Editorial: Judgments in Medicine: The Assumption of Risk

PERSPECTIVES IN BIOLOGY AND MEDICINE Volume 11 · Number 4 · Summer 196S EDITORIAL: JUDGMENTS IN MEDICINE: THE ASSUMPTION OF RISK Science has rules for testing claims to knowledge. These are aids to developing a body ofknowledge and concepts which are accepted as verities . But not even science could function if communication and decisions werelimited to knowledge which has been verified by experiment. In the practice ofmedicine the physician must daily make judgments based on information that does not always meet strict scientific standards ofvalidity . Certainty is not required when to withhold a decision or advice involves greater risk of harm to the individual or society. There is risk oferror inmakingadiagnosis; there is risk intheadministration ofany drug and in any major surgical intervention. The risk ofintervention is balanced against the risk assumed by not intervening. Thus, a surgeon would not undertake the replacement ofa diseased but effectively functioning organ but might assume high risk offailure of a transplant ifthe alternative was almost certain death of the patient. There is risk of error in most decisions on human welfare made by man. This should be remembered by agencies of government, which gain increasing control over therapeutic agents and medical and surgical interventions by promising to reduce risks to near zero but thereby increase the likelihood of impeding medical progress. This should be remembered in the field of jurisprudence, where overzealous efforts to abolish the risk oferror in enforcement oflaw gravely impair the protection ofthe property and person of the good citizen. This should be remembered by scientists and all others who claim that any proposition must be established beyond doubt before it can be used as a basis for ajudgment. The student ofscience must learn to live in the atmosphere of "probably so." The claim that the smoking ofcigarettes causes diseases in man is a case in point. Men and women who smoