Background The numbers of medicines being prescribed has increased in recent years and this presents a challenge to clinicians in balancing the use of indicated medicines against potential overtreatment and its associated risks. This study aims to examine: (i) changes in rates of polypharmacy from 1997 to 2012 and; (ii) changes in potentially inappropriate prescribing (PIP) prevalence and the relationship between PIP and polypharmacy in individuals aged ≥65 years over this 15 year period in Ireland.

Methods This repeated cross-sectional study using pharmacy claims data included patients eligible for the means-tested HSE-PCRS General Medical Services scheme in the Eastern Health Board region of Ireland in 1997, 2002, 2007, and 2012 (range 338,025–539,752 individuals). Rates of polypharmacy (prescribed ≥5 regular medicines) and excessive polypharmacy (≥10 medicines) in all individuals and PIP prevalence in the elderly (as determined by 30 Screening Tool for Older Persons’ Prescriptions criteria) were determined for each year. The association of study year with polypharmacy and PIP was analysed using negative binomial regression and logistic regression respectively.

Results Between 1997 and 2012 there was a substantial increase in the prescribing of regular medicines, particularly in older adults, with a four-fold increase in polypharmacy (adjusted incident rate ratio (IRR) 4.16) and a ten-fold increase in excessive polypharmacy (adjusted IRR 10.53), independent of age and gender. In those aged ≥65 years, prevalence of PIP rose from 32.6% in 1997 to 37.3% in 2012. However after adjusting for polypharmacy and gender changes, the odds of having any PIP were lower in 2012 compared to 1997, adjusted odds ratio 0.39 (95% CI 0.39–0.4). Although most PIP criteria decreased in prevalence with time, prescription of long-term proton pump inhibitors (PPIs) at maximal dose increased at each time point (from 0.8% to 23.8% over the period). Males, patients with polypharmacy or prescribed aspirin, warfarin, or long-term steroids were more likely to be prescribed a long-term PPI at maximal dose, while older patients and those prescribed an NSAID were less likely.

Conclusion Accounting for the marked increase in polypharmacy, prescribing quality appeared to improve with a reduction in the odds of having PIP from 1997 to 2012. The exception to this is long-term use of maximal dose PPIs, and such prescribing was not consistently associated with risk factors of gastrointestinal bleeding (e.g. NSAID use, older age). Although prescribing of most PIP medicines has improved, strategies to address the related challenges of polypharmacy and PIP may further improve quality of care.