NEJM Interview: Eric Schneider on the limitations of pay-for-performance programs and alternative options to support quality improvement.

For more than two decades, quality improvement has been a central focus of US healthcare policy. These efforts have often relied on a pay-for-performance paradigm. I'm Stephen Worsey, managing editor of the New England Journal of Medicine, and I'm talking with Eric Schneider, an adjunct professor of Health Policy and Management at the Harvard T.H. Chan School of Public Health, and a member of the journal's prospective advisory board. Dr. Schneider has co-authored a prospective article about the role of pay-for-performance in healthcare quality programs. Dr. Schneider, what were the origins of this focus on quality improvement in US healthcare?

The early days of quality improvement really latched onto the notion of measurement in the 1990s as a way of driving improvement, measuring quality, and, and then driving improvement through the measurement. And then in the early 2000s, pay-for-performance came into vogue, the idea of measuring, scoring providers, and then attaching dollar figures to the results. And we've been on that journey ever since.

And what are the primary components of these pay-for-performance approaches, and how have they been incorporated into federal healthcare quality programs?

The primary components really are the notion of quality measurement and the development of quality measures, which was done mostly with federal investment, but also by many professional societies and others participating in measure development. And then the scoring and publication of reports, so publishing the results to provide transparency for consumers, and then finally attaching payment to the scoring, and that can be done through a variety of different mechanisms, but basically the idea is to reward better quality providers, whether they're hospitals, physician groups, or individual clinicians.

And what are the results of pay-for-performance looked like? To what extent have they improved care processes, improved health outcomes?

I think in the early 2000s we saw some-- In fact, there was some evidence emerging that sort of supported the idea of measurement as a mechanism for improvement. Once you measure something and make those results transparent, you see that providers do change their behavior. For diabetes in particular, I, my own primary care practice, once we knew what our results were across our, our population of patients living with diabetes, we made adjustments and actually did improve the control of hemoglobin A1C, which is one of the early successes, I think, with performance measurement. What we've seen, though, in the last twenty years or so is a kind of expansive use of performance measurement by the federal government in CMS programs such as the Hospital Value-Based Purchasing Program, the Merit-based Incentive Payment System, or MIPS for physicians, and the Star Rating system for health plans. Tying performance payments in those programs, what some people call value-based purchasing, hasn't really produced the types of improvement results that I think people have been hoping for. On the other hand, that has introduced a whole set of administrative burden and costs of reporting the measures to practice.

In fact, you say in your