Opioid use: Case-control analyses of worker's compensation data

Abstract

Working-age adults are disproportionately impacted by opioid misuse. Factors associated with opioid misuse in people with workers compensation (WC) claims are not well studied. WC in some states is a "captured" market making it a more efficient site for researching the opioids epidemic. A pilot study was conducted to identify factors associated with opioid use using a large WC insurer's claims in Utah. This was a case-control study using a large WC insurer's database. We conducted secondary data analyses of a de-identified dataset originally obtained from the WC insurer. Cases were defined as claims with a morphine equivalent dose (MED)≥50 mg/d in the 30 days after the claim was filed while controls = 0 mg/d. A total of 76 patient's claims (28 cases and 48 controls) were included in the final data analyses. The majority of claimants were male (N = 50, 65.8 percent), worked full time (N = 58, 76.3 percent) and had a mean age of 37.0±11.4 years. The majority of controls filed medical only claims (N = 40, 83.3 percent) while the majority of cases filed indemnity claims (N = 19, 67.9 percent). Cases were prescribed a mean MED of 126.4 (SD = 93.3) within the first month after filing the claim. Most cases visited>3 medical providers (N = 13, 46.4 percent) in the first month after filing the claim while the majority of controls only visited one provider (N = 28, 58.3 percent). Remarkably, the mean number of providers visited within the first month for the cases was 3.8, which was 2-fold greater than the control group. Exploratory multivariate analyses showed that cases were 4.6 times more likely to have visited 2-3 medical providers (p = 0.025), and 41.8 times more likely to have visited more than three medical providers (p < 0.001). Cases had 3.6 higher odds of having been prescribed nonsteroidal anti-inflammatory prescription within the first month as compared to controls (p = 0.014). This pilot study found risk factors, some of which may be modifiable. We aim to conduct a large study using existing WC data to create a scoring system that identifies those claimants at higher risk of adverse opioid-related events that may have preventive applications at a systems-level.