Daniel Schwartz

Abstract

Pioneer of evidence-based medicine in France. Born in Paris, France, on Jan 30, 1917, he died in Paris on Sept 6, 2009, after a series of strokes, aged 92 years. “Before 1950, medicine was an art. Daniel Schwartz helped to turn it into a science”, said his long-time colleague Joseph Lellouch, past director of research at the Institut National de la Santé et de la Recherche Médicale (INSERM). Schwartz helped to launch the discipline of epidemiology in France and led the statistical training of many French medical researchers and doctors, thus preparing the ground in France for evidence-based medicine. Schwartz's family was, and remains, intellectual and illustrious, in the fields of medicine, literature, art, politics, and mathematics in France. His parents were refugees from the Franco-Prussian war in 1870–71. Schwartz himself played a part in the French resistance in the south of France during World War II. Trained as a mathematician in the École Polytechnique, he began work in the French public tobacco monopoly. There he applied statistics to the highly variable incubation period of tobacco mosaic virus, developing a skill in relating mathematics to life. “And then the head of the Institut Gustave Roussy, a very bright surgeon called Pierre Denoix, said you are a smart Polytechnician, you understand numbers, and you are crunching numbers in the tobacco monopoly, so come and work in my cancer hospital, we need people like you—and no-one was thinking that way then”, explains Catherine Hill, an epidemiologist at Biostatistics and Epidemiology Unit, Institut Gustave Roussy, and a former student of Schwartz. Schwartz taught statistics at the Institut Henri Poincaré in Paris and, in 1959, he became head of research at the French National Institute of Hygiene. Schwartz went on to become director of research statistics for INSERM from 1960 to 1985 and founder and director of the Centre d'Enseignement de la Statistique Appliquée à la Médecine et à la Biologie Médicale. As Schwartz described in his popular 1994 paperback on the fundamentals of statistical medicine, Le Jeu de la Science et du Hasard (The Game of Science and Chance), in the 1950s bronchial cancer was growing “explosively and spectacularly…Yet the cause was unknown, and the projections were dramatic”. Schwartz says “modern epidemiology was born with the studies of Doll and Hill” on smoking and cancer in that period. Indeed, Schwartz visited Doll in the UK during his epoch-making studies on cancer and smoking. Alfred Spira, who worked for 40 years with Schwartz, and was his successor in the School of Medicine at the Université de Paris Sud, said “Just a few weeks before he died, I discussed with him exactly what he did”, after his visit to Doll in the early 1950s. “First of all he spotted a way to improve the sensitivity of a case-control study on the inhalation of tobacco smoke, as opposed to smokers who did not inhale”, said Spira. The study began in 1953, and was reported (with Denoix) in 1961. “He also published a paper concerning tobacco and bladder cancer, and to my knowledge, he was the first to show this relationship. But—unfortunately he published it in French, so it was not so widely read.” Schwartz co-authored a landmark paper with Lellouch on the crucial distinction between explanatory clinical trials and pragmatic ones (J Chron Dis 1967; 20: 637–48). Their paper described this distinction and its critical consequences for trial design and interpretation, and was recently recognised as an important paper—it was republished last year in the Journal of Clinical Epidemiology. In Le Jeu de la Science et du Hasard, Schwartz bemoaned the difficulty of establishing a statistical mentality in France. “The French are first of all rigorous—Cartesian” he said, so the admission of randomness is hard in French culture. Nevertheless, says Philippe Lazar, a past Director-General of INSERM—who worked for 20 years with Schwartz—“Daniel Schwartz found a language that uses mathematics, and uses statistical models, without having to master a very high level of mathematics to understand them. I think that was a fabulous contribution. So he had a major impact on the spirit of medical research in France, in almost all fields. Most French doctors have been trained by himself or his colleagues, directly or indirectly.” Schwartz was a constant humorist and Hill recalls how he would amuse audiences by “constantly pulling notes out of his socks, his shoes, his pockets”. Schwartz leaves a brother, Bertrand; two sons, Maxime and Yves; and a daughter Irène.