Trends in Market Share Among Hospitals Participating in ACOs and Nonparticipants

ABSTRACT

Objectives: To assess trends in market share (MS) over time among hospitals participating in Medicare accountable care organizations (ACOs) and non-ACO participants, and evaluate correlates of differences.

Study Design: Longitudinal study of US hospital and ACO data for 3534 short-term acute care hospitals ever participating in Medicare ACOs or not across hospital referral regions (HRRs) from Milliman Torch Insight (2011-2020).

Methods: Using HRR as the local market, we calculated 3 hospital MS measures using annual net patient revenue, discharges, and beds, and we calculated market concentration using a modified Herfindahl-Hirschman Index. We compared hospital characteristics by Medicare ACO participation. In longitudinal, multivariable ordinary least squares regressions, we examined the association among hospital MS, ACO participation, and market concentration.

Results: Thirty percent of hospitals (n = 1048) reported ever participating in ACOs. Across MS measures, ACO participants had larger MS, with a mean MS of 11.5 (vs 8.5 for nonparticipants) for net patient revenue, 11.4 (vs 8.4) for discharges, and 10.9 (vs 8.6) for beds. The difference in MS between ACO participants and nonparticipants was larger in more concentrated markets relative to less concentrated markets. ACO participation was associated with increases in mean MS of 1.7 percentage points (PP) for net patient revenue, 2.0 PP for discharges, and 1.6 PP for beds. ACO participation was associated with additional MS gains in more (vs less) concentrated markets. More importantly, baseline (2011) MS emerged as the most important predictor of MS growth.

Conclusions: Hospitals participating in ACOs still surpass nonparticipants on MS. Our finding that baseline MS is most predictive of future MS growth suggests that competition should be managed prior to ACO entry and participation should be closely monitored in concentrated markets. With competition fostering quality and improving patient welfare, MS gains associated with hospital ACO participation may suggest the need for future assessments to ensure that quality and patient welfare do not decrease over time.

Am J Manag Care. 2025;31(7):In Press

Takeaway Points

We assessed trends in market share over time among short-term acute care hospitals participating in accountable care organizations (ACOs) and non-ACO participants.

• ACO participation was associated with an increase in mean market share.
• ACO participation was associated with additional market share gains in more (vs less) concentrated markets.
• More importantly, baseline (2011) market share was the most important predictor of market share growth.

Accountable care organizations (ACOs) are groups of doctors, hospitals, and other health care providers that accept financial risk for the entire continuum of patient care. ACOs seek to promote care coordination to lower the cost of care while improving quality. From the ACO contract perspective, including a hospital may have various benefits, such as maintaining a referral network of hospital care and reducing costs (eg, from unnecessary hospitalizations).1,2 Prior studies have found that hospitals participating in ACOs varied in characteristics compared with nonparticipants. Early program research found that ACO participants were likely to be large, part of a health care system, teaching hospitals, not for profit, and located in more competitive and urban areas with higher income per capita and to have prior experience with risk-based payments and care management programs.3,4

As ACOs expand, there have been concerns that they may incentivize provider consolidation, particularly between physicians and hospitals and among physician groups. This may be worrisome because although consolidation reduces competition and increases market share and power, it was found to be associated with higher prices and lower care quality and patient satisfaction.5-10 Earlier studies investigating the first 3 years of the Medicare ACO implementation found that specialty practices grew in size after joining ACOs, and large practices increased in counties with higher ACO penetration vs counties without ACO penetration.11,12

ACO participants can change the list of their participating providers each year. Consequently, the provider landscape and market share (MS) of ACOs can change over time. Thus, it is crucial to assess the relationship between ACO participation and provider MS over the first decade of the program’s implementation. We expand on the existing work by using longitudinal data over 9 years of Medicare ACO implementation to examine trends in MS for hospitals participating in Medicare ACO programs and analyze the relationship among MS, ACO participation, and market concentration over time using national data.

STUDY DATA AND METHODS

Data

The main data source for this longitudinal study was the Milliman Torch Insight hospital and ACO data from 2011 to 2020 (eAppendix A [eAppendices available at ajmc.com]).13 The data included US short-term acute care hospital financial and other metrics collected from the CMS Healthcare Cost Report Information System.14 These Medicare-reported data provided information on hospital size with annual net patient revenues overall and from different payers (Medicare, Medicaid, private/self-pay) and annual discharges and beds. The Medicare ACO data in Torch Insight were collected from public reports. Also included in the data were hospital organizational structure, ownership, and participation in a Medicare ACO during the year.

