Potential prescribing issues among older HIV‐infected subjects in a Mediterranean cohort: Does the current prevalence give cause for concern?

Aims To determine the prevalence of potential prescribing issues (PPI) in HIV‐infected subjects aged ≥65 years according to the Beers and STOPP/START criteria and antiretroviral drug–drug interactions (Liverpool website). Secondary objectives were to assess the concordance between Beers and STOPP/START criteria in our population, and to identify the drugs most frequently involved in PPI. Methods Cross‐sectional cohort study based on a systematic review of the electronic drug prescriptions confirmed by an interview of 91 HIV‐infected patients aged ≥65 years. Discrepancies between prescription criteria were assessed using crosstabs and compared using the χ2 test or Fisher exact test. Results The mean age was 72.1 (5.6) years, 75.8% had ≥3 comorbidities and 59.3% polypharmacy. PPI were identified in 87.9%: 71.4% by STOPP/START and 45.1% by Beers. Comparing both criteria, 56.9% of PPI by STOPP/START were detected by Beers, while 92.5% of those detected by the Beers criteria were detected by STOPP/START (P < .001). Amber/red flag interactions between antiretrovirals and comedications were found in 45.1%: 3 severe (red) in 2 patients (2.2%). The most frequent drugs involved in PPI were benzodiazepines (>30%). Cobicistat was the drug most frequently involved in potential interactions (42.2%). Conclusion The prevalence of PPI among older HIV‐infected persons gives cause for concern, as it is almost 90%. Optimization strategies, including a critical review of the treatment plan, should be implemented in clinical routine by a multidisciplinary team, in particular in patients with multiple comorbidities and polypharmacy. The STOPP/START criteria seem to detect more PPI, mainly for European populations.