A Regional Analysis of Low Back Pain Treatments in the Military Health System
Jon D Lurie1, Christopher G Leggett2, Jonathan Skinner3
1Geisel School of Medicine, Lebanon, NH.
Study Design:
Claims-based analysis of cohorts of TRICARE Prime beneficiaries.
Objective:
To compare rates of utilization of 5 low back pain (LBP) treatments (physical therapy (PT), manual therapy, behavioral therapies, opioid, and benzodiazepine prescription) across catchment areas and assess their association with the resolution of LBP.
Summary Of Background:
Guidelines support focusing on nonpharmacologic management for LBP and reducing opioid use. Little is known about patterns of care for LBP across the Military Health System.
Patients And Methods:
Incident LBP diagnoses were identified data using the International Classification of Diseases ninth revision before October 2015 and 10th revision after October 2015; beneficiaries with "red flag" diagnoses and those stationed overseas, eligible for Medicare, or having other health insurance were excluded. After exclusions, there were 159,027 patients remained in the final analytic cohort across 73 catchment areas. Treatment was defined by catchment-level rates of treatment to avoid confounding by indication at the individual level; the primary outcome was the resolution of LBP defined as an absence of administrative claims for LBP during a 6 to 12-month period after the index diagnosis.
Results:
Adjusted rates of opioid prescribing across catchment areas ranged from 15% to 28%, physical therapy from 17% to 39%, and manual therapy from 5% to 26%. Multivariate logistic regression models showed a negative and marginally significant association between opioid prescriptions and LBP resolution (odds ratio: 0.97, 95% CI: 0.93-1.00; P = 0.051) but no significant association with physical therapy, manual therapy, benzodiazepine prescription, or behavioral therapies. When the analysis was restricted to the subset of only active-duty beneficiaries, there was a stronger negative association between opioid prescription and LBP resolution (odds ratio: 0.93, 95% CI: 0.89-0.97).
Conclusions:
We found substantial variability across catchment areas within TRICARE for the treatment of LBP. Higher rates of opioid prescription were associated with worse outcomes.
