Introduction

In the US, health insurance is an important determinant of health, enabling access to diagnostic and therapeutic services that can improve outcomes. Researchers and policy makers have long focused on increasing enrollment in health insurance as a means to improve health. Enrolling in insurance may have more limited effects, however, if individuals are unable to maintain their insurance. Research on insurance churn, defined as the movement of individuals between or out of insurance plans, has highlighted the importance of maintaining continuous insurance coverage and minimizing insurance transitions. Gaps in insurance coverage have been linked to delayed medical care, decreased medication adherence, and increased financial risk. Even without a gap in coverage, switching insurance plans may result in treatment disruptions due to differences in benefits or clinician practices. In contrast, insurance transitions could potentially lead to improvements in clinician network breadth and to more generous benefits.

Opioid use disorder (OUD) is a chronic condition associated with significant morbidity and mortality, with more than 80 000 opioid-related overdose deaths in the US in 2021 alone. There are multiple effective medications for OUD (MOUD) associated with decreased morbidity and mortality, although they remain underused. Insurance benefits are crucial for accessing life-saving MOUD, as the cost of these medications can be prohibitive for patients with OUD with disproportionately low income. Furthermore, treatment retention in MOUD is associated with decreased mortality, but insurance instability may limit retention if patients cannot afford services. Despite the potential importance of insurance stability to treatment access and retention for OUD, there has been limited research investigating insurance stability in this patient population.

Using data from the Massachusetts Public Health Data Warehouse, we sought to estimate the cumulative incidence of insurance transitions among patients with newly diagnosed OUD covered by Medicaid or commercial insurance in the first year after their diagnosis. We also estimated the probability of insurance transitions across various sociodemographic groups, given known inequities in access to MOUD.

Methods

Study Design and Data Source

We performed a retrospective cohort study using data from the Massachusetts Public Health Data Warehouse. In the Public Health Data Warehouse, records from the All-Payer Claims Database (APCD) are linked longitudinally to individual-level administrative records from more than 30 sources using a probabilistic linkage which has been described previously. For this study, we used data from the APCD, Acute Care Hospital Case Mix, the Prescription Monitoring Program, the Bureau of Substance Addiction Services, the Registry of Vital Records and Statistics, the state Department of Correction (state prisons), and county Houses of Corrections (county jails) (eTable 1 in Supplement 1. The Public Health Data Warehouse suppresses the output for all counts between 1 and 10 to protect privacy. This study was determined to be not human participant research by the Boston University Medical Campus Institution Review Board. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Reporting of Race and Ethnicity

The Department of Public Health sought to identify race and ethnicity for as many individuals as possible included in the Massachusetts Public Health Data Warehouse. For each data set with race and ethnicity, individuals were assigned 1 of 5 mutually exclusive race categories of Asian/Pacific Islander or Native Hawaiian non-Hispanic, Black non-Hispanic, Hispanic, White non-Hispanic, non-Hispanic American Indian, Alaska Native, or other. The Massachusetts Public Health Data Warehouse then extracted a single race and ethnicity for each individual using a hierarchy of which data sets most reliably used self-reported race and ethnicity. In the case of more than one race being reported, we used the following approach: take the most frequently reported race and ethnicity for that individual, and if there was a tie for the most frequently reported race and ethnicity for that individual, and one of those tied values was Hispanic, then the individual was identified as Hispanic. Otherwise, individuals were assigned as the least common race in Massachusetts that they had listed as one of their tied race categories with the following configuration: the least common were American Indian/Alaska Native/Other, non-Hispanic; second least common, Asian, Pacific Islander, Native Hawaiian, non-Hispanic; third least common, Black, non-Hispanic, and most common, White, non-Hispanic.

