Did Medicare Coverage for Methadone Increase Treatment for Opioid Use Disorder?

Agniel et al. (2026) asked whether Medicare’s 2020 decision to cover methadone treatment for opioid use disorder changed how opioid treatment programs operated and whether it expanded access to care. They analyzed national administrative data on treatment facilities (2015–2023) and treatment episodes using difference-in-differences methods. They found that opioid treatment programs were 45.4 percentage points more likely to accept Medicare after the policy, nearly doubling participation. Methadone treatment episodes among Medicare-age adults also roughly doubled. They also found spillover effects: programs became more likely to accept commercial insurance and Medicaid, offered more charity care, and changed the mix of services they provided.

Why This Article Was Selected for The Policy Scientist

Opioid use disorder remains one of the most consequential public health challenges in the United States, making policies that improve access to effective treatment an enduring policy concern. As the Medicare population grows and opioid-related mortality among older adults continues to rise, understanding how insurance coverage influences treatment delivery has become increasingly important. Agniel et al. (2026) have published extensively on substance use disorder treatment, health insurance, and provider behavior, and this study extends that body of work by examining both direct and spillover effects of Medicare’s methadone coverage expansion. Published in the Journal of Policy Analysis and Management, one of the leading journals in public policy, the article makes a substantial contribution by employing a rigorous difference-in-differences causal inference design with high-quality national administrative data. The findings are likely relevant to other jurisdictions where public insurance expansions influence provider participation and treatment markets, although institutional differences should be considered.

Full Citation and Link to Article

Agniel, D., Cantor, J., Maclean, J. C., Taylor, E. A., & Simon, K. (2026). Insurance coverage and provision of opioid disorder treatment: Evidence from Medicare. Journal of Policy Analysis and Management, 45, e70106. https://doi.org/10.1002/pam.70106

Central Research Question

This study examines whether Medicare’s 2020 decision to cover methadone treatment for opioid use disorder (OUD) changed the behavior of opioid treatment programs (OTPs) and expanded access to care for older adults. Before 2020, Medicare generally did not reimburse methadone treatment for OUD, despite methadone having long been an approved and effective medication. The SUPPORT Act created a new Medicare benefit that reimbursed OTPs for providing methadone and related services. The authors investigate whether this policy increased Medicare participation among OTPs, expanded treatment availability, affected the number of patients receiving care, altered the services offered by treatment facilities, and generated spillover effects into other insurance markets. They also examine whether the policy affected treatment capacity and whether impacts differed across patient and community characteristics. Rather than focusing solely on Medicare beneficiaries, the study evaluates how a major insurance expansion influenced the broader market for opioid treatment services.

Previous Literature

The authors build on two related bodies of research. The first examines the effectiveness of medications for opioid use disorder. Previous medical research consistently demonstrates that methadone, buprenorphine, and naltrexone are effective treatments, although individual patients may benefit more from one medication than another. Methadone is particularly important because it has high treatment retention rates and remains the preferred therapy for many patients, especially those using fentanyl. However, methadone can only be dispensed through certified opioid treatment programs, making provider availability an important determinant of access to care.

The second literature examines how insurance policies influence healthcare providers. Previous studies have shown that changes in Medicare and Medicaid reimbursement can affect provider participation, practice patterns, and treatment availability. Researchers have also documented spillover effects in multi-payer healthcare markets, where changes affecting one insurer influence services provided to patients with other forms of insurance. Within substance use disorder treatment, earlier studies found that insurance expansions sometimes increased provider participation across multiple insurance markets, although evidence has been mixed. The existing literature on Medicare’s methadone benefit has largely been descriptive, documenting changes in billing or dispensing without identifying causal effects. The authors position their study as the first to estimate the causal effects of Medicare’s methadone coverage on provider behavior, treatment utilization, service offerings, and spillovers across insurance markets.

Data

The analysis combines two national administrative datasets covering the years 2015 through 2023. The first is the Mental Health and Addiction Treatment Tracking Repository (MATTR), which contains information on more than 5,700 substance use disorder treatment facilities that continuously reported throughout the study period. MATTR includes information on whether facilities operate as opioid treatment programs, which insurance plans they accept, whether they provide charity care, ownership characteristics, and the treatment services they offer. These data allow the authors to examine how provider behavior changed following Medicare’s policy expansion.

