Effect of Implementation of a Moral Case Delibration Meetings on the Prevalence of Burn-out Syndrome in the ICU

Studies have shown that healthcare professionals working at the ICU are exposed to greater risk of developing burnout than colleagues working in other fields of medicine (Moss, Good, Gozal, Kleinpell, \& Sessler, 2016). It is widely suggested that moral distress, resulting from morally problematic situations particular for the ICU, is an important risk factor for burnout (Fumis, Junqueira Amarante, de Fatima Nascimento, \& Vieira Junior, 2017). One suggested way to help caregivers mitigate moral distress is moral case deliberation. Moral case deliberation is a structured dialogue between healthcare providers about ethically problematic situations in their daily practice. It has been widely held that frequent moral case deliberation may help healthcare providers to better appreciate moral problems, to learn about their own and others' moral viewpoints, to foster a sense cohesion within their team and to mitigate moral distress (Haan, van Gurp, Naber, \& Groenewoud, 2018). Moral case deliberation may particularly help professionals if it is based on professionals' own objectives in and experiences and expectations of moral case deliberation. It is therefore needed to firmly root moral case deliberation in the ICU practice together with/with the help of ICU-professionals (Weidema, van Dartel, \& Molewijk, 2016). Healthcare professionals themselves are best able to evaluate, learn from and adjust the practice of moral case deliberation on the ICU.

This research sets out to assess the claim that moral case deliberation can help mitigate moral distress among healthcare providers and as such reduce the risk of burnout. It does so by employing Mixed Methods Action Research (MMAR), through which the research will be conducted through and with ICU-professionals, creating co-ownership of moral case deliberation among participants. The study hopes to achieve that, instead of developing burnout as a consequence of moral distress, moral case deliberation will help IC professionals learn from morally distressing situations, potentially leading to cultural an organizational improvement and improvement of quality of care.

Research questions to be answered:

1. How does moral distress, as a supposed risk factor for burnout of IC-professionals, relate to other risk factors of burnout, personality, negative work-home interactions and the ICU context?
2. How do IC-professionals experience preparation, participation and the impact on daily practice of particular moral case deliberations?
3. Does moral case deliberation lead to a reduction of moral distress, burnout and team cohesion?
4. How can IC-professionals establish cumulative learning from moral case deliberation throughout their department?

Methods:

Quantitative methods Quantitative methods will be used to assess questions 1 and 3. Survey data will be gathered in a stepped wedge trial. Surveys will be sent to around 450 ICU-professionals which are nested in six ICUs, nested in two hospitals. The data thus has a multilevel structure (individuals in units in hospitals). The study aims for a response rate of 50% or higher, including more than 230 participants. This is expectation is formulated on the basis of the departments' experience of response rates in previous studies done in this population. During the trial, the different ICUs will receive the intervention in a stepwise manner. Within a stepped wedge design, what is randomized is the time point at which a unit receives the intervention. One unit starts in January 2020 and will receive the intervention for the full two years. Two units will start participating as of July 2020 and will then participate until December 2021. In January 2021, two other units will start participating. The last unit will receive the intervention as of July 2021 and participate for the remaining six months.

Survey data will be gathered at five measurement points, after every 6 months, the first being a baseline measurement. The baseline measurement will take place before January 2020, the second measurement will take place before July 2020, the third measurement will take in the months preceding January 2021, the fourth measurement will take place in the months preceding July 2021 and the last measurement will take place in December 2021. The survey consists of validated measurements for burnout syndrome (UBOS-C), moral distress (MDS-R), work-home interaction (SWING), personality (BFI), departmental culture (Culture of Care Barometer) and team cohesion (a set of six questions taken from the Safety Attitudes Questionnaire).

The baseline data will inspire answering research question 1. Hierarchical linear regression analysis will be used to assess the association between several concepts (e.g. moral distress, work-home interaction, personality). With hierarchical regression analysis, the explaining variance of moral distress on burn-out, controlled for burnout risk factors such as work-home interaction and personality, will be analysed. Hierarchical linear regression is a framework for model comparison: several regression models are built by adding variables to previous model at each new step and thus, later models always include the previous smaller models. The question is whether the newly added variable shows a significant improvement in the proportion of explained variance of burnout (the R2).

The survey data gathered at all five measurement points is suited for analysis which allows for assessment of question 3. The data gathered in the stepped wedge design, in which clusters (units) are randomized, is well suited for interrupted time series analysis. Interrupted time series analysis is used to evaluate the effect of an intervention in longitudinal data. The key assumption of interrupted time series analysis of this data is that moral case deliberation, as an intervention, changes the pre-intervention trend of burnout prevalence among ICU-professionals; in other words, it assumes that the pre-intervention trend would continue unchanged into the post-intervention period if there had not been an intervention. Training, as well as the advice of an expert statistician (Steven Teerenstra), will be sought to gain better insight in conducting interrupted time series analysis.