A population-based retrospective cohort study to examine the risk of psychiatric side effects associated with glucocorticoid use in adults in the United Kingdom (2015-2023): a Clinical Practice Research Datalink study

Research question To what extent do Glucocorticoids (GCs) increase the risk of mental health side effects when prescribed in primary care and what patient factors affect this risk. Background Glucocorticoids (GCs) are amongst the most frequently prescribed medicines worldwide. In 2020 over one million UK prescriptions for prednisolone were made. GCs are associated with multiple side effects, including mental health associated risks. Although it is accepted that these side effects occur, there is little evidence of the absolute size of this risk, who is at increased risk or whether risk is related to the dose/duration of treatment. It is important to address this gap so that patients can be fully informed of the risk associated with their treatment, supporting shared decision making. Aims and Objectives The aim of this study is to determine the mental health risks associated with glucocorticoid use. Specific objectives are to determine: 1. the risk of mental health adverse events with systemically administered glucocorticoid prescriptions in adults in UK primary care, compared to the general population. 2. patient and prescription-level factors associated with an increased risk of adverse mental health events. Methods The study uses data from the Clinical Practice Research Datalink (CPRD) and linked datasets. We will carry out an emulated target trial. Eligible individuals will be aged ≥18 years, permanently registered with a contributing practice (01/01/2015-31/03/2023) and contributing research-quality data. The incidence of adverse mental health events will be compared between people prescribed short term GCs on the same date as a coded consultation with a recognised indication, and people who did not receive GCs during the study period. Analysis will allow for demographic and clinical confounders, decided using directed acyclic graphs. The primary outcome is a composite of anxiety, neurosis, cyclothymia, depression, psychosis, schizophrenia, bipolar affective disorder, delirium, hallucinations, acute confusion, insomnia, cognitive impairment or dementia, self-harm or death due to suicide within 3 months of the treatment date. Secondary outcomes will involve assessing each of these outcomes separately. Analyses will be stratified by ethnicity to ascertain whether individuals from any ethnic groups are at greater risk of GC-related mental health side effects. Our prognostic study includes all individuals prescribed a GC during the study period. We will investigate potential prognostic factors that modify the risk of mental adverse events relating to the prescription or individual. Potential prognostic factors will include mental health diagnosis in the 12 months prior to index date (other than the outcome of interest), previous mental health diagnosis, multi-morbidity, and cognitive impairment, polypharmacy, Index of Multiple Deprivation (IMD) quintile, sex, smoking status, previous GC use, GC indication, BMI, age and ethnicity. Timelines for delivery This study will be delivered in 24 months. Anticipated Impact and Dissemination The results of this study will be of direct relevance to patients requiring GCs and the clinicians providing care. Key outputs include: • New evidence regarding the safety of GCs and factors associated with an increase in the risk of mental health side effects. • information to support clinician and patient decision making.