The second dataset is the Treatment Episode Dataset (TEDS), which contains approximately two million substance use disorder treatment episodes annually. Using patient age, treatment setting, medications, and diagnosis information, the authors identify treatment episodes likely involving methadone therapy for opioid use disorder. They compare treatment episodes among adults aged 65 years and older, who became eligible for the Medicare benefit, with adults aged 55 to 64 years, who serve as a comparison group. Combining facility-level and patient-level administrative data enables the authors to study both provider responses and changes in treatment utilization. Although each dataset has limitations, together they provide unusually comprehensive national evidence on the policy’s effects.

Methods

The study employs a difference-in-differences research design, one of the strongest quasi-experimental approaches for estimating causal policy effects when randomized experiments are infeasible. The authors compare outcomes before and after the 2020 Medicare policy change between affected and unaffected groups. In the facility analysis, opioid treatment programs serve as the treatment group, while other substance use disorder treatment facilities that were not directly affected by the policy serve as the comparison group. In the patient-level analysis, adults aged 65 years and older are compared with adults aged 55 to 64 years. Both models include fixed effects that account for stable differences across facilities or states and for national trends over time.

Recognizing that the Medicare policy took effect during the COVID-19 pandemic, the authors devote substantial attention to addressing this potential source of bias. They estimate event-study models to evaluate pre-policy trends, conduct robustness analyses excluding 2020, control for state-level pandemic severity and unemployment, account for stay-at-home orders and telemedicine reimbursement policies, and examine sensitivity to possible violations of the parallel trends assumption. These analyses strengthen confidence that the estimated effects reflect the Medicare policy rather than pandemic-related disruptions. The methodological approach therefore represents a rigorous application of modern causal inference methods and substantially improves upon descriptive analyses that cannot isolate policy effects.

Findings/Size Effects

The policy substantially increased Medicare participation among opioid treatment programs. Following implementation of Medicare coverage, OTPs became 45.4 percentage points more likely to accept Medicare, representing nearly a doubling in Medicare participation. This finding indicates that the policy achieved its immediate objective of expanding Medicare beneficiaries’ access to methadone treatment.

The authors also identify important spillover effects extending beyond Medicare patients. Following the policy change, opioid treatment programs became more likely to accept commercial insurance and Medicaid and were more likely to provide charity care. At the same time, facilities became less likely to accept self-paying patients. These results suggest that changes in Medicare reimbursement influenced broader organizational decisions regarding insurance participation and treatment financing.

Provider behavior changed in additional ways. OTPs reduced the range of services they offered, including lower provision of buprenorphine and certain other treatment services. The authors suggest that these changes may reflect capacity constraints or shifts toward services emphasized under Medicare’s bundled payment system. They also report evidence that patients experienced longer waiting times following the policy change, consistent with increased demand for treatment.

At the patient level, treatment episodes among adults aged 65 years and older approximately doubled following implementation of Medicare coverage. However, most of this increase reflected greater treatment among patients already receiving services rather than entirely new patients entering treatment. This finding suggests that insurance coverage increased treatment intensity and continuity rather than substantially expanding the number of individuals initiating care. The authors conclude that increased demand likely tightened existing capacity constraints within opioid treatment programs.

Collectively, these findings indicate that insurance expansions influence both the quantity of care delivered and the organizational behavior of treatment providers. Rather than producing isolated effects for Medicare beneficiaries, the policy reshaped multiple aspects of the opioid treatment market.

Conclusion

The study provides rigorous causal evidence that Medicare’s decision to cover methadone treatment substantially changed the operation of opioid treatment programs and expanded access to treatment for older adults. The policy nearly doubled Medicare participation among providers and approximately doubled treatment episodes among Medicare-age patients. At the same time, it generated spillover effects across other insurance markets, altered facility service offerings, and revealed evidence of capacity constraints within the treatment system. These broader market responses illustrate that insurance policies can affect healthcare delivery well beyond the populations they directly target.

The article makes an important contribution by moving beyond descriptive analyses to estimate causal effects using a carefully implemented difference-in-differences design supported by extensive robustness testing. The combination of national administrative datasets and modern causal inference methods provides strong evidence regarding how provider organizations respond to insurance expansions. More broadly, the findings demonstrate that changes in public insurance reimbursement can reshape provider behavior across an entire healthcare market, offering insights that are likely relevant to other insurance expansions and specialized treatment systems in which providers serve patients covered by multiple payers.