Medicine by the Numbers
Health economist David Meltzer does the math on cost-effective medicine, hospitalists, and more.
David Meltzer was 30,000 feet in the air—en route, as a third-year medical resident, to give an auditorium full of physicians a controversial lesson in cost-effective medicine—when he decided to rework his whole presentation. His slides were all in calculus, indigestible and uninviting, and he realized he could simplify everything to algebra. “So I drew it out on the plane,” Meltzer says, “and then later drove around trying to find a place that would make up new slides.”
“Hospitalists produce a change in culture,” Meltzer says, “and create a collective body of knowledge and set of habits.”
An audience of hundreds awaited him in Phoenix, where in October 1995 the Society for Medical Decision Making was holding its annual conference. Meltzer, U-High’82, AM’87, PhD’92, MD’93, had never attended a scientific meeting, but he was slated to give a plenary lecture on medical treatments’ future costs. “Say you save someone’s life today and they live to be 90,” he says. “And then at 90 they get some other health problem that’s totally unrelated to the treatment you gave them before, except they’re living longer. Should you count that as a cost?”
Until a decade ago, the answer was no. Physicians regarded a patient’s subsequent heart attack or pneumonia as entirely separate from the cancer therapy that preceded it. But to Meltzer—who earned concurrent Chicago doctorates in economics and medicine and who is now an associate professor in medicine, the Harris School of Public Policy, and economics—that reasoning doesn’t work. Getting one disease, he figures, is the natural, eventual consequence of getting rid of another. “If you’re going to count all the benefits, you should count all the costs, and the costs of extending a life are”—and here he shifts to mathematics—“consumption plus medical expenditures minus earnings. You even need to count all the hamburgers people eat in added years of life, because that’s part of the cost of being alive.” A staggering calculation, he concedes, but a necessary one. Otherwise society spends too much money on simply prolonging life and not enough on making it better. “Are you going to give an older person chemotherapy, or are you going to give them a hip replacement, which won’t make them live longer but greatly improves quality of life? These are the kinds of trade-offs I’m talking about.”
Meltzer’s research on medical care’s future costs is now widely cited, taught, and, for the most part, agreed upon. Eleven years ago, however, it sounded pretty improbable to Society for Medical Decision Making members. “So I get up there and give this talk half in calculus with too many slides,” Meltzer recalls, “and then there’s silence in the room. And this is the premier organization in this area.” Finally, three people raised their hands. He remembers their questions word-for-word. One was from someone who hadn’t understood the lecture, he says, and another from someone who worried Meltzer’s analysis would render every treatment cost-ineffective and therefore demanded to know, “Don’t you believe in medicine?” (It doesn’t, Meltzer says, and he does.) “The third question was this: ‘I have an 80-year-old father. If I understand your argument, you’re saying I should kill him.’ I came home and thought, ‘Oh, my god. This is not a good start.’”
To check his bearings—and his research (“at that point I wasn’t totally sure I had nailed it”)—Meltzer paid a visit to Chicago economics and sociology professor Gary Becker, AM’53, PhD’55. Becker had been his dissertation adviser, and he’d worked as Becker’s research assistant. “I told him, people aren’t buying this.” No stranger to professional controversy, Becker checked Meltzer’s equations and theories and pronounced them solid. “He said, ‘If you’re not making people just a little bit angry, you’re not doing anything important.’ He told me to hang in there.”
For Meltzer, Chicago has always been both sanctuary and springboard. He first landed in Hyde Park at age 4, when his parents, Sharon and Herbert Meltzer, joined the faculty—she in English and he in psychiatry. Meltzer entered the Lab Schools’ nursery program. “I think my fascination with research dates to being a kid,” he says. “I remember my father pulling me in to help him do data analysis at night. Some families play basketball—and we occasionally did too—but more likely I was sitting there helping him type in numbers.”
The Lab Schools, meanwhile, gave Meltzer a sampling of the social sciences. He read Thomas Malthus and joined the debate team. He took an intro course in economics. During Meltzer’s freshman year of high school, his social-studies teacher concocted a series of exercises requiring the class to reconstruct prehistoric societies using nothing but data—a map, for instance, scattered with bones, charred campsites, and bits of tools. “We had to figure out, were the people hunters? Gatherers? Traders? Did they live in families? Were they peaceful or warlike?” Once his teacher handed out a passage from the Iliad and asked students to decipher ancient Greek life. “Only nobody told us it was from the Iliad,” Meltzer says. “It was just data. This was one of the most exciting experiences I’d ever had in my life.”
