Patient characteristics associated with adverse drug reactions in older community-dwelling adults attending general practice: a prospective cohort study

Problem

Adverse drug reactions (ADRs) are associated with adverse outcomes for patients including emergency admissions. Older people are vulnerable to ADRs due to ageing-related physiological changes and increasing multimorbidity and polypharmacy. While some ADRs can be prevented, identifying those most at risk is challenging and there is limited primary care research to support this. This study aimed to examine patient characteristics associated with ADRs in a prospective cohort of older community-dwelling adults.

Approach

Study design and participants: Prospective cohort study (2010-2016), 904 community-dwelling adults aged ≥70 years taking ≥1 regular medication recruited from 15 general practices. Outcome measures: review of GP medical record, detailed patient questionnaire, and linked dispensed medications from a national pharmacy claims database. Baseline patient characteristics: age, sex, deprivation, health insurance, marital status, number of drug classes, medication adherence (Medication Possession Ratio) and multimorbidity (Charlson Comorbidity Index). Primary outcome: ADRs, measured by GP medical record review for the six-year follow-up period. ADR severity: mild-no treatment required, moderate-required treatment or resulted in non-permanent disability, severe-resulting in permanent disability/death. Statistical analysis: Descriptive statistics and logistic regression (Odds Ratios [OR], 95% Confidence Intervals [CIs]) are presented.

Findings

Of 904 study participants, median age was 78 years (Interquartile Range [IQR] 73, 81), 489 (54%) were female with a median of 6 regular drugs (IQR 3, 8). A total of 205 (23%) participants experienced ≥1 ADR with 61 (7%) experiencing ≥2 ADRs. Of 266 ADRs identified, 236 (88%) were mild, 28 (11%) moderate and 2 were severe. A total of 26 ADRs (10%) resulted in an emergency admission. Cardiovascular system drugs resulted in 30% of all ADRs. In multivariable analysis, female sex (OR 1.7, 95% CI 1.2-2.5 p 0.003) and being in receipt of 5-10 drug classes (compared to <5) (OR 1.8, 95% CI 1.2-7.4 p 0.003) and >10 drug classes (OR 2.0, 95% CI 1.1-3.7 p 0.017) were associated with ADRs.

Consequences

ADRs are common in older people attending general practice with one in four study participants experiencing at least one ADR. One in 10 identified ADRs resulted in an emergency admission. Cardiovascular drugs were most commonly implicated, while females and those with polypharmacy were more likely to experience an ADR. This study highlights the importance of proactive medication review for older people taking multiple medications to reduce their risk of ADRs. Future research could examine risk stratification to identify those older people at highest risk of experiencing an ADR and consider strategies to reduce the overall drug burden including dose reduction and deprescribing.