Abstract
Introduction
Safety net hospitals (SNHs) disproportionately serve uninsured, Medicaid, and low-income Medicare patients, yet no universally accepted SNH definition exists. Existing binary measures frequently identify divergent hospital sets, creating research inconsistency and risking misaligned policy protections, concerns heightened by recent federal Medicaid cuts and site-neutral payment reforms.
Methods
Using Medicare cost report data for community hospitals, we introduce and validate a continuous, multidimensional SNH Index constructed as a composite of Medicaid revenue share, uncompensated care burden, and low-income Medicare share.
Results
Hospitals in the highest SNH Index quartile had substantially lower operating margins (−2.0% vs 3.1% in the lowest quartile), fewer days cash on hand (91 vs 158 days), and higher rates of financial distress (36% vs 21%). The SNH Index showed stronger correlations with financial performance than 5 binary SNH definitions and a proxy for MACPAC’s Medicare SNH Index. Within each quartile, system-affiliated hospitals significantly outperformed independent hospitals financially.
Conclusion
These findings suggest that policies protecting SNHs should incorporate composite, burden-based measures, and account for system-level financial resources rather than relying on binary classifications alone.
Keywords: hospitals, health care safety net, uninsured, uncompensated care, health care systems
Introduction
Safety net hospitals (SNHs) disproportionately provide medical care to individuals regardless of their insurance status, ability to pay, or immigration status across both urban and rural settings. SNHs’ financial health can be vulnerable as they often deliver high levels of uncompensated care, offer essential services with thin profit margins, and have an unfavorable payer mix with limited access to commercial reimbursement. Given their unique function in the US health care system, SNHs are the focus of a large body of research examining access to care, health outcomes, health care quality, and hospital financial performance.1-3
Numerous public policies have also hinged on supporting SNHs, including Medicare and Medicaid Disproportionate Share Hospital (DSH) payments, distribution of provider relief funds under the Coronavirus Aid, Relief, and Economic Security Act, and Medicaid managed care state-directed payments. Recent cost containment proposals, such as a site-neutral payment reform in New York,4 and the 2026 Hospital Outpatient Prospective Payment System (OPPS) final rule,5 seek to exempt SNHs or certain rural and sole community hospitals, many of which operate as SNHs by serving vulnerable populations.
Despite their importance, there is not a universally accepted approach to identifying or characterizing SNHs. Hundreds of empirical studies have examined SNHs in some capacity, often using inconsistent and non-overlapping definitions depending on the policy or payer context.1 A 2025 study explored 9 hospital-level and 4 area-level metrics to identify SNHs, and showed that different measures identify different hospitals as SNHs and that there is considerable variation in the extent of overlap and stability across metrics.6 Although that study focused on the classification of SNHs, other research has identified SNHs using a range of methods and showed significant variation in their financial condition, ownership status, size, geographic location, and case mix.6-9 To address this fragmentation, researchers have considered the need for a continuum-based approach incorporating multiple factors, such as Medicaid and uninsured payer share, Medicare DSH percentage, geographic indicators of socioeconomic disadvantage, and hospital financial health.2,6
In response to the inability of a single metric to identify SNHs, MedPAC created the Medicare Safety Net Index (MSNI) to capture multiple dimensions of an SNH in a single measure. The MSNI combines the hospitals’ share of Medicare patients receiving Part D low-income subsidies, uncompensated care costs, and overall Medicare volume. MedPAC's goal is to identify SNHs based on the types of vulnerable Medicare patients they serve and to more effectively target Medicare DSH and uncompensated care payments to hospitals facing financial risk.10 MedPAC's MSNI focuses on Medicare payment redistribution and while useful, does not incorporate the extent to which a hospital serves vulnerable patients under 65, a group that often relies on an SNH for care. This group would typically include those covered by Medicaid—typically a low-reimbursement payer—as well as those lacking any coverage.
In this study, we build on MedPAC's approach to develop a new SNH Index that would allow policymakers to measure the degree to which a hospital functions as an SNH for a broader vulnerable population than just Medicare enrollees. This would allow policymakers to consider a gradient of SNH function as opposed to “either or” binary choices that have been used in the literature. Developing a broad-based SNH metric would be valuable because it could be used to prioritize the distribution of any available SNH funding at this time when the safety net is at increased risk due to recent federal policy changes that are likely to increase the number of uninsured Americans. The One Big Beautiful Bill Act (OBBBA), the 2025 budget reconciliation bill, includes over $900 billion in federal Medicaid spending reductions alone over a decade and caps provider taxes and state-direct payments starting in 2028, disproportionately impacting SNHs that rely on these payments to support patient care.11-13 These changes are in addition to the expiration of the enhanced Marketplace subsidies that had been introduced during the Covid pandemic.