Measures

Outcomes. Each hospital’s zip code was used to identify its hospital referral region (HRR), serving as regional market. The Dartmouth Atlas of Health Care defines 306 HRRs nationally, each representing a set of hospital service areas grouped according to referral patterns to hospitals in the region.15 HRRs were used to identify geographic distribution of care delivery.

We calculated 3 hospital MS measures based on the Medicare-reported hospital-level annual (1) net patient revenue, (2) discharges, and (3) beds. MS for a hospital was calculated as the ratio of the individual hospital’s MS measure (ie, net patient revenue, discharges, beds) to the sum of the MS of all hospitals within an HRR in a given year (eAppendix B).

Explanatory variables. The main exposure variable was defined dichotomously based on whether a hospital ever participated in a Medicare ACO (ie, hospital ACO participant) during our study period. We measured market concentration with a modified Herfindahl-Hirschman Index (HHI),16 resulting in 3 modified HHIs for net patient revenues, discharges, and beds (eAppendix B). HHI is frequently used to measure the level of health care competition in markets. The HHIs range from 0 to 10,000, indicating whether care in a market was concentrated or equally divided among hospitals. To avoid having a hospital’s market HHI as a function of that hospital’s MS (the outcome variable), the modified HHI was defined as the sum of squares of all hospitals’ MSs within a market (HRR), excluding the hospital of interest. Market concentration indicated whether a market was competitive (HHI < 1500), moderately concentrated (1500 ≤ HHI < 2500), or highly concentrated (HHI ≥ 2500).17 Our HHI ranges were based on the most updated guidelines for the study period (2010 guidelines).17

Covariates.We used binary indicators to characterize hospital ownership into 3 ownership categories (private for-profit, private not-for-profit, and other) and to indicate whether a hospital was affiliated with medical schools. We also reported the payer mix, reflected by the proportion of hospital revenue from different payers (eg, Medicare, Medicaid, and private/self-pay).

Statistical Analysis

We analyzed data for 3534 short-term acute care hospitals reporting net patient revenues, number of discharges, beds, their characteristics, and participation in Medicare ACOs from 2011 through 2020, for a total of 32,494 hospital-year observations.

Longitudinal analysis. Using longitudinal, multivariable ordinary least squares regression analyses, we examined associations of MS measures with hospital participation in the ACO overall and across different levels of market concentration (competitive, moderately concentrated, and highly concentrated), controlling for hospital characteristics, payer mix, and year effects (eAppendix B). Other regression models included an interaction between ACO participation and market concentration to assess variation across market competition levels.

We assessed heterogeneous relationships by evaluating whether the hospital MS changed at a different rate with ACO participation based on the hospital’s initial MS (larger vs smaller baseline MS). To achieve this, hospitals were categorized into quartiles based on their baseline MS in 2011, with the first/bottom quartile as smallest MS (reference), second quartile as medium-small MS, third quartile as medium-large MS, and fourth/top quartile as largest MS. Baseline MS categories were added to the model, and regressions were estimated, excluding 2011 data (eAppendix B). SEs were clustered at the hospital level in all regressions. Analyses were conducted using Stata 17 (StataCorp LLC).

In sensitivity analyses, we reestimated the main regression models after stratifying hospitals by their location into urban, suburban, and rural areas according to 2010 rural-urban commuting area codes.18 In these analyses, the statistical tests were 2-sided, with P values less than .05 considered to be statistically significant and P values of at least .05 but less than .10 considered to be marginally statistically significant.

RESULTS

Descriptive Analyses

Of the 3534 hospitals in our sample, 30% of hospitals (n = 1048) had ever participated in Medicare ACOs. Figure 1 shows that ACO participants had relatively larger MS than nonparticipants and that the difference in MS between participants and nonparticipants tended to be larger in more concentrated markets. The 3 outcome measures—MS based on net patient revenue, discharges, and beds—were highly and positively correlated (pairwise correlations ρ > 0.9; P < .01) (eAppendix Table 1).

We compared hospitals’ characteristics, market concentration level, and outcome measures by Medicare ACO participation across years. Results suggested significant heterogeneity across hospitals by ACO participation (Table). Hospitals ever participating in ACOs were statistically significantly more likely to be private not-for-profit (58.5% vs 38.9%) or have medical school affiliation (42.0% vs 30.9%) than nonparticipants.

