This review indicates that African American disparities in health access, treatment, or health outcomes—with the important exception of maternal mortality rates—remain largely unchanged by Medicaid expansion. However, whether Medicaid eligibility expansion reduced African American-white health coverage disparities remains an open question: Absolute disparities in coverage appear to have declined in expansion states, although exceptions have been reported. Future research addressing key weaknesses or oversights in existing research may help to uncover sources of continuing disparities and clarify the impact of Medicaid expansion on changes in health coverage disparities.
Improving research precision and rigor
Improved research efforts can clarify the answer to this question—and identify structural sources of continuing disparities—by more carefully targeting Medicaid eligibility expansion as a source of disparity reduction and accounting for, or specifically examining, the role of variation in broader ACA-related health system changes. Further studies should examine changes in relative health disparities as well as absolute health disparities and must examine disparity changes for African Americans separately from disparity changes for other racial or ethnic groups. Deeper investigations of African American-white disparity reductions in healthcare access, treatment, and health outcomes—which appear to be relatively unchanged by Medicaid expansion—should consider community and provider-level treatment contexts that may impact African Americans especially and have sometimes been impacted by the ACA’s health reforms.
Testing the triple interaction. To test disparity reduction directly, studies need to document significant reductions in the differences between 1) African American and whites’ coverage, access, utilization, and health outcome rates, 2) before and after Medicaid expansion, 3) in expansion versus non-expansion states. Only 11 out of 26 studies tested for the significance of all three differences. Of these studies, no study examining the general population found significant disparity reductions in health coverage, treatment, access, or health outcomes associated with Medicaid expansion. However, coverage disparity reductions were found for young adults and patients with acute myocardial infarctions. Less than half of the sample studies used a full triple difference analysis, and to overcome present uncertainty, investigators must routinely comply with this requirement.
Absolute versus relative disparities. Relative disparities target equity, and only one study examined relative disparities (14). The investigators found that African American-white relative disparities were not significantly changed from 2013 to 2014 in Medicaid expansion states but were significantly reduced in non-expansion states. Absolute disparities can close when African Americans’ rates begin far enough away from whites’ that larger absolute gains are required merely for African Americans’ gains to keep pace with whites’ gains (39).
Disaggregating Medicaid expansion from other ACA elements. The ACA ushered in many innovations apart from the Medicaid eligibility expansion. The Medicaid application process was streamlined as online filing options increased and verification and certification procedures capitalized on new technologies (40). Individuals with incomes between 100% and 400% FPL became eligible for “Premium Tax Credits” on a sliding scale to purchase private, non-group coverage through state or federally-operated healthcare exchanges (41), and persons with incomes between 100–250% FPL became eligible for cost-sharing subsidies. Gains were concentrated among those with incomes between 138–250% of the FPL—those who were eligible for the ACA’s cost-sharing reductions and among whom African Americans are also over-represented (42, 43). In non-expansion states, premium tax credits and subsidies could offset denial of access to expanded Medicaid for persons with incomes above 100% FPL.
Marketplaces, which informed inquiring persons about Medicaid eligibility, actively sought enrollees through vigorous outreach efforts. Community targeted advertising raised awareness, and marketplaces provided individual counseling on eligibility and options, sometimes facilitated by culturally sensitive enrollment assistors (44). Safety net hospitals faced new incentives to avoid hospital readmission and reduce lengths of stay by shifting newly eligible patients to Medicaid-funded outpatient care (44). Funding was increased for new Federally Qualified Health Centers, which disproportionately support African Americans through targeting services for the poor (45). These and other developments promised to reduce barriers to coverage and access for non-white, low-income adults—lessening healthcare disparities throughout the United States as many previously eligible people become aware of Medicaid eligibility and enrolled (“woodwork effect”) (46). New research must examine the impacts of ACA policy elements on disparities in specific types of health insurance coverage rather than on the all-inclusive “un-insurance.”
Advancing knowledge: Beyond Medicaid expansion’s eligibility requirements
Additional advances in research should examine variation in state implementation of Medicaid expansion. This includes attention to the role the Section 1115 Medicaid waivers have played in expanding Medicaid eligibility—both before and after the ACA’s implementation—and the extent to which changes in health coverage disparities are attributable to enhanced awareness of health coverage possibilities resulting from vigorous outreach and health coverage enrollment efforts in both expansion and non-expansion states.