Cohort Selection

We defined a cohort of Massachusetts residents with incident OUD using International Classification of Diseases, Ninth Revision (ICD-9) and International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) diagnosis codes for insurance claims submitted to the Massachusetts APCD and Acute Care Hospital Case Mix files (eTables 1 and 2). We included adults aged 18 to 63 years to exclude children and individuals who would become eligible for Medicare during follow-up. We defined an incident diagnosis as those without a diagnosis or OUD-related claim in the previous 180 days. Because most individuals with OUD in Massachusetts are insured by either Medicaid or commercial insurers (80.1% during the study timeframe), we focused on these 2 types of insurance at the time of diagnosis. The primary cohort included individuals diagnosed between July 1, 2014, and December 31, 2014, with follow-up through December 31, 2015. At the start of 2016, APCD reporting requirements were removed for self-funded employer plans governed by the Employee Retirement Income Security Act. These changes resulted in an abrupt 27% decrease in commercial claims submitted to the APCD in early 2016, which stabilized after the start of the year. To minimize this change in reporting requirements in our estimates, we excluded individuals with incident diagnoses after 2014 and restricted our follow-up period to 2015.

Exposures and Outcomes

The exposure of interest was having commercial or Medicaid insurance at the time of diagnosis. Our main outcome of interest was the incidence of insurance transitions in the 12 months after diagnosis. Incident insurance transition was defined as any transition away from baseline insurance type at time of diagnosis (either Medicaid or commercial) into a different insurance type or having no reported insurance during follow-up. We used the Massachusetts APCD to classify mutually exclusive insurance types each month. Our secondary outcome was the type of insurance transition experienced. While individuals were restricted to having commercial or Medicaid insurance at the time of diagnosis, they could transition to any insurance type, including commercial (from Medicaid), Medicaid (from commercial), Medicare Advantage, Emergency Medicaid/other (including Veterans Affairs and worker’s compensation), or uninsured/missing. The uninsured/missing category includes individuals who became uninsured, switched to an insurer that does not report to the APCD (eg, fee-for-service Medicare), moved out of state, died, or became incarcerated. We lacked data on the components of the missing category except for death and incarceration.

To characterize the study cohort, we abstracted the following information at the time of diagnosis: age category (aged 18-25, 26-34, 35-44, and 45-63 years); sex; race and ethnicity (with categories as given in the Massachusetts Public Health Data Warehouse as Asian/Pacific Islander non-Hispanic, Black non-Hispanic, Hispanic, White non-Hispanic, other non-Hispanic, missing); pregnancy status (yes/no); number of hospitalizations and emergency department visits in the 180 days prior to diagnosis (0, 1, 2, ≥3); incarceration in the 180 days prior to diagnosis (yes/no); whether an individual experienced an opioid overdose in the 180 days prior to diagnosis (yes/no); and whether they started MOUD (buprenorphine, methadone, naltrexone) in the 30 days after diagnosis (yes/no). We included race and ethnicity as a marker of experiences of discrimination and structural racism that individuals face.

Statistical Analysis

We calculated descriptive statistics at the time of diagnosis stratified by commercial vs Medicaid insurance. We then calculated the unadjusted cumulative incidence of experiencing a first insurance transition during the 12 months after diagnosis by insurance type using the Kaplan-Meier estimator. We then tabulated the frequency and percentage of insurance transition types (transitions from commercial or Medicaid at baseline to commercial, Medicaid, Medicare Advantage, other, or uninsured/missing).

To characterize patient characteristics associated with insurance transitions in the 12 months after diagnosis, we used logistic regression models to generate adjusted estimated probabilities of insurance transitions using a priori selected estimators based on the literature: age category, race and ethnicity, pregnancy status, and whether individuals started MOUD in the 30 days after diagnosis. All logistic regression models were run separately by insurance type, except for models investigating the probability of insurance transitions by baseline insurance type, which pooled all patients together.

