Pharmacoepidemiology
Psychiatric Co-prescribing and Prescribing Characteristics of Older Adults Initiated on Sertraline, Citalopram, or Mirtazapine in Clinical Practice Research Datalink Aurum: Descriptive Analyses to Inform a Causal Study Design Aaron Jun Yi Yap* Aaron Jun Yi Yap Yap Yap Yap Yap Yap MRC Integrative Epidemiology Unit, Population Health Sciences, Bristol Medical School, University of Bristol, Bristol, UK
Antidepressants, antipsychotics, benzodiazepines and Z-hypnotics are commonly prescribed to treat depression and anxiety; however, polypharmacy of these medications have been associated with increase risks of falls, fractures and cardiovascular events in older adults. Challenges of designing pharmacoepidemiological observational studies examining safety of polypharmacy using electronic health records are (1) specifying sufficiently-well defined interventions where switching and combination use are prevalent, (2) identifying suitable time (grace) periods to apply methodological approaches to reduce time-related biases, and (3) understanding the completeness of indication-based sampling. We aimed to describe in detail prescription and co-prescribing characteristics of adults aged 50 years and above who are initiated on Sertraline, Citalopram or Mirtazapine in GP surgery practices in England registered with Clinical Practice Research Datalink Aurum between 1 January 2010 to 31 March 2023. The outcomes are: (1) proportion of prevalent and new users of Sertraline, Citalopram and Mirtazapine, (2) time to first prescription of a second antidepressant, (3) proportion with coded evidence of depression and/or anxiety, (4) starting dose of Sertraline, Citalopram and Mirtazapine and time to first increase in dose, (5) proportion with co-prescriptions of antidepressants, antipsychotics, benzodiazepines and Z-hypnotics, and (6) average initiation gap and treatment duration of combination therapy for each drug-class pair. Data analysis is underway and will enhance understanding of the feasibility of studying drug combinations. We anticipate discussing applications of findings to the target trial framework.