The Projected Effects of the Arkansas Medicaid Work Requirement Demonstration on Community Health Centers

In June 2018, Arkansas became the first state in the nation to implement a Medicaid work demonstration. Two recent blogs examined the enrollment and economic[1] impact of the work demonstration. A companion piece to this blog examining the demonstration’s initial impact on beneficiaries subject to its work/community engagement and reporting rules estimates that between 19 percent and 30 percent of current beneficiaries subject to the requirements – or 30,700 to 48,300 people – will lose coverage over the course of 12 months.[2] Because Arkansas’ policy “locks out” those who fail to comply with the rules for three months for the remainder of the calendar year, it will inevitably lengthen coverage gaps during which people are uninsured.This analysis assesses the consequences of this widespread loss of coverage on community health centers, which depend extensively on Medicaid revenue. We conclude that coverage losses of this magnitude will result in a loss to the state’s community health centers of $1.5 to $2.3 million in Medicaid revenue. This revenue loss translates into a drop in patient care capacity of between 1,811 and 2,859 patients and a reduction of 6,827 to 10,779 patient visits.Background This assessment of the impact of large-scale Medicaid coverage losses is designed to consider the downstream consequences of loss of coverage on health care capacity in low-income, medically underserved urban and rural communities, where health care providers depend extensively on Medicaid to insure their patients and provide essential operating revenue. This analysis focuses on community health centers in Arkansas, which provide comprehensive primary health care in medically underserved communities throughout the state and whose patients rely heavily on Medicaid and have benefitted from the expansion.[3] Two considerations cause us to focus on community health centers. First, as in all states, Arkansas health centers rely on Medicaid as an important source of operating revenue and represent a major source of health care for Medicaid patients. In 2017, the state’s health centers cared for one in 11 Medicaid/CHIP enrollees.[4]Second, the Special Terms and Conditions (STCs) established by the federal Centers for Medicare and Medicaid Services (CMS) and that govern the Arkansas work demonstration expressly point to health centers as a source of care for people losing coverage under the demonstration. The STCs specifically instruct the state to inform beneficiaries losing coverage as a result of the new rules that they can obtain free or low-cost primary and preventive care from community health centers.[5] In other words, CMS not only assumes widespread coverage losses, but further assumes that health centers will be able to absorb these losses and will continue to provide care at current capacity. CMS does not indicate what evidence it relies on for these assumptions. The question is whether its assumptions are sound.Arkansas community health centers According to data from the Uniform Data System (UDS) – the official federal health center reporting system maintained by the Health Resources and Services Administration (HRSA) within the United States Department of Health and Human Services – twelve federally-funded community health centers served 210,380 Arkansas patients in 2017.[6] This figure includes one in 14 state residents [7] and one in six low-income residents that year.[8] Over the past five years, the number of sites has increased by about one-third and the number of patients by 46,583, or 28 percent (Table 1). In the wake of Arkansas’ Medicaid expansion, the number of uninsured health center patients declined dramatically -- by 41 percent over the 2013-2017 time period.Arkansas elected to expand Medicaid under the Affordable Care Act (ACA) through a so-called “private option” model. Under this option, most newly eligible enrollees (excluding those classified as medically frail because of greater medical needs, who remain enrolled in traditional Medicaid) are covered through the same private health insurance plans sold in the Marketplace that are available to residents whose subsidies are derived from advance premium tax credits.