We also performed several sensitivity analyses. First, because individuals may lose insurance temporarily due to lapses in paperwork, we performed additional analyses where we allowed individuals to transition from their baseline insurance to the uninsured/missing category for a month, and then return to their baseline insurance without counting it as an insurance transition. Using this 2-month definition, we recalculated the unadjusted cumulative incidence of insurance transitions and the types of insurance transitions that occurred. Second, to clarify whether individuals who transitioned to the uninsured/missing category later regain insurance, we calculated the frequency with which these individuals transitioned back to another insurance category by the end of 12-month follow-up. Third, to assess how insurance transitions may have changed over time, we expanded the cohort to use more recent data, including individuals diagnosed with OUD from July 2014 to March 2019 (with follow-up through March 2020). In this expanded cohort, we excluded individuals diagnosed with OUD in 2015 whose follow-up would have spanned the time when APCD reporting requirements changed. Fourth, to help contextualize the insurance transition patterns in patients with OUD, we compared insurance transition incidence and estimated probabilities to a propensity score–matched cohort of individuals with an incident diagnosis of type 2 diabetes (T2D), another chronic condition requiring routine visits and consistent medication.

All analyses were performed using SAS Enterprise Edition software version 3.81 (SAS Institute Inc) and took place from November 10, 2022, to May 6, 2024.

Results

Baseline Characteristics

There were 20 768 individuals diagnosed with incident OUD between July 1, 2014, and December 31, 2014 (Table 1; eTable 3). Most individuals with newly diagnosed OUD were male (78.4% of those with Medicaid, 87.4% of those with commercial insurance). Just over three-quarters of individuals with OUD (75.4%) had Medicaid insurance at baseline. Of those with Medicaid insurance, 0.4% were Asian/Pacific Islander non-Hispanic, 7.1% were Black non-Hispanic, 12.9% were Hispanic, 78.4% were White non-Hispanic, and 0.9% were other non-Hispanic, and of those with commercial insurance, 0.5% were Asian/Pacific Islander non-Hispanic, 2.6% were Black non-Hispanic, 3.2% were Hispanic, 87.4% were White non-Hispanic, and 2.1% were other non-Hispanic. Individuals with OUD receiving Medicaid also had more hospitalizations and emergency department visits in the 180 days prior to diagnosis compared with those with commercial insurance. Less than one-third of patients with OUD started MOUD within 30 days after diagnosis (28.7% in commercial insurance and 32.7% in Medicaid insurance).

Incidence of Insurance Transitions

In the first 12 months after diagnosis, 30.4% of patients with OUD experienced a transition in insurance. The unadjusted incidence of insurance transitions was similar for those with Medicaid at baseline (30.7%) compared with those with commercial insurance at baseline (29.4%) Figure 1 and Table 2. The cumulative incidence of insurance transitions increased most rapidly in the first 6 months after diagnosis (Figure).

The most common type of insurance transition was to the uninsured/missing category, which occurred in 16.7% of individuals with commercial insurance and 24.4% of individuals with Medicaid (Table 2). Transitions to other forms of insurance were less common, although 9.5% of patients with OUD and commercial insurance transitioned to Medicaid. Among the 24.4% of individuals with OUD who transitioned from Medicaid to the uninsured/missing category, 6.6% died or were incarcerated.

Estimators of Insurance Transitions

In adjusted logistic regression models, the overall estimated probability of experiencing an insurance transition in the 12 months after OUD diagnosis was 31.3% (95% CI, 30.5%-32.0%) in Medicaid and 27.9% (95% CI, 26.6%-29.1%) in commercial insurance (table 3, eFigure 1). The probability of insurance transitions was generally higher among younger individuals, particularly for those aged 18 to 24 years with Medicaid (39.0%; 95% CI, 37.0%-40.9%) and those aged 26 to 35 years with commercial insurance (37.5%; 95% CI, 34.8%-40.3%). In those with commercial insurance, Black non-Hispanic individuals had the highest probability of insurance transitions at 40.1% (95% CI, 31.8%-48.4%) followed by Hispanic individuals at 33.6% (95% CI, 26.5%-40.8%). For individuals covered by Medicaid, Black non-Hispanic and Hispanic individuals had lower probabilities of insurance transitions than White non-Hispanic individuals (Table 1; eFigure 1). The probability of an insurance transition was lower for individuals who started MOUD at the time of diagnosis when covered by Medicaid (27.0% in those starting vs 32.5% in those not starting MOUD) but not when covered by commercial insurance (28.9% in those starting vs 29.7% in those not starting MOUD).

