Doctor Know

Michael E. Porter is the Bishop William Lawrence university professor at Harvard Business School. Elizabeth Olmsted Teisberg is associate professor of business administration at the University of Virginia's Darden School of Business. This article is adapted from their book, Redefining Health Care: Creating Value-Based Competition on Results.Strokes are the leading cause of long-term disability in the United States and the third-most common cause of death. It is also extremely costly: Conservative estimates suggest it costs the nation almost $58 billion per year. Successful recovery depends partly on luck, of course. But it also depends significantly on how and where care is delivered. Treatment for strokes varies markedly from one hospital to another due to different professional opinions as well as different capabilities and equipment. Nationally, less than 5 percent of ischemic stroke patients are given intravenous drugs to clear the blocked blood vessel in the brain. But some hospitals treat up to 15 percent of such patients with this approach. Even for large ischemic strokes, which account for most of the costs, some hospitals do little to intervene. Others use mechanical intervention via a catheter in the brain or a drug delivered via a catheter directly to the clot. Not only do the hospitals' methods vary widely, so do their results. A patient with the same condition could fully recover if treated in one hospital and not survive if treated in another. But emergency medical technicians don't consider this information when taking stroke victims to the hospital. Instead, they are often required to take every stroke victim to the nearest hospital. Why? Because, in most states, nobody collects and disseminates information about which hospitals achieve the best results. In a normal functioning market, producers of a good or providers of a service compete on results--their ability to provide the highest quality at the lowest prices. Excellent competitors prosper and grow, weaker rivals are restructured or go out of business. Value--the quality received per dollar spent--increases. Health care could not be more different. Most people have better information about their restaurants, cars, and sellers on eBay than they do about their medical care. And the information that is available--health plan ratings, subscriber satisfaction surveys, and rankings based on doctor or hospital reputation--do not really measure health outcomes. Even the most basic data, such as how many patients with a particular diagnosis a hospital or physician group has treated, is rarely available. U.S. News & World Report makes a start with its annual rankings of hospitals. But, just to be considered in twelve specialty areas, hospitals must offer broad service lines and have a wide array of equipment and facilities, be affiliated with a medical school, or be a member of the Council of Teaching Hospitals. These criteria exclude many specialty hospitals and community hospitals. This whole approach reinforces the bias that community hospitals are not as good. Yet studies reveal that, for some medical conditions, community hospitals often achieve equal or better outcomes--and at lower costs--than teaching hospitals. Without true results-measurement, doctors, hospitals, and health systems have traditionally competed in unproductive ways--shifting costs to others, attempting to accumulate bargaining leverage with insurance companies, or restricting services. This has eroded quality, fostered inefficiency, and driven up administrative costs. Fortunately, we can fix this. The key lies in redefining the nature of health care competition. And the first step is measuring results, so the power of competition will be directed at driving dramatic improvements in the value of health care. In the last few years, government and the business community--the two largest purchasers of health care--have tried to improve quality by compiling and publishing information about whether doctors and hospitals comply with established treatment guidelines. Some guidelines are generic, such as whether hospitals use computerized order entries (to reduce errors) and have procedures in place to control infections. Some guidelines apply to particular medical conditions, such as whether a hospital administers aspirin to a patient arriving with a heart attack. Practice standards make sense, to a point: Appropriate practice will reduce the high rates of medical errors and inappropriate care. But process-compliance is not the same as high quality. Different providers using the same processes achieve very different results, because guidelines cover only some aspects of care. Good medical practice requires adjustments to address each patient's particular circumstances. Also, it is difficult to keep process guidelines up to date with the latest medical knowledge, since adopting new guidelines often depends on developing consensus among experts, which happens slowly. Ironically, too much attention to process-compliance can actually impede innovation rather than encourage it. Results are what really matter. But measuring results isn't as simple as tracking death rates. In spine care, for example, it's important to measure pain reduction, improvement in range of movement, functional ability, and the time between the beginning of care and the return to work or restoration of normal activity. Outcome measures should also cover the full cycle of care, not just individual interventions: It is important to know the combined results of surgery and physical therapy, because the two together drive patient outcomes and insight about how to improve care. Only by measuring results over the full cycle of care is it possible to track complications leading to high rates of repeat surgery or re-hospitalization. To be meaningful, outcome measures must also be adjusted for risk: If a hospital takes on the hardest cases, its results will be affected. And there must be a major role for experts. Doctors and medical societies should participate in defining measures; experts should also be able to check and correct the accuracy of data before (and after) it is published. The feasibility of measuring results and their impact is not in doubt. There is extensive outcome data, for example, about organ transplants. In 1984, Congress responded to widespread concern over the fairness of organ allocation by establishing the Organ Procurement and Transplantation Network. The Network is operated under federal contract by the United Network for Organ Sharing (unos), a nonprofit organization. Hospitals that receive organs must report data, so reporting is effectively mandatory. The data is national, includes all providers, and goes beyond the sum of individual hospitals' reports. Unos keeps track of whether each patient ends up needing re-transplantation (of the same organ) at another institution, even if the original provider is not aware of the repeat surgery. The registry of transplant data now contains information on over 371,000 transplant recipients. The transplant outcome data is publicly available on the Web, sorted by organ type and by patient age, gender, ethnicity, and blood type. Unos adjusts these results for risk and presents the information in clear and well-annotated charts. For each transplant center, it is possible to calculate how long its patients live--and how long the organs it transplants typically last. Although many patients still need help interpreting the information, and although not all referring doctors use the data, the transplant surgery teams are wellaware of where they stand--and how they stack up against teams at other centers. Not coincidentally, performance in transplants has continued to improve, including successful transplants in sicker patients. Public outcome data has also driven dramatic improvement in heart surgery. The state of New York reports outcome data for coronary artery bypass graft (cabg) procedures. A group of cardiothoracic surgical experts devised the measures, along with methods to adjust them for patient circumstances. The program publicly reports riskadjusted mortality rates of cabg procedures by hospital and by surgeon. In the first four years of New York's reporting program, deaths from cardiac surgery fell by 41 percent. While many patients and referring doctors stuck in the old mindset do not use the data, even uninformed patients benefited from the significant outcome improvement. Moreover, the Society of Thoracic Surgeons (STS) was motivated by the public reporting to collect and improve outcome measurement for many types of heart surgery. Superb outcomes determine which methods are taught to other surgeons at STS meetings. While the STS data is not public, even uninformed patients benefit, once again, by the innovations in care enabled by measuring and analyzing outcomes. Measuring results is the single-most important step in transforming health care, not only in the United States, but in other countries, too. So what's standing in the way of developing more such measures? Some providers have resisted outcome reporting, which means that mandatory reporting might be necessary. Government needs to ensure the establishment of a structure to collect and publish information about medical conditions, the way unos does for transplants. Another impediment has been the fear, by insurers, among others, that publicizing outcome measures would lead patients to demand care from the most expensive doctors and hospitals. But, by and large, the best health care providers are the most efficient. They make fewer errors and enable faster recovery and less long-term disability. They also tend to manage diseases better, so that fewer patients have full-blown, acute crises. And this, ultimately, is the mindset that should define our health care system: Better health outcomes do not have to cost more money. On the contrary, they may end up costing less.

By michael e. porter and elizabeth olmsted teisberg