403 General Discussion BOX 3 The risk of a DRA Examples 1 and 2 combined A s a result of the variability in screening and implementation, each signal for potentially inappropriate prescribing can result in at least eight possible outcomes, as shown in Figures 2.1 and 2.2. A medication review might only prevent a DRA if the signal for potentially inappropriate prescribing was accepted after clinical evaluation at the individual patient level, followed by implementing the recommended medication change (1 of 8 possible outcomes). If unindicated ASA was successfully discontinued in the example patient, bleeding might have been prevented. However, the number needed to harm (NNH) of unindicated use of ASA combined with a DOAC is 24 over a mean period of 21 months according to literature [61]. Hence, 24 patients with unindicated ASA should be discontinued to prevent one bleeding. In OPERAM, unindicated ASA was discontinued in only 18 patients. In addition, if ASA was discontinued, but the use of a BZD in the same patient was not (e.g. due to patient disagreement or re-prescription at discharge), the patient was still at risk for a DRA (i.e. a fall related to benzodiazepine use). This risk of falls with BZD use could either result in an actual fall during the follow-up period of a clinical trial or may not (yet) have occurred. In either case, continued inappropriate prescribing would be unfavourable because the BZD would still contribute to an increased risk of falls (OR 1.6) [62]. The two decision trees presented in Figure 2.1 and 2.2 illustrate how the variability in decisions during the medication review process can impact the outcome ‘DRA’, based on the aforementioned examples (Boxes 1–3). 5
RkJQdWJsaXNoZXIy MTk4NDMw