Overdose risk profiles in pregnancy: Latent class analysis of pregnant individuals with opioid use disorder.

Introduction: Given the continued rise in opioid exposed pregnancies and overdose during the postnatal period, it is critical to identify risk characteristics among this population to enable clinicians to better tailor interventions. This exploratory study sought to develop a deeper understanding of overdose risk characteristics among pregnant people with opioid use disorder and which characteristics may contribute to differing risk profiles.

Methods: Design and participants. This exploratory secondary analysis utilized baseline data from a large-scale national multi-site randomized controlled trial that compared two buprenorphine formulations among treatment seeking pregnant individuals with opioid use disorder.

Assessments: For risk group identification, the Personal Opioid-Overdose Risk Survey was used. Trauma history experience was assessed using the Trauma History Screen and substance use history was captured using the DSM-5 Checklist and Treatment Services Review V6.

Analyses: Latent class analysis identified unique subgroups of participants based on overdose risk factors. Latent class group membership was associated with trauma history and substance use characteristics using logistic and stepwise logistic regression.

Results: Three distinct classes of overdose risk emerged: the tolerance and polysubstance/alcohol use (HIGH-ALC) class (n = 14, 10 %), synthetic opioid and polysubstance use (LOW-ALC/FENT) class (n = 65, 46.4 %), and the low risk (LOW-RISK) class (n = 61, 43.6 %). The HIGH-ALC class reported the most (non-opioid) substance use in the last 12 months with 6 times higher odds of marijuana use (95 % CI, 1.01-35.67) and 17.48 times higher odds of cocaine use (95 % CI, 3.45-88.48) compared to the LOW-RISK class. The LOW-ALC/FENT class (n = 65, 46.4 %) had the highest reports of childhood physical abuse, greater odds of experiencing intimate partner violence regarding recovery (OR = 4.82, 95 % CI = 1.90-12.26), and greater odds of a threat to safe living (OR = 3.35, 95 % CI = 0.72-15.66). The LOW-RISK class (n = 61, 43.6 %) had the lowest reports of polysubstance use in the last 12 months and the least reports of both childhood sexual trauma and adulthood sexual trauma.

Conclusions: Through better understanding distinct patient overdose risk profiles, healthcare providers can deliver more targeted prevention interventions to address individual needs and improve maternal outcomes.

Trial registration: NCT03918850.