[9] Thus the same qualified health plans that cover subsidized residents with incomes above 138 percent of the federal poverty level also enroll Medicaid beneficiaries who are part of the expansion population and are not medically frail.Community health centers serve all residents of their service areas and participate extensively in health plan provider networks. As a result, they enroll both publicly and privately insured community residents, particularly low-income residents whose coverage is subsidized either by Medicaid or premium tax credits.Table 1 shows that between 2013 and 2017, the number of Medicaid and privately insured patients served by Arkansas’ community health centers increased by 89 and 71 percent, respectively. Following implementation of the Medicaid expansion, the number of adult Medicaid patients increased by 221 percent (note that the UDS classifies Medicaid patients enrolled in private insurance plans as covered by Medicaid).[10]Table 1. Arkansas health centers’ sites, patients, and health insurance, 2013-2017Source: GW analysis of 2013-2017 HRSA UDS dataFigure 1 compares the health insurance status of Arkansas health center patients in 2013 and 2017. The share of uninsured patients fell from 40 percent to 18 percent, while the share of privately insured patients rose from 20 percent to 27 percent.  The share of Medicaid-covered children remained the same; in contrast, the share of Medicaid-covered adults ages 18 and older nearly tripled, from eight percent in 2013 to 21 percent in 2017.Figure 1. Health insurance coverage of Arkansas health center patients, 2013 and 2017Source: GW analysis of 2013 and 2017 HRSA UDS dataTable 2 shows changes in Arkansas health centers’ staffing, visits, and revenue over the 2013-2017 time period. As Medicaid and private insurance revenue increased, health centers were able to increase their full-time equivalent (FTE) medical and mental health staff by 49 percent and 223 percent, respectively. Correspondingly, the number of medical and mental health visits increased by 45 and 106 percent, respectively. Total revenue increased over this time period from $111 million to $169 million, with Medicaid revenue increasing by 86 percent, and private insurance revenue by 307 percent. Growth in both Medicaid and private insurance revenue eclipsed that of federal health center grant revenue (received by all federally-funded health centers), which increased by 47 percent over the same time period.Table 2. Arkansas health centers: staff, visits, and revenue, 2013-2017Note: Other revenue includes other public insurance, other federal grants, non-federal grants, and other revenue. Source: GW analysis of 2013-2017 UDS data, HRSA.Source: GW analysis of 2013 and 2017 HRSA UDS dataFigure 2 illustrates how health center revenue sources have shifted over time. While the share of revenue from federal Bureau of Primary Health Care (BPHC) grants stayed stable from 2013 to 2017, the share from Medicaid increased from 24 percent to 29 percent, and the share from private insurance increased from five percent to 14 percent.Figure 2. Arkansas health center revenues, by source, 2013 and 2017Source: GW analysis of 2013 and 2017 HRSA UDS dataProjected impact of Arkansas’ Medicaid work requirements on health centers The companion analysis to this report estimates that during the initial year of the Medicaid work demonstration, between 19 percent and 30 percent of Medicaid beneficiaries will lose their benefits because of their inability to meet work and/or reporting rules. To estimate the spillover impact of these losses on health center revenue, patients, and staffing, we applied this estimate to Arkansas health centers using HRSA UDS data for 2017.Table 3 shows the potential impact on health center Medicaid patient volume if 19 to 30 percent of health center patients subject to the work requirement were to lose coverage. As table 3 shows, among nearly 84,000 Medicaid patients served by Arkansas health centers, we estimate that just under 13,000, or 15 percent, will be subject to the work requirements. A loss of Medicaid coverage by 19 percent to 30 percent of this patient group translates to between 2,453 and 3,873 health center Medicaid patients losing coverage in the first year of the demonstration. This figure equates to a reduction in the health center Medicaid patient population overall of 2.9 percent to 4.6 percent.Table 3. Potential loss of Medicaid coverage of patients at Arkansas health centersSource: 2017 UDS and Brantley Ku, 2018To further estimate the impact of the loss of Medicaid patients on Arkansas health center operations, we used this estimated Medicaid coverage loss to estimate the accompanying revenue losses. Medicaid revenue plays a critical role in health centers’ ability to hire staff, maintain multiple service sites, expand the range of services offered, and reach additional patients.[12] As in other states, while Arkansas health centers rely on a mix of revenue sources to maintain health care services, Medicaid remains an important driver of health center growth and viability.