Introduction
The OPAL trial (n = 347) revealed that an opioid (oxycodone/naloxone) was not superior to placebo in reducing pain among an acute neck/back pain population. But at 1-year follow-up, the opioids group reported slightly worse pain compared to the placebo. Mediation analysis is a quantitative approach to investigate how an intervention causes an effect on an outcome. The analyses aim to explore the underlying mechanisms of why a short course of opioids led to worse pain in the long term.
Methods
Proposed mediators were short term pain severity, poorer mental health, poorer physical function, and the presence of opioid-related adverse events. We controlled for confounders (age, sex, BMI, pain duration, pain location, previous episodes, work status, income, baseline pain, function, and mental health). We used the R package CMAverse to fit two regression models - the mediator model and the outcome model - to estimate the intervention-mediator and mediator-outcome effects considering each mediator individually.
Results
A total of 346 participants provided their characteristic at baseline, 289 participants (84%) provided data on mediators at the end of the 6-week prescription period, and 256 participants (74%) had the data at 52 weeks. Mean age was 45 (16) years and 51% were female. Proposed mediators (mental health, pain severity, opioid related adverse events and physical function) at week 6 did not mediate the worse pain severity at 52 weeks.
Conclusion
It remains uncertain why a short course of opioids would cause worse pain in the long term.