The proportion of ACO participants and nonparticipants varied across market concentration levels. Additionally, the mean percentage of hospital revenue from Medicare was statistically significantly lower among ACO participants (22.6% vs 23.5% among nonparticipants). Compared with non-ACO participants, ACO participants had a mean MS of 11.5 (vs 8.5 for nonparticipants) for net patient revenue, 11.4 (vs 8.4) for discharges, and 10.9 (vs 8.6) for beds (Table).

Regression Analyses

Longitudinal regression analyses showed that ACO participation among hospitals was associated with increases of 1.7 percentage points for revenue MS, 2.0 percentage points for discharge MS, and 1.6 percentage points for bed MS (Figure 2 and eAppendix Table 2 [panel 1; columns 1, 3, and 5]). Higher market concentration was associated with higher MS across all measures. When considering the interaction between ACO participation and market concentration (eAppendix Table 2 [panel 1; columns 2, 4, and 6] and eAppendix Figure 1), ACO participation was associated with an increase in revenue and discharge MS of 2.4 to 2.5 percentage points in moderately concentrated markets and in discharge and bed MS of 2.4 to 2.8 percentage points in highly concentrated markets. In general, a higher share of hospital revenue from public payers was associated with lower MS.

Hospital baseline MS was highly predictive of MS post ACO implementation. Compared with hospitals with the smallest MS in 2011 (first/bottom quartile), having medium-small baseline MS (second quartile) was associated with an increase in subsequent MS of 0.67 to 1.35 percentage points (Figure 2, eAppendix Table 2 [panel 2; columns 1, 3, and 5], and eAppendix Figure 2) and having baseline MS in the largest MS category (fourth/top quartile) was associated with an increase in subsequent MS of 20.7 to 23.1 percentage points. These relationships were accentuated in concentrated markets (eAppendix Table 2 [panel 2; columns 2, 4, and 6]). Additional interactions between baseline MS, market concentration, and ACO participation showed no further evidence of heterogeneity across markets (eAppendix Table 2 [panel 2; columns 2, 4, and 6]).

Stratification analyses showed heterogeneity across geographical areas, with ACO participation being consistently associated with higher revenue and discharge MS in urban hospitals than in suburban or rural hospitals (eAppendix Table 3).

Consistent with expectations about the organizational design of ACOs, which creates a network of providers and facilitates integration among these providers, hospitals ever participating in ACOs had consistently higher MS across all measures relative to nonparticipating hospitals.

DISCUSSION

Longitudinal multivariable regression analyses revealed that ACO participation was associated with higher MS during our study period, further accentuated in more concentrated markets. This underscores the importance of competition as key to lowering costs.19 Potential mechanisms for ACO participants to increase their MS include the participants optimizing their network composition and reducing system leakage to outside providers (including non-ACOs) by improving alignment of providers and coordinating patients within the system.20 These strategies may be easier to implement in concentrated markets, where participants may already benefit from greater market power relative to more competitive markets, but will likely decrease overall competitiveness in the market and increase market power for these participants. ACOs can also use current and historical data from CMS to detect existing and new market opportunities. Additionally, hospitals joining ACOs could increase their competitive advantage by increasing the network breadth of their health information exchange partners, particularly in markets with high ACO penetration.21

An important finding from our study is the role played by the hospitals’ baseline MS in growth of MS. Our results on baseline MS suggest that regulating agencies such as the Federal Trade Commission22 that may be concerned about anticompetitive behavior of new ACO participants should be more concerned about hospital MS prior to entering ACOs and ACO participation in concentrated markets, as baseline MS appears to be the factor with the strongest correlation with MS growth.

Some hospitals may still find it difficult to join ACOs for several reasons. Prior research in the first year of the Medicare ACO programs indicated that the fear of potential losses in revenues and the lack of willingness to make up-front investments are reasons for hospitals not participating in ACOs.23 Moreover, the ACO participation process initially attracted participants who had high quality and existing processes to coordinate patient care across the continuum (and thus were in more favorable positions to succeed), and the model design makes delivering care in a densely populated market more efficient. Nonparticipating hospitals may be lacking these capabilities, may be located in less densely populated areas, or may need more time to adapt to new alternative payment models such as ACOs.3 Nonetheless, policy makers should provide additional incentives to encourage more ACO participation by hospitals—particularly targeted at smaller hospitals—and physician practices and increase competition among existing participants. Incentives to stimulate competition in hospital markets could help address health care affordability but could also face some challenges because some areas cannot support competition and many markets are highly concentrated.8

We also found that higher proportions of hospital revenues coming from public payers were associated with lower MS. Lower reimbursement rates of public payers could be a reason for these findings.

