Non-benzodiazepine Hypnotics and Police-Reported Motor Vehicle Crash Risk among Older Adults: A Sequential Target Trial Emulation

Andrew R Zullo,Marzan A Khan,Melissa R Pfeiffer,Seth A Margolis,Brian R Ott,Allison E Curry,Thomas A Bayer,Melissa R Riester,Nina R Joyce
DOI: https://doi.org/10.1093/aje/kwae168
2024-07-02
American Journal of Epidemiology
Abstract:Abstract Non-benzodiazepine hypnotics ( “Z-drugs”) are prescribed for insomnia, but might increase risk of motor vehicle crash (MVC) among older adults through prolonged drowsiness and delayed reaction times. We estimated the effect of initiating Z-drug treatment on the 12-week risk of MVC in a sequential target trial emulation. After linking New Jersey driver licensing and police-reported MVC data to Medicare claims, we emulated a new target trial each week (July 1, 2007 – October 7, 2017) in which Medicare fee-for-service beneficiaries were classified as Z-drug-treated or untreated at baseline and followed for an MVC. We used inverse probability of treatment and censoring weighted pooled logistic regression models to estimate risk ratios (RR) and risk differences with 95% bootstrap confidence limits (CLs). There were 257,554 person-trials, of which 103,371 were Z-drug-treated and 154,183 untreated, giving rise to 976 and 1,249 MVCs, respectively. The intention-to-treat RR was 1.06 (95%CLs 0.95, 1.16). For the per-protocol estimand, there were 800 MVCs and 1,241 MVCs among treated and untreated person-trials, respectively, suggesting a reduced MVC risk (RR 0.83 [95%CLs 0.74, 0.92]) with sustained Z-drug treatment. Z-drugs should be prescribed to older patients judiciously but not withheld entirely over concerns about MVC risk.
public, environmental & occupational health
What problem does this paper attempt to address?