Meltzer went to Yale for college, where he embarked on two separate paths—economics and chemistry—thinking eventually he’d choose between them. He never did. Instead, as a junior he began writing to admissions directors at MD/PhD programs across the country. Usually applicants planned to combine a medical degree with a basic science doctorate; making his way toward a career in health economics, however, Meltzer wanted to study econ. “I just wrote letters saying, ‘Would you consider this?’” Most schools said no, but a few took the bait. One of them was Chicago. “Later I found out that no one had actually done this here, but nevertheless I came.”
All of the above: not choosing between medicine and economics earned Meltzer a career in both.
He came—passing up offers from established health-economics programs at Harvard and MIT—because of work Becker and others were doing in labor economics, human capital, fertility, and economics of the family. “That was before the U of C had won this incredible stream of Nobel Prizes,” Meltzer says. “It wasn’t just fame; it was the work itself that was so exciting. And there was a great medical school on the same campus, so it just made sense. Plus, I was coming home.”
Meltzer enrolled in 1986, alternating two years of economics with two years of medicine. “David was excellent at dealing with data,” Becker says. “When he first said he was going to combine medicine and economics, I was skeptical. They’re both so hard and time-consuming. But he was right and I was wrong. ... We’ve had a few others do the same thing since then, and he’s encouraged a few. Certainly it’s still a small number, but it’s a growing number, and people like David have blazed the way.”
As a third-year med student in 1991, Meltzer began teaching health economics at the Harris School, filling a departed instructor’s slot. Two years later he headed to Harvard-affiliated Brigham and Women’s Hospital for an internal-medicine residency. “I knew I needed to get out of Chicago,” he says, “because I thought someday I’d want to come back.” In Boston he solidified his future-costs hypothesis, indirectly inspired by graduate-school research he’d done with Chicago urologist Gerald Chodak examining the usefulness of prostate-cancer screenings. “Chodak was skeptical,” Meltzer says, about the practice of universally testing older men, for whom an indolent tumor is often of least concern. “He basically thought we were finding a lot of cancers that were never going to kill anyone—and causing a lot of impotence and incontinence with radiation and chemotherapy.”
With his future-costs theory hammered out, Meltzer flew to Phoenix. He returned feeling a little shaken.
By the following fall, however, his research was gaining ground. He had delivered his lecture to seven more audiences, including ones at Harvard, Chicago, the Society of General Internal Medicine, and the National Bureau of Economic Research. He reprised the presentation for a smaller session at the Society for Medical Decision Making’s 1996 meeting in Toronto, where it won a Lee B. Lusted Prize for original research (named after the former Chicago radiologist and medical decision-making pioneer). Meltzer’s future-costs article in the Journal of Health Economics won an outstanding-paper award from the Society for Medical Decision Making and a paper-of-the-year citation from the National Institute for Health Care Management. “So that got me into the field,” he says, “and I’ve been kind of doing cost-effectiveness analysis since then.”
That’s putting it lightly. Since 1996, when Meltzer finished his residency and joined Chicago’s faculty full time, he’s studied cost-effectiveness relating to patient preferences, simultaneous health problems, medical-research priorities, Medicare policies, physicians’ learning processes, technological advances, and the “spillover effects” that illness imposes on patients’ families. More than 15 years after he began studying prostate-cancer screening, Meltzer and his colleagues continue to gather data and refine their conclusions. For several years he’s asked a similar question of intensive treatment for type 2 diabetes: for whom is it worthwhile? Intensive treatment requires more pills and insulin shots than conventional therapy—and can cause more frequent low blood-sugar episodes. “On average, the people who choose it benefit from it,” Meltzer says. But for many seniors, the side effects supersede the remedy. “Applying it indiscriminately to older people, ignoring their preferences,” he says, doesn’t make medical or economic sense.
“One of the things David has contributed to the field,” says Marshall Chin, associate professor of internal medicine, codirector of internal-medicine research, and one of Meltzer’s closest collaborators, “is an individualized approach to care. Most societal cost-effectiveness analyses just use population averages for how a patient values a treatment or avoiding a complication like blindness. David is one of the first people to start utilizing data from individuals.”
