Tyler
The year after its inception in 1966, Medicaid spending was less than $1 billion. By 2023—57 years later—that number had increased to $871.7 billion, accounting for a whopping 18% of total national health expenditure in 2023. This substantial increase in cost can be attributed, in large part, to the gradual expansion of federal and state Medicaid policies.
Although Medicaid coverage for low-income adults has steadily expanded, few studies meaningfully examined the impact of enrollment on patient outcomes until the Oregon Health Insurance Experiment (OHIE). The OHIE, a natural randomized controlled trial (RCT) that began in 2008, was quickly recognized as a pivotal study in assessing the effectiveness of Medicaid coverage.
Background
The Oregon Health Plan Standard is the Medicaid program for all low-income, uninsured, able-bodied adults who aren’t eligible for other public insurance in Oregon. Primarily delivered through managed care organizations, the plan provides comprehensive medical benefits (including prescription drugs) with no patient cost-sharing and low monthly premiums. To qualify, individuals were required to be Oregon residents aged 19 to 64, U.S. citizens or legal immigrants, uninsured for at least six months, ineligible for other public insurance, as well as receiving incomes below 100% of the federal poverty line and holding assets of less than $2,000 in value.
The program’s enrollment was closed in 2004 but reopened as a waiting list in 2008, quickly garnering over 90,000 applications and far surpassing the number of available spots. The state conducted eight random lottery drawings that year to expand coverage to over 30,000 individuals. Those who were selected won the opportunity—for themselves and any householder member—to apply for the Oregon Health Plan Standard.
Compared to quasi-experimental designs of the past, Oregon’s lottery drawings presented researchers with a unique opportunity to track health outcomes of participants enrolled in Medicaid using a randomized controlled trial (RCT) format, where participants were randomly assigned to an experimental group or a control group.
The study compared in-person interviews of 2 groups of adults—6,387 who were randomly selected to receive plan coverage (the experimental group) and 5,842 who were not (the control group). These interviews occurred, on average, 25 months from the date of the first lottery drawing. Because Medicaid is a social program with pre-determined eligibility requirements, opportunities to evaluate its impact in a controlled, RCT setting are rarely available. Measuring experimental outcomes using this type of study design reduces bias and provides a rigorous tool to examine cause-effect relationships.
The goal of the Oregon Health Insurance Experiment was to evaluate the effects of expanded Medicaid coverage for low-income adults in Oregon. The measures utilized to evaluate the health of the population included blood-pressure, cholesterol, and glycated hemoglobin levels; screenings for depression; medication inventories; and self-reported diagnoses of health status, health care utilization, and out-of-pocket spending for such services.
Results
Results from the Oregon Health Insurance Experiment were mixed. The study found no effect of Medicaid coverage on the diagnosis or use of medication for hypertension or high cholesterol. There was, however, a greater probability of receiving a diagnosis of diabetes (3.83 percentage points) and using medications for diabetes (5.84 percentage points) for those covered by Medicaid.
Screenings for depression utilized the Patient Health Questionnaire-8 (PHQ-8) where those classified as depressed received a score of 10 or greater. Although there was no significant increase in the use of antidepressant medication, Medicaid coverage resulted in a decrease in the rate of depression by 9.15 percentage points and an increase in the probability of receiving a diagnosis after the lottery of 3.81 percentage points.
Medicaid coverage led to a 7.84 percentage point increase in the proportion of individuals who reported that their health was the same or better as compared to their health one year prior. Despite this improvement, the study found no meaningful improvements across physical- or mental-component scores using the Medical Outcomes Study 8-Item Short-Form Healthy Survey (SF-8). Stated differences in quality of life for self-reported levels of pain and happiness also saw no significant change.
The study signaled a reduction of financial strain for medical costs, according to a number of self-reported measures. Catastrophic expenditures, defined as out-of-pocket expenses exceeding 30% of income, were substantially reduced with a relative decrease of more than 80% compared to the control group. This was an expected outcome given lack of patient cost-sharing and low monthly premiums that come standard with Medicaid programs.
While changes to the number of hospital admissions and emergency room visits were insignificant, Medicaid coverage did result in an increase in the number of office visits and prescription drugs received in the previous year. As a result, it was estimated that annual medical spending increased by about 35% compared to the control group. The study also found increases in some preventive care and screening services, as well as improvement to the perceived access to care. Medicaid coverage did not correspond to a change in the likelihood of being a smoker or obese.
Considerations
Before diving into the key takeaways from the study, it is important to note several limitations to the generalizability of the findings:
Discussion
The largest takeaway from the OHIE: health insurance cannot operate as the sole mechanism for change. Simply providing access to care is not enough. The study found that access increased levels of utilization, but measured changes in key health markers were minimal. Utilization alone is insufficient—improving patient outcomes should be the primary focus of any social program.
While some may argue that the financial security derived from Medicaid is a leading indicator of its success, increased utilization does not lower overall spending. Rather, the financial burden tends to shift to taxpayers of the state who help fund these programs.
Excessive spending without meaningful improvements in patient outcomes is an obvious issue—one that likely results from the traditional, reactive approach to providing Medicaid services. Instead of only treating patients when urgent health needs arise, states must focus on being more proactive in the care they provide. Options for policymakers to consider include pairing Medicaid coverage with care coordination programs, patient education, and incentives for preventive care that drives long-term health improvements. States may also consider piloting value-based care programs that link payments to outcomes, rather than simply incentivizing greater utilization.
Conclusion
The OHIE was highly relevant in the context of major changes to the health policy landscape from 2008 to 2012, coinciding with the major expansion of Medicaid enrollment through the Affordable Care Act (ACA) of 2010. Despite fluctuations in enrollment efforts since the ACA, policymakers can still leverage various aspects of the study’s findings to craft targeted Medicaid expansions that balance cost, access, and outcomes for low-income adults.
Overall, the Oregon Health Insurance Experiment demonstrated the effects of expanding healthcare benefits to low-income adults on a randomized basis, a phenomenon rarely seen in the context of social insurance research. This study reinforced the idea that insurance is necessary but inadequate by itself—true healthcare improvement requires a combination of coverage, education, incentives, and proper care delivery. U.S. policymakers should leverage the results of research like this to strengthen plan design and improve healthcare utilization practices for their respective patient populations.
CMS HCC-28 is vital for Medicare Advantage risk adjustment, aligning payments with enrollee health needs. How does this impactful model refine patient data and shape today’s healthcare landscape? Learn more about the model’s updated framework and its impact on accurate reimbursement.
How has the 340B Drug Discount Program influenced healthcare? In what ways has a program built for the poor come to serve the rich, and what does that mean for patient care quality? Learn about the controversial history of 340B, how it has shaped the American healthcare system, and the legal challenges it faces.
What happens to the demand for healthcare goods when costs change? The RAND Health Insurance Experiment was a groundbreaking study that attempted to address this question and more, and its results still shape healthy policy today. Learn more about why this study still remains a game-changer more than 40 years later.