SNHs have also been part of the market consolidation that is playing an increasingly important role in the US health care system—the share of community hospitals that are part of a larger health system increased from 53% in 2005 to 68% in 2022.14,15 Cross-sectional data have also shown that system affiliation is associated with stronger financial performance for Critical Access Hospitals.16 While existing studies have explored the link between SNH status of a given facility and its financial health, little is known about how this relationship is related to a facility's system affiliation. This analysis highlights how system affiliation is another dimension to consider in SNH definitions and how it relates to hospital financial performance.
Data and Methods
We construct an SNH Index that is more generalized than MedPAC's Medicare-focused MSNI. Like the MSNI, our SNH Index uses the uncompensated care share and a proxy for low-income Medicare share, but also incorporates hospitals’ Medicaid revenue share, a key component missing from the MSNI. Our SNH Index equally weights the following components: (1) the hospitals’ share of operating revenue from Medicaid; (2) total uncompensated care costs as a share of operating expenses; and (3) Supplemental Security Income (SSI)-recipient patient days as a share of all Medicare Part A patient days as a proxy for the low-income Medicare share. These 3 metrics capture distinct dimensions of how SNHs serve vulnerable populations and are not highly correlated with each other (ie, none of the pairwise correlations are greater than 0.35; Appendix Table S1). We also explored various combinations of different measures (eg, including Medicare share or the SVI) and using unequal weights—where some measures are more heavily weighted than others, but found that our key findings and conclusions were consistent throughout. While our main analytic index uses these 3 measures with equal weights, researchers and practitioners can modify the index components to focus on their specific questions.
By incorporating both uncompensated care costs and Medicaid revenue shares, the SNH Index avoids misclassifying states based on Medicaid expansion status—that is, by capturing states with high Medicaid shares and low uncompensated care (due to expansion) as well as those with the opposite pattern in non-expansion states. We use the SSI share, which is a lower-bound proxy of the actual low-income Medicare share, as opposed to MedPAC's low-income subsidy share because the latter is not available in the cost report data. Incorporating the low-income Medicare measure is important because low-income Medicare beneficiaries are more likely to be disabled, have end-stage renal disease, and live in rural areas compared to the full fee-for-service Medicare population.10 With respect to uncompensated care costs, we recognize that there will be some measurement error because hospitals do not account for these costs consistently (eg, the extent to which some hospitals classify bad debt as charity care).
To assess the added value of the new SNH Index, we compare it to indicators of SNH status that have been commonly used in the literature. These measures include Medicare DSH, Medicaid revenue, uncompensated care costs, public hospitals, and area attributes. The Medicare DSH, Medicaid revenue, and uncompensated care binary measures are set to 1 for hospitals in the top quartile of that measure by state. We used state-specific cutoffs because Medicaid coverage and the percentage of the population that is uninsured (which affects uncompensated care costs) differ across states based on their Medicaid eligibility criteria. For example, defining Medicaid quartiles at the national level would disproportionately include hospitals located in states that expanded Medicaid.
Our assessment of the SNH Index compares binary measures of SNH status, hospital characteristics, and financial performance across quartiles of this index. (See Appendix Table S1 SNH measures.) Hospitals in the bottom quartile of the SNH Index are hospitals least likely to serve a safety net function, while those in top quartile are hospitals that play the most critical SNH roles. We focus on variation in financial performance and distress across the quartiles because that signals which hospitals likely have the resources to continue operation and fulfill their role as an SNH.
See the online Appendix and Appendix Table S2 for additional details on commonly used SNH definitions we identified in the literature and our approach to categorizing these studies; Appendix Table S3 for additional details on the construction of the binary measures; and Appendix Table S4 for the overlap across these binary measures.
Hospital cost reports
Our main analysis uses 2023 hospital cost report data processed by RAND.17 We restrict our sample to short-term general hospitals that are not owned by the federal government, ie, community hospitals. We use the cost report data to construct the SNH Index, measures of hospital financial performance and most of the binary SNH measures, as well as various hospital characteristics, including geographic indicators, ownership type, bed size, system affiliation, and payer mix. We also use 2022 cost report data to explore the stability of the SNH Index across years.
