Abstract
Self-controlled study designs can be used to assess the association between exposures and acute outcomes while controlling for important confounders. Using routinely collected health data, a self-controlled case series design was used to investigate the association between opioid use and bone fractures in 2008-2017 among adults registered in the United Kingdom Clinical Practice Research Datalink. The relative incidence of fracture was estimated, comparing periods when these adults were exposed and unexposed to opioids, adjusted for time-varying confounders. Of 539,369 people prescribed opioids, 67,622 sustained fractures and were included in this study. The risk of fracture was significantly increased when the patient was exposed to opioids, with an adjusted incidence rate ratio of 3.93 (95% confidence interval (CI): 3.82, 4.04). Fracture risk was greatest in the first week of starting opioid use (adjusted incidence rate ratio: 7.81, 95% CI: 7.40, 8.25) and declined with increasing duration of use. Restarting opioid use after a gap in exposure significantly increased fracture risk (adjusted incidence rate ratio: 5.05, 95% CI: 4.83, 5.29) when compared with nonuse. These findings highlight the importance of raising awareness of fractures among patients at opioid initiation and demonstrate the utility of self-controlled methods for pharmacoepidemiologic research.