Sensitivity Analyses

Sensitivity analyses allowing individuals to enter the missing category for 1 month and then return to their prior insurance without counting it as a transition decreased the rate of insurance transitions more for those with Medicaid (23.5% vs 30.7% in main analyses) than for those with commercial insurance (27.4% vs 29.4% in main analyses) (eTable 4 and eFigure 2). In analyses tracking individuals who entered the uninsured/missing category, 74.6% of those with commercial insurance at diagnosis remained uninsured/missing throughout follow-up while 20.5% returned to a commercial plan (eTable 5). For those with Medicaid, 47.7% remained uninsured/missing throughout follow-up, and 49.6% reenrolled in Medicaid.

Using the expanded cohort with data from July 2014 through March 2020 resulted in lower 12-month rates of insurance transitions for patients with Medicaid (28.3%) and higher rates for patients with commercial insurance (34.0%) (eTable 6 and eTable 7). The estimated probabilities of insurance transitions in this expanded cohort showed patterns similar to the main analyses (eTable 8).

In the propensity score–matched cohorts of patients with OUD or T2D, rates of insurance transitions were higher for patients with OUD than T2D, particularly for those with commercial insurance (eTable 9, eTable 10, and eFigure 3). For patients covered by Medicaid at baseline, 28.3% of individuals with OUD experienced an insurance transition compared with 25.3% of individuals with T2D (eTable 10). For patients with commercial insurance at baseline, 26.8% of patients with OUD experienced an insurance transition compared with 17.5% with T2D. The estimated probabilities of insurance transitions by sociodemographic factors also differed by diagnosis (eTable 11).

Discussion

In this study using longitudinal individual-level data, individuals with OUD had high rates of insurance transitions in the 12 months after their diagnosis. Nearly one-third of individuals experienced an insurance transition, with particularly high rates of transition among young individuals and among individuals with commercial insurance who were Black non-Hispanic and Hispanic. Rates were slightly higher in Medicaid compared with commercial insurance although some Medicaid transitions appeared to be brief interruptions with rapid reenrollment.

To our knowledge, this is the first study to investigate insurance instability in individuals with newly diagnosed OUD. While not directly comparable, Nguyen et al. previously found high rates (24.8%) of insurance disenrollment among patients with OUD actively receiving treatment with buprenorphine in an integrated health insurance and delivery system. The rates of insurance transition among patients with OUD appear to be higher than in the general population in Massachusetts, where an estimated 26% of individuals switched insurers over a 2-year period. They are also higher than insurance transition rates for individuals with schizophrenia in Massachusetts, of whom 18.6% have at least 1 insurance transition per year.

We found that insurance transitions for patients with OUD varied considerably by age and race and ethnicity. Younger individuals were most likely to experience insurance transitions whether in commercial insurance or Medicaid. Individuals aged 26 to 35 years with commercial insurance were at particularly high risk, perhaps due to less stable employment. Similar age patterns have been observed in studies of pregnancy and psychosis. In terms of race and ethnicity, Black non-Hispanic individuals with OUD covered by commercial insurance had a notably high rate of insurance transitions at 40.1% followed by Hispanic individuals at 33.6%. This finding is consistent with prior research about racial and ethnic inequities in insurance transitions in the peripartum period. However, differences were less pronounced for patients with OUD covered by Medicaid. Future research should investigate the clinical impact of insurance transitions among individuals with OUD, particularly among racially minoritized individuals who transition from commercial insurance at a substantially higher rate.