Sometimes, says Meltzer, giving individualized care involves a purely clinical judgment: “A patient has terrible, incurable heart disease and would die on the operating table. Well, then we shouldn’t do that to them.” More often, though, it also means taking a patient’s wishes and worries into account—“doctors shouldn’t be assuming what patients want”—and rarely is any situation straightforward. As a rule, 80-year-olds expect shorter lives than younger patients. “Does that mean older people shouldn’t get treatment?” Meltzer asks. “No. It means you should think about the capacity to benefit, to get meaningful improvements, when you make a decision. It may be that somebody’s grandfather has a longer life expectancy than a very sick young person. Or maybe an older person has a bunch of other problems that make solving this one even more important.”
Attention to individuals isn’t only a more mindful way to practice medicine; it can also be a tool, Meltzer says, for controlling costs by forcing wiser, shrewder decisions about technology. More than any other factor, technology adds to medicine’s expense. “One of the problems over the last half-century is that we’ve paid for whatever we’ve developed. There’s something insidious about that—there’s less incentive to create cost-saving technologies.” The solution, he says, is to be more selective about how and when to use medical innovations. An example is the Pap smear. Done every three years, it costs little and provides effective cervical-cancer screening, but used annually, as it most often is, the Pap smear adds only a few extra hours of life expectancy and costs more than $1 million per life-year saved. “This doesn’t mean,” Meltzer says, “that all new technology should save money—that isn’t at all the point—but there should be significant incentive to develop new technologies that make us healthier, but at a lower cost. Or at less of an increase in cost.”
Doctors and patients, says Meltzer, need to be more selective about how and when to use medical innovations.
One of Meltzer’s most far-reaching cost-effectiveness projects began as a suggestion for saving money on Medical Center inpatient care. “We were in a hospital meeting and [former internal-medicine section chief] Wendy Levinson said she wanted to try hospitalists,” Meltzer recalls. “And it was like, what’s a hospitalist?” That was 1997; the term had been coined only a year earlier to describe hospital-based general physicians whose sole responsibility consists of caring for inpatients from admission through discharge. (Conventionally, primary-care and attending physicians make periodic hospital rounds, seeing hospitalized patients once or twice a day.) Chicago hired two hospitalists to cover the general-medicine service, and Meltzer launched a study of their cost-effectiveness. “We hired a medical student who was taking the year off and an undergraduate work-study student, and we started collecting data,” he says. The research team interviewed patients at admission and one month after discharge; they reviewed patient charts and studied administrative data; they surveyed primary-care physicians, attending physicians, residents, and interns. In nine years, the study has enrolled 70,000 patients at Chicago and 20,000 more across the country. In a 2002 article in the Annals of Internal Medicine, Meltzer analyzed the first two years of data, reporting that hospitalists saved $800 per admission, cut length of stay by half a day, and lowered patient mortality—“but not really until the second year,” he notes, when hospitalists’ performance had improved. “They were getting better over time.”
They were also burning out. Hospitalists’ “intensity of experience”—the daily litany of pneumonia, heart failure, heart attacks, strokes, and renal disease; weeklong on-call shifts; and weekends without rest—“makes these doctors more effective clinicians,” Meltzer says, but it also exhausts them. At Chicago hospitalists started quitting after two years, to be replaced by newcomers. In 2001 Meltzer convinced Medical Center higher-ups to change the job description. Four hospitalists would share the patient load previously carried by two, and as part of an academic program, they’d also do quality improvement and research. “We said, ‘Let them build academic careers,’” Meltzer says. “We went from having 30 to 40 percent turnover a year to almost none.” This past summer the hospitalist program grew, adding six physicians; expanding its coverage to the lung-transplant, liver-transplant, cardiology, and cancer wards; and stationing a physician in the hospital around the clock. “And we’ve built a group of people,” he says, “who are publishing, getting grants, getting promoted, becoming leaders in the institution.”
Hospitalist Chad Whelan is one of those people. He arrived at Chicago six years ago as a clinician and educator; Meltzer helped him become a researcher and, eventually, an administrator. Now he directs the year-old Hospitalist Scholars Training Program, which hires internists as junior faculty for two years and combines their clinical work with research and social-sciences coursework. He also sits on several Medical Center committees, chairing the health-information management committee and serving as a “physician champion” for implementing electronic medical records. He researches pain measurement and management, gastrointestinal bleeding, and the decision-making resources available to physicians.
“I would never have predicted this was the direction I would take,” Whelan says. “When I came here I had no research training.” Meltzer stoked his academic curiosity and coached him on his first research project, an evaluation of pain on the general-medicine ward. “Many traditionally trained researchers would have said, ‘It’s really hard for you to do this, you don’t have research training, you might want to focus on something else,’” Whelan says. “But David’s view is, if you’re excited by it, let’s find a way to get you immersed in it.”
