Title : Association between patient characteristics and low-dose naltrexone use among palliative care patients: A retrospective analysis using Epic Cosmos (2022–2024)
Abstract:
Background: Low-dose naltrexone (LDN) is increasingly prescribed for chronic pain and inflammatory conditions, yet utilization patterns among palliative care populations remain poorly characterized. Older adults receiving palliative care represent a clinically distinct group in whom insurance coverage, neighborhood disadvantage, and geographic factors may substantially shape access to emerging analgesic therapies. This study evaluated demographic, socioeconomic, and geographic factors associated with LDN use among palliative care patients using a large, multi-institutional electronic health record dataset.
Methods: This retrospective observational study used de-identified data from Epic Cosmos, a federated electronic health record dataset containing records from over 304 million patients across more than 2,133 U.S. hospitals. Palliative care patients were identified between 2022 and 2024 using ICD-10-CM/ICD-9-CM diagnosis codes and palliative care-specific CPT procedure codes. LDN use was defined as naltrexone ≤4.5 mg/day or a documented compounded LDN formulation. A one-year lookback period (2021) was used for baseline comorbidity ascertainment. Multivariable patient-year logistic regression with cluster-robust standard errors evaluated associations between LDN utilization and age, sex, insurance type, U.S. Census region, Social Vulnerability Index (SVI) quintile, and calendar year.
Results: Among 1,912,564 palliative care patients identified, 385 received LDN (prevalence ∼0.02%). LDN utilization increased markedly over the study period, with adjusted odds 3.82-fold higher in 2023 (95% CI 2.27–6.44) and 6.58-fold higher in 2024 (95% CI 3.98–10.89) relative to 2022. Male sex (OR 0.515, 95% CI 0.389–0.682) and increasing age (OR 0.969 per year, 95% CI 0.964–0.974) were independently associated with lower odds of LDN receipt (both p<0.001). Medicaid (OR 0.292, 95% CI 0.174–0.489) and Medicare coverage (OR 0.456, 95% CI 0.320–0.650) were each associated with significantly lower utilization compared with commercial insurance. Patients residing in the most socially vulnerable neighborhoods (SVI Q5) had approximately one-third the odds of LDN receipt versus those in the least vulnerable quintile (OR 0.332, 95% CI 0.210–0.524). Residence in the Western United States was associated with approximately twice the odds of LDN use compared with the South (OR 1.907, 95% CI 1.265–2.875). Fibromyalgia (58.2%), refractory low back pain (29.6%), and peripheral neuropathic pain (28.8%) were the most prevalent comorbidities among LDN users. No significant interaction was observed between insurance type and neighborhood social vulnerability (χ²=13.78, df=12, p=0.315).
Conclusions: LDN utilization among palliative care patients increased substantially between 2022 and 2024 but remained rare overall and highly unequal. Older age, male sex, public insurance, and residence in socially vulnerable neighborhoods were each independently associated with lower utilization. These findings suggest that structural access barriers, particularly insurance coverage and neighborhood disadvantage; may limit equitable uptake of emerging analgesic options in aging palliative care populations. Targeted prescribing support and formulary access interventions should be explored to reduce these disparities.