Hospital integration and consolidation have rapidly increased over the past decades with new market incentives.24,25 However, our results should not be interpreted as ACO participation being associated with higher consolidation because ACOs do not necessarily entail changing ownership, unlike consolidation via mergers. Nonetheless, ACO affiliations can be considered soft forms of consolidation and can still create similar problems as mergers.8 Our results on the role played by baseline MS on MS growth indicate that the Federal Trade Commission should continue enforcing antitrust laws and encourage and facilitate more hospital participation in ACOs in markets with lower penetration (eg, highly concentrated markets) to increase competition among participants.

Limitations

Our study has several limitations. We assessed changes in individual hospital MS, but broader hospital system–level MS, likely spanning multiple HRRs, could yield different findings. Additionally, nonrandom ACO participation and nonreporting of outcome input variables may bias our results, but we mitigated these issues by controlling for several hospital and market factors that influence hospital MS. Furthermore, we did not account for ACO participation duration.

Our results cannot be treated as causal, but rather they inform about a potential association between ACO participation and MS. It could be that hospitals with greater MS are more likely to participate in ACOs.

Our results also may not be generalizable to other types of hospitals such as critical access hospitals, which are smaller (≤ 25 beds). Nonetheless, short-term hospitals represent the majority of hospitals26 and were far more likely to participate in ACOs.3

Lastly, we identified hospital participation in ACOs but were unable to identify the role played by these hospitals in the ACO arrangement, although a considerable proportion of ACOs are hospital led.6 Therefore, we could not assess the difference in MS associated with hospital leadership in ACOs compared with regular participation.

CONCLUSIONS

We found that ACO participation was associated with small increases in hospital MS on average. Nonetheless, the most important predictor of hospital MS growth was the hospital’s baseline MS. Policies enforcing antitrust laws for ACO participants should continue targeting potential participants’ MS prior to participation and promote participation in markets with lower ACO penetration. With competition fostering quality and improving patient welfare, MS gains associated with hospital ACO participation may suggest the need for future assessments to ensure that quality and patient welfare do not decrease over time.

Acknowledgments

The data set forth as supplemental data (2019 zip code to health service area to hospital referral region crosswalk)15 were obtained from the Dartmouth Atlas Data website, which was funded by the Robert Wood Johnson Foundation, the Dartmouth Clinical and Translational Science Institute under award number UL1TR001086 from the National Center for Advancing Translational Sciences of the National Institutes of Health, and in part by the National Institute on Aging under award number U01 AG046830.

Author Affiliations: Department of Population Health Sciences, University of Wisconsin School of Medicine and Public Health (MHO, XL), Madison, WI; Health Management Associates (DBM), Lansing, MI; now with Simple Healthcare (DBM), Orlando, FL.

Source of Funding: None.

Author Disclosures: Dr Muhlestein worked at Health Management Associates, which consults with accountable care organizations, during part of the work on this paper. 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 (MHO, XL, DBM); acquisition of data (DBM); analysis and interpretation of data (MHO, XL, DBM); drafting of the manuscript (MHO, XL); critical revision of the manuscript for important intellectual content (MHO, DBM); statistical analysis (MHO, XL); administrative, technical, or logistic support (MHO); and supervision (MHO).

Address Correspondence to: Mariétou H. Ouayogodé, PhD, University of Wisconsin School of Medicine and Public Health, 610 Walnut St, Madison, WI 53726. Email: marietou.ouayogode@wisc.edu.

REFERENCES

1. Berwick DM, Hackbarth AD. Eliminating waste in US health care. JAMA. 2012;307(14):1513-1516. doi:10.1001/jama.2012.362

2. Chukmaitov A, Harless DW, Bazzoli GJ, Muhlestein DB. Preventable hospital admissions and 30-day all-cause readmissions: does hospital participation in accountable care organizations improve quality of care? Am J Med Qual. 2019;34(1):14-22. doi:10.1177/1062860618778786

3. Colla CH, Lewis VA, Tierney E, Muhlestein DB. Hospitals participating in ACOs tend to be large and urban, allowing access to capital and data. Health Aff (Millwood). 2016;35(3):431-439. doi:10.1377/hlthaff.2015.0919

4. Audet AM, Kenward K, Patel S, Joshi MS. Hospitals on the path to accountable care: highlights from a 2011 national survey of hospital readiness to participate in an accountable care organization. Issue Brief (Commonw Fund). 2012;22:1-12.

