Book Description

Opioid use disorder (OUD) among pregnant and non-pregnant women of reproductive age continues to be a significant public health concern in the U.S. Using opioids during pregnancy poses significant health risks, including fatal overdose, infectious disease exposure, and other major health concerns. Among individuals who use opioids during pregnancy, co-occurring substance use is common, specifically alcohol, stimulant and cannabis use. Data from the Centers for Disease Control and Prevention indicate that 7 percent of pregnant women reported using opioids during pregnancy in 2019, a figure which has trended upward for the past two decades. Despite long-standing recognition of medications for opioid use disorder (MOUD) as the gold standard of care to treat OUD among pregnant and reproductive-aged women, most will not receive treatment. Additionally, Medicaid pays the largest portion of OUD-related healthcare costs for pregnant women, solidifying substance use during pregnancy as a critical policy issue for the Federal government. The 21st Century Cures Act, passed in 2016, dedicated $1 billion toward prevention, treatment, and recovery efforts, with states receiving funding through State-Targeted Response (STR) grants to increase state-specific opioid funding. Funded programs include Washington State's hub and spoke model to increase access to medication treatment for OUD, which began in 2017. While increasing MOUD receipt is the core of these efforts, it is unclear to what extent these investments have been successful for pregnant and non-pregnant women of reproductive age. Further, less is known more broadly about how pregnancy affects treatment access and healthcare utilization, if and how treatment continues post-pregnancy, and what healthcare utilization more holistically looks like for those who receive treatment. Using Washington state as its setting, this dissertation had three aims: 1) to understand what MOUD receipt and utilization look like for pregnant and non-pregnant women of reproductive age with co-occurring alcohol, stimulant and cannabis use disorder, explore differences of setting diagnosis for those with OUD and comorbid alcohol use disorder (AUD), and explore other OUD-related service utilization; 2) to explore pregnant women's pregnancy complications and healthcare utilization outcomes of receiving consistent MOUD versus inconsistent MOUD or no MOUD; and 3) to explore, among pregnant women who receive MOUD, the impact that receiving treatment from Washington's hub and spoke program has on their outcomes. Significant differences in MOUD outcomes among pregnant women were further examined by race/ethnicity, age, and locality, in addition to co-occurring alcohol, stimulant, and cannabis use, as well as comorbid mental health conditions. To explore these aims, the dissertation used Washington Medicaid claims data of 564 pregnant women and 564 non-pregnant women aged 18-44 from 2016-2019, while employing the Andersen behavioral model of healthcare utilization, to understand how predisposing and enabling factors impact MOUD utilization and subsequent outcomes. The methodology used was a combination of descriptive data analysis and multivariate regression analyses to understand the relationship between these factors with MOUD utilization and outcomes. Results from Aim 1 showed that while pregnant women with OUD were more likely than non-pregnant women with OUD to receive MOUD, most pregnant and non-pregnant women did not receive MOUD, including in the postpartum period. Comorbid mental health disorders were high overall. Co-occurring stimulant use disorder was more common among non-pregnant and pregnant women with OUD than AUD or cannabis use disorder; co-occurring AUD was less common among pregnant women than non-pregnant women. Both groups of women had low rates of residential, inpatient, and detoxification services, while both groups had high rates of outpatient psychosocial service utilization. Pregnant women were more likely than non-pregnant women to be diagnosed with OUD in an inpatient hospital setting, whether they had AUD or not. Aim 2 results showed that those who received MOUD had better healthcare utilization measures (e.g., more likely to receive certain health screenings) compared to those who did not receive MOUD. Pregnancy complications among all women were low and receiving consistent MOUD (i.e., six monthly medication visits prior to delivery) resulted in greater healthcare utilization. Aim 3 results were limited due to a small sample of pregnant women with OUD who received MOUD from a hub and spoke network, although it seems that hub and spoke did not affect the overall findings. The small sample size may have curtailed meaningful conclusions about how the hub and spoke system affected pregnant women's healthcare utilization and pregnancy outcomes. Results from this dissertation have important clinical and policy-related implications. Further evidence that substantiates the prevalence of polysubstance use may be useful to providers in treating reproductive aged and pregnant women with OUD, and low MOUD rates may be additional reason for universal OUD screening and brief referral to treatment. Specific findings around both low MOUD rates and low rates of certain healthcare utilization measures, such as HIV screening, may also be useful to medical associations, as they consider how to update their quality measures and guidelines, and state agencies. State policymakers and substance use programs could tailor programming for subsets of pregnant women that are most at risk of treatment underutilization. Lastly, understanding the impact of delivery system reform for substance use treatment, such as the hub and spoke model, and its impact on vulnerable, priority populations such as pregnant women is especially imperative in today's fiscally challenging policy environment.