Limitations and risk of bias

The first and second author of this protocol are involved in the trial both as researchers and as clinical pharmacists conducting the intervention. As previously mentioned, we will undertake several measures to reduce the risk of bias that could emerge from these multiple roles. For example, we will consult external staff for the primary outcome assessment and statistical analyses. Other clinical outcomes in terms of all-cause readmissions, emergency department visits and mortality are electronically documented data with no room for subjectivity. Therefore, this risk of bias mainly applies to MARS-5 and EQ-5D, which should be considered in the interpretation of our findings. Moreover, all interviews within the qualitative part of the study will be conducted by another person than the clinical pharmacist who performed the interventions on those specific participants. Due to the nature of the intervention, neither participants nor clinical pharmacists are blinded. We do not consider the lack of blinding to be a major issue for the study, but bias related to special attention and the knowledge of being observed could possibly affect self-reported adherence and quality of life among the participants.