Drug-drug interactions and the risk of adverse drug reaction-related hospital admissions in the older population.

Aims: The aims of this study were to estimate potentially clinically important drug-drug interaction (DDI) prevalence, and the average causal effect of DDI exposure on adverse drug reaction (ADR)-related hospital admission, and to examine differences in health-related quality of life (HRQoL) and length of stay (LOS) per DDI exposure in an older (≥65 years) population acutely hospitalized.

Methods: This was a cross-sectional study conducted among 798 older individuals acutely admitted to hospital in Ireland between 2016 and 2017. Medication (current/recently discontinued/over-the-counter) and clinical data (e.g., creatinine clearance) were available. DDIs were identified using the British National Formulary (BNF) and Stockley's Drug Interactions. Causal inference models for DDI exposure on ADR-related hospital admission were developed using directed acyclic graphs. Multivariable logistic regression was used to estimate the average causal effect. Differences in HRQoL (EQ-5D) and LOS per DDI exposure were examined non-parametrically. DDI prevalence, adjusted odds ratios (aOR), and 95% confidence intervals (CIs) are reported.

Results: A total of 782 (98.0%) individuals using two or more drugs were included. Mean age was 80.9 (SD ± 7.5) years (range: 66-105); 52.2% were female; and 45.1% (n = 353) had an ADR-related admission. At admission, 316 (40.4% [95% CI: 37.0-43.9]) patients had at least one DDI. The average causal effect of DDI exposure on ADR-related hospital admission was aOR = 1.21 [95% CI: 0.89-1.64]. This was significantly increased by exposure to: DDIs which increase bleeding risk (aOR = 2.00 [1.26-3.12]); aspirin-warfarin (aOR = 2.78 [1.37-5.65]); and esomeprazole-escitalopram (aOR = 3.22 [1.13-10.25]. DDI-exposed patients had lower HRQoL (mean EQ-5D = 0.49 [±0.39]) compared those non-DDI-exposed (mean EQ-5D = 0.57 [±0.41]), (P = .03); and greater median LOS in hospital (8 [IQR5-16]days) compared those non-DDI-exposed (7 [IQR 4-14] days),(P = .04).

Conclusions: Potentially clinically important DDIs carry an increased average causal effect on ADR-related admission, significantly (two-fold) by exposure to DDIs that increase bleeding risk, which should be targeted for medicine optimization.