401 General Discussion patient-physician relationships are one of the facilitators for successful shared decision-making [50]. Thus, insufficient, long-term trustful relationshipsmay negatively influence the implementation of medication optimisation recommendations, which may also explain lower implementation rates usually found in multicentre clinical trials compared to single-centre trials. While the OPERAM trial was ongoing, the results of a European large randomised controlled study were published investigating the effectiveness of a CDSS-assisted deprescribing tool used in older adults (≥75 years) with polypharmacy (≥8 chronic drugs) in a primary care setting versus usual care. After 24 months, no betweengroup difference was found according to the intention-to-treat analysis for the composite endpoint of unplanned hospital admissions or death. However, the perprotocol analysis restricted to patients who fully followed the trial protocol favoured the intervention (odds ratio 0.82, 95% CI 0.68–0.98), possibly indicating that the non-implementation of the intervention affected trial results significantly [52]. This tool resulted in a reduction of 0.42 drugs per patient. In line with the results from OPERAM, it appears to be safe to withdraw drugs in older people, which is a positive result in itself by reducing the treatment burden and prescribing costs. Furthermore, unintentional non-implementation can occur at any stage during a follow-up period, and many possible factors can impede the actual and persistent implementation of agreed recommendations. For example, unintentional medication discrepancies at the transition of care are common, posing an increased risk of avoidable patient harm [53–55]. Canadian research found that almost half (44%) of all included patients (n = 2,655; mean age 69.5 years [SD = 14.7]) were not adherent to some or all changes made to their medications after hospital discharge. Patients who were not adherent to any medication changes had a significantly higher risk (35%) of adverse events compared than thosewhowere adherent to all medication changes [56]. Unfortunately, no data was available in the OPERAM study to distinguish between the intentional and unintentional non-implementation of recommendations. Thus, concise documentation and the transfer of reasons for medication changes will likely prevent unintentional (dis)continuation of medication across care settings and will be an informative source for future research. 3. Outcome measures in medication review The previous section discussed the causes of the high variability in the medication review process impacting potential outcomes. The next paragraphs reflect on 1) the risk of drug-related hospital readmissions (the primary outcome of the OPERAM trial) and 2) how to measure the impact of medication review. 5
RkJQdWJsaXNoZXIy MTk4NDMw