates some effects without altering the results. PRACTICE IMPLICATIONS: Ignoring PE potentially amounts to forgoing a potential source of additional safety. On the other hand, despite the abovementioned substitution effect and conjectures of SC being inert, PE should not be considered as a replacement for SC.