1115 Medicaid Waivers. Medicaid 1115 waivers were issued to 14 states between 2004 and 2012 for early Medicaid expansion, and, in some states, early expansion significantly affected coverage rates (47). Two studies excluded these states from consideration (19, 25), but others failed to account for the possible pre-ACA reduction in coverage increase and disparity. Investigators may have underestimated ACA expansion’s impact on disparities by neglecting early expansion. Medicaid waivers played a dual role in Medicaid eligibility expansion. In addition to the 1115 waivers approved prior to the ACA Medicaid expansion, after the ACA Medicaid eligibility expansion, several initially rejecting states expanded Medicaid eligibility through Section 1115 waivers.
These states used these waivers to customize eligibility standards to accommodate better ideological and fiscal reservations (48). Some states expanded Medicaid with restrictions—requiring premium payment to begin coverage, using health savings accounts, tying healthy activities to waived premiums, or including work requirements. These are complex to implement and present grave administrative challenges (49, 50), reducing uptake of Medicaid coverage (51). Arkansas’ coverage gains did not differ in gains from traditional Medicaid expansion (52), but Arkansas’ addition of work requirements in June 2018 resulted in thousands losing coverage—reportedly due to administrative complexity (53). African Americans have experienced race-related aversive experiences with bureaucratic programs (54), and waiver-imposed barriers may deter African Americans especially.
More research is needed to identify the impact of waivers on disparities. This knowledge is critical to informing future approvals for state maneuvers to expand Medicaid conditionally or partially through these policies. Currently, 63 waivers have been approved across 45 states, and 28 applications in 22 states are currently pending decisions from the Centers for Medicaid and Medicare Services (CMS) (55).
Outreach and Enrollment Assistance. Disparity reduction in non-expansion states points to the possibility that some states reduced enrollment barriers for African Americans especially. Advertising, enrollment assistance, and greater enrollment incentives for FQHCs and safety net hospitals to maximize enrollment likely increased Medicaid uptake. Conceivably, previously uninsured African Americans who were eligible for Medicaid prior to Medicaid expansion disproportionately responded to ACA messages about coverage possibilities, were less deterred by burdensome enrollment procedures due to streamlining efforts under the ACA or were disproportionately gaining enrollment through newly available Federally Qualified Health Centers or in safety-net hospitals as they encouraged covered outpatient care.
Populations with Chronic or Critical Conditions
Among eight studies focusing on populations with specific illnesses, one study found Medicaid expansion to be associated with African American-white disparity reduction in coverage (31). However, none of these studies report significant reductions in disparities in access to treatment, survival rates, or health outcomes. Coverage disparity reductions in populations with critical or chronic conditions, and associated changes in access to care, must be considered considering the presence of strong incentives to find insurance coverage for costly medical procedures. Providers are motivated to facilitate enrollment to avoid the burden of uncompensated care—the very “adverse selection” that concerns insurers and necessitated the ACA’s requirement that persons with pre-existing conditions not be denied coverage (56). Opportunities for gaining coverage are likely more available in expansion states, and thus coverage disparity reductions observed under strong incentives to enroll must be understood on their own terms and may not be generalized to the wider population.
Access, Treatment, and Health Outcomes
This review also highlights that there is limited evidence supporting the expectation that disparity reductions in coverage translated into disparity reductions in access, utilization, or health outcomes. Refinements are needed to determine better whether such reductions occurred and how. Studies assessing access and treatment utilization should consider other non-cost-related barriers to healthcare access—including barriers that may impact African Americans especially. Size and location of provider supply, program outreach and cultural responsiveness, and other determinants of receiving care may be relevant. An expansion of Community Health Centers funded by an ACA-created trust fund, where African Americans disproportionately are treated, is particularly ripe for study as an ACA-related trigger for change in provider supply. Focusing directly on access and treatment disparities is indicated, taking us beyond inconclusive findings from present approaches measuring only the onset of the ACA and its immediate impacts on coverage disparities.
Examinations of the ACA’s impact on disparities in health outcomes and health status—which may result from higher health insurance rates but will likely take longer to emerge—should also be examined in the coming decade. Due to the impact of a wide range of social determinants upon health—and the disproportionate exposure of African Americans to determinants that negatively impact health status and health outcomes (57)—the impacts of the ACA on health outcomes will be complex to untangle and likely more difficult to detect.