Literature Collection

Introduction: Access to prenatal care offers the opportunity for providers to assess for substance use disorders (SUDs) and to offer important treatment options, but utilization of treatment during pregnancy has been difficult to measure. This study presents pre-COVID trends of a subset of SUD diagnosis at the time of delivery and related trends in treatment utilization during pregnancy. Materials and Methods: A retrospective cohort design was used for the analysis of West Virginia Medicaid claims data from 2016 to 2019. Diagnosis of SUDs at the time of delivery and treatment utilization for opioid use disorder (OUD) and non-OUD diagnosis during pregnancy across time were the principal outcomes of interest. This study examined data from n = 49,398 pregnant individuals. Results: Over the 4-year period, a total of 2,830 (5.7%) individuals had a SUD diagnosis at the time of delivery. The frequency of opioid-related diagnoses decreased by 29.3%; however, non-opioid SUD diagnoses increased by 55.8%, with the largest increase in the diagnosis of stimulant use disorder (30.9%). Treatment for OUD increased by 13%, but treatment for non-opioid SUD diagnoses during pregnancy declined by 41.1% during the same period. Conclusions: Interventions enacted within West Virginia have improved access and utilization of treatment for OUD in pregnancy. However, consistent with national trends in the general population, non-opioid SUD diagnoses, especially for stimulants, have rapidly increased, while treatment for this group decreased. Early identification and referral to treatment by OB-GYN providers are paramount to reducing pregnancy and postpartum complications for the mother and neonate.

BACKGROUND: The COVID-19 pandemic drove significant disruptions in access to substance use disorder (SUD) treatment and harm reduction services. Healthcare delivery via telemedicine has increasingly become the norm, rendering access to a phone essential for engagement in care. METHODS: Adult patients with SUD who lacked phones (n = 181) received a free, pre-paid phone during encounters with inpatient and outpatient SUD programs. We evaluated changes in healthcare engagement including completed in-person and telemedicine outpatient visits and telephone encounters 30 days before and after phone receipt. We used descriptive statistics, where appropriate, and paired t-tests to assess the change in healthcare engagement measures. RESULTS: Patients were predominantly male (64%) and white (62%) with high rates of homelessness (81%) and opioid use disorder (89%). When comparing 30 days before to 30 days after phone receipt, there was a significant increased change in number of telemedicine visits by 0.3 (95% CL [0.1,0.4], p < 0.001) and telephone encounters by 0.2 (95% CL [0.1,0.3], p = 0.004). There was no statistically significant change in in-person outpatient visits observed. CONCLUSIONS: Pre-paid phone distribution to patients with SUD was associated with an increased healthcare engagement including telemedicine visits and encounters.

Medications such as buprenorphine-naloxone are among the most effective treatments for opioid use disorder, but limited retention in treatment limits long-term outcomes. In this study, we assess the feasibility of a machine learning model to predict retention vs. attrition in medication for opioid use disorder (MOUD) treatment using electronic medical record data including concepts extracted from clinical notes. A logistic regression classifier was trained on 374 MOUD treatments with 68% resulting in potential attrition. On a held-out test set of 157 events, the full model achieved an area under the receiver operating characteristic curve (AUROC) of 0.77 (95% CI: 0.64-0.90) and AUROC of 0.74 (95% CI: 0.62-0.87) with a limited model using only structured EMR data. Risk prediction for opioid MOUD retention vs. attrition is feasible given electronic medical record data, even without necessarily incorporating concepts extracted from clinical notes.