We primarily rely on complementary financial measures—operating margins and days cash on hand (DCOH) to explore the association between SNH status and financial outcomes. Operating margins capture hospitals’ profitability from operating activities, total margins include profitability from all activities and revenue sources, and whereas DCOH is a measure of liquidity. We exclude for-profit hospitals in all estimates of DCOH because they typically return excess cash to shareholders or owners as dividends or through stock buybacks when the cash cannot be productively invested in business assets.18
To supplement these financial measures, we use the cost report data to construct an indicator for hospitals being in financial distress based on having a modified Altman Z-score of less than 1.8. The modified Altman Z-score has been used in previous studies and combines ratios of hospital liquidity, profitability, efficiency, and solvency to categorize hospitals as distressed and at high risk for bankruptcy.19
See the online Appendix for additional information on constructing the binary SNH measures, the Altman Z-score, and how we address certain limitations associated with the cost report data.18
Social Vulnerability Index
We use the CDC/ATSDR Social Vulnerability Index (hereafter SVI) from AHRQ's 2018 Social Determinants of Health file to assess the geographic context of SNHs.20 The SVI is further described in the Appendix.18
Results
Table 1 presents the overall average of the SNH Index and average SNH Index values for hospitals broken out by quartiles of the SNH Index, where the bottom quartile captures hospitals with the lowest values of the index and top quartile captures hospitals with the highest values. On average, hospitals have a Medicaid revenue share of 10.6%, a low-income Medicare share of 4.7%, and an uncompensated care cost share of 6.9%, resulting in an overall SNH Index value of 7.4. These values vary substantially across quartiles: The Medicaid share ranges from 4.2% in the bottom quartile to 19.3% in the top, the low-income Medicare share from 1.0% to 10.2% and uncompensated care burden from 4.1% to 10.1%. As a result of this variation, the SNH Index ranges from 3.1 in the bottom quartile to 13.1 in the highest quartile.
Table 1.
Overlap between safety net hospital index and binary safety net hospital definitions, 2023.
| Mean (SD) | |||||
|---|---|---|---|---|---|
| SNH Index quartiles | |||||
| All hospitals | 1 (bottom quartile) | 2 | 3 | 4 (top quartile) | |
| Hospitals, No. | 4494 | 1081 | 1081 | 1081 | 1080 |
| Average SNH Index (%) | 7.4 (0.04) | 3.1 (0.01)a | 5.6 (0.01)a | 7.7 (0.01)a | 13.1 (0.14) |
| Medicaid revenue share (%) | 10.6 (0.09) | 4.2 (0.03)a | 7.9 (0.04)a | 10.8 (0.05)a | 19.3 (0.13) |
| Low-income Medicare share (%) | 4.7 (0.06) | 1.0 (0.02)a | 2.9 (0.03)a | 4.9 (0.04)a | 10.2 (0.08) |
| Uncompensated care share (%) | 6.9 (0.05) | 4.1 (0.02)a | 6.1 (0.03)a | 7.6 (0.04)a | 10.1 (0.08) |
| Binary SNH definitions (%) | |||||
| Medicare DSH | 24.2 (0.43) | 5.6 (0.23)a | 16.1 (0.37)a | 26.5 (0.44)a | 50.0 (0.50) |
| Medicaid revenue | 24.5 (0.43) | 2.8 (0.16)a | 11.1 (0.31)a | 26.9 (0.44)a | 57.1 (0.50) |
| Uncompensated care costs | 24.5 (0.43) | 6.7 (0.25)a | 20.5 (0.40)a | 29.6 (0.46)a | 42.6 (0.49) |
| Public hospitals | 21.1 (0.41) | 22.9 (0.42) | 19.7 (0.40) | 21.1 (0.41)a | 22.4 (0.42 |
| Average SVI | 0.52 (0.27) | 0.36 (0.25)a | 0.46 (0.25)a | 0.57 (0.24)a | 0.68 (0.22) |
Source: 2023 RAND Medicare Cost Report, 2018 CDC/ATSDR SVI.
Abbreviations: DSH, Disproportionate Share Hospital, SNH, Safety Net Hospital; SVI, Social Vulnerability Index.