Whelan is also immersed in Meltzer’s hospitalist research, which continues to collect data. The latest cost-effectiveness findings show that hospitalists still save money, but not as much as before. Meltzer says the numbers don’t tell the whole story. “What we’re finding is that hospitalists are making the care provided to nonhospitalist patients better.” He cites one study tracking the use of a new blood thinner that enables patients to leave the hospital days sooner. “We found that hospitalists tried it first and taught it to the interns and residents, who in turn taught it to other physicians.” Rigorous workdays provide hospitalists physical and mental dexterity, give them an incentive to learn better practices, and introduce them to a broad network of specialists who can offer advice. Eventually, other physicians reap those benefits, too. “So the differences we measured between hospitalist and nonhospitalist services underestimate the real effects,” Meltzer says. “What we’re arguing is that these hospitalists basically produce a change in culture, that they change the people around them and create a collective body of knowledge and set of habits.”
Although it is one of the nation’s fastest-growing specialties—on its way, Meltzer believes, to becoming the dominant model for inpatient care—hospital medicine is still charting its course. Academic medical centers, as training grounds, have an essential responsibility, he says. “If we convey hospital medicine as clinical tedium and [only] nonacademic pursuits, then that’s what we will get. But if we instead convey hospital medicine as exciting, where patient care is valued but integrated with quality improvement and research and teaching, then we’re going to get people who hold those values. Right now I think it could go either way, and what we’re doing at the U of C is a crucial part of making sure it goes the right way. We’re helping this profession define itself.”
It’s a role Meltzer plays often. In the past decade he has helped give definition to health economics at Chicago and coaxed more than one scholar into the field and onto the faculty. Geriatrician William Dale, AM’94, PhD’97, MD’99, an assistant professor of medicine and collaborator in Meltzer’s research on prostate-cancer screening and simultaneous health problems, says he “never envisioned” himself with a doctorate until, as a restless Pritzker student, he met with Meltzer to discuss Chicago’s MD/PhD program. “I still remember sitting in the hospital cafeteria, at those fold-down tables, and saying to David, ‘Should I consider this?’” Four years later, Dale had earned a Harris School PhD; now he serves under Meltzer as associate director of the MD/PhD program in medicine and the social sciences. “I’m more of a behavioral economist; I do the psychology of decision-making—the emotional factors that affect decisions, like anxiety,” he says. “That’s not David’s thing, but he’s encouraged me to pursue it. He’s opened a lot of opportunities for me.”
Anirban Basu, PhD’04, an Indian-educated pharmacist with a biostatistics master’s from the University of North Carolina, enrolled in the Harris School in 1999, in part to study with Meltzer. The two have worked together on three separate projects: uncertainty in medical decision-making, the relative value of research—“should we develop a screening program versus a treatment for children with genetic disorders like muscular dystrophy?”—and illness’s spillover effects within families. After graduation Basu took a job as the only non-MD in Chicago’s internal-medicine section, turning down an offer from Yale’s health-policy department. “It was hard because there I would have been surrounded by health economists,” Basu says. But he was convinced that Chicago’s “environment”—which Meltzer helped shape—“was right. Even though the senior faculty is not here, even though you are almost alone in your department, you can always look around campus and reach out.”
Meltzer is trying to make that reach even easier. Tucked into a cluster of first-floor conference rooms along the hospital’s southern wall is the Center for Health and the Social Sciences (CHeSS), a fledgling effort at buttressing cross-curricular cooperation. Last year Meltzer helped organize CHeSS, which recruits scholars—about 100 so far, including Chin, Basu, Whelan, Dale, and Becker—from the biological and social sciences, the University’s professional schools, and the National Opinion Research Center. The organization unites them around common causes (as well as common data sets, research staffs, and grants) like health disparities, cancer risk factors, pharmaceutical policy, and academic research training for gifted high-schoolers. Faculty attached to each of CHeSS’s ten programs meet regularly to brainstorm ideas and share findings. They arrange workshops and coordinate resources. “The idea is to institutionalize interdisciplinary research while keeping people in touch with their original disciplines,” say Meltzer, who serves as CHeSS’s director. “People clear their schedules every Tuesday morning or whatever to come here, and it doesn’t require an act of Congress to get them together. It brings a degree of agility to these problems.”
Whether convening meetings between scientists and sociologists or converting calculus to algebra, Meltzer brings his own agility to medical problem-solving. Small wonder that this October, 11 years after his contentious presentation to the Society for Medical Decision Making, he begins a term as the group’s president.