Related to treatment for OUD, we found that individuals with Medicaid were more likely to initiate MOUD at diagnosis compared with those with commercial insurance, although rates of treatment were low. Initiating MOUD had no effect on insurance transitions for those with commercial insurance but was associated with a 5.5% reduction in the probability of insurance transition for those receiving Medicaid. This may partly reflect differences between insurers in cost-sharing practices that could affect treatment retention, although findings on this topic are mixed.

Our sensitivity analyses revealed several noteworthy findings. First, when altering our definition for insurance transition to allow for a 1-month lapse in insurance coverage, we found that the rates of insurance transitions decreased considerably for individuals covered by Medicaid but not by commercial insurance. We also found that nearly 50% of individuals with Medicaid who entered the uninsured/missing category reenrolled in Medicaid during the 12-month follow-up period (compared with only 20% of individuals with commercial insurance). This is consistent with known lapses in Medicaid enrollment due to annual eligibility redeterminations. Whether these brief lapses or transitions in insurance interrupt addiction treatment is a question that merits further inquiry.

Second, when we included more contemporary data, the insurance transition rates appeared higher for patients with commercial insurance than for Medicaid. The increase in insurance transitions for those with commercial insurance may be due to the Employee Retirement Income Security Act–related changes in reporting requirements to the APCD. The rate of insurance transitions in Medicaid was lower using more contemporary data (28.3% vs 30.7%), indicating that lapses in Medicaid insurance may have become less common over time.

Third, compared with a propensity score–matched cohort of patients with T2D, individuals with OUD had higher rates of insurance transitions, particularly for those with commercial insurance. The discrepancy in insurance transitions between individuals with OUD and T2D covered by commercial insurance may be associated with changes in employment associated with an OUD diagnosis. While changes in employment and insurance instability have been observed in other chronic conditions such as cancer, employer-based insurance may be particularly difficult to maintain for individuals with OUD. Active OUD may interfere with the ability to perform job functions, and individuals may also face stigma from employers due to drug use and treatment with MOUD.

Limitations

This study has several limitations. First, our analysis focused on insurance categories rather than individual insurance plans. As such, our insurance transition rates may be underestimates, as we would not detect individuals switching between plans within the same category of insurance; in addition, the uninsured/missing category represents a heterogenous group, and the data limited our ability to distinguish insurance loss from other sources of transition, such as moving out of state. Second, we lacked data on important insurance characteristics that likely affect the treatment implications of insurance switching, such as network breadth and medication benefits. Third, our data are from a single state that expanded Medicaid eligibility and has high rates of insurance coverage, which may not generalize to other states. Fourth, while we can track longitudinal transitions in insurance, we do not assess whether these transitions affect treatment episodes, including disrupting or expanding access to MOUD. Fifth, we used a 180-day lookback period to define incident OUD, which may have captured some prevalent cases for individuals with limited contact with the medical system.

Conclusions

In this longitudinal analysis of patients with newly diagnosed OUD, high rates of insurance transitions in the year after diagnosis, with differences by insurance type and by age and race and ethnicity were found. Given that insurance transitions may result in changes or loss of insurance coverage for highly efficacious medications for OUD, tracking changes in treatment outcomes related to insurance instability in this population is needed.

Summary

A retrospective cohort study utilizing data from the Massachusetts Public Health Data Warehouse examined insurance transitions among individuals with newly diagnosed opioid use disorder (OUD) within the first year post-diagnosis. The study population comprised Massachusetts residents aged 18-63 with incident OUD and Medicaid or commercial insurance. Researchers analyzed the cumulative incidence of insurance transitions, identifying associated sociodemographic factors and comparing findings to a propensity score-matched cohort with type 2 diabetes.

Study Design and Data Source

This retrospective cohort study leveraged data from the Massachusetts Public Health Data Warehouse, integrating information from multiple sources including the All-Payer Claims Database (APCD). Data linkage employed probabilistic methods, and privacy protections were implemented. The study adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.

Reporting of Race and Ethnicity

Race and ethnicity categorization followed a hierarchical approach based on data source reliability, prioritizing self-reported information. In cases of multiple reported races, a predefined algorithm determined the assigned category.

