• Center on Health Equity & Access
  • Clinical
  • Health Care Cost
  • Health Care Delivery
  • Insurance
  • Policy
  • Technology
  • Value-Based Care

Experience Incentivizing Reduction of Racial and Ethnic Disparities in a Medicaid Hospital Quality Incentive Program

Publication
Article
The American Journal of Managed CareApril 2023
Volume 29
Issue 4

Construction of a composite measure, use of a summary disparity statistic, and measure selection are key considerations in the design of equity-focused payment programs.

ABSTRACT

Objectives: We aimed to describe the experience of a state Medicaid agency incentivizing reduction of racial and ethnic disparities in a hospital quality incentive program (QIP).

Study Design: Retrospective review of a decade of experience implementing a hospital health disparity (HD) composite measure.

Methods: Observational analysis of programwide trends in missed opportunity rates and between-group variance (BGV) for the HD composite from 2011 to 2020 and subanalysis of 16 metrics included in the HD composite for at least 4 years over the decade.

Results: Programwide missed opportunity rates and BGV fluctuated widely from 2011 to 2020, likely due to variation in measures included in the HD composite. When the 16 measures that were included in the HD composite for at least 4 years were collapsed into a hypothetical 4-year period, missed opportunity rates decreased across the 4 consecutive years, from 47% in year 1 to 20% in year 4. Differences among racial and ethnic subgroups also decreased across the 4-year period, as reflected in the BGV decrease from 7.85 × 10–4 in year 1 to 5.10 × 10–4 in year 4.

Conclusions: Construction of a composite measure, use of a summary disparity statistic, and measure selection are key considerations in the design and interpretation of equity-focused payment programs. This analysis revealed improved aggregate quality performance and a modest reduction in racial and ethnic disparities for measures included in the HD composite for at least 4 years. Further research is needed to evaluate the association between equity-oriented incentives and health disparities.

Am J Manag Care. 2023;29(4):e124-e128. https://doi.org/10.37765/ajmc.2023.89353

_____

Takeaway Points

The Massachusetts Medicaid and Children’s Health Insurance Program (administered by MassHealth) is one of the few payers with longitudinal experience holding provider entities financially accountable for health equity, specifically through its hospital quality incentive program (QIP).

  • Construction of a composite measure and use of a summary disparity statistic are key considerations in the design of equity-oriented QIPs.
  • This analysis revealed modest reductions in racial and ethnic differences for targeted quality measures across 4 years.
  • Observed trends do not suggest an association between equity incentives and health disparities reduction. Further research is needed to evaluate the impact of equity-oriented QIPs on health equity.

_____

Several public and private payers—including the Center for Medicare and Medicaid Innovation and various state Medicaid programs—have implemented or are designing value-based payment programs that incentivize the reduction of health disparities as a pillar of value alongside high-quality care.1-4 This shift has been spurred by recognition that existing value-based payment programs have had mixed effects on health equity and in fact may unintentionally exacerbate health disparities.4-10 However, experience to inform the design of equity-oriented value-based payment models remains sparse.

The Massachusetts Medicaid and Children’s Health Insurance Program (administered by MassHealth) is one of the few payers in the nation that has longitudinal experience holding provider entities financially accountable for disparities reduction. In 2007, as part of its acute hospital quality incentive program (QIP), MassHealth began requiring its hospital providers to report individual-level race and ethnicity data to support stratification of clinical quality measures. In 2009, a subset of these measures was aggregated into a health disparity (HD) composite measure. A composite was used to overcome small sample sizes for individual measures. Measures were selected for inclusion in the program if they were in priority quality areas and had sufficient variation and opportunity for improvement. Measures included in the HD composite evolved with changing clinical measure slates, with new measures introduced and topped-out measures retired each year.

Hospitals receive payment for the HD composite based on their composite between-group variance (BGV): a summary disparity statistic that measures the consistency of care provided across racial and ethnic groups for targeted measures. BGV is a nondirectional disparities statistic that measures variability in quality of care compared with the reference group (all racial and ethnic groups). Quality was calculated using missed opportunity rates, or the percentage of opportunities in which quality care was not delivered. Larger variations in missed opportunity rates among racial and ethnic subgroups yields higher BGVs. The BGV approach was selected to account for variability in sample sizes across hospitals that made it challenging to reliably show disparities with other approaches.11

BGV of the HD composite determined between 5% and 23% of the total annual hospital QIP from 2011 to 2020, with associated total dollars eligible to be earned ranging from $1.5 million to $17 million across 60 acute care hospitals. A hospital with a lower BGV than other hospitals earned higher payments based on decile attainment thresholds. As part of the overall QIP, hospitals also received payment for each individual measure in the HD composite based on improvement in missed opportunity rates alone. Thus, for measures included in the HD composite, hospitals were dually incentivized to provide both high-quality and consistent care across racial and ethnic groups.

