Editor’s Note: Today, The Health Affairs Blog concludes 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. Arnold Milstein (below) and Howard Beckman contribute posts today. Last Thursday, James Robinson and Tom Williams weighed in.

Using the estimation methods published by the Institute of Medicine, and recently updated by the Urban Institute, more than 24,000 of the 48 million uninsured Americans will die as a result of uninsurance next year. Commonwealth Fund surveys indicate that another 12 million or so insured Americans will be financially stressed by their health care and health insurance spending. For most of these predominantly low-to-middle-income Americans, the underlying problem is not lack of desire for health insurance; rather. their income is insufficient to reasonably afford a health insurance policy or pay its deductible and other out-of-pocket costs. The life-and-death struggle of the sick among this expanding group of 60 million Americans is detailed in Uninsured in America: Life and Death in the Land of Opportunity. If you disagree with the title of my comments, scan this book. It illustrates the most toxic personal consequences of insufficient effort by health care leaders to remove clinical waste estimated to constitute around 35% of U.S. health care spending.

Options for Broad Waste Removal

In theory, there are three solutions to the growing unaffordability of U.S. health care: (1) withhold expensive treatments from nonaffluent people; (2) ask affluent people to pay more for the care of others; or (3) accelerate improvement in the efficiency of health care delivery. Option 1 violates a long-standing American social contract to care for one another when we’re sick. Option 2 is inherently unstable: as long as health care spending continues to outgrow our economy, it’s only a matter of time before there’s no one left to pay for the care of others. Option 3 suffers no inherent limitations and is technically feasible.

Technical Feasibility of Accelerating Clinical Efficiency Improvement

Evidence that it is technically feasible to improve the ratio of health produced to dollars spent on U.S. health care comes from two primary sources. The first is consensus of expert opinion based on available evidence such as the Dartmouth Atlas. Both the National Academies’ estimate in its 2005 Engineering/Health Care Partnership report and the Congressional Budget Office’s 2007 expert consensus estimate concluded that around 35% of current health care spending represents waste that could be eliminated with no loss of health. Waste includes eliminating clinical services that offer no likely net health gain, as well as failing to produce valuable clinical services at benchmark levels of cost-efficiency. Consensus estimates of potential savings from other sources of efficiency capture, such as lowering health insurance administrative costs, show additional but lower potential yield.

Beyond these static assessments of accumulated waste in U.S. care delivery are dynamic estimates. Richard Hillestad’s recent Health Affairs article examined shortfalls in the annual rate of efficiency gain in the health care industry compared to other U.S. service industries, such as retail stores, commercial airlines, financial services, and telecommunications services. As in health care, other U.S. service industries continue to adopt new and more costly technologies, such as mobile phones, that deliver more consumer benefit every year; unlike health care, these other industries simultaneously discover efficiency improvements sufficient to lower the cost of their services per unit of benefit to consumers by 2.5 – 6.5% annually and thereby offset the cost-additive impact of these new technologies. In contrast, roughly estimated rates of annual efficiency gain achieved by the U.S. health care industry are much lower, leaving an average annual 2.5-percentage-point gap between U.S. care spending growth and growth in gross domestic product (GDP). The cumulative consequence of this gap is an increase in the percentage of Americans who don’t earn enough to pay even a 25% share of average annual health care per capita spending in the U.S.

How Can We Induce Our Health Care Industry To Shrink Its 35% “Waste-line” And Thereafter Improve Its Efficiency By An Additional 2.5% Annually, While Simultaneously Improving Quality Of Care?

This question was addressed directly in the National Academies’ 2005 report, Engineering/Health Care Partnership, and more generally in its Quality Chasm and P4P reports. These reports recommended (1) educating health industry workers in the science of service engineering; (2) boosting health professionals’ sense of responsibility to lead efforts to improve “value,” defined as the ratio of health gain to health care spending; (3) public reporting of performance comparisons of treatment options, providers and communities on all six IOM-specified clinical performance domains; (4) incentivizing consumers to choose higher-value treatment options and providers; and (5) via P4P and payment reform, paying more for higher-value providers and treatment options, and less for others.

Why Is Measuring And Incentivizing Superior Provider Value Proving To Be Difficult, Especially For Individual Physicians Who Exercise The Most Authority Over Health Care Resource Use Decisions?

As discussed in the New England Journal of Medicine, December 27, 2007, efforts to measure and incentivize superior provider value, especially for individual physicians who under state laws govern some 90% of health care resource use, is proving to be difficult for multiple reasons. Most difficulties are not unique to the health care industry; they arise in parallel efforts to measure and improve the value of restaurant chefs, travel agents, school teachers, and individual employees of most large U.S. companies. They include insufficient or inaccurate data capture, rough edges in performance measurement methods, small measurement samples, and impact on performance by parties other than those being measured. For example, the performance of employees in most industries can be affected by the behavior of coworkers and/or customers.

Should Physician And Hospital P4P Programs Encompass Efficiency Measures Now Or Await More-Refined Efficiency Measures?

The answer to this question hinges on how much weight we attach to the primary consequences of moving forward and of delay: (1) the injustice suffered by providers from rewarding or penalizing the wrong providers due to measurement flaws; and (2) the injustice suffered by low- and middle-income Americans from postponing greater mobilization of health industry effort to improve value.

Most providers and health services researchers attach more weight to the first source of injustice, while most consumer and purchaser leaders attach more weight to the second source. This is not surprising; most providers and health services researchers can afford health insurance and easily network informally to higher-value providers when they become ill.

This leaves private and public policymakers with a difficult choice. As a clinician and purchaser advocate who feels empathy with those exposed to both sources of injustice, I favor moving forward now with current efficiency measurement methods and their incorporation into public reporting, P4P, and tiered health insurance benefit plans at multiple health care system levels, including individual physicians because:

* Attaching economic consequences to imperfect performance measures usually speeds improved performance measures and improved performance.

* Insurers that have used provider efficiency and quality measurements to tier health insurance benefits have lowered their premiums without evidence of lower quality of care.

* Since American clinicians and health care executives are better paid than their counterparts in other wealthy nations, it seems fair to ask them to accept greater accountability for better clinical and financial outcomes.

* After more than 20 years of peer-reviewed published evidence of double-digit percentage waste in the U.S. care delivery, American clinicians and their educators have not yet prioritized efficiency improvement. Elimination of half of our 35% wasted health care spending would be more than enough to pay for excellent care for uninsured low- and middle-income Americans and spare them 24,000 deaths annually.