Notes: 171 hospitals are excluded from SNH Index sample because they have missing values for at least 1 of the SNH Index components.
Medicare DSH: hospitals in the top quartile of Medicare DSH adjustment, by state.
Medicaid Revenue: hospitals in the top quartile of Medicaid discharges, by state.
Uncompensated care costs: hospitals in the top quartile of UCC as a share of operating expenses, by state.
Public hospitals: all government owned hospitals.
a P < 0.01, where the top SNH Index quartile is the reference group.
The bottom panel of Table 1 demonstrates that the SNH Index captures multiple dimensions of a hospital's safety net role and, therefore, is less likely to narrowly identify an SNH. Except for public ownership, the share of hospitals classified as SNHs under various binary definitions and the average SVI increases steadily across index quartiles.
Table 2 compares hospital characteristics across SNH Index quartiles. Compared to those in the bottom quartile, hospitals in the top quartile are more likely to be in urban areas (65.7% vs 52.2%) and are more concentrated in the South (43.7% vs 27.0%) and West (28.4% vs 15.0%). Government ownership rates are relatively consistent across quartiles, but hospitals in the top quartile are more likely to be for-profit than those in the other 3 quartiles, which is consistent with for-profit hospitals being concentrated in the South. Additionally, hospitals in the top 3 quartiles tend to be larger and more likely to be part of multi-hospital systems than those in the bottom quartile.
Table 2.
Hospital characteristics, overall and by SNH Index quartile, 2023.
| Mean (SD) | |||||
|---|---|---|---|---|---|
| SNH Index quartile | |||||
| All hospitals | 1 (bottom quartile) | 2 | 3 | 4 (top quartile) | |
| Hospitals, No. | 4494 | 1081 | 1081 | 1081 | 1080 |
| Geographic indicators (%) | |||||
| Urban | 59.4 (0.49) | 52.2 (0.50)a | 55.7 (0.50) | 61.2 (0.49) | 65.7 (0.47) |
| Rural | 40.6 (0.49) | 47.8 (0.50)a | 44.3 (0.50) | 38.8 (0.49) | 34.3 (0.47) |
| Medicaid expansion state | 67.4 (0.47) | 62.2 (0.49)a | 70.0 (0.46) | 66.4 (0.47) | 70.8 (0.45) |
| Region (%) | |||||
| Northeast | 12.1 (0.33) | 11.0 (0.31)b | 12.6 (0.33) | 10.6 (0.31) | 14.4 (0.35) |
| Midwest | 29.8 (0.46) | 47.0 (0.50)a | 37.7 (0.50) | 24.2 (0.43) | 13.5 (0.34) |
| South | 38.1 (0.49) | 27.0 (0.44)a | 31.6 (0.47) | 45.9 (0.50) | 43.7 (0.50) |
| West | 20.0 (0.40) | 15.0 (0.36)a | 18.1 (0.39) | 19.2 (0.39) | 28.4 (0.45) |
| Ownership (%) | |||||
| Non-profit | 61.6 (0.49) | 60.2 (0.49)a | 68.9 (0.46) | 63.9 (0.48) | 54.0 (0.50) |
| For profit | 17.3 (0.38) | 16.8 (0.37)a | 11.4 (0.32) | 15.0 (0.36) | 23.6 (0.42) |
| Government | 21.1 (0.41) | 22.9 (0.42) | 19.7 (0.40) | 21.1 (0.41) | 22.4 (0.42) |
| Bed size (%) | |||||
| Small (<100 beds) | 57.1 (0.50) | 74.2 (0.44)a | 56.3 (0.50) | 50.9 (0.50) | 46.1 (0.50) |
| Medium (100–299 beds) | 27.7 (0.45) | 18.0 (0.38)a | 27.6 (0.45) | 30.1 (0.46) | 36.7 (0.48) |
| Large (300+ beds) | 15.2 (0.36) | 7.8 (0.27)a | 16.1 (0.37) | 19.1 (0.39) | 17.2 (0.38) |
| Affiliated health system, No. | 3329 | 692 | 846 | 874 | 835 |
| Affiliated health system (%) | 74.1 (0.44) | 64.0 (0.48)a | 78.3 (0.41) | 80.9 (0.39) | 77.3 (0.42) |
Source: 2023 RAND Medicare Cost Report, 2018 CDC/ATSDR SVI, 2018 AFS.
Abbreviation: SNH, Safety Net Hospital.