This review examines experience of the program over the past decade. It reports on HD composite performance aggregated at the programwide level to describe observed racial and ethnic variability in missed opportunity rates (or BGV) across the hospitalized MassHealth population. To facilitate interpretation of the equity-oriented QIP, it also presents the BGV for measures included in the HD composite for at least 4 years between 2011 and 2020. Although this observational analysis cannot suggest causality between the equity-oriented QIP and health disparity trends, the program design, evaluation approach, and experience with a summary disparity statistic may offer useful information for payers that are incorporating health equity incentives into their value-based payment programs.

METHODS

Programwide HD composite performance was evaluated from 2011 to 2020 using HD composite BGV and missed opportunity rates for the hospitalized MassHealth population aggregated across 60 acute hospitals. Quality performance was calculated using chart-abstracted clinical and administrative data submitted annually by hospitals for MassHealth members. Race and ethnicity data were self-reported and validated through chart review. Programwide missed opportunity rates were calculated as the sum of all instances in which the desired care was not provided to a patient (in the numerator) divided by the total number of opportunities to provide desired care to patients (in the denominator). Programwide BGV was calculated as the sum of the difference between missed opportunity rates for each racial and ethnic subgroup and the overall missed opportunity rates (weighted by population of each subgroup). The BGV statistic is displayed in 6 decimal points and ranges from 0 to 1, with a BGV of 0 representing no variation among racial and ethnic groups in the care being measured and a BGV of 1 representing maximal variation. Further details on missed opportunity rates and BGV calculations are described in eAppendix A (eAppendices available at ajmc.com).12,13

The evolving measures included in the HD composite made it difficult to interpret year-to-year changes in that composite (eAppendix B). Therefore, we performed a subanalysis to better represent individual-level measure trends obscured by the BGV statistic. For this subanalysis, performance data for 16 measures included in the HD composite for at least 4 consecutive years (eg, 2011-2014 or 2016-2019) (eAppendix B) were combined into a single hypothetical 4-year period of analysis (“performance years 1-4”). A threshold of 4 years was chosen to include as many measures as possible while allowing for trend evaluation. For these 16 measures, we calculated programwide missed opportunity rates and BGV for the HD composite for performance years 1 to 4. Finally, we stratified programwide missed opportunity rates and BGV results for these 16 measures by thematic categories (eAppendix B).12,13 These data are a descriptive analysis of trends, and statistical analyses were not performed.

RESULTS

From 2011 to 2020, the HD composite was composed of 5 to 15 individual measures per year (with a total of 24 different measures across the 10-year period). Included measures represented care in 7 major domains: pediatric asthma care (n = 3), maternity care (n = 6), care coordination (n = 3), surgical infection care (n = 3), pneumonia care (n = 4), newborn care (n = 2), and tobacco use screening and treatment (n = 3) (eAppendix B). Measures contributed proportionally to the HD composite according to the total number of opportunities for quality care that they provided (the denominator). Care coordination measures made up the majority of total opportunities because those measures crossed all service lines (eAppendix C).

The programwide missed opportunity rate for the HD composite fluctuated from 2011 to 2020. In 2011 and 2012, the missed opportunity rate was 5% and increased to 41% in 2013. From 2013 to 2018, the missed opportunity rate decreased steadily to 17%, and from 2018 to 2020, it steadily increased to 25% (Figure 1 and eAppendix D). The programwide HD composite BGV also varied across the 10 years. In 2011 and 2012, BGV was lowest, at 2.9 × 10–5 and 8 × 10–6, respectively. The BGV increased to 2.124 × 10–3 in 2013 and steadily decreased to 1.53 × 10–4 by 2017. From 2017 to 2020, the BGV increased again to 1.804 × 10–3 (Figure 1 and eAppendix D).

Subanalysis of performance data for 16 individual measures collapsed into hypothetical performance years 1 to 4 revealed an overall decrease in missed opportunity rates, from 47% in year 1 to 20% in year 4 (Figure 2 [A]). When stratified by race and ethnicity, each subgroup also had a steady decrease in its missed opportunity rate during this period. Differences among racial and ethnic subgroups also decreased across the hypothetical 4-year period, as reflected in the BGV decrease from 7.85 × 10–4 in year 1 to 5.10 × 10–4 in year 4 (Figure 2 [A]). For each year, White and Asian members had higher missed opportunity rates compared with Black and Hispanic members (Figure 2 [B] and eAppendix E). The maximum absolute missed opportunity rate difference among racial and ethnic subgroups ranged from 6% to 12% each year (eAppendix E).

