Predictors of opioid use for rib fractures in a regional Australian hospital

Abstract

BackgroundRib fractures (RFs) are the leading type of single serious injury in New South Wales trauma patients. Uncontrolled pain drives the sequelae of atelectasis, pneumonia, respiratory failure, and death in severe cases. Opioids are the mainstay of management; however, they carry numerous adverse effects. Understanding patient or injury factors which predict opioid requirement is important to tailor management. Existing evidence is limited to metropolitan trauma centres (MTCs). MethodsWe conducted an observational, retrospective, single-centre cohort study of all admissions to Albury Wodonga Health diagnosed with one or more RFs and discharged between January 1st, 2017, and December 31st, 2022, inclusive. Data collected included demographics, injury characteristics, and management, including analgesia. LASSO regression was performed to determine predictors of average daily opioid use for the first five days of admission in oral morphine equivalents (mg). R2 and root mean square error (RMSE) were calculated to assess model performance. ResultsWe included 624 patients. LASSO selected number of RFs, fracture displacement score, pulmonary contusion, new injury severity score, age, chest tube use, chronic pain history, opioid history and upper or middle lateral RF location categories as predictors. Sex, middle anterior, middle posterior, and lower RF location categories were excluded by LASSO. The out of sample R2 was 28.6 %. On the scale of log OME, the RMSE was 1.08. ConclusionThe model is effective at identifying predictors of opioid use in this regional centre, which are similar to those described in evidence from MTCs. However, the low R2 with wide prediction intervals limits its utility on an individual level.