Notes: 171 hospitals are excluded from SNH Index sample because they have missing values for at least 1 of the SNH Index components. System definition based on affiliation with multi-hospital health system.
a P < 0.01, for comparisons between the bottom and top SNH Index quartiles.
b P < 0.05, for comparisons between the bottom and top SNH Index quartiles.
Since hospital payer and patient mix are not fixed over time, we also examined the stability of the SNH Index across years. Appendix Table S5 compares the distribution of hospitals across quartiles in 2023 to the distribution computed with 2022 data. We see that 81.6% of the hospitals in the highest quartile in 2023 were also in that quartile the previous year. Another 14.2% of hospitals in the highest SNH quartile in 2023 appeared in the third quartile in 2022. Although the findings show the quartiles are not static, we see that there is not very much movement across the entire SNH Index range between years.
Table 3 shows how the SNH Index is negatively related to hospitals’ financial performance. Three key findings emerge. First, profitability declines along the SNH Index continuum. Operating margins fall from 3.1% and 2.7% in the bottom 2 quartiles to 1.4% in the third quartile and −2.0% in the top quartile. Total margins follow a similar pattern, dropping from 5.4% in the bottom quartile to 1.2% in the top. Second, hospital liquidity deteriorates across the SNH Index continuum. Average DCOH declines from 158 days in the bottom quartile to 136, 117, and 91 days in the second, third, and top quartiles, respectively. Third, the SNH Index is positively related to the likelihood of a hospital being in financial distress. The average modified Altman Z-score, and each of its components (Appendix Table S6), generally decreases as you go from the bottom to the top quartile of the SNH Index.16 This corresponds with an increased likelihood of being in financial distress: Just over 1 in 5 hospitals in the bottom 2 quartiles are in financial distress compared to 27.2% in the third quartile and 36.4% of hospitals in the top quartile.
Table 3.
Hospital financial characteristics, overall and by SNH Index quartile, 2023.
| Mean (SD) | |||||
|---|---|---|---|---|---|
| SNH Index quartile | |||||
| All hospitals | 1 (bottom quartile) | 2 | 3 | 4 (top quartile) | |
| Hospitals, No. | 4494 | 1081 | 1081 | 1081 | 1080 |
| Profitability | |||||
| Operating margins (%) | 1.10 (0.16) | 3.10 (0.14)a | 2.70 (0.14)a | 1.40 (0.16)a | −2.00 (0.19) |
| Total margins (%) | 3.60 (0.14) | 5.40 (0.13)a | 4.90 (0.13)a | 3.40 (0.14)a | 1.20 (0.16) |
| Liquidity: days cash on hand | 127 (171.5) | 158 (176.4)a | 136(176.1)a | 117 (174.8)a | 91 (149.4) |
| Financial distress | |||||
| Altman Z-score | 6.4 (10.8) | 7.2 (10.4)a | 7.3 (10.3)a | 6.3 (10.7) | 5.4 (11.6) |
| Share in financial distress (%) | 27.8 (0.45) | 21.3 (0.41)a | 23.9 (0.45)a | 27.2 (0.45)a | 36.4 (0.48) |
Source: 2023 RAND Medicare Cost Report.
Abbreviation: SNH, Safety Net Hospital.
Notes: 171 hospitals are excluded from SNH Index sample because they have missing values for at least 1 of the SNH Index components. Days cash on hand excludes for-profit hospitals. The Altman Z-score incorporates measures of hospital liquidity, profitability, financial efficiency, and solvency to categorize hospitals as distressed and at high risk for bankruptcy. The share in financial distress is an indicator of hospitals having an Altman Z-score of less than 1.8.
a P < 0.01, where the top SNH Index quartile is the reference group.
Consistent with these findings, Appendix Table S7 shows the SNH Index is negatively correlated with operating margins, total margins, DCOH, and the modified Altman Z-score (which represents being in a better financial position), and positively correlated with the indicator for being in financial distress.16 The SNH Index and modified MSNI are more strongly correlated with the financial measures compared to binary SNH measures, with the SNH Index having the largest absolute correlation coefficients among all measures.
