MIPS is meant to encourage quality care, but is it preventing it?

September 29, 2020

Measuring high-quality care cannot be a barrier to providing it, notes John T. Thompson, MD.

This article was reviewed by John T. Thompson, MD

The Medicare Merit-based Incentive Payment System (MIPS) is a flawed system that cannot accurately distinguish between good and poor performing ophthalmologists, according to a presentation during the 2020 American Society of Retina Specialists virtual annual meeting.

John T. Thompson, MD, a partner at Retina Specialists, Baltimore, MD, and Chair of Health Policy at The American Society of Retina Specialists evaluated the limitations of MIPS—a program that was created in 2015 to incentivize healthcare providers to provide high-quality, cost-effective care—by modeling the expected differences between potential typical performance for all retina-related clinical quality outcome measures.

MIPS: A primer

Under MIPS, high-quality physicians are rewarded with higher Medicare reimbursements, and low-quality physicians receive reduced Medicare payments for all Medicare patients. But what defines a “high-quality physician” under MIPS?

“What makes a physician high quality? Is it diagnostic acumen, surgical prowess, patient satisfaction, excellent electronic medical record documentation, or inexpensive patient care,” Thompson said.

Early on in MIPS, quality measures were actually “process measures,” which, according to Thompson, meant “checking a box indicating good clinical care.” Examples include asking patients if they’ve received a flu shot, recommending AREDS supplements in patients with intermediate age-related macular degeneration, or examining the macula in diabetic patients.

That changed when Centers for Medicare & Medicaid Services (CMS) transitioned from process measures to outcome measures to assess quality because “process measures were becoming topped out with most retina specialists achieving high scores with bonuses,” Thompson said. “CMS believed measuring patient outcomes would allow them to distinguish good from bad physicians.”

Outcome measures for the 2020 performance year include such things as:

  • Return to the operating room within 90 days for retinal detachment;
  • Visual acuity (VA) improvement within 90 days of retinal detachment;
  • Losing less than 0.3 logMAR in patients with diabetic macular edema treated with anti-VEGF agents; and
  • Improved VA by 20% within 120 days of vitrectomy for epiretinal membrane.

Flawed rules, data collection

During his presentation, Thompson noted several critical shortcomings under MIPS.

“CMS takes the individual retina specialist outcomes and divides them into deciles, awarding 3 to 10 points. The problem is that the rules CMS has applied make it impossible for most retina specialists to use many of these measures,” Thompson said.

For example, data collection can only occur in a single calendar year, limiting the surgical timeframe to January through September for several outcome measures in order to complete the required followed-up period. Further, CMS requires a minimum of 20 completed cases during the measurement period.

“Most retina specialists don't perform 20 of these surgeries in the first 8 to 9 months of the calendar year,” Thompson said. “Even if they do, dividing the results into deciles and comparing two physicians leads to statistically meaningless results due to [the] small sample size.”

Thompson discussed other challenges as well, including a lack of risk adjustment for case complexity for any of the retina outcome measures, which encourages retina specialists to select easier cases. He also found errors in detecting eligible cases by registries and in categorizing outcomes that compromise the accuracy of the results.

“The noise introduced by these two sources of error may make comparisons between individual physicians inaccurate, even if the quality measures themselves were perfect,” he said.

Improvement is possible

Thompson suggested several actions CMS can take to improve MIPS, including pilot testing outcome measures, capturing outcomes through a 5-year rolling average, and re-evaluating the value of outcome measures as opposed to process measures.

Importantly, measuring high-quality care cannot be a barrier to providing it. Thompson expressed concern that the rules were inadvertently encouraging retinal specialists to refer out difficult cases and discouraging them to take on disadvantaged patients.

“Their outcomes are often not as good since they present with more advanced disease and may have more difficulty returning for appropriate follow-up care,” he said. “Quality of care is meaningless with the current rules and certainly does not justify paying one physician better than the other.”