TY - JOUR
AU - Berkson, Joseph
AB - SMOKING AND LUNG CANCER: SOME OBSERVATIONS ON TWO RECENT REPORTS JOSEPH BERKSON Mayo Clinic, Rochester, Minnesota "I do not reject the use of statistics, but 1 condemn not trying to go beyond them .... " CLAUDE BERNARD H E opinion that smoking is a cause of cancer of the lung, expressed by a number of writers and reflected in some recent public announcements [4, 16], rests to a considerable degree, though not entirely, on two now cele­ brated "prospective" statistical studies, one reported by Doll and Hill [5, 6], the other by Hammond and Horn [10, 11]. In a commentary article in 1955, [1] I pointed out that the results presented in the first report of Hammond and Horn [11] exhibited an association of smoking with increased death rates, not only for cancer of the lung, but also for each of the other several categories of cause of death into which all causes had been segregated, namely, cancer ex­ clusive of lung, diseases of the coronary arteries, and other diseases. In the sec­ ond report of these authors, the association of smoking with the death rates from causes other than lung cancer is shown even more clearly. In the 
TI - Smoking and Lung Cancer: Some Observations on Two Recent Reports
JF - Journal of the American Statistical Association
DO - 10.1080/01621459.1958.10501421
DA - 1958-03-01
UR - https://www.deepdyve.com/lp/taylor-francis/smoking-and-lung-cancer-some-observations-on-two-recent-reports-ClC4PaKIQ4
SP - 28
EP - 38
VL - 53
IS - 281
DP - DeepDyve
ER -