Table 4 examines the role of health systems in hospital financial performance by splitting hospitals in each SNH Index quartile into system vs non-system affiliation. Within each SNH Index quartile, hospitals affiliated with health systems have higher profitability than independent hospitals. In the top quartile, system-affiliated hospitals have an average operating margin of −0.7%, compared to −6.2% for independent hospitals and total margins are 1.7% for system-affiliated hospitals vs −0.5% for independents. Similar gaps in financial performance between system-affiliated and independent hospitals persist in the other 3 quartiles of the SNH Index. System-affiliated hospitals tend to have lower average DCOH than their independent counterparts within each quartile of the index, possibly reflecting centralized financial management and resource pooling within systems.
Table 4.
Hospital financial characteristics, overall and by SNH Index quartile and system status, 2023.
| Mean (SD) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| SNH Index quartile | ||||||||||
| All hospitals | 1 (bottom quartile) | 2 | 3 | 4 (top quartile) | ||||||
| System | Non-system | System | Non-system | System | Non-system | System | Non-system | System | Non-system | |
| Hospitals, No. | 3329 | 1165 | 692 | 389 | 846 | 235 | 874 | 207 | 835 | 245 |
| Profitability | ||||||||||
| Operating margins (%) | 3.0 (0.15) | −4.3 (0.17) | 5.9 (0.13) | −2.0 (0.15) | 4.3 (0.13) | −3.2 (0.14) | 3.0 (0.15) | −5.7 (0.17) | −0.7 (0.19) | −6.2 (0.18) |
| Total margins (%) | 4.7 (0.14) | 0.3 (0.14) | 7.3 (0.13) | 2.1 (0.12) | 5.8 (0.13) | 1.3 (0.11) | 4.4 (0.14) | −1.0 (0.15) | 1.7 (0.16) | −0.5 (0.15) |
| Liquidity: days cash on hand | 114 (173.8) | 157 (163.1) | 157 (200.9) | 158 (131.6) | 125 (178.6) | 163 (166.6) | 102 (166.5) | 159 (190.3) | 70 (134.4) | 144 (171.3) |
| Financial distress | ||||||||||
| Altman Z-score | 7.6 (60.7) | 6.8 (20.4) | 11.1 (21.2) | 9.4 (29.7) | 5.9 (90.3) | 7.1 (8.5) | 7.2 (69.0) | 5.0 (18.1) | 6.8 (29.8) | 6.2 (10.5) |
| Financial distress (%) | 16.7 (0.37) | 17.5 (0.38) | 10.6 (0.31) | 10.9 (0.31) | 13.0 (0.34) | 12.3 (0.34) | 16.9 (0.38) | 17.5 (0.38) | 24.9 (0.43) | 24.7 (0.43) |
Source: 2023 RAND Medicare Cost Report, 2018 Audited Financial Statements.
Abbreviation: SNH, Safety Net Hospital.
Notes: 171 hospitals are excluded from SNH Index sample because they have missing values for at least 1 of the SNH Index components. Days cash on hand excludes for-profit hospitals. The Altman Z-score incorporates measures of hospital liquidity, profitability, financial efficiency, and solvency to categorize hospitals as distressed and at high risk for bankruptcy. The share in financial distress is an indicator of hospitals having an Altman Z-score of less than 1.8.
Additionally, across all SNH quartiles, hospitals affiliated with a health system are larger, more likely to be in urban areas, and are more likely to be not-for-profit and are less likely to be government owned compared to hospitals that are not affiliated with a health system (Appendix Table S8).18
Discussion
This study highlights the complexity of identifying SNHs and the implications of this challenge for policy design. Our analysis underscores this point and offers an alternative. We developed a continuous SNH Index that captures the degree to which hospitals provide uncompensated care and serve Medicaid and low-income Medicare beneficiaries. High SNH Index values correlate strongly with a range of binary SNH indicators and SVI scores but capture multiple dimensions in a single metric. For example, half of hospitals in the top quartile of our index receive Medicare DSH payments, serve large shares of Medicaid patients and are major providers of uncompensated care, indicating that the index captures a broader picture of SNHs than any of these binary measures alone.
