The heart of the matter

Mark Ostow

Krumholz, at left, with cardiology fellows Shahnaz Punjani and Teferi Mitiku at Yale–New Haven Hospital. View full image

Lesson 5: keep it real

Raised in Ohio, Krumholz majored in biology at Yale before going to Harvard for medical and public health degrees. He returned to Yale as a member of the medical faculty in 1992 because “the opportunity here was wide open,” he says. “I would have the opportunity to be entrepreneurial and creative.”

In fact, he would have to be entrepreneurial and creative: Yale had no outcomes research program, “no group to join and no critical mass for what I was interested in.” After submitting his first grant application, Krumholz realized he had included “a stupid mistake” that might have been caught if he’d had colleagues to look over the application. He didn’t get the grant.

And there was no revolution in sight. “I published a lot of papers and nothing happened,” he says. “The world wasn’t changing”—at least not fast enough.

Then, around 1994, came a big break. Medicare was planning a massive data-collection project on thousands of heart attack patients. The standard academic approach was to wait until the government finished a study, then conduct independent research based on the data. Krumholz saw a different opportunity: to “get in on the inside of the process and help shape the project.”

“I wasn’t thinking ‘I’m going to do something different,’” he says. “It just seemed natural. Why wouldn’t you work with the people who are responsible for the policies for how people 65 and over get care?” His efforts helped determine what data Medicare collected, and eventually, those data led to the D2B project. They also answered such critical questions as what percent of patients get aspirin after a heart attack. And they showed that beta blockers are helpful for heart attack patients over 65 (who had been mostly excluded from the clinical trials of the drugs)—but were prescribed for fewer than half of them.

“We have vastly increased the amount of money we’re spending on health care in the last 20 years,” Krumholz says. “The only thing we should be asking is, what do we get out of it? Our life expectancy has risen. But it’s risen faster in Canada, for a lot less money. It becomes even more relevant as we think about health care systems and the pressures on us to allocate resources appropriately and to ensure that as new technologies get involved and medications get promoted, the right choices are being made that are in the best interests of patients.”

Lesson 6: don’t believe everything you read

Buried in the diplomatic verbiage of that last sentence is a critique of Big Pharma. The critique becomes sharper when you read Krumholz’s scholarly work about Vioxx, a now-defunct painkiller marketed by Merck, and sharper still when you hear him lecture on the subject.

Merck withdrew Vioxx from the market in 2004, after it was shown to cause heart attacks. Krumholz served as an expert witness for people suing the company, on the condition that he could use internal Merck documents for his own research and publication. Among his findings: more than three years before it stopped selling the drug, Merck had studies showing the increased risk of heart attacks. Also, the company hired ghostwriters—prominent university researchers—to put their names on papers it commissioned from medical writers. And Merck had tested Vioxx on people with Alzheimer’s, without a data safety and monitoring board, and continued to do so even as the deaths mounted. (“Why isn’t this Tuskegee?” Krumholz demands of that study.)

Krumholz was paid a considerable sum for working on the lawsuit, and he also took considerable professional flak for it. A friend and junior colleague, Yale medical professor Joseph Ross ’06MHS (then a Robert Wood Johnson Clinical Scholar), was hired to help plow through the thousands of documents and coauthored the resulting journal articles. Ross learned from that experience “the importance of being open about your motivation.” Krumholz, he says, made it clear to the plaintiffs’ lawyers that “his motivation was to discover the truth, that he wanted to fairly and objectively examine the data and then testify to what he had found.” What’s more, Ross says, Krumholz told the lawyers he wanted to use the Merck documents “to improve research conduct and scientific integrity for the medical profession.”

Krumholz’s testimony in the Vioxx trials (now documented in Snigdha Prakash’s new book, All the Justice Money Can Buy)ended in February 2007. A year later, he made news again. Speaking at the American College of Cardiology, Krumholz argued that a cholesterol-drug combo marketed by Merck and Schering-Plough had not been proven to work. The drugs do lower bad cholesterol, but, with sales of $5 billion a year, they may perform no better than generic statins in slowing actual heart disease. The speech set off a bomb on Wall Street. Merck’s stock dropped by 15 percent, Schering-Plough’s by 26 percent.

The saga underscores one of Krumholz’s scientific principles: lab tests of surrogate measurements, such as cholesterol or blood sugar levels, don’t necessarily tell you how a drug will affect real patients in the real world.

Lesson 7: there’s no “I” in team

Krumholz has no swagger. His words tumble out with a low-key intensity. He’ll gladly talk about his work for hours, but he says he “feels funny” being the subject of a magazine profile: “It’s not me, it’s the team.”

That’s not just Midwestern modesty. It is part of Krumholz’s core (and CORE) approach. Collaborating with a broad network of colleagues, mentors, and students is part of his strategy for building “a national movement for outcomes research.” “The culture I grew up in was: someone was going to make a discovery in a lab—one brilliant scientist,” he says. But “most of our problems are much more complex.” They require a team.

That attitude makes Krumholz and his team stand out, observes Lein Han, a senior technical adviser at the Centers for Medicare and Medicaid Services who has worked closely with him. In the highly competitive world of research contractors, “Harlan is very different,” Han says. “He’s never being very selfish or very possessive about his intellectual property. He always wants to share”—even with other contractors—“because he’s a tireless advocate for improving patient care.”

What’s true for the researchers is also true for their subjects, it seems. Krumholz and the ACC are rolling out a new campaign: Hospital to Home, or H2H. The idea is to reduce the number of heart attack patients who end up back in the hospital within 30 days of being sent home. Krumholz and CORE developed the national parameters for measuring readmission, and they are now studying how successful hospitals achieve low rates. They reported this spring that “hospital culture”—including teamwork, or the absence thereof—is one key factor.

The project addresses a major new federal goal: cutting preventable readmissions for common diagnoses, which cost an estimated $12 billion a year. “It’s a sweet spot,” Krumholz says. “It’s good for patients and it saves money. Everybody should be happy if we do that.” But he has encountered some opposition. He counts as a readmission any hospital stay within 30 days of the first, and at least one cardiologist criticizes that definition. Hospitals shouldn’t be blamed, she says, when patients return for problems unrelated to their heart conditions.

“It’s a system responsibility,” responds Krumholz. “A lot of the reasons [the patients] are being readmitted have nothing to do with the heart. But they were just in the hospital, and we had the opportunity to treat them as a whole patient.”

That argument goes to the heart of Krumholz’s diagnosis—not only of cardiac readmissions but of the entire health care system. From one angle, his career looks like a frantic patch job, jumping from one urgent fix to the next. But the seemingly disparate pieces fit into an overall scheme: reframing medical culture. “Success in the future is going to be about systems and teams,” he says. “It’s not just about the person who’s traditionally at the top, which is the physician. But it’s about how people break down that traditional hierarchy and work together” for the benefit of the person who’s traditionally at the bottom of the hierarchy: the patient.