"The pump don't work, 'cause the vandals took the handles".

Abstract

A t the request of EPIDEMIOLOGY, a group of epidemiologists celebrated the anniversary of the removal of the Broad Street pump handle by paying a visit to the Pub, just next to the famous pump. We gathered in the upstairs bar to sign the visitors' book (under the names of many other epidemiologists through the years) and inspected the memorabilia. We dedicated the first round of drinks to the controversy between the miasma theory and the germ theory of disease. With our second round of drinks, we celebrated the centenary of the 1904 report that explored why so many young British working-class men were physically unfit for military service.2 This report ushered in a new era that emphasized the childhood origins of adult health. Scientific controversy in this era centered around the relative importance of nature and nurture.3 The pub hosting our celebrations had marked the centenary of the Broad Street pump in 1954 by changing its name from the Newcastle-upon-Tyne to the John Snow (an ironic commemoration, given that himself was a teetotaller). Sir Austin Bradford-Hill had unveiled the new inn sign. So, with our third round of drinks, we toasted the jubilee of Doll and Hill's 1954 first report from their prospective study of smoking and cancer.4 Described as one of the intellectual levers of modern epidemiology,5 this study exemplified the adult-lifestyle model of disease causation. Like the connections made between contaminated water and cholera, the link between cigarette smoking and lung cancer was challenged by medical and statistical scientists and took time to be accepted by the general public. Scientific controversies, the emergence of scientific facts, and the closure of scientific debates are all social processes.6 The title of this piece (courtesy of Bob Dylan) serves as a reminder that public health interventions are not always appreciated by the people they are intended to benefit. The etiologic models of disease causation represented by these 3 anniversaries of 1854, 1904, and 1954 all remain relevant today. What model emerging in 2004 might be singled out by future epidemiologists as the most influential in 50 or 100 years? Our nonrandom sample of U.K. epidemiologists dragged from the Pub for a photograph around the famous pump would not be drawn on these vexing questions. It was left to George Davey Smith, absent for once from a drinking session, to propose Mendelian randomization as a current major advance for epidemiology!