Our analysis also underscores the financial vulnerability of hospitals with high SNH Index values. We show that in 2023, prior to the passage and implementation of OBBBA, hospitals’ financial performance worsened as their SNH roles increased. These findings suggest that across-the-board payment reductions—such as site-neutral reforms or limits on Medicaid supplemental payments—would disproportionately affect hospitals that play a critical safety net role. Although the OBBBA included a $50 billion rural hospital fund to offset Medicaid cuts, our findings suggest that urban SNHs are more likely to face financial distress, echoing concerns raised by recent analyses.21
Our results also indicate that the consolidation that has reshaped the hospital landscape and created more health care systems has also emerged as an important factor in mitigating financial vulnerability among SNHs. System-affiliated hospitals now outnumber independents by 3 to 1. Understanding system-level financial resources is critical, as systems often retain substantial income, assets, and liabilities at the parent level, and may report multi-billion-dollar investment portfolios and related income only at the system level. This suggests that system-level efficiencies—such as centralized purchasing and staffing—or internal revenue transfers may buffer SNHs from the most severe financial strains. While current policies that protect the SNHs focus on hospital-level metrics to identify those in need, policymakers should not assume uniform vulnerability across SNHs and overlook the potentially mitigating effect of system membership.
Efforts to protect SNHs should move beyond binary classifications and adopt composite measures that reflect multiple dimensions of safety net functions. For example, CMS exempted Sole Critical Access Hospitals in the 2026 OPPS rule that will reduce reimbursement for drug administration services when provided in off-campus hospital outpatient departments.5 Our results suggest that this exemption criterion could be too narrow and that the SNH Index offers a complementary tool for providing targeted financial protection.
Moreover, under future potential site-neutral payment reforms, policymakers could exempt additional providers based on SNH Index values or use the index to calibrate subsidies, eg, fully offset revenue losses for hospitals with the highest SNH burden while providing smaller offsets for those with lower index values. Such an approach would promote equity and efficiency, avoiding both the under-protection of high-burden hospitals and overcompensation of those with limited safety net roles.
Supplementary Material
Contributor Information
Fredric Blavin, Policy Development and Research, Blue Cross Blue Shield Association, Washington, DC 20001, United States.
Avani Pugazhendhi, Health Policy Division, Urban Institute, Washington, DC 20024, United States.
Stephen Zuckerman, Health Policy Division, Urban Institute, Washington, DC 20024, United States.
John Holahan, Health Policy Division, Urban Institute, Washington, DC 20024, United States.
Supplementary material
Supplementary material is available at Health Affairs Scholar online.
Funding
This work was funded by Arnold Ventures, Grant ID 24-12119.
Conflicts of interest
Please see ICMJE form(s) for author conflicts of interest. These have been provided as supplementary materials.
None of the authors have conflicts of interest related to this work. The necessary ICMJE forms have been submitted.
Notes
- 1. Hefner JL, Hogan TH, Opoku-Agyeman W, Menachemi N. Defining safety net hospitals in the health services research literature: a systematic review and critical appraisal. BMC Health Serv Res. 2021;21(1):278. 10.1186/s12913-021-06292-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Chatterjee P, Sommers BD, Joynt Maddox KE. Essential but undefined—reimagining how policymakers identify safety-net hospitals. N Engl J Med. 2020;383(27):2593–2595. 10.1056/NEJMp2030228 [DOI] [PubMed] [Google Scholar]
- 3. Powell WR, Hansmann KJ, Carlson A, Kind AJH. Evaluating how safety-net hospitals are identified: systematic review and recommendations. Health Equity. 2022;6(1):298–306. 10.1089/heq.2021.0076 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. New York State Senate . Bill S705: An act to amend the public health law, in relation to regulating commercial reimbursement rates for certain health care services. Albany (NY): New York State Senate; 2025. Accessed April 22, 2026. https://nysenate.gov/legislation/bills/2025/S705