[13] Assuming that the loss of Medicaid-covered patients would lead to a proportional loss in Medicaid revenue, we estimate that Arkansas health centers would lose between $1.5 million and $2.3 million in Medicaid revenue, or 0.9% to 1.4% of total revenue (Table 4). This loss, in turn, would reduce patient capacity by 1,811 to 2,859 patients, while total visits would fall by 6,827 to 10,779.Table 4. Projected Medicaid revenue losses and resulting reductions in patient and visit capacity at Arkansas health centersSource: 2017 UDS and Brantley Ku, 2018These estimates are conservative. First, Arkansas currently applies the work requirement policy to expansion adults with incomes below 100% of poverty, even though CMS permits it to apply to those with incomes up to 138% of poverty. If the state expands the scope of the work policy, far more people could be subject to work requirements. Second, it is likely that UDS data underestimate the number of people who gained coverage under Arkansas’s Medicaid expansion. Arkansas took an innovative approach in which the Medicaid expansion population was in fact enrolled in the private health insurance marketplaces, not the traditional Medicaid program. Thus, some of the substantial gains in private insurance observed in UDS data actually may represent Medicaid enrollees who could be subject to work due to the mis-categorization of these Medicaid patients as privately insured. These two factors mean that the actual effect of Arkansas’ work requirements may be more severe for health centers than we estimate.Implications As CMS has recognized, Arkansas’ community health centers, like those in all states, represent a critical source of primary health care for low-income and medically underserved populations, regardless of their health insurance coverage. At the same time, while federal health center grants represent the largest proportion of operating revenue, health centers rely heavily on Medicaid for the funds they need hire to staff, expand services and locations, and grow their treatment capacity. Our analysis of the spillover effects flowing from widespread coverage loss among Arkansas Medicaid beneficiaries shows that a major loss of coverage can be expected to translate into a substantial loss of Medicaid financing, leading to an overall decline in operating revenue. With this loss will come a notable decline in patient care capacity, as staffing is reduced, sites are shuttered, and services are rolled back in order to absorb losses. Our previous work examining how health centers respond to threatened funding reductions suggests that sites and operating hours could be targeted for reductions, with staffing reductions increasingly likely as losses remain sustained.[14]A loss of patient care capacity would, of course, affect the very communities that have most benefitted from the Arkansas Medicaid expansion and whose residents are at risk for Medicaid losses under the work experiment. These are communities where economic prospects are lower, where jobs are scarcer, and whose residents have limited education, more limited experience with the use of internet information systems, and significantly lower access to the online technology needed to navigate Arkansas’ online reporting systems. They also are more likely to lack the transportation needed to reach sources of online reporting assistance, such as the kind of assistance often found at health centers, which offer help with online insurance enrollment and renewals. A unique aspect of community health centers also has important implications for the impact of the work requirements on the broader population served by health centers. Under federal rules, health centers must seek to serve all patients, regardless of their insurance status. They cannot selectively choose to reduce access for former Medicaid enrollees who lost coverage due to work requirements, but they will lose Medicaid revenue. This will, in turn, force them to reduce staffing levels and curtail operating hours or close sites, thereby reducing overall patient care capacity. This means that the broad cross-section of health center patients, which includes those on Medicare, private insurance and those who retain Medicaid, will have more limited access. It is plausible that service cutbacks will have the most serious consequences in smaller and rural sites, reducing access for many people who are not the immediate targets of Arkansas work requirements policy.Health centers’ role in providing health care for medically underserved residents is well-documented and clearly recognized by CMS. At the same time, this analysis suggests that, contrary to CMS assumptions, health centers may not be in a position to mitigate the detrimental impact of the work experiment. Indeed, our estimates point in the opposite direction.These