Blending the clinical and the statistical

For more than two decades physicians have carried in their pocket copies of the Goldman Index, a list of factors to determine whether a patient undergoing surgery for noncardiac reasons is at risk of a heart attack or another major cardiac complication. The index is named for Lee Goldman, M.D. ’73, M.P.H. ’73, FW ’78, who developed it during his medical residency and published it during his cardiology fellowship at Yale by applying techniques derived from epidemiology and clinical practice.

In recent years Goldman has also achieved renown for his leadership of the department of medicine at the University of California, San Francisco (UCSF), which he joined in 1995. In July 2006 Goldman moved to Columbia University’s College of Physicians and Surgeons as executive vice president for health and biomedical sciences and dean of the faculties of health sciences and medicine.

Goldman traces his interest in statistics to a class he took during his first semester at the School of Medicine with John D. Thompson, R.N., M.S., the legendary director of the program in hospital administration at the Department of Epidemiology and Public Health. That course convinced Goldman to enroll in Yale’s master of public health program, and a study he undertook there cemented his love of numbers and other data. In 1971, President Richard Nixon was pushing a plan for national health insurance. Goldman surveyed Yale medical students and faculty, deconstructing this bill and competing pieces of legislation to determine how his colleagues should perceive each proposal. Goldman then took his programming textbook to the beach over Labor Day weekend and came back ready to crunch numbers in the giant computers of the day. As the analysis came together, he was hooked. “I ended up publishing five papers out of my thesis,” he said, “and kind of got the bug.”

He soon began to apply his new-found statistical and programming savvy to studies of cardiac risk. After an internship and residency at UCSF and a second residency at Massachusetts General Hospital, Goldman became a cardiology fellow at Yale. As his fellowship ended, Goldman said, his interest in epidemiology propelled him toward nontraditional applications of cardiology. “Cardiology divisions weren’t really interested in me, and I didn’t really want to run an echo [cardiogram] lab,” he said.

In the division of general internal medicine at Harvard’s Peter Bent Brigham (now Brigham and Women’s) Hospital, where he stayed for almost two decades, Goldman continued to focus on combining epidemiology with clinical care in cardiology. “Most of my work used the same kinds of methods an epidemiologist used, but applied it to patients we actually touched,” he said, describing his efforts to determine risk factors for heart attacks after surgery and among patients who came to emergency rooms with chest pain. “There were so many factors to consider that clinicians did not have a way of determining which ones were the best predictors of a patient’s risk,” he said. “The concept was always to gather comprehensive data and to whittle it down to what might be important.”

One of Goldman’s most enduring contributions was the Goldman Index. “It was the first systematic approach to that question,” said Lawrence S. Cohen, M.D., HS ’65, special advisor to the dean. “It has remained a benchmark for the care of patients that are undergoing noncardiac surgery, and it has stood the test of time over three decades.”

Goldman and his colleagues also started one of the first chest pain evaluation units at Harvard. Today, these are common at many hospitals nationwide. And Goldman developed the Coronary Heart Disease Policy Model, which sets priorities for preventing and treating heart disease. At Harvard, he also codeveloped the Program in Clinical Effectiveness, which has trained over 1,000 young physicians in the basics of clinical research.

“He really has been a pioneer,” said Harlan M. Krumholz, M.D., M.Sc., the Harold J. Hines Jr. Professor of Medicine and professor of epidemiology and public health, and of investigative medicine, one of Goldman’s advisees at Yale. “He was one of the first people to develop large observational studies that drew knowledge from the real world and could be fed back into practice.”

After becoming the chair of medicine at UCSF, Goldman and his colleague Robert M. Wachter, M.D., in 1996 coined the term hospitalist for inpatient physicians, a concept that quickly produced a major movement in medicine. (Goldman and Wachter started the first hospitalist program in the country at UCSF.) With inpatient care changing rapidly, they argued, office-based doctors simply do not have time to monitor their hospital patients the necessary three or four times a day. Establishing a separate specialty of hospital medicine, they argued, would increase hospital efficiency and benefit patients. “The board is about to recognize it as a distinct discipline,” he said, referring to the American Board of Internal Medicine, which certifies subspecialties in internal medicine.

“Columbia is fortunate,” said his former advisee Krumholz, recalling his mentor’s seemingly limitless capacity to inspire students and colleagues. “When he walked into a room, the energy would just increase, the quality of conversation would go up. I think a whole generation of people was drawn to research because of their association with him.”