5. Dunn A, Shapiro AH. Do physicians possess market power? J Law Econ. 2014;57(1):159-193. doi:10.1086/674407

6. Austin DR, Baker LC. Less physician practice competition is associated with higher prices paid for common procedures. Health Aff (Millwood). 2015;34(10):1753-1760. doi:10.1377/hlthaff.2015.0412

7. Dafny L. Addressing consolidation in health care markets. JAMA. 2021;325(10):927-928. doi:10.1001/jama.2021.0038

8. Levinson Z, Godwin J, Hulver S, Neuman T. Ten things to know about consolidation in health care provider markets. KFF. April 19, 2024. Accessed July 24, 2024. https://www.kff.org/health-costs/issue-brief/ten-things-to-know-about-consolidation-in-health-care-provider-markets/

9. Liu JL, Levinson ZM, Zhou A, Zhao X, Nguyen P, Qureshi N. Environmental scan on consolidation trends and impacts in health care markets. Rand Health Q. 2023;10(3):2.

10. Beaulieu ND, Dafny LS, Landon BE, Dalton JB, Kuye I, McWilliams JM. Changes in quality of care after hospital mergers and acquisitions. N Engl J Med. 2020;382(1):51-59. doi:10.1056/NEJMsa1901383

11. Neprash HT, Chernew ME, McWilliams JM. Little evidence exists to support the expectation that providers would consolidate to enter new payment models. Health Aff (Millwood). 2017;36(2):346-354. doi:10.1377/hlthaff.2016.0840

12. Kanter GP, Polsky D, Werner RM. Changes in physician consolidation with the spread of accountable care organizations. Health Aff (Millwood). 2019;38(11):1936-1943. doi:10.1377/hlthaff.2018.05415

13. Our data. Torch Insight. Accessed April 11, 2023. https://torchinsight.com/data.html

14. Cost reports. CMS. Updated September 6, 2023. Accessed September 6, 2023. https://www.cms.gov/data-research/statistics-trends-and-reports/cost-reports

15. Supplemental data. Dartmouth Atlas Data. Accessed November 16, 2022. https://data.dartmouthatlas.org/supplemental/

16. Rhoades SA. The Herfindahl-Hirschman Index. Fed Res Bull. 1993;79:188-189.

17. Horizontal Merger Guidelines. US Department of Justice; Federal Trade Commission. August 19, 2010. Accessed August 19, 2024. https://www.ftc.gov/system/files/documents/public_statements/804291/100819hmg.pdf

18. Rural-urban commuting area codes. US Department of Agriculture Economic Research Service. Updated September 25, 2023. Accessed July 15, 2024. https://www.ers.usda.gov/data-products/rural-urban-commuting-area-codes

19. Baker LC, Bundorf MK, Kessler DP. Vertical integration: hospital ownership of physician practices is associated with higher prices and spending. Health Aff (Millwood). 2014;33(5):756-763. doi:10.1377/hlthaff.2013.1279

20. Sahni NR, Groh R, Nuzum D, Chernew M. The math of ACOs. McKinsey & Company. August 19, 2020. Accessed November 19, 2024. https://www.mckinsey.com/industries/healthcare/our-insights/the-math-of-acos#/

21. Apathy NC, Holmgren AJ, Werner RM. Growth in health information exchange with ACO market penetration. Am J Manag Care. 2022;28(1):e7-e13. doi:10.37765/ajmc.2022.88815

22. Federal Trade Commission. Statement of antitrust enforcement policy regarding accountable care organizations participating in the Medicare Shared Savings Program. US Department of Justice. October 28, 2011. Accessed January 2, 2024. https://www.govinfo.gov/content/pkg/FR-2011-10-28/pdf/2011-27944.pdf

23. Larking H. ACO or not? Hosp Health Netw. 2014;88(5):26-31.

24. Cutler DM, Morton FS. Hospitals, market share, and consolidation. JAMA. 2013;310(18):1964-1970. doi:10.1001/jama.2013.281675

25. Nikpay SS, Richards MR, Penson D. Hospital-physician consolidation accelerated in the past decade in cardiology, oncology. Health Aff (Millwood). 2018;37(7):1123-1127. doi:10.1377/hlthaff.2017.1520

26. Liu JB, Kelz RR. Types of hospitals in the United States. JAMA. 2018;320(10):1074. doi:10.1001/jama.2018.9471