(224) Patient characteristics associated with new community prescription of benzodiazepine and z-drug hypnotics after critical illness in the UK, a retrospective cohort study.

Background: Initiation of benzodiazepines and z-drug hypnotics after critical illness is not uncommon, despite the physiologic vulnerability of critical illness survivors and known possible adverse events (e.g., falls, fractures, and dependency). Understanding which patients are at a higher risk of prescription may help target post-discharge treatments and avoid future morbidity.

Objectives: To estimate the proportion of critical care survivors that received a new benzodiazepine or z-drug prescription after hospital discharge and to identify associated patient characteristics.

Methods: This retrospective cohort study used linked general practice, hospital, and mortality datasets from the UK Clinical Practice Research Datalink. Our study included all adults (18+ years) admitted for critical care in 2010 or 2018 that survived to hospital discharge, who were not prescribed a benzodiazepine or z-drug in the 180 days prior to hospitalisation. The primary outcome was receipt of a community benzodiazepine or z-drug prescription within 90 days of index hospital discharge. Multivariable logistic regression models were used to examine associations between patient characteristics and outcomes.

Results: There were 2758/53,386 (5.2%) naïve survivors prescribed a new benzodiazepine or z-drug within 90 days of discharge (median time to prescription: 21 days, IQR 7-49). Zopiclone was most commonly prescribed (50%), followed by diazepam (18%). Among patients prescribed benzodiazepine or z-drugs, 47.3% (1305/2758) received 2+ prescriptions within 180 days of discharge (median 3 prescriptions, IQR: 2-5). Patients with history of insomnia (aOR 1.99; 95%CI 1.77-2.23), prescription opioid use (aOR 1.51; 95%CI 1.38-1.65), and anxiety/depression (aOR 1.41; 95%CI 1.29-1.54) had the highest risks of new benzodiazepine or z-drug prescribing when compared to patients without a history of insomnia, prescription opioid use, or anxiety/depression, respectively.

Conclusions: One in twenty benzodiazepine or z-drug-naïve critical illness survivors were prescribed a new benzodiazepine or z-drug within 90 days of discharge. Prescribing was associated with history of insomnia, recent opioid prescription, and history of anxiety/depression. Future research should focus on understanding clinical decision making around initiation, and quantifying the risk of adverse events of benzodiazepine or z-drug prescription in survivors of critical illness, to help guide policy for clinical practice.