Does the Merit-Based Incentive Payment System Disproportionately Affect Safety-Net Practices?

Three years since the introduction of the nationwide Merit-Based Incentive Payment System (MIPS), the US Centers for Medicare & Medicaid Services (CMS) continues using the program to incentivize clinicians to improve health care value for Medicare beneficiaries. The MIPS is a pay-for-performance program that increases payment rates for clinicians who provide high-value care, while penalizing clinicians who do not by decreasing their payment rates. The program’s design has been largely similar since launching in 2017. As in earlier years, participating clinicians are evaluated on performance in the domains of (1) quality (as assessed through clinical quality measures), (2) improvement activities (initiatives that can improve care), (3) costs (patients’ resource use and spending), and (4) promotion of interoperability (use of health information technology to improve care delivery). Performance in each domain is incorporated into an overall composite performance score that ultimately dictates whether clinicians receive upward, downward, or no payment rate adjustments.

Aggregate 2017 and 2018 results from CMS suggest that most individual clinicians have fared well thus far in MIPS, with 93% to 98% receiving an upward payment adjustment. However, those data do not account for the fact that many clinicians participate in MIPS through group practices rather than as individuals and MIPS appeared to disproportionately penalize small and rural practices in 2017.

Unfortunately, little is currently known about practice-level MIPS performance. In particular, it is unknown how safety-net practices (ie, those in which large proportions of care are provided to individuals with vulnerabilities) fared in MIPS. This dearth of knowledge is especially concerning given the history of safety-net practices being disproportionately penalized under other national Medicare payment reforms, such as the Hospital Readmissions Reduction Program and the Value-Based Purchasing Program.^1,2 Safety-net practices may face more barriers and possess fewer strategies for responding to such programs if large proportions of their care are for patients with social risk factors.³

The Payment Insights Team and the Value & Systems Science Lab track MIPS policy, including group practice participation,⁴ as part of a portfolio monitoring and evaluating value-based payment programs. We present an analysis of safety-net practices in MIPS and discuss key policy implications.

We combined year 1 MIPS data from CMS with data from the US Census Bureau, Physician Compare, and the Dartmouth Atlas to evaluate whether similar dynamics existed for safety-net practices under MIPS (ie, if they fared worse than other practices). Among 22 659 practices, we categorized 4845 (21.4%) as safety-net practices based on their location in counties defined by 2 or more of the following characteristics: low education levels (a population in the bottom quartile nationwide of individuals with some college education or more); low income (in the bottom quartile nationwide of median household income level); and high housing burden (in the bottom quartile nationwide of the proportion of individuals spending more than 50% of their income on housing).

In line with Medicare’s composite performance score thresholds and prior analysis,⁴ we categorized practices into performance groups described as low (those receiving neutral or negative adjustments), high (those receiving positive adjustments), and exceptional (those eligible for additional performance bonuses). Low performers were more likely to be safety-net practices: 1955 of 8148 low-performing practices (24.0%) were such, compared with 1189 of 5963 high-performing practices (19.9%) and 1701 of 8548 practices with exceptional performance (19.9%). Other practice characteristics associated with lower composite scores included a higher case mix, smaller practice size, and rural location. Regression results controlling for multiple practice characteristics—including practice size, urban or rural status, case mix, and patient population size—produced results that were similar in direction and corroborated the association between safety-net status and lower composite performance score.

The association between safety-net status and lower composite score performance extends concerns raised in other payment programs to MIPS and underscores the need to create an even playing field for practices serving patients who have social and/or clinical complexities and vulnerabilities. That these associations existed under the lenient year 1 MIPS rules, in which reporting alone was sufficient to avoid downward payment adjustments, also suggest that low awareness among programs may be a problem requiring greater clinician education or that barriers to participation (eg, data reporting) are onerous.

Policymakers could consider implementing several strategies if differences by safety-net status persist or widen in later years, when composite scores depend increasingly on performance, not just reporting, in the 4 MIPS domains. First, CMS could extend the customized technical assistance it provides to certain practices in MIPS. Currently, the agency offers several levels of support ranging from program-level support (eg, eligibility determination, data submission, activity and performance measure selection) to practice support (eg, readiness assessment, creation of partnerships with local and regional stakeholders) for small or rural practices in medically underserved areas. The CMS could expand this assistance to practices in which large proportions of clinical care are for socially or economically vulnerable groups.

Second, policymakers could adopt strategies used in other payment programs, such as the Hospital Readmissions Reduction Program, to compare MIPS performance within practice peer groups defined by safety-net status rather than across the entire MIPS program. Other potential policy solutions that account for outsized resource or capacity constraints among safety-net practices could include applying similar scoring flexibilities currently offered to small or rural practices, for instance providing partial scores for practices that do not meet data completeness requirements or incorporating socioeconomic factors into risk adjustment.

Insight from granular MIPS year 1 performance data suggests that safety-net practices may perform more poorly than their non safety-net counterparts. Policymakers should monitor for such dynamics and consider ways to adjust MIPS policy to ensure these practices are not inappropriately penalized by the program.

Open Access: This is an open access article distributed under the terms of the CC-BY License.

Corresponding author: Amol S. Navathe, MD, PhD, Department of Medical Ethics and Health Policy, Perelman School of Medicine, University of Pennsylvania, 1108 Blockley Hall, 423 Guardian Dr, Philadelphia, PA 19104 (amol@pennmedicine.upenn.edu).

Conflict of Interest Disclosures: Dr Liao reports textbook royalties and an honorarium from Wolters Kluwer and personal fees from Kaiser Permanente Washington Research Institute outside of the submitted work. Dr Navathe reported receiving grants from Hawaii Medical Service Association, Anthem Public Policy Institute, Commonwealth Fund, Robert Wood Johnson Foundation, Donaghue Foundation, Pennsylvania Department of Health, Healthcare Research and Education Trust, Cigna, and Oscar Health; personal fees from Navvis Healthcare, Agathos Inc, University Health System (Singapore), Embedded Healthcare, Social Security Administration (France), and Medicare Payment Advisory Commission; personal fees and equity from NavaHealth; equity from Embedded Healthcare; speaking fees from the Cleveland Clinic; service as a board member of Integrated Services Inc without compensation; an honorarium from Elsevier Press; and personal fees for service as a Commissioner on the Medicare Payment Advisory Commission outside of the submitted work.