Cohort Selection and Outcomes

The study cohort included adults aged 18-63 with incident OUD diagnoses between July 1, 2014, and December 31, 2014, with follow-up through December 31, 2015. The primary outcome was the incidence of insurance transitions within 12 months of diagnosis, categorized by type. Secondary outcomes included characterization of the study cohort via sociodemographic data, prior healthcare utilization, and MOUD initiation.

Statistical Analysis

Descriptive statistics, Kaplan-Meier estimation, and logistic regression modeling were used to analyze the data. Logistic regression models incorporated a priori selected variables to assess associations between patient characteristics and insurance transitions. Several sensitivity analyses were conducted to address potential biases and refine the interpretations.

Baseline Characteristics and Incidence of Insurance Transitions

The cohort (n=20,768) comprised predominantly male individuals, with a majority insured by Medicaid. Significant disparities in baseline characteristics were observed between Medicaid and commercially insured individuals. Approximately 30% of the cohort experienced an insurance transition within the first year, with the uninsured/missing category representing the most frequent transition type. The cumulative incidence of transitions increased sharply within the first six months post-diagnosis.

Estimators of Insurance Transitions and Sensitivity Analyses

Adjusted logistic regression analyses revealed higher probabilities of insurance transitions among younger individuals and specific racial/ethnic groups, particularly within the commercially insured population. Sensitivity analyses modifying the definition of insurance transitions, expanding the timeframe, and comparing to a T2D cohort provided further insights into the patterns and potential underlying factors.

Discussion and Limitations

The study findings highlight substantial insurance instability among individuals with OUD, exceeding rates observed in other populations and potentially impacting access to MOUD. Significant disparities by age and race/ethnicity were noted. The limitations of the study include the potential underestimation of transitions, the heterogeneity of the uninsured/missing category, and the state-specific nature of the data.

Conclusions

High rates of insurance transitions were observed in the year following OUD diagnosis, underscoring the need for further research into the impact of insurance instability on treatment outcomes and access to effective OUD medications.

Summary

This study investigated insurance transitions among Massachusetts residents with newly diagnosed opioid use disorder (OUD) within the first year post-diagnosis. Researchers used data from the Massachusetts Public Health Data Warehouse, focusing on individuals covered by Medicaid or commercial insurance. The study examined the cumulative incidence of transitions and explored sociodemographic factors influencing these changes. Comparisons were made to a cohort of individuals with type 2 diabetes (T2D) to contextualize findings.

Methods

The study employed a retrospective cohort design using data from the Massachusetts Public Health Data Warehouse. This warehouse links various administrative records, including claims data, hospital records, and prescription monitoring program data. The cohort included adults (18-63 years) with incident OUD diagnoses between July 1, 2014, and December 31, 2014, followed through December 31, 2015. The primary outcome was the incidence of insurance transitions within 12 months of diagnosis. Logistic regression models were used to analyze factors associated with transitions, controlling for age, race/ethnicity, pregnancy status, and initiation of medication for OUD (MOUD). Sensitivity analyses were conducted to address potential limitations, including temporary insurance lapses and changes in APCD reporting requirements. A propensity score-matched comparison group with T2D was included.

Results

The study included 20,768 individuals. Approximately 30% experienced an insurance transition within 12 months of OUD diagnosis. Transitions to the uninsured/missing category were most frequent. Adjusted analyses revealed higher transition probabilities among younger individuals and disparities across racial and ethnic groups, particularly among those with commercial insurance. Initiating MOUD was associated with a reduced probability of transition only among Medicaid recipients. Sensitivity analyses indicated that temporary insurance lapses, particularly within Medicaid, significantly impacted the observed transition rates. Comparison with the T2D cohort showed higher transition rates among those with OUD, especially with commercial insurance.