When stratifying this subanalysis by measure category, measures related to pediatric asthma and maternity care had trends similar to those of the 16-measure subanalysis (eAppendix F [A and B]). The newborn care measure (exclusive breastfeeding) had opposite trends, with the missed opportunity rate and BGV increasing across the 4 years. It was also among the few measures in which minorities had higher missed opportunity rates than White members and showed the most variation in care (eAppendix F [C]). For care coordination measures, the missed opportunity rate decreased across the 4 years but BGV fluctuated (eAppendix F [D]). For surgical infection and pneumonia care measures, missed opportunity rates were less than 5%, making missed opportunity rates and BGV trends difficult to interpret (eAppendix F [E and F]).

DISCUSSION

For more than a decade, MassHealth’s acute hospital QIP has included an HD composite that measures differences in receipt of high-quality care by patient race and ethnicity. As interest in incorporating measures of equity in value-based purchasing programs increases around the country, this program’s experience with measuring and incentivizing health disparities reduction in a QIP may offer some useful lessons.

The program’s approach to disparities measurement was not conducive to longitudinal interpretation of progress on equity. The BGV trend was not easily comparable year to year because the combination of quality measures included in the HD composite evolved annually (Figure 1 and eAppendix B [A]). Further, the financial incentives tied to disparities reduction also varied over time. The use of a consistent measurement approach with consistent financial incentives in a future program would aid in evaluating the program’s impact.

Although summary disparity statistics (such as BGV) have gained favor among payers due to their simple representation of variation, they have several limitations.2,14 The BGV measures uniformity of care and thus produces similar results regardless of which subgroup receives better care. In the subanalysis of 16 measures included in the HD composite for 4 consecutive years, White members had higher missed opportunity rates than members of racial and ethnic minority groups at baseline and thus benefited most from reductions in subgroup differences (Figure 2 [B]). If health equity programs aim to redress historical and contemporary injustices toward racial and ethnic minority groups, payers should intentionally select measures in which such groups fare worse or otherwise alter design to ensure that goals are met. This may require payers to have access to baseline measure data stratified by race and ethnicity or evidence of disparities for similar populations prior to measure selection.15 Moreover, BGV is sensitive to racial or ethnic case mix, meaning that hospitals with more minority group patients can receive worse results than those with fewer minority group patients because they will inherently have more opportunities for variation. Future health equity performance assessment efforts may benefit from selection of disparity statistics and other program design elements that safeguard against disadvantaging safety-net hospitals.

Limitations

This analysis has several limitations. To overcome the limitations of BGV interpretation outlined earlier, we performed a subanalysis of 16 measures included in the HD composite for at least 4 years, and trends revealed modest disparities reduction across the hypothetical 4-year period. However, for several reasons, these trends do not suggest association or causality between the inclusion of a measure in the HD composite and reductions in disparities: Given that all hospitals participated in the QIP, comparison to a logical control group and rigorous statistical analyses were not possible; given that measures in the subanalysis were collapsed into a hypothetical 4-year period, they were subject to different external influences, including varying incentive amounts and unmeasured population traits. Although the implications of this analysis are limited, they point to important trends in the theoretical value of QIP in reducing disparities that require well-controlled investigation. Another limitation is that programmatic data from 2009 and 2010 were excluded from the trend analysis because they were not available. The HD composite was tied to financial incentives for the first time in 2009 (pay for reporting) and 2010 (pay for performance). Because 10 of the 16 analyzed measures in the 16-measure subanalysis were part of the HD composite in 2009-2010, this analysis does not capture the baseline subgroup variations for these 10 measures (eAppendix B [A]). Finally, hospital data from 2018 were drawn from only 2 rather than 4 quarters. Thus, the number of potential opportunities was reduced by half, limiting the eligible denominator for analysis.

CONCLUSIONS

The recent collective imperative for action on health equity is increasingly prompting payers and health systems to consider incorporating health equity as an explicit component of value. This observational analysis identifies relevant considerations for equity-oriented QIP design and evaluation, including measure selection, construction of a composite measure, and use of a summary statistic. Further research is needed to evaluate the impact of equity incentives on disparities reduction.

Author Affiliations: Harvard Medical School (PE), Boston, MA; Telligen, Inc (CS), Waltham, MA; UMass Chan Medical School, Commonwealth Medicine (LS, CEF), Quincy, MA; Massachusetts Medicaid (MassHealth) (LS, CEF), Boston, MA; Department of Medicine and Center for Health Economics and Policy, Washington University School of Medicine (KEJM), St Louis, MO.

Source of Funding: Dr Joynt Maddox receives research support from the National Heart, Lung, and Blood Institute (R01HL143421) and National Institute on Aging (R01AG060935, R01AG063759, and R21AG065526).

Author Disclosures: Dr Joynt Maddox serves on the Health Policy Advisory Council for Centene Corp and has previously done contract work for HHS. 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 (PE, CS, LS, CEF); acquisition of data (CS, LS); analysis and interpretation of data (PE, CS, LS, KEJM, CEF); drafting of the manuscript (PE, CEF); critical revision of the manuscript for important intellectual content (CS, LS, KEJM); statistical analysis (PE, CS); administrative, technical, or logistic support (CS, KEJM, CEF); and supervision (KEJM, CEF).

