Christopher Schmid

Image

Christopher Schmid

Professor of Biostatistics

Frank Mullin/Brown University
Math meets medicine: By combining the data from more than 18 studies involving more than 6,000 patients, Christopher Schmid and colleagues devised and validated new equations for estimating a key but tricky metric of kidney performance.

In his youth, Christopher Schmid would watch and play sports, even manning third base for Haverford College’s JV team one year (and batting over .300). But for Schmid the game was always about more than the crack of the bat. He loved the numbers, too, and in the context of medicine and science, he has made them his career.

This summer he joined the Brown roster both as a professor of biostatistics and as a founding member of the University’s Center for Evidence-Based Medicine, along with fellow professors Tom Trikalinos and Joseph Lau.

When Schmid wasn’t in the Haverford dugout he pursued a math degree. Then he earned a master’s and a Ph.D. in statistics at Harvard. Health care wasn’t on his mind until he landed his first job, a statistical consultant position that happened to be at the Tufts University School of Medicine.

“I really had no intention of going into medicine, but a woman at Tufts told me they had an opening,” Schmid said. “In fact it was her job. She was leaving it.”

Ever since, Schmid has been applying math to medicine. He has developed a particular specialty of devising and implementing clever ways of merging data from disparate sources, such as clinical studies or databases, into a more informative single repository of knowledge, a field called meta-analysis. Lau, a well-known pioneer in meta-analysis for bringing computing to bear on problems in the field, introduced Schmid to meta-analysis about two decades ago and they have been working together ever since.

The power of successfully merging data from multiple sources can be found in a seemingly endless number of medical applications. Two meaningful examples seem to bookend Schmid’s career so far (though there have been many other applications and a number of more fundamental papers in between). In July, Schmid co-authored a paper in the New England Journal of Medicine that could have an important impact on diagnosing chronic kidney disease. By combining data from more than 18 studies involving more than 6,000 patients, he and his colleagues were able to devise and validate new equations for estimating a key but tricky metric of kidney performance. A “glomerular filtration rate” reading below a certain level indicates disease, but the estimation methods doctors have relied upon have had shaky accuracy. The new math is more accurate and could therefore prevent people from being misdiagnosed.

Schmid made a similarly important contribution in one of his first professional projects. In collaboration with colleagues at Tufts, Schmid helped develop the “thrombolytic predictive instrument.” That set of statistical models integrated data including electrocardiogram readings to help doctors predict a suspected heart attack patient’s near-term risk for further cardiovascular problems, such as stroke. Hewlett-Packard included it as a feature of their ECG machines.

“This was my very first time I had worked with putting databases together,” Schmid said.

As rewarding as his work has been over the last 20 years, Schmid was always in a medical center at Tufts, rather than in a more academic setting. “I worked at a hospital,” Schmid said. “There wasn’t a lot of methods research done. I had always been looking for a statistics or biostatistics department.”

Now that he’s come to the biostatistics department in Brown’s Program in Public Health, he’ll continue to influence medicine mathematically, but he can also interact more readily with methodological colleagues in computer science, math, and applied math, as well as engage new application areas such as ecology. In May he attended the University’s “Day of Data,” in which colleagues from many fields talked about how they grapple with huge amounts of quantitative information.

“There are obviously a lot of people in a lot of different areas here, and maybe someday one of our methods will be useful to them,” he said.

In addition to working with a broader set of colleagues, Schmid said he is also looking forward to working with likeminded students. At Tufts he mentored plenty of future physicians, but they were not primarily interested in statistics. At Brown he’ll run the biostatistics department’s master’s program.

“This enables me to continue my research but also to stay involved in education,” he said.

After all, in baseball terms, professors are not only players but also coaches.