Although safety net organizations are eligible for some two-thirds of federal payment reform programs, fewer than 20% of these programs directly target the safety net.
Objectives: Safety net organizations face challenges under new value-driven payment models. More than 70 payment reform programs were created through the CMS Center for Medicare and Medicaid Innovation (CMMI). These programs were analyzed to determine the extent to which they target safety net organizations.
Study Design: Systematic review of publicly available CMMI program descriptions.
Methods: We reviewed all CMMI programs’ stated goals, eligibility criteria, participating organizations (if applicable), and all other relevant information publicly available on the CMMI website. We classified each program as (1) specifically targeting the safety net, (2) applying to the safety net but not specifically targeting it, or (3) not applying to the safety net. For safety net initiatives, we also reviewed each program to determine which portion(s) of the safety net it was relevant to.
Results: Only 13 CMMI programs directly targeted safety net providers, 32 included safety net providers, and 22 programs did not apply to or target the safety net. No trends were found to suggest that this proportion was increasing over time. Safety net—relevant programs tended to target critical access hospitals rather than primary care or behavioral health.
Conclusions: Programs that specifically target the safety net may represent an important opportunity to overcome unique payment reform challenges facing this portion of the healthcare sector. However, safety net organizations are inconsistently represented in CMMI programs. There is currently no expedient approach to determine whether CMMI programs apply to specific safety net subsectors. Additional clarity regarding safety net relevance may help organizations understand and participate in relevant CMMI programs.
The American Journal of Accountable Care. 2019;7(1):17-23One potentially promising approach to improving value in healthcare is the use of alternative payment models (APMs) to fee-for-service (FFS)—based models.1 APMs include FFS payments that are linked to quality and value, gain sharing and/or downside risk, condition-specific population-based payment, and comprehensive population-based payment.2 Success under an APM requires transforming the manner in which care is delivered in a healthcare organization.1,3
Payment and care delivery reform pose significant technical and operational challenges for healthcare delivery organizations.4,5 However, not all healthcare delivery organizations are equally well positioned to respond to such challenges.6-9 Study results suggest that safety net organizations may not be well positioned to succeed under new payment models.6,9-11 A safety net healthcare organization is one that provides a significant level of care to low-income, uninsured, and vulnerable populations and has a legal mandate to serve patients regardless of their ability to pay.12 The safety net sector provides care to Americans who live in underserved areas.6,12
Safety net organizations face unique challenges that limit the success of payment reform initiatives. In addition to the challenges of patient attribution and risk adjustment, safety net organizations also have exceptionally high-cost patients.13-15 These challenges are further complicated by policy changes, such as Disproportionate Share Hospital funding streams that changed substantially for many safety net hospitals with the passage of the Affordable Care Act.9,11,16 These and other payment reform initiatives have traditionally put pressure on safety net hospitals owing to their unique missions, patient populations, and funding streams.7 Despite declining uninsurance rates, safety net providers persist as an integral source of care for many.17 Understanding the challenges and opportunities that payment reform presents for safety net organizations is of immense importance due to the size and scope of the safety net sector.6,12
Payment reform presents significant opportunities for achieving improved population health outcomes, improved patient experience of care, and decreased per capita costs.18 However, new payment models often carry high financial, technological, and personnel needs.11 Safety net organizations tend to have more difficulty meeting these needs compared with non—safety net organizations, especially the large start-up costs for new programs.6,9-11 Furthermore, advanced payment models often rely on strong, interoperable data systems, which may not exist in safety net organizations.19,20 Advanced payment models frequently place greater demands on personnel who may already be operating at maximum capacity within safety net organizations.13,21,22 Complex programs also require qualified staff, and safety net organizations often have limited access to the needed talent pool because of budget restrictions and geographic location.11 Because of these challenges, safety net organizations stand to benefit most from advanced payment models that not only apply to the safety net sector but are also specifically designed with these organizations in mind.23
Little empirical evidence has been compiled regarding systemwide trends in payment reform across the safety net.17 Safety net organizations generally derive large proportions of their revenues from public payers, such as state Medicaid programs, so recent payment reform initiatives from large private payers may not directly apply to safety net organizations.24 Likewise, the last remaining members of the recently concluded Pioneer Accountable Care Organization (ACO) program were not safety net organizations. Research suggests that although ACOs may be establishing new linkages with specific safety net providers, as of 2014 only 28% of ACOs included community health centers within their care networks.25 A centerpiece of payment reform efforts in the United States is CMS’ Center for Medicare and Medicaid Innovation (CMMI), which is home to some 70 payment reform programs. However, the overall extent to which safety net organizations are represented in these programs is not known, nor is it clear which of these programs are relevant for or directly target safety net organizations.
