Association of Pharmacologic and Nonpharmacologic Management of Acute Low Back Pain with Overdose Hospitalizations: A Nested Case-Control Study

Background: Noninvasive nonpharmacologic therapies are recommended for managing acute low back pain (aLBP) and have the potential to mitigate opioid-related harms. However, little is known about whether incorporating nonpharmacologic therapies into aLBP management affects adverse outcomes. The objective was to determine if receiving nonpharmacologic pain therapies, alone or combined with pharmacologic options, is associated with drug-related overdose hospitalizations among Medicare beneficiaries with aLBP. Methods: A nested case-control study was conducted using 2016-2019 Medicare claims to identify fee-for-service beneficiaries with new episodes of aLBP (i.e., LBP lasting <3 months). Cases had inpatient claims for drug overdoses within 90 days of aLBP diagnosis. The exposure was mutually exclusive categories for pain therapies: (1) pharmacologic only (opioids and/or gabapentinoids), (2) nonpharmacologic only (physical therapy and/or spinal manipulation therapy), (3) both pharmacologic and nonpharmacologic, and (4) none of these. The outcome was hospitalization involving drug overdose. We conducted conditional logistic regression adjusting for baseline sociodemographic, clinical, and geographic covariates. Results: There were 3,042 cases and 12,168 matched controls. One-third (33.7%) of cases versus 26.8% of controls received pharmacologic therapies only compared with 6.7% (cases) and 10.2% (controls) for nonpharmacologic therapies only. Receipt of both pharmacologic and nonpharmacologic therapies was 7.3% (cases) and 3.2% (controls). Compared with exclusively receiving pharmacologic therapies, receiving nonpharmacologic therapies only was associated with lower odds of overdose-related hospitalization (adjusted odds ratio [aOR] = 0.56, 95% confidence interval [CI]: 0.47-0.66), whereas pharmacologic and nonpharmacologic treatments combined were associated with nearly twofold increased odds of overdose-related hospitalization (aOR = 1.87, 95% CI: 1.55-2.27). Conclusions: Among Medicare beneficiaries with new episodes of aLBP, treatment with only nonpharmacologic therapies was protective of overdose hospitalizations. However, any treatment with opioids and/or gabapentinoids, alone or combined with nonpharmacologic therapies, was associated with increased odds of overdose hospitalization. Implementation research is needed to inform successful adoption of nonpharmacologic pain therapies especially in subgroups with increased risk of adverse outcomes.