Address Correspondence to: Parsa Erfani, MD, Harvard Medical School, 25 Shattuck St, Boston, MA 02115. Email: perfani@bwh.harvard.edu.

REFERENCES

1. Blue Cross Blue Shield of Massachusetts becomes first health plan in market to incorporate equity measures into its payment models. News release. Blue Cross Blue Shield of Massachusetts; September 23, 2021. Accessed February 6, 2022. https://bit.ly/40b7OC1

2. Comprehensive health care program for the Michigan Department of Health and Human Services. State of Michigan. 2021. Accessed February 6, 2022. https://bit.ly/3lx6Zo7

3. Spaan M, Faulkner D, Morales D. Webinar Q&A: addressing social determinants of health through Medicaid accountable care organizations: early state efforts. Center for Health Care Strategies. February 14, 2018. Accessed February 6, 2022. https://www.chcs.org/media/SDOH-webinar-QA_follow-up_20180501.pdf

4. Gondi S, Maddox KJ, Wadhera RK. “REACHing” for equity—moving from regressive toward progressive value-based payment. N Engl J Med. 2022;387(2):97-99. doi:10.1056/NEJMP2204749

5. Hsu HE, Wang R, Broadwell C, et al. Association between federal value-based incentive programs and health care–associated infection rates in safety-net and non–safety-net hospitals. JAMA Netw Open. 2020;3(7):e209700. doi:10.1001/jamanetworkopen.2020.9700

6. Khullar D, Schpero WL, Bond AM, Qian Y, Casalino LP. Association between patient social risk and physician performance scores in the first year of the Merit-based Incentive Payment System. JAMA. 2020;324(10):975-983. doi:10.1001/jama.2020.13129

7. Aggarwal R, Hammond JG, Joynt Maddox KE, Yeh RW, Wadhera RK. Association between the proportion of Black patients cared for at hospitals and financial penalties under value-based payment programs. JAMA. 2021;325(12):1219-1221. doi:10.1001/jama.2021.0026

8. Song Z, Rose S, Chernew ME, Safran DG. Lower- versus higher-income populations in the alternative quality contract: improved quality and similar spending. Health Aff (Millwood). 2017;36(1):74-82. doi:10.1377/hlthaff.2016.0682/

9. Nundy S, Cooper LA, Mate KS. The quintuple aim for health care improvement: a new imperative to advance health equity. JAMA. 2022;327(6):521-522. doi:10.1001/jama.2021.25181

10. Sandhu S, Saunders RS, McClellan MB, Wong CA. Health equity should be a key value in value-based payment and delivery reform. Health Affairs. November 25, 2020. Accessed February 6, 2022. https://www.healthaffairs.org/do/10.1377/forefront.20201119.836369/full/

11. Blustein J, Weissman JS, Ryan AM, Doran T, Hasnain-Wynia R. Analysis raises questions on whether pay-for-performance in Medicaid can efficiently reduce racial and ethnic disparities. Health Aff (Millwood). 2011;30(6):1165-1175. doi:10.1377/hlthaff.2010.1022

12. Massachusetts Executive Office of Health and Human Services. EOHHS technical specifications manual for MassHealth acute hospital quality measures (version 15.0). Commonwealth of Massachusetts. 2021. Accessed February 6, 2022. https://www.mass.gov/doc/ry2022-eohhs-technical-specifications-manual-for-masshealth-hospital-quality-measures-150-0/download

13. 20LCEHSACUTEHOSPITAL: rate year 2020 acute hospital request for application. COMMBUYS Operational Services Division. October 4, 2019. Accessed February 6, 2022. https://bit.ly/3lEjePu

14. Biggs TP, Maksut J, Wilson B. Health equity as a “new normal”: CMS efforts to address the causes of health disparities. Presented at: 2021 CMS Quality Conference; March 2-3, 2021. Accessed February 6, 2022. https://bit.ly/3lAj4Zw

15. Filice CE, Joynt KE. Examining race and ethnicity information in Medicare administrative data. Med Care. 2017;55(12):e170-e176. doi:10.1097/MLR.0000000000000608

Related Videos
Screenshot of Adam Colborn, JD during an interview
Screenshot of Stephanie Hsia, PharmD
Cesar Davila-Chapa, MD
Screenshot of an interview with Nadine Barrett, PhD
Female doctor in coat with stethoscope on blue background - Pixel-Shot - stock.adobe.com
Io Hui, PhD, researcher at The University of Edinburgh
Jonathan Kurman, MD
Scott Manaker,MD
Juan Carlos Martinez, MD
Related Content
© 2024 MJH Life Sciences
AJMC®
All rights reserved.