Given the large role that the safety net plays in delivering care to a range of underserved populations in the United States, it is crucial to consider how payment reforms may or may not be targeting safety net organizations. CMMI payment reforms are particularly relevant to safety net organizations because of these organizations’ strong linkages with public programs. However, no summary evidence is available to guide policy makers’ ability to determine whether current payment reform efforts generally target or include safety net organizations at levels that are comparable with their role in the US healthcare delivery system. Likewise, safety net organizations themselves may benefit from a comprehensive review of the applicability of all CMMI programs to their organizations. The goal of this analysis is to explore the landscape of CMMI payment reform initiatives to better understand the range of payment reform efforts currently under way and the implications for how safety net organizations are funded in the future.
Defining the Safety Net
Safety net providers organize and deliver a significant level of healthcare and other health-related services to the uninsured, Medicaid enrollees, and other vulnerable patient populations.12 Although there is a lack of consensus on what specifically constitutes a safety net organization,26 we identified the safety net as including organizations serving vulnerable or underserved patients across 6 distinct but interrelated subsectors: federally qualified health centers (FQHCs), critical access hospitals, public teaching hospitals, community mental health centers, tribal health centers, and oral health providers who care for the underserved. The 6 subsectors are outlined in Table 1. These are not the only safety net organizations (others would include, for example, freestanding charity clinics), but they represent the largest portion of safety net providers.
Data Sources and Approach
To identify federal payment reform opportunities available to safety net organizations, we identified all CMMI payment reform initiatives. We carefully reviewed each program and its stated goals, eligibility criteria, terms and conditions, and other relevant information available on the CMMI public website. We then classified the program according to a 3-level, mutually exclusive categorization: (1) specifically targeting the safety net, (2) applying to the safety net but not specifically targeting it, or (3) not applying to the safety net. Second, we classified each initiative as past (no longer active as of 2017), current, or future (under development, announced, accepting applications, applications under review, or participants announced). Third, for initiatives that applied to or specifically targeted the safety net, we reviewed the full description of the initiatives to determine which safety net subsector the payment reform initiative was relevant to.
To ensure the reliability of our categorizations, 2 members of the research team (N.C. and M.G.) each independently reviewed and coded all programs. When assessments differed, a third member of the research team (J.M.M.) reviewed the program and adjudicated resolution for whether that program specifically targeted, applied to, or did not apply to the safety net. For programs that applied to or specifically targeted safety net organizations, we again reviewed the program eligibility criteria, functions, and targeted stakeholders. We then classified applicable programs as applying to 1 or more of the 6 safety net sectors.
We conducted univariate and bivariate analyses to assess the breadth and identify patterns of CMMI payment reform programs applicable to safety net organizations. We specifically examined (1) whether the proportion of payment reform programs applicable to the safety net differed for past, current, and future CMMI programs and (2) whether any individual safety net subsector was over- or underrepresented in CMMI payment reform programs. All analyses were performed using Excel (Microsoft; Seattle, Washington) and Stata 13.1 (StataCorp; College Station, Texas).
Our review of CMMI initiatives revealed a total of 76 programs. Nine programs that were not directly related to payment reform (ie, initiatives to improve quality of care that did not incorporate any payment reform components) were excluded. The final sample size was 67 CMMI payment reform initiatives.
A total of 45 (67%) of these initiatives applied to safety net organizations, whereas 22 (33%) did not apply. Of the 45 initiatives that applied to safety net organizations, 13 (19%) specifically targeted safety net providers; the remaining 32 (48%) did not directly target safety net providers and only included safety net organizations more generally. Complete, initiative-by-initiative findings can be seen in the eAppendix (available at ajmc.com).
Of the 67 CMMI payment reform initiatives, we also identified 20 as past (no longer active) initiatives, 29 as current initiatives, and 18 as planned or anticipated initiatives. The Figure shows that the proportion of initiatives targeting the safety net has not varied over time and that the change in the proportion of programs that targeted, applied to, or did not involve the safety net was not statistically significant (P = .41).
