Abstract
BACKGROUND: Various analgesics are frequently prescribed by physicians and used by patients with low back pain (LBP) despite limited effect on pain and disability and risk of side effects. Current knowledge on how psychological measures and self-management interventions influence analgesic use is limited. We investigated if analgesic use changed after participating in a patient education and exercise therapy program (GLA:D® Back), to what extent analgesic use and self-efficacy at baseline were potential determinants of analgesic use at the end of the program, and, to what extent improvement in self-efficacy from before to after the intervention modified the relationship between analgesic use at baseline and follow-up. METHODS: We used data from the Danish GLA:D(®) Back registry collected from March 28, 2018, until October 16, 2023. Potential determinants were self-reported baseline analgesic use and self-efficacy (the Arthritis Self Efficacy Scale pain subscale). The outcome was analgesic use at 3 months follow-up. We used logistic regression to investigate associations and effect modification. RESULTS: Among 4721 included participants, 34% of those using analgesics at baseline (n = 942) discontinued this at 3 months follow-up. Analgesic use at baseline was associated with increased odds of analgesic use at follow-up (odds ratio [OR]: 9.79, 95% confidence interval [CI]: 7.88, 12.15), and higher levels of self-efficacy at baseline was associated with decreased odds of analgesic use at follow-up (OR: 0.85, 95% CI: 0.81, 0.89). Improved self-efficacy, obtained during the program, reduced the risk of analgesic use at follow-up from 15 to 6% and from 76 to 54% among participants with and without baseline analgesic use respectively. CONCLUSIONS: Patients using analgesics when initiating care were more likely to use analgesics three months later, while those having high levels of self-efficacy were less likely. Improved self-efficacy during the program reduced the absolute risk of analgesic use following the intervention to a larger extent among those using analgesics at baseline compared to those without baseline use. Further investigation is needed to confirm whether these findings reflect causal effects.