Health Care Financing – McCanne Health Justice Monitor

Mortality Effects and Choice Across Private Health Insurance Plans, The Quarterly Journal of Economics, May 6, 2021, Jason Abaluck, Mauricio Caceres Bravo, Peter Hull, and Amanda Starc.

*The abstract is available at https://doi.org/10.1093/qje/qjab017 While the full text is behind a paywall, an earlier version of the full paper is available here.

From the Abstract: “We develop and apply a novel instrumental variables framework to quantify the variation in causal mortality effects across plans and how much consumers attend to this variation. We first document large differences in the observed mortality rates of Medicare Advantage plans within local markets. We then show that when plans with high (low) mortality rates exit these markets, enrollees tend to switch to more typical plans and subsequently experience lower (higher) mortality. . . .

We then extend our framework to study other predictors of plan mortality effects and estimate consumer willingness to pay. Higher-spending plans tend to reduce enrollee mortality, but existing quality ratings are uncorrelated with plan mortality effects. Consumers place little weight on mortality effects when choosing plans. Good insurance plans dramatically reduce mortality, and redirecting consumers to such plans could improve beneficiary health.”

From the text: “We find that the most widely used measure of plan quality, CMS star ratings, is uncorrelated with plan mortality effects. Higher premium plans have better mortality effects, as do plans with more generous prescription drug coverage and higher medical-loss ratios. Thus, in every way we measure, plans that spend more tend to reduce enrollee mortality.”

***

Comment by David Himmelstein and Steffie Woolhandler

A simple message emerges from this complex econometric analysis: insurers that skimp on paying for care (as measured by a low medical-loss ratio – the share of premiums devoted to patient care) or provide highly restrictive drug coverage, are killing their enrollees.

The study examined the mortality rates of patients covered by different Medicare Advantage (MA) plans. Not surprisingly, mortality varied among plans. That could just reflect differences in how sick the enrollees were to begin with, not anything the insurers did.

The compelling finding is that when plans with high mortality rates stopped offering coverage in a particular county – forcing their enrollees to switch to a different plan – the mortality rate for those forced to switch went down. (Similarly, when low-mortality plans shut down, the mortality rate of the enrollees forced to look elsewhere for coverage went up.) In sum, the study provides quite convincing, quasi-experimental evidence that MA plans’ quality influences the mortality of their enrollees. Indeed, the effect is so large that shifting MA enrollees from the worst 5% of plans to the average (not even the best) plan would save about 12,000 lives each year.

The researchers then looked at what differed between high and low mortality MA plans. They found no correlation with the CMS star rating – a one star-rated plan was just as likely to do well by its enrollees as a five star plan. And, not surprisingly, seniors, who often look to these ratings were bad at choosing a plan with low mortality.

What made a difference – and a big one – was how much a plan spent on its enrollees. Plans that had a higher medical loss ratio (i.e. they spent more of their premiums on care, and less on overhead and profit) had lower mortality rates. (The medical loss ratio of MA plans averaged 83% in 2020 vs. traditional Medicare’s 97.7%). Additionally morality was lower in plans with better drug coverage and those with higher premiums.

As is often true of economists, these researchers drew the wrong conclusions from their important findings, clinging to the pro-market notion that additional tweaks could finally make commercially-driven health insurance work to patients’ advantage. They recommend that CMS improve its rating system to take account of MA plans’ mortality performance. As the authors note, however, private insurers would game any mortality-based rating system, just as they have successfully gamed CMS’ star system. The authors further suggest that health and life insurance be merged so that MA plans that kill people by squeezing care would have to pay out more in life insurance benefits.

They ignore the obvious better choice: a single payer system that would have a loss ratio of 97.7%, as well as better drug coverage than the best MA plan. The data indicate that would save tens of thousands of seniors’ lives annually.

Time and Financial Costs for Physician Practices to Participate in the Medicare Merit-based Incentive Payment System: A Qualitative Study, JAMA Health Forum, May 14, 2021, By Dhruv Khullar, Amelia M. Bond, Eloise May O’Donnell, Yuting Qian, David N. Gans, and Lawrence P. Casalino

“Participating in the MIPS program results in substantial financial and time costs for physician practices. We found that, on average, it cost practices $12 811 per physician to participate in MIPS in 2019. We found that physicians themselves spent a considerable amount of time to participate in MIPS. In 2019, physicians spent more than 53 hours per year on MIPS-related activities, which translates to nearly $7000 per physician. If physicians see an average of 4 patients per hour, then these 53 hours could be used to provide care for an additional 212 patients a year—equal to more than a full week’s work for a physician.”

***

Comment by Adam Gaffney

Pay-for-performance (P4P) is an increasingly central part of the American healthcare landscape. The Affordable Care Act added a multitude of new P4P programs to Medicare, including the Hospital Readmissions Reductions Program (HRRP) and the Hospital Value-Based Purchasing Program (HVBP). Then, the Medicare Access and CHIP Reauthorization of 2015 gave us the Merit-based Incentive Payment System (MIPS), a new P4P program that imposes financial sticks and carrots on individual clinicians across the country based on a slew of complicated performance metrics.

Much research suggests that these programs have little effect on patient outcomes. The HRRP was much lauded for apparently reducing readmissions, but later research attributed much (or all) of this apparent reduction to changes in diagnostic coding. There is also some evidence HRRP may have harmed some cardiac patients. Meanwhile, studies of the HVBP have found virtually no impact. Fewer studies, however, have examined the costs of such programs.

That’s what makes this study, led by Dr. Dhruv Khullar at the Weill Cornell Medical College, so valuable. The researchers interviewed the leaders of 30 physician practices across the nation who participated in the MIPS, and quantified the costs of participation in the program. Overall, they found that we spend more than $12,000 per physician annually to cover the administrative costs of participation in MIPS. Additionally, “MIPS-related activities” suck up over 200 hours of labor per year from practice staff, including 53.6 hours from frontline clinicians. And this is merely for a single P4P program.

There is little evidence, in other words, that P4P programs substantively improve care — and growing evidence that they further inflate our already enormous administrative costs while sapping the time and energy of practicing doctors. For these reasons, P4P should not be included in a Medicare for All reform. Notably, the House Medicare for All Bill excludes this payment mechanism. The underlying political idea of P4P is a fundamentally neoliberal one: the idea that we are all motivated only by pursuit of the dollar. Instead, doctors want to provide the best care they can. That it is not to say that there isn’t room for quality improvement in our healthcare system — far from it — but a paucity of profit incentives is not the culprit. Further, an increasing number of studies show that P4P is redistributive — shifting funds from providers that care for poorer patients (who tend to have worse outcomes) to the providers of the wealthy.