As shown in Table 2, CMMI payment reform initiatives most commonly targeted safety net hospitals, whereas far fewer initiatives targeted oral health safety net providers. Of the 45 initiatives that applied to safety net organizations, 11 were applicable to safety net hospitals but to no other safety net subsectors (data not shown in table). There were 26 initiatives that included multiple safety net subsectors (eg, hospitals plus FQHCs or community mental health providers). Of the 22 initiatives not applicable to the safety net, most were determined to be targeted to either hospitals (n = 8) or nursing homes/home health care (n = 8).
Reimbursement models that link payment to quality or value are becoming increasingly common and important in the US healthcare delivery system. In 2015, HHS set a goal of tying 30% of traditional FFS Medicare payments to quality or value through APMs by 2016 and 50% by 2018.27 The Health Care Payment Learning & Action Network, an organization sponsored by HHS, has since expanded this goal to encompass the entire US health system, including public and private sectors.2 However, the extent to which organizations serving vulnerable populations are represented in these initiatives is unknown. This analysis quantified past, current, and future CMMI payment reform initiatives relevant to the safety net.
We found that two-thirds of CMMI initiatives were applicable to safety net organizations, but the majority of these initiatives did not directly target the safety net. This could mean, for example, that only a single safety net hospital or FQHC participated in the initiative. Additionally, several large initiatives included safety net organizations, but the designs of these initiatives were not explicitly tailored to address the challenges that safety net organizations face in caring for underserved populations, nor were these programs tailored to the specific features of various safety net organization payment models. Initiatives that directly target safety net organizations are far less common, representing less than 20% of all CMMI payment reform programs. Among initiatives that are applicable to the safety net, nearly half included organizations from only a single portion of the safety net (eg, safety net hospitals, FQHCs, or behavioral health clinics). This is despite the well-known importance of integrating care across settings and coordinating among inpatient providers, outpatient providers, and community resources for service delivery.10,28
Results from previous studies have shown that safety net hospitals may not be well positioned to succeed under APMs.6,7,9-11 Safety net organizations may therefore require more specialized or targeted support, including opportunities such as CMMI programs that enable healthcare delivery organizations to test and refine innovative payment reform approaches. However, we did not find evidence that CMMI initiatives disproportionately target the safety net or that the number of programs targeting safety net organizations is increasing over time. Although the majority of CMMI initiatives considered allowing safety net organizations to participate, these programs do not account for the challenges that safety net organizations must confront. Limited financial resources may make it difficult for safety net organizations to invest in the up-front costs, such as data systems or care coordination resources, that are required to successfully undertake the shift toward value-based care.29
Payment reform can provide safety net organizations with the opportunity to improve the health of the nation’s most vulnerable populations, but a one-size-fits-all approach to the diffusion of APMs across the entire healthcare system may not be ideal. Specifically, different safety net sectors need different types of support to succeed in value-based reimbursement approaches, given the unique financing mechanisms for each sector. Particularly at the outset, implementing payment reform programs may require additional financial and labor inputs.30 In the safety net, these resources are typically in shorter supply in comparison with non—safety net organizations.29,31
Our findings also reveal that certain safety net sectors appear to be disproportionately targeted by payment reform programs. For example, public county hospitals and critical access hospitals, 2 settings that provide mainly acute care, were eligible for the largest number of CMMI programs. In contrast, FQHCs and community mental health centers, 2 settings that provide mainly primary care, were eligible for participation in fewer payment reform programs. Given the importance of coordinating care across settings,32 linking acute with primary care under new payment models,3,33 and continuing to strengthen the primary care system in the United States,34-36 the underrepresentation of primary care settings in CMMI programs is an area of specific concern for policy makers. Improving quality and reducing costs are important goals in all sectors, but this imbalance could fuel the mismatch between primary and preventive care compared with acute care.
Some non-CMMI payment reform initiatives are also under way within the safety net. For example, as of 2016, Medicaid ACOs and Delivery System Reform Incentive Payment programs had each been established in 8 states.37 However, it appears that CMMI represents the greatest opportunity for safety net payment reform work. CMMI plays a vital role in providing opportunities to address the challenges that these safety net organizations face under new payment and care delivery models. Payment reform models that include or specifically target the safety net sector should address the unique challenges presented by underserved patients, such as lack of resources to hire specialized, highly trained administrative and clinical personnel and to implement and coordinate data infrastructure systems.
