Advance care planning with people with dementia: a process evaluation of an educational intervention for general practitioners.

Background: General practitioners (GPs) are advised to offer advance care planning (ACP) to people with dementia (PWD). In a randomized controlled trial, an educational intervention for GPs aimed at initiating and optimizing ACP proved to be effective. During the intervention most GPs were accompanied by their practice nurse (PN). To provide insights into the intervention's successful components and what could be improved, we conducted a process evaluation and explored implementation, mechanisms of impact and contextual factors.

Methods: We used the Medical Research Council guidance for process evaluations. Implementation was explored identifying reach and acceptability. We performed descriptive analyses of participants' characteristics; selection, inclusion and intervention attendance; a GP post-intervention survey on initiating ACP; a post intervention focus group with trainers of the intervention. Mechanisms of impact were explored identifying adoption and appropriateness. We used: participants' intervention ratings; a GP post-intervention survey on conducting ACP; ACP documentation in PWD's medical files; post-intervention interviews with PWD/FC dyads. All data was used to identify contextual factors.

Results: The intervention was implemented by a small percentage of the total Dutch GP population invited, who mostly included motivated PWD/FC dyads with relatively little burden, and PWD with limited cognitive decline. The mechanisms of impact for GPs were: interactively learning to initiate ACP with training actors with a heterogeneous group of GPs and PNs. For PWD/FCs dyads, discussing non-medical preferences was most essential regarding their SDM experience and QoL. Some dyads however found ACP stressful and not feasible. Younger female GPs more often initiated ACP. Male PWD and those with mild dementia more often had had ACP. These characteristics and the safe and intimate training setting, were important contextual facilitators.

Conclusion: We recommend Interventions aimed at improving ACP initiation with PWD by GPs to include interactive components and discussion of non-medical preferences. A safe environment and a heterogeneous group of participants facilitates such interventions. However, in practice not all FC/PWD dyads will be ready to start. Therefore, it is necessary to check their willingness when ACP is offered.