- 5. Centers for Medicare & Medicaid Services . Medicare program: hospital outpatient prospective payment and ambulatory surgical center payment systems; quality reporting programs; overall hospital quality star rating; hospital price transparency; and notice of closure of a teaching hospital and opportunity to apply for available slots. Fed Regist. 2025;90:53448–. Published November 25, 2025. Accessed April 28, 2026. https://www.federalregister.gov/documents/2025/11/25/2025-20907/medicare-program-hospital-outpatient-prospective-payment-and-ambulatory-surgical-center-payment
- 6. Chatterjee P, Liao JM, Amagai K, Zhao Y, Shirk T, Navathe AS. Variation, overlap, and stability in defining safety net hospitals. JAMA Netw Open. 2025;8(7):e2523923. 10.1001/jamanetworkopen.2025.23923 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. McHugh M, Kang R, Hasnain-Wynia R. Understanding the safety net: inpatient quality of care varies based on how one defines safety-net hospitals. Med Care Res Rev. 2009;66(5):590–605. 10.1177/1077558709334895 [DOI] [PubMed] [Google Scholar]
- 8. Popescu I, Fingar KR, Cutler E, Guo J, Jiang HJ. Comparison of 3 safety-net hospital definitions and association with hospital characteristics. JAMA Netw Open. 2019;2(8):e198577. 10.1001/jamanetworkopen.2019.8577 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. McNeill E, Cronin C, Puro N, Franz B, Silver D, Chang J. Variance of US hospital characteristics by safety-net definition. JAMA Netw Open. 2023;6(9):e2332392. 10.1001/jamanetworkopen.2023.32392 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Medicare Payment Advisory Commission . Report to the Congress: Medicare Payment Policy. Chapter 3: Hospital acute inpatient services payment system. Washington (DC): MedPAC; 2023 March. Accessed April 22, 2026. https://www.medpac.gov/document/march-2023-report-to-the-congress-medicare-payment-policy/
- 11. Congressional Budget Office . Estimated budgetary effects of H.R. 1, the One Big Beautiful Bill Act. Washington (DC): CBO; 2025. Accessed April 22, 2026. https://www.cbo.gov/publication/61570
- 12. Kaiser Family Foundation . Allocating CBO's estimates of federal Medicaid spending reductions across the states: enacted reconciliation package. San Francisco (CA): KFF. Accessed April 22, 2026. https://www.kff.org/medicaid/allocating-cbos-estimates-of-federal-medicaid-spending-reductions-across-the-states-enacted-reconciliation-package/
- 13. Blavin F, Buettgens M, Simpson M. Hospital revenue losses and increased uncompensated care if Medicaid funding is cut. Washington (DC): Urban Institute; 2025 March. Accessed April 22, 2026. https://www.urban.org/research/publication/hospital-revenue-losses-and-increased-uncompensated-care-if-medicaid-funding-cut
- 14. Levinson Z, Godwin J, Hulver S, Neuman T. Ten things to know about consolidation in health care provider markets. San Francisco (CA): Henry J. Kaiser Family Foundation. 2024 April 19. Accessed April 22, 2026. https://www.kff.org/health-costs/ten-things-to-know-about-consolidation-in-health-care-provider-markets/
- 15. James J, Moscovice IG, Harrington M, Platonova E. The rural hospital and health system affiliation landscape: a brief review. Iowa City (IA): RUPRI Center for Rural Health Policy Analysis; 2024 April. Accessed April 22, 2026. https://rupri.org/wp-content/uploads/Health-System-Affiliation-Landscape.Finalversion.April-2024.pdf
- 16. Bozovich L, Knocke K, Pink GH, Howard A, Reiter KL. The association between system affiliation and financial performance in critical access hospitals. Chapel Hill (NC): Flex Monitoring Team; 2021 October. Accessed April 22, 2026. https://www.flexmonitoring.org/publication/association-between-system-affiliation-and-financial-performance-critical-access
- 17. Friedson AI, Jung J, Nicosia N, Karaca-Mandic P. RAND hospital data: web-based tool. Santa Monica (CA): RAND Corporation; 2018. Accessed April 22, 2026. https://www.rand.org/pubs/tools/TL303.html
- 18. Kane N, Berenson BA, Blanchfield B, Blavin F, Arnos D, Zuckerman S. Why policymakers should use audited financial statements to assess health systems’ financial health. J Health Care Finance. 2021;48(1):1–26. [Google Scholar]
- 19. Harvard T.H. Chan School of Public Health, Health Quality . Medicaid cuts likely to affect urban safety-net hospitals. Boston (MA): Harvard T.H. Chan School of Public Health; 2025 November 18. Accessed April 22, 2026. https://hsph.harvard.edu/health-quality/news/medicaid-cuts-likely-to-affect-urban-safety-net-hospitals/
- 20. Agency for Healthcare Research and Quality . Social determinants of health database. Rockville (MD): Agency for Healthcare Research and Quality. Accessed April 22, 2026. https://www.ahrq.gov/sdoh/data-analytics/sdoh-data.html
- 21. Kliff S, Upbin B. Urban hospitals may be hit especially hard by Medicaid cuts. New York Times. 2025 November 18. Accessed April 22, 2026. https://www.nytimes.com/2025/11/18/upshot/urban-hospitals-medicaid-cuts.html
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.