In this study, our team carefully reviewed each CMMI program to assess its relevance to the safety net; this required more than 100 hours to analyze published CMMI programs in order to determine their relevance to the safety net in general and to specific subsector(s) therein, and additional time would be necessary to determine how to apply for upcoming programs. Programs that are applicable to the safety net should be easily identifiable. Safety net organizations already face resource constraints and may not have spare personnel resources to dedicate to identifying appropriate programs. Clearly highlighting applicable programs could concretely support the safety net in participating in payment reform initiatives.
Our findings should be viewed in light of our study’s limitations. First, there is not a specific universal acknowledgment of what constitutes the safety net. We operationalized the most widely used definition—that of the Institute of Medicine—yet this definition may be broader or narrower than others.12 We categorized which subsector(s) of the safety net each CMMI program fit into in order to facilitate further reanalysis of safety net sector applicability if desired. Second, our review reflects applicability of CMMI payment reform programs only. Given the involvement of state-level funding sources across the safety net, additional work to categorize the full extent of safety net payment reform programs may generate different point estimates for the number of relevant programs that are active or planned. Our findings can still be interpreted to reflect federal initiatives that reach safety net populations that span multiple states (eg, Medicare—Medicaid dual enrollees) or underserved populations in hospital referral regions that span multiple states.
Changing healthcare payment to emphasize value and de-emphasize volume presents an important opportunity to improve performance of the healthcare system. However, the safety net sector may face challenges in transitioning to value-based payment approaches, and it is likely to need special consideration and focus if HHS and private payers are to achieve ambitious targets for widespread dissemination of value-based payments. To date, very little summary evidence is available to guide policy makers and practitioners with respect to the extent to which the safety net is involved in payment reform initiatives.
Our analysis reveals that although the safety net is broadly eligible for some two-thirds of CMMI payment reform programs, less than 20% of these programs directly target the safety net. This is in spite of the fact that the safety net, broadly defined, serves more than 100 million patients annually. Safety net—relevant CMMI payment reform programs tend to be focused more on hospitals than on primary care settings, and the proportion of safety net–relevant CMMI programs is not increasing over time. Moreover, there is no expedient approach to determine the relevance of each CMMI program to specific subsectors of the safety net. Additional consideration of ways to enhance the clarity of safety net relevance may help provider organizations to more fully understand and participate in relevant CMMI programs. Programs that specifically target the safety net may represent an important opportunity to overcome unique payment reform challenges that this portion of the healthcare sector must face. The safety net sector is underaddressed by past, current, and future CMMI payment reform programs.Author Affiliations: National Safety Net Advancement Center (JMM, NC, MG, AR, KL, WR), Phoenix, AZ; School for the Science of Health Care Delivery, Arizona State University (JMM, NC, MG, AR, KL, WR), Phoenix, AZ.
Source of Funding: Funding in support of this work was provided to the National Safety Net Advancement Center by the Robert Wood Johnson Foundation. The funder had no editorial review or approval of this work. The views expressed are those of the authors and do not necessarily reflect the views of the Robert Wood Johnson Foundation.
Author Disclosures: Ms Love reports receiving grants from the Robert Wood Johnson Foundation. The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.
Authorship Information: Concept and design (JMM, NC, AR, KL, WR); acquisition of data (NC, MG); analysis and interpretation of data (JMM, NC, MG, WR); drafting of the manuscript (JMM, NC, AR, KL, WR); critical revision of the manuscript for important intellectual content (JMM, MG, AR, KL, WR); statistical analysis (JMM); obtaining funding (JMM, AR, KL, WR); administrative, technical, or logistic support (AR, KL); and supervision (JMM, KL, WR).
Send Correspondence to: J. Mac McCullough, PhD, MPH, Arizona State University, 550 N 3rd St, Phoenix, AZ 85004. Email: email@example.com.REFERENCES
1. Porter ME. What is value in health care? N Engl J Med. 2010;363(26):2477-2481. doi: 10.1056/NEJMp1011024.
2. Alternative payment model (APM) framework refresh. Health Care Payment Learning & Action Network website. hcp-lan.org/apm-refresh-white-paper. Published July 11, 2017. Accessed April 21, 2018.
