411 General Discussion [16,52,65]. In addition, the low implementation of the investigated, complex interventions impeded drawing firm conclusions about their potential effectiveness. Therefore, investigating the effectiveness of domains that more directly relate to medication review interventions may be more suitable as outcomes in clinical trials, along with efforts to ensure actual implementation of interventions. Examples of domains from the aforementioned core outcome sets that may be more directly related to the medication review process are measurements of medication-related outcomes, patient-related outcomes and patient knowledge about treatment. Previous research has observed that a medication review focused on patient’s personal preferences and goals in older patients (> 70 years) with polypharmacy (≥ 7 chronic drugs) in a primary care setting improved older persons’ self-reported quality of life measured with EQ-VAS (but not with EQ-5D) and decreased the number of health problems impacting daily life [66]. Such ‘step-by-step’ results are highly relevant to gaining knowledge for the future development of the medication review process. The impact of a deprescribing intervention on older patients’ priorities and goals is currently being investigated in a Canadian RCT conducted in a primary care setting [67]. If such interventions demonstrate effectiveness on several medication-related and patient-related outcomes and implementation is feasible and scalable, we can further explore their potential benefits in reducing ‘hard’ clinical (adverse) outcomes. 4. Medication optimisation in older people – who, when, where and how? Given the vulnerable sustainability of our current healthcare system by increasing healthcare expenditures, making choices in healthcare (and research) is unavoidable [68]. Delivering the right care to the right patient at the right time is more important than ever. In the next section, we will reflect on 1) selecting patients with the highest risk of drug-related harm (the who) and 2) when, where and how to perform medication optimisation interventions. 4.1. Selecting patients with the highest risk of drug-related harm (the who) Risk prediction models and screening tools based on high-risk medications could support the triage of patients with the highest risk of drug-related harm and, therefore, may benefit most from a structured medication review (Figure 3). The applicability of the STOPP/START criteria and the ADR trigger tool have predictive 5
RkJQdWJsaXNoZXIy MTk4NDMw