Pay-For-Performance: Does it Really Work?

In theory, pay-for-performance programs seem simple enough—pay physicians and hospitals based on health outcomes or value of patient care, rather than on quantity of medical services provided. Congress, for example, passed the Medicare Access CHIP Reauthorization Act of 2015 (MACRA) that moves Medicare from a volume-based payment system to one that rewards value. The first performance period under MACRA took effect on January 1, 2017. The 2010 Affordable Care Act has put a lot of time and effort into a variety of new programs on pay-for-performance such as value-based purchasing and the Hospital Readmissions Reduction Program (HRRP). Yet, the evidence supporting pay-for-performance is mixed at best. Studies have shown marginal gains or failure of programs.

The Evidence

Cost of health care in the United States continues to escalate, and one expectation of pay-for-performance is that it will reduce costs. Although cost estimates are limited, regulations aimed to incentivize physicians for quality and efficiency cost physicians millions, according to a 2016 study published in Health Affairs. The researchers found that each year US physician practices in four common specialties, including primary care, spend, on average, 785 hours per physician and more than $15.4 billion just for documenting quality measures. While there is much to gain from quality measurement, “the current system is far from being efficient and contributes to negative physician attitudes toward quality measures,” they concluded.
______________________________________________________________________________________________________________________________________
Related Content
AGS CONTINUES EFFORTS TO SHAPE PAY-FOR-PERFORMANCE POLICY
Betting on Pay-for-Performance
_______________________________________________________________________________________________________________________________________

A four-group, multicenter, cluster, randomized trial investigated the effect of financial incentives to physicians, patients, or both on reducing low-density lipoprotein cholesterol (LDL-C) among patients with high cardiovascular disease risk. Physicians in the physician incentives group were eligible to receive up to $1024 whenever a patient met target cholesterol levels. Physicians in the control groups received no economic incentives to reach target cholesterol goals. The findings published in JAMA in 2015 found no real difference in reduction of LDL-C levels at 12 months between the two groups.

“Many prior evaluations of pay-for-performance programs have found limited program impact, and several authors have speculated the reason was the programs had too little money on the table to spark changes in the way physicians practice,” said Jessica Greene, PhD, professor and Luciano chair of health care policy, Marxe School of Public and International Affairs, Baruch College, in an interview with First Report Managed Care. Dr Greene and colleagues have tested the efficacy of an ambitious physician incentive program at Minnesota-based Fairview Health Services, an accountable care organization.

In 2011, Fairview began implementation of a team-based quality-focused compensation model that tied 40% of physician pay-for-performance on clinical-level quality outcomes in diabetes, cardiovascular disease, and certain evidence-based cancer screenings. The researchers examined how much incentivized quality metrics improved over 2 years and published the findings in Health Affairs in 2015.

“While we observed some improvement in quality outcomes, particularly for providers with low baseline performance, pay-for-performance did not prove to be a magic solution for improving quality,” explained Dr Greene. “There was greater improvement in other Minnesota delivery systems without very large pay-for-performance programs than in Fairview.”

Are hospitals fairing any better with pay-for-performance? A range of different pay-for-performance programs has shown that the impact has been “really disappointing,” according to Ashish Jha, MD, MPH, a physician, health policy researcher, and advocate for global health care reform. Dr Jha was one of three panelists who participated in a recent webinar on pay-for-performance sponsored by the Center for Health Journalism, University of Southern California Annenberg School for Communication & Journalism.

He cited a 2016 study published in the New England Journal of Medicine that examined the impact of HRRP on 30-day readmission rates among Medicare beneficiaries. From 2007 to 2015, readmission rates for targeted conditions declined from 21.5% to 17.8%, and rates for nontargeted conditions declined from 15.3% to 13.1%. “I would argue that readmissions is one of the very few good news stories [on pay-for-performance] and suggest incentives can move the needle,” Dr Jha said. He noted that “reducing readmissions is not the same as actually making care better. We don’t understand what has happened to allow the readmissions to drop. Are we doing a better job of care coordination or are we just denying people admission to the hospital if they come back to the emergency department within the first 30 days?”

Dr Jha shared the outcomes of a research he coauthored that looked at the impact of the Hospital Value-Based Purchasing program on 30-day mortality in US hospitals. The findings reported in 2016 in The BMJ showed no impact on mortality rates from 2011 to 2013 (11.2% vs 11.1%, respectively). During the same period, patient experience did not change significantly (69% vs 71%, respectively).

Why Doesn’t Pay-for-Performance Work?

Behavioral economists and other experts cite complex compensation designs, poor alignment of goals, and lack of clearly defined, actionable measures that can lead to failed efforts and unintended consequences, according to a Modern Healthcare article by Sabriya Rice, who participated in the webinar.

“There is a tension in designing pay-for-performance programs to make them comprehensive, but in doing so, the programs can become complex, which can undermine their effectiveness,” said Dr Greene.

Will increasing monetary incentives for pay-for-performance programs make them more effective? Dr Greene has her doubts. “More money does not seem to make pay-for-performance models more effective. At Fairview, there was greater improvement in quality outcomes that were easier to impact, like cancer screening, which had a smaller financial incentive than for vascular care, which was harder to impact but had a larger incentive,” she said. 

“Unfortunately, it’s not yet clear what will make the programs more effective. There is a lot of theorization about making the models more closely aligned with principles of behavioral economics, but it’s not yet clear how effective such changes will make pay-for-performance programs,” she continued.

Robert Berenson, MD, a fellow at the Urban Institute who conducts research and provides policy analysis on health care delivery issues, agrees that significant conceptual concerns and operational challenges exist with pay-for-performance. “At this point, we cannot put the genie of performance measurement back in the bottle, even though some us would like to,” he said during the webinar.

Dr Greene said pay-for-performance programs make several assumptions that she believes are unfair. “First, physicians are principally motivated by money, and second, physicians know how to improve quality but don’t do so because there is no financial reward for doing so. In our research on Fairview clinicians, we found that those who were providing the highest quality of care, were doing so before there was a financial incentive to do so as well as afterwards. They practiced differently, and it was not about the money.”

Going Forward

If pay-for-performance is to succeed, Dr Berenson said improved policies are needed on measures. For example, use measures strategically as part of a major quality improvement initiative, and invest more in the basic science of measurement development. “To a large extent, we’re proceeding with performance measurement and pay-for-performance without establishing the validity of what we’re doing.”

Dr Greene noted that lessons from the Fairview program can be beneficial in reforming pay-for-performance programs. “Fairview’s takeaway was to dramatically scale back their pay-for-performance program,” said Dr Greene. “Notably, they did not end the program, as they wanted incentives to continue to be aligned with quality. But, they realized that improving quality cannot solely rely on pay-for-performance. There has to be a much broader strategy.”

Dr Jha believes stakeholders need to stop asking if pay-for-performance works. Instead, they should ask: How do we get pay-for-performance to work. “The bottom line is that there is broad agreement that the measurement scheme has gone awry. We’re over measuring [and] that has absolutely little to no value. We’re not measuring things that matter enormously to patients.
These little incentive programs are somewhere between ineffective and insulting. I think the consensus is coming together that
we’re not doing it right.”