Editor’s Note: Today, The Health Affairs Blog begins a series of four posts on trends in performance measurement and performance-based payment in health care. The series focuses particularly on the increasing emphasis being placed on measuring and rewarding cost-efficiency. James Robinson and Tom Williams (below) contribute posts today. On Monday, Arnold Milstein and Howard Beckman weigh in.

In his blog posting, James Robinson points out the shifting emphasis from quality to cost within the evolving world of pay-for-performance in health care. Attention to cost is not new to pay-for-performance programs. In fact, a recent survey by the Leapfrog Group and Med-Vantage Inc. reported that about 23% of P4P programs nationally already include measures of the efficiency or cost of care. Nonetheless, quality has played the undisputed lead role in pay-for-performance to date. In the future, it will increasingly share the stage with efficiency and cost.

Robinson also points out that although the emphasis on quality emerged as a reaction to heavy-handed managed care approaches to cost control in the late 1990s, this shift was also triggered by the seminal Institute of Medicine (IOM) reports on health care quality and patient safety, which galvanized a call to action. Since then, we have made significant progress in measuring quality, and mercifully put the tired claim, “You can’t measure quality in health care,” to rest. With an important foothold in quality measurement established, and the political environment again receptive to cost management, attention has turned to the question of how best to measure and reward efficient health care. But how do we measure and manage health care costs without returning to the perceived blunt-edge utilization practices that drove managed care into the background?

This question has inspired active debate and a flurry of pilot projects with measurement schemes ranging from the most elementary to approaches of impractical complexity. Such experimentation is critically important if we have any hope for payment reform. Why? Because if we cannot define what we want and develop a workable scheme to pay for it, we will continue to get what we don’t what: spiraling costs with questionable quality and outcomes.

So, the debate about how best to measure costs and efficiency continues. It currently revolves around a core set of questions and technical challenges. Is efficiency simply a measure of cost, or a calculus that incorporates both cost and quality? Can we really measure efficiency in health care without comparing costs to medical outcomes? Is it enough to measure resource use, or should actual prices be included? If so, how do we avoid punishing physicians affiliated with higher-cost hospitals? Is the purpose of this measurement public reporting, incentive payments, or a new basis for payment?

Fortunately, we have more than one place to start. In terms of a framework, one of the most helpful was developed by the IOM National Roundtable on Health Care Quality. It offers the simple notion that we should begin by focusing on medical care that is underused, overused, and misused. We know that patients should receive more of services that are underused, such as screenings. However, the business case demonstrating that providing more of these services improves efficiency is unclear. In fact, this approach may lead to increased costs, at least in the short run. Eliminating services that are overused and misused offers the clearest and most immediate way to lower costs, improve quality, and drive waste out of the system.

Finding Ways To Measure Overuse And Misuse Of Health Care Services

Of course, like everything in health care, measuring overuse and misuse is more easily said than done. Unlike widgets, no two patients are alike, and there is that pesky problem of a paper-based system, which makes measurement a daunting challenge.

Fortunately, there are places we can start. Measures of overuse and misuse already exist. The National Committee for Quality Assurance (NCQA) and other respected organizations have developed, specified, and tested a variety of overuse and misuse measures. A simple example is the overuse of antibiotics for upper respiratory infections. At the Integrated Healthcare Association (IHA) in California, we are working with a collaboration of health plans and physician organizations to implement as many measures targeting overuse and misuse as possible.

Unfortunately, the pace and scope of this investigation has been slow. Implementation of these measures on a large scale is limited to areas for which reliable, electronic data are accessible. Like the drunk looking for his keys under the lamppost, we limit our search to areas where we have light–in the form of accessible data.

Another place to start is episodes of care. This approach has been largely used by health plans applying proprietary software to a segment of the patient population. This has caused providers at times to cry foul, amid complaints of a “black box” approach. At IHA we are testing episodes of care using data aggregated from multiple payers in order to get a full view of the patient population. As this approach is brought into the light, its deficits are exposed, including challenges with attribution to providers; matching data to episodes; and properly defining episodes, causing one medical director to exclaim that this approach is “still somewhere between alchemy and science.”

Nonetheless, despite current shortcomings, episodes of care do offer a mechanism to examine practice variation, identify opportunities for quality improvement, and provide measures that lend themselves to P4P incentives and new payment methods. Taking hip replacement as an example, our current diagnosis-related group (DRG) system pays once for a hip replacement and again for a revision if there is a complication. A single payment based on a total episode of care — including both procedures — adjusted for various risks and demographics, potentially offers a payment scheme promoting the most efficient, safest, and highest-quality care.

Moving from alchemy to science will be hard work for organizations and collaborations like IHA across the country. These efforts to work out the challenges by testing different approaches in diverse settings are essential to move from theory to practice. Equally important will be efforts led by national organizations such as the NCQA and the National Quality Forum to design more-refined measures of efficiency and cost. This combination of efforts is our best hope for the development of a measurement system that can support Version 2.0 of P4P.

Finally, pay-for-performance is not a panacea or ultimate solution to align payment incentives with our goals for better quality and efficiency, but it does offer a path to find that solution in payment reform. Pay-for-performance is driving the creation of a system of measurement and a rational, effective method of data collection. These are the necessary foundations of payment reform.