Abstract
Judith Johnson et al.'s paper sets out to investigate whether workplace bullying and discrimination have a negative effect on healthcare safety through affected employees’ experience of burnout using structural equation modelling. Statistical associations have been previously reported between burnout and patient safety, bullying and discrimination with burnout and bullying with patient safety. Most of these past studies had been conducted outside the United Kingdom. The authors assert that there is a lack of research into possible associations between discrimination and patient safety.
As well as gaps in the evidence for correlational links among the constructs of interest, this paper makes an important point about hypothesis generation based on such associations and the need to design interventions to test causal paths between such constructs. They note, for example, that although the literature had reported direct statistical associations between burnout and safety, interventions designed to address burnout did not produce strong results; similarly, it remains unclear whether addressing discrimination could improve patient safety.
The current study showed there is a likely impact path from bullying and discrimination to patient safety (professional reported) that operates via perceptions of burnout. Also interesting was that the relationship between burnout and safety was the strongest for both patient- and ward-level perceptions of safety. The authors conclude that reducing bullying and discrimination at an organisational level may be one way to reduce burnout and could be useful targets for patient safety initiatives to address.
The extent to which structural equation modelling with non-experimental data can actually demonstrate causality has been debated in the literature (Tarka, 2018), therefore the associations found in this study should be treated as such and the possibility of a common underlying factor influencing both cannot be excluded. As it has been discussed in the context of ethical leadership (CIPD, 2019), unethical behaviours such as bullying and discrimination can be due to systemic problems, in particular organisational culture or ingrained norms of behaviour or the result of business models, such as ones in which there is overwhelming pressure to compromise ethical principles to meet business needs. If the latter is the case in this study, burnout, safety compromises and bullying may have a common cause, for example pressure to cut corners. Pointing to the small size of the associations between bullying and burnout and discrimination and burnout, the authors suggest further investigation into whether there are factors that moderate the strength of these relationships. Recent studies have identified moderators such as job autonomy and occupational self-efficacy resources (Livne and Goussinsky, 2018) and social support (Rossiter and Sochos, 2017).
Structural equation modelling generally assumes latent constructs and attempts to unpick relationships between them and their manifestations in the form of measured variables. Such an approach usually requires more measures per construct, whereas in this study single items were used, which poses reliability risks. Other quality risks include common method variance with all measures being taken simultaneously by the same raters and validity, as all measures are subjective perceptions. Of note, however, is that bullying, discrimination and burnout are naturally subjective experiences, with the first two even recognised as such in law, so perception measures should be appropriate.
Overall, this well-conducted study added to our understanding that patient safety is not only a technical or skills matter but is linked to organisational behaviours, employee relations and wellbeing (Martin and Manley, 2017).