Discussion

This study highlights substantial insurance instability among individuals with OUD, exceeding rates observed in other chronic conditions. Disparities in transition rates based on age, race/ethnicity, and insurance type were identified. The findings suggest the need for interventions to improve insurance stability and continuity of care for OUD patients, particularly among vulnerable populations. Further research is warranted to understand the clinical impact of these transitions on treatment outcomes.

Limitations

Several limitations were acknowledged. The study focused on insurance categories rather than specific plans, potentially underestimating transitions. The heterogeneous "uninsured/missing" category limited detailed analysis. Data were from a single state and may not generalize. The impact of transitions on treatment episodes was not directly assessed. Finally, the 180-day lookback for incident OUD might have included some prevalent cases.

Conclusions

This study demonstrates substantial insurance instability in individuals with newly diagnosed OUD. The observed disparities highlight the need for policies and interventions to ensure continuous insurance coverage and access to effective treatment for this population. Further research on the clinical implications of these transitions is necessary.

Summary

This study examined insurance transitions among individuals in Massachusetts newly diagnosed with opioid use disorder (OUD) during the year following their diagnosis. Researchers used data from the Massachusetts Public Health Data Warehouse, focusing on those with Medicaid or commercial insurance. The study aimed to determine the frequency of insurance changes, identify factors associated with these transitions, and compare these findings to a similar group with type 2 diabetes.

Methods

The study used a retrospective cohort design analyzing data from various Massachusetts databases. Researchers identified a cohort of adults (18-63 years) with a first-time OUD diagnosis between July 1, 2014, and December 31, 2014. The primary outcome was the incidence of insurance transitions within 12 months of diagnosis. Statistical analysis included Kaplan-Meier estimation and logistic regression to identify factors associated with insurance transitions. Sensitivity analyses addressed potential limitations of the data.

Results

Approximately 30% of individuals with OUD experienced an insurance transition within the first year. The most frequent transition was to the "uninsured/missing" category. Younger individuals and those identifying as Black or Hispanic (with commercial insurance) had higher transition probabilities. Starting medication for OUD (MOUD) was associated with a reduced likelihood of transition, particularly among Medicaid recipients. Sensitivity analyses showed that many Medicaid transitions were short-term interruptions. Compared to a similar group with type 2 diabetes, the OUD group had higher rates of insurance transitions.

Discussion

The study highlights the significant insurance instability experienced by individuals with OUD, particularly among younger individuals and those from minority groups with commercial insurance. These transitions could negatively impact access to vital MOUD. The high rate of transitions among those with OUD suggests a need for interventions to improve insurance stability and access to consistent care. Further research should examine the clinical implications of insurance transitions on treatment outcomes.

Limitations

The study's limitations include the inability to distinguish between different plans within the same insurance category, a diverse "uninsured/missing" category, limited data on insurance plan details, a focus on a single state, and the lack of data on treatment disruptions caused by insurance transitions. Additionally, the definition of incident OUD might include some prevalent cases.

Summary

Health insurance is very important for getting good healthcare. People with opioid use disorder (OUD) need insurance to get the medicine and treatment they need to recover. This study looked at how often people with newly diagnosed OUD changed their health insurance in the first year after their diagnosis.

Study Design and Data

Researchers used information from Massachusetts' health records. They looked at adults (ages 18-63) diagnosed with OUD who had Medicaid or private insurance. They tracked how often these people switched insurance plans or lost their insurance during the year after their diagnosis.

Who Was Studied?

The study included over 20,000 people newly diagnosed with OUD. Most were men, and most had Medicaid. Researchers also looked at things like age, race, and whether people started getting medication for OUD.

What Happened?

About 30% of people changed their insurance within a year. Many people ended up with no insurance for a while. Younger people and some racial groups changed insurance more often. Starting medication for OUD sometimes helped keep people's insurance.

What the Study Means

Losing or changing health insurance is a big problem for people with OUD. It makes it harder to get the medicine that helps them get better. This study shows that it's important to make sure people can keep their insurance, especially younger people and those from some racial groups.