Evaluating the impact of medication review and deprescribing on prescribing appropriateness and clinical outcomes in older people residing in long-term care facilities: a systematic review and meta-analysis.

BACKGROUND

Polypharmacy is a major concern among older adults in long-term care facilities (LTCFs), as it increases the risk of potentially inappropriate medications (PIMs) and related adverse outcomes. Medication review and deprescribing interventions may help optimise therapy and reduce harm.

DESIGN

Systematic review and meta-analysis.

METHODS

This study was conducted according to PRISMA guidelines (PROSPERO: CRD42023486056). PubMed, Embase and Scopus were searched up to 27 August 2024, for experimental studies evaluating the impact of medication review/deprescribing interventions in older LTCF residents with polypharmacy. Outcomes included medication appropriateness indexes, falls, hospitalisations and mortality. We calculated risk ratios for dichotomous data and mean differences for continuous data [with 95% confidence intervals (CIs)]. The quality of the studies was assessed using RoB 2 for the randomised controlled trials (RCTs) and the ROBINS-I for non-randomised studies.

RESULTS

From 3548 records, 38 studies (22 RCTs, 16 quasi-experimental) were included. Pooled analyses demonstrated significant reductions in the number of drugs per patient [within 12 months: -0.89 (95% CI -1.46, -0.32); at ≥12 months: -1.60 (95% CI -2.68, -0.52)] and in PIMs [at 6 months: -0.48 (95% CI -0.74, -0.22); at ≥12 months: -0.26 (95% CI -0.40, -0.13)]. No significant effects were observed on falls, hospitalisations or mortality. Studies showed wide methodological heterogeneity and had moderate to high risk of bias (23 moderate, 14 high, 1 low).

CONCLUSIONS

Comprehensive medication review interventions improved prescribing appropriateness in older LTCF residents with polypharmacy but did not significantly affect clinical outcomes (i.e. falls, hospitalisations and mortality). Further high-quality studies using standardised approaches are needed.