Medicine is flying blind. Thousands of medical journal articles are published every month on potential new treatments and diagnostic tests. Precious few of them measure how well doctors are doing in the real world, outside of controlled trials--what they are doing right, what they are doing wrong and what they are forgetting to do entirely. No wonder our medical system wastes billions of dollars a year. Yale cardiologist Harlan Krumholz, age 52, has spent two decades shining a light into this broken system. When he started, the kind of "outcomes" research he does was deeply unfashionable. Doctors figured that it would merely confirm the obvious fact that some doctors and hospitals are better than others. But Krumholz proved the skeptics wrong. By figuring out what to measure and how, he showed that even top hospitals were systematically underperforming, largely because no one was tracking the results. In 2004 he proved that only one-third of American hospitals were treating heart attack patients quickly enough. His work laid the groundwork for the system the Medicare program now uses to compare hospitals. Another line of research proved that heart-failure patients are frequent flyers: They end up back in the hospital almost as soon as they leave. This result led to a provision in Obama's new health reform law that will allow Medicare to dock hospitals' pay starting in 2012 if that revolving door is moving too fast. The new law even creates a Patient-Centered Outcomes Research Institute that aims to extend the type of work Krumholz does in cardiology to other medical arenas such as cancer and psychiatry. "Every day millions of patients are being treated, and the lessons from their experiences are lost because there is no systematic effort to learn from them," says Krumholz. "If I'm sitting down with a patient, I should be able to take advantage of everything we have learned up until yesterday to treat them." Krumholz's basic idea is that if you ask the right question and pick the right measurement, you can figure out a way to get the answer, often using billing records or existing databases. This frequently involves partnering with insurers or Medicare. He has a knack for focusing on performance metrics that hold hospitals accountable. For heart attack patients, Krumholz picked "door to balloon time," or how fast the patient is treated upon arrival. Hospitals that are doing badly on this measure can't blame it on how long the ambulance took to get there. For big corporations the idea that "you can't manage it if you can't measure it" is an old chestnut. General Electric, Toyota and other companies have had data-driven quality-improvement efforts for years. But medicine--supposedly a more scientific profession--has been slow to measure itself. Krumholz has been one of a small handful of pioneers behind the scenes pushing to do this. "There's no one else who has been more of a leader in allowing us to understand what we are doing" right and wrong, says Ralph Brindis, the chair of the American College of Cardiology. Adds Memorial Sloan-Kettering Cancer Center pulmonologist Peter Bach: "He's one of the most prominent quality-of-care researchers in the world. He's trained countless researchers everywhere." Krumholz was inspired to be a doctor by his father, a pulmonologist who took him on rounds when he was a child. "It's the most amazing profession," he says. "You're making decisions that are so important to people, and they have to trust you to help them." As a biology student at Yale in the late 1970s he studied how the rural poor in America and foreign countries like India and China often didn't get care that was available to richer folks in cities. He learned low-tech solutions often could help. At Harvard Medical School in the 1980s he rode the subway to housing projects to teach residents there to give basic emergency care. His career got off to a quick start. On his first day as a Yale professor in 1992, he published a study in the New England Journal of Medicine defending a then controversial Genentech clot-buster drug. But after that he wasn't sure what to do next until he heard about Qualidigm, a nonprofit that paid moonlighting doctors to go over patient records and find mistakes for Medicare. He went to one of its meetings and met the group's chief executive, Marcia K. Petrillo. She was talking about using the Medicare billing system to identify ways patient care was being hurt. As she was talking, the idea that the medical system itself keeps doctors from adequately caring for patients snapped into Krumholz's head. "In medical school you never thought, Gee, how will the system affect me?" Krumholz says. "The lesson of medical school was that you are a singular crusader. It's about how the individual operates. If you were lucky a smart nurse would tell you if you were about to do something really dumb." Krumholz convinced Medicare in 1995 to give him access to detailed patient data it had just started collecting from heart doctors in Alabama, Connecticut, Iowa and Wisconsin. The result was a database of 16,000 patients that became the basis for Medicare's hospital comparison projects, including its public site, hospitalcompare.hhs.gov. The same year he showed that aspirin was being given to only 75% of heart patients who should get it. He studied blood pressure pills called beta blockers and found that only half of the heart patients who needed them were getting them. In 1997 he found that half the patients hospitalized for heart failure returned to the hospital within six months after their first visit. Krumholz's biggest hit came in 2004. Angioplasty, a procedure in which a balloon-tipped catheter is used to reopen a clogged artery, saves lives if performed soon after a heart attack. Guidelines recommend doing it within 90 minutes of a patient's arrival at the hospital. Using two drug company databases, Krumholz showed that only 33% of patients were getting angioplasty in that time frame. He also showed that the delays increased the death rate. Working with Yale management researcher Elizabeth Bradley, his team visited the fastest and slowest hospitals. He found speed was often the result of taking simple steps like letting the emergency room doc call for an angioplasty without consulting a cardiologist or putting electrocardiograms in ambulances so the heart attack would be confirmed before the patient arrived. He led an effort by the ACC and the American Heart Association to push hospitals to change their ways. Today 90% of heart attack patients have their arteries opened within 90 minutes of hospital arrival (unless they end up at a hospital that can't perform angioplasty). Krumholz is branching out beyond heart disease. He played a key role in a 2009 study showing 4 million Americans each year get high radiation doses from routine scans, potentially boosting cancer risk. Krumholz's hunger for new data sources can get him into trouble. He was paid $225,000 over three years to be a witness for plaintiff lawyers suing Merck over its arthritis drug Vioxx after it was recalled for causing heart attacks. This gave him access to Merck's e-mails and other data. Krumholz used them to publish articles that show Merck might have identified Vioxx's risks earlier. Merck says the payments are a conflict of interest. But doctors trust him because he speaks his mind and puts patients first. When Merck's anticholesterol drug Zetia failed to clear artery plaque in a key study in 2008, Krumholz addressed a big conference of cardiologists and bluntly told them to use better-proven drugs. Merck lost $22 billion in market cap that afternoon. Source