3. Rittenhouse DR, Shortell SM, Fisher ES. Primary care and accountable care—two essential elements of delivery-system reform. N Engl J Med. 2009;361(24):2301-2303. doi: 10.1056/NEJMp0909327.
4. Miller HD. From volume to value: better ways to pay for health care. Health Aff (Millwood). 2009;28(5):1418-1428. doi: 10.1377/hlthaff.28.5.1418.
5. Fisher ES, McClellan MB, Safran DG. Building the path to accountable care. N Engl J Med. 2011;365(26):2445-2447. doi: 10.1056/NEJMp1112442.
6. Andrulis DP, Siddiqui NJ. Health reform holds both risks and rewards for safety-net providers and racially and ethnically diverse patients. Health Aff (Millwood). 2011;30(10):1830-1836. doi: 10.1377/hlthaff.2011.0661.
7. Gilman M, Adams EK, Hockenberry JM, Wilson IB, Milstein AS, Becker ER. California safety-net hospitals likely to be penalized by ACA value, readmission, and meaningful-use programs. Health Aff (Millwood). 2014;33(8):1314-1322. doi: 10.1377/hlthaff.2014.0138.
8. Mohan A, Grant J, Batalden M, McCormick D. The health of safety net hospitals following Massachusetts health care reform: changes in volume, revenue, costs, and operating margins from 2006 to 2009. Int J Health Serv. 2013;43(2):321-335. doi: 10.2190/HS.43.2.h.
9. Cole ES, Walker D, Mora A, Diana ML. Identifying hospitals that may be at most financial risk from Medicaid disproportionate-share hospital payment cuts. Health Aff (Millwood). 2014;33(11):2025-2033. doi: 10.1377/hlthaff.2014.0109.
10. Katz MH. Future of the safety net under health reform. JAMA. 2010;304(6):679-680. doi: 10.1001/jama.2010.1126.
11. Felland L, Cunningham P, Doubleday A, Warren C. Effects of the Affordable Care Act on Safety Net Hospitals. Washington, DC: Mathematica Policy Research; 2016. aspe.hhs.gov/system/files/pdf/255491/SafetyNetHospital.pdf. Accessed April 21, 2018.
12. Lewin ME, Altman S, eds; Institute of Medicine. America’s Health Care Safety Net: Intact but Endangered. Washington, DC: The National Academies Press; 2000. doi: 10.17226/9612.
13. Blegen MA, Goode CJ, Spetz J, Vaughn T, Park SH. Nurse staffing effects on patient outcomes: safety-net and non-safety-net hospitals. Med Care. 2011;49(4):406-414. doi: 10.1097/MLR.0b013e318202e129.
14. Friedberg MW, Chen PG, White C, et al. Effects of health care payment models on physician practice in the United States. Rand Health Q. 2015;5(1):8.
15. Joynt KE, Orav EJ, Jha AK. Thirty-day readmission rates for Medicare beneficiaries by race and site of care. JAMA. 2011;305(7):675-681. doi: 10.1001/jama.2011.123.
16. Neuhausen K, Davis AC, Needleman J, Brook RH, Zingmond D, Roby DH. Disproportionate-share hospital payment reductions may threaten the financial stability of safety-net hospitals [erratum in Health Aff (Millwood). 2014;33(7):1304. doi: 10.1377/hlthaff.2014.0688]. Health Aff (Millwood). 2014;33(6):988-996. doi: 10.1377/hlthaff.2013.1222.
17. Ku L, Shin P, Regenstein M, Mead H. Promoting the integration and coordination of safety-net health care providers under health reform: key issues. The Commonwealth Fund website. commonwealthfund.org/publications/issue-briefs/2011/oct/promoting-integration-and-coordination-safety-net-health-care. Published October 13, 2011. Accessed January 29, 2019.
18. Berwick DM, Nolan TW, Whittington J. The triple aim: care, health, and cost. Health Aff (Millwood). 2008;27(3):759-769. doi: 10.1377/hlthaff.27.3.759.
19. Adler-Milstein J, DesRoches CM, Kralovec P, et al. Electronic health record adoption in US hospitals: progress continues, but challenges persist. Health Aff (Millwood). 2015;34(12):2174-2180. doi: 10.1377/hlthaff.2015.0992.
20. Torda P, Han ES, Scholle SH. Easing the adoption and use of electronic health records in small practices. Health Aff (Millwood). 2010;29(4):668-675. doi: 10.1377/hlthaff.2010.0188.