estimates also underscore another concerning aspect of the HHS-approved work experiment in Arkansas. No evaluation design has yet been approved, and no evaluation plan is in place. In allowing the demonstration to move forward in the absence of an evaluation, CMS is permitting an experiment that carries great risk of adverse impact – the elimination of coverage – without ensuring that an objective, well-designed evaluation is in place and fully implemented. The experiment is being permitted to proceed without producing reliable evidence regarding the extent of coverage loss, the factors underlying such losses, or the spillover effects that coverage losses may have on entire, underlying community health systems. Finally, the STC provision related to notification about health centers clearly suggests that CMS does not assume alternative coverage pathways for those who lose Medicaid. This, of course, is at odds with the justifications for the work experiments that the administration has put forth – a strategy for moving people into alternative coverage arrangements such as Marketplace coverage or employer-sponsored plans. To the extent that the STCs require Arkansas to notify people experiencing Medicaid loss about where they can turn to for free or low-cost care, this suggests that, in fact, preserving and improving insurance coverage is not an experimental outcome the administration is contemplating.References[1] Glied, S. A. (October 31, 2018). How a Medicaid Work Requirement Could Affect Arkansas’ Economy. Commonwealth Fund Blog. https://www.commonwealthfund.org/blog/2018/medicaid-work-requirement-arkansas-economy[2] Brantley, E. Ku, L. (October 31, 2018). Arkansas’ Early Experience with Work Requirements Signals Larger Losses to Come.  Commonwealth Fund Blog. https://www.commonwealthfund.org/blog/2018/arkansas-early-experience-work-requirements[3] The remaining low-income uninsured were also more likely to have access to care in expansion states due to a more robust and financially viable safety net compared to their counterparts in non-expansion states.  See U.S. Government Accountability Office (2018), Access to Health Care for Low-Income Adults in States with and without Expanded Eligibility. https://www.gao.gov/products/GAO-18-607[4] Sharac, J., Shin, P., Gunsalus, R., Rosenbaum, S. (2018). Community health centers continued to expand patient and service capacity in 2017. Geiger Gibson/RCHN Community Health Foundation Research Collaborative, George Washington University. Policy Research Brief No. 54. https://www.rchnfoundation.org/?p=7172[5] See section 54q. https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/ar/ar-works-ca.pdf[6] Bureau of Primary Health Care. (2018). 2017 Health Center Data: Arkansas Data. Health Resources and Services Administration. https://bphc.hrsa.gov/uds/datacenter.aspx?q=tallyear=2017state=AR[7] U.S. Census Bureau. Current Population Survey, Annual Social and Economic Supplement, 2018.  https://www.census.gov/cps/data/cpstablecreator.html[8] Estimated number of low-income patients based on the sum of the low-income percentage of patients for each health center multiplied by the total number of patients for all health centers in Arkansas from 2017 UDS data.[9] Maylone, B. Sommers, B.D. (2017). Evidence from the private option: the Arkansas experience. The Commonwealth Fund. https://www.commonwealthfund.org/publications/issue-briefs/2017/feb/evidence-private-option-arkansas-experience[10] See p. 145 of the 2017 UDS manual, https://bphc.hrsa.gov/datareporting/reporting/2017udsreportingmanual.pdf[11] We estimate that roughly 30% of adult Medicaid enrollees at Arkansas Medicaid enrollees are subject to work requirements, based on the state’s policy under which work requirements are imposed on Medicaid expansion enrollees with incomes below 100% of poverty who are not otherwise exempt from the policy, e.g., children, the elderly, the medically frail, etc.[12] See also, https://www.healthaffairs.org/do/10.1377/hblog20180412.955253/full/[13]  Shin P, Sharac J, Rosenbaum S.  Community Health Centers and Medicaid at 50: An Enduring Relationship Essential for Health System Transformation.  Health Affairs, 2015, 34(7): 196-1104.[14] Geiger Gibson/RCHN Community Health Foundation Research Collaborative and the Kaiser Family Foundation. (2018). How Are Health Centers Responding to the Funding Delay?https://www.kff.org/medicaid/fact-sheet/how-are-health-centers-responding-to-the-funding-delay/"The Projected Effects of the Arkansas Medicaid Work Requirement Demonstration on Community Health Centers" was written by Peter Shin, Jessica Sharac and Sara Rosenbaum of the Milken Institute School of Public Health’s Department of Health Policy and Management.