A Cohort Study in the Clinical Practice Research Datalink to explore whether potentially inappropriate prescribing (PIP) in middle-aged adults is associated with increased healthcare utilisation and mortality

Problem

Potentially inappropriate prescribing (PIP) is the prescribing of medication where the risks are likely to outweigh the benefits. It is common in middle-aged adults (45-64 years), with a recent systematic review and meta-analysis showing that 38% of middle-aged adults are exposed to PIP annually. However, all studies included were conducted in Ireland, so PIP prevalence in the UK population is yet to be established. PIP has been shown to be associated with Adverse Drug Events (ADEs) in hospitalised middle-aged adults, but associations with healthcare utilisation or other adverse clinical outcomes have yet to be explored.

Approach

A retrospective cohort study will be carried out, routinely collected data from patient records of adults aged 45-64 (N=1,100,000), using Clinical Practice Research Datalink (CPRD) Aurum, will be examined. The cohort will include all adults registered on Jan 1st 2019 with at least 12 months prior registration. PIP will be defined using the Prescribing Optimally in Middle-aged People’s Treatment (PROMPT) criteria, exposure to PIP will be established in the period 1st Jan 2018-31st December 2018. Covariates for analysis will be established including, gender, age, number of repeat medications in the previous year (2018), ethnicity, multimorbidity (defined using Cambridge Multimorbidity Score), socio-economic status (Index of Multiple Deprivation), and the number of GP clinic attendances in previous year.

Outcomes will be examined for the period 1st Jan 2019-31st December 2019. These will include the number of GP attendances, number of hospitalisations, and all-cause mortality. A multi-level Poisson regression model will be used to assess association between PIP and count outcomes (e.g., number of GP attendances, hospital admissions). A Cox proportional hazard model will be used to assess the association between PIP exposure and all-cause mortality. Appropriate methods will be used to account for missing data.

Findings

The number and percentage of those participants experiencing PIP in the exposure period will be reported.

The number of patients experiencing the pre-specified outcomes will be reported, as well as the association of PIP with each outcome. We will report hazard ratios/risk ratios, 95% confidence intervals and p-values.

Consequences

The use of PIP criteria in RCTs for medicines optimisation in the elderly has been shown to reduce PIP prevalence, falls, length of hospital stays, care-visits, and medication costs.

Understanding the relationship between PIP in middle-aged adults and important clinical outcomes (e.g. healthcare utilisation, mortality) will facilitate the development of medicines optimisation interventions in this middle aged group where polypharmacy and multimorbidity are increasingly common.