21. Bodenheimer T, Sinsky C. From triple to quadruple aim: care of the patient requires care of the provider. Ann Fam Med. 2014;12(6):573-576. doi: 10.1370/afm.1713.
22. Vahey DC, Aiken LH, Sloane DM, Clarke SP, Vargas D. Nurse burnout and patient satisfaction. Med Care. 2004;42(suppl 2):II57-II66. doi: 10.1097/01.mlr.0000109126.50398.5a.
23. Medicare value-based payment models: participation challenges and available assistance for small and rural practices. US Government Accountability Office website. gao.gov/products/GAO-17-55. Published December 9, 2016. Accessed April 24, 2018.
24. Chernew ME, Mechanic RE, Landon BE, Safran DG. Private-payer innovation in Massachusetts: the ‘alternative quality contract.’ Health Aff (Millwood). 2011;30(1):51-61. doi: 10.1377/hlthaff.2010.0980.
25. Lewis VA, Colla CH, Schoenherr KE, Shortell SM, Fisher ES. Innovation in the safety net: integrating community health centers through accountable care. J Gen Intern Med. 2014;29(11):1484-1490. doi: 10.1007/s11606-014-2911-0.
26. Phillips RL Jr, Bazemore A, Miyoshi TJ. Mapping tools for monitoring the safety net. In: Weinick RM, Billings J, eds. Tools for Monitoring the Health Care Safety Net. Rockville, MD: Agency for Healthcare Research and Quality; 2003.
27. Better care. Smarter spending. Healthier people: paying providers for value, not volume [news release]. Washington, DC: CMS; January 26, 2015. cms.gov/newsroom/fact-sheets/better-care-smarter-spending-healthier-people-paying-providers-value-not-volume. Accessed April 24, 2018.
28. Cunningham P, Felland L, Stark L. Safety-net providers in some US communities have increasingly embraced coordinated care models [erratum in Health Aff (Millwood). 2012;31(10):2356. doi: 10.1377/hlthaff.2012.0983]. Health Aff (Millwood). 2012;31(8):1698-1707. doi: 10.1377/hlthaff.2011.1270.
29. Katz MH, Brigham TM. Transforming a traditional safety net into a coordinated care system: lessons from Healthy San Francisco. Health Aff (Millwood). 2011;30(2):237-245. doi: 10.1377/hlthaff.2010.0003.
30. McGinnis T, Small DM. Accountable care organizations in Medicaid: emerging practices to guide program design. Center for Health Care Strategies, Inc, website. chcs.org/resource/accountable-care-organizations-in-medicaid-emerging-practices-to-guide-program-design. Published February 2012. Accessed January 29, 2019.
31. Lewin ME, Baxter RJ. America’s health care safety net: revisiting the 2000 IOM report. Health Aff (Millwood). 2007;26(5):1490-1494. doi: 10.1377/hlthaff.26.5.1490.
32. Peikes D, Chen A, Schore J, Brown R. Effects of care coordination on hospitalization, quality of care, and health care expenditures among Medicare beneficiaries: 15 randomized trials. JAMA. 2009;301(6):603-618. doi: 10.1001/jama.2009.126.
33. Porter ME, Pabo EA, Lee TH. Redesigning primary care: a strategic vision to improve value by organizing around patients’ needs. Health Aff (Millwood). 2013;32(3):516-525. doi: 10.1377/hlthaff.2012.0961.
34. Donaldson MS, Yordy KD, Lohr KN, Vanselow NA, eds; Institute of Medicine. Primary Care: America’s Health in a New Era. Washington, DC: The National Academies Press; 1996. doi: 10.17226/5152.
35. Davis K, Schoenbaum SC, Audet AM. A 2020 vision of patient‐centered primary care. J Gen Intern Med. 2005;20(10):953-957. doi: 10.1111/j.1525-1497.2005.0178.x.
36. Starfield B, Shi L, Macinko J. Contribution of primary care to health systems and health. Milbank Q. 2005;83(3):457-502. doi: 10.1111/j.1468-0009.2005.00409.x.
37. Chokshi DA, Chang JE, Wilson RM. Health reform and the changing safety net in the United States. N Engl J Med. 2016;375(18):1790-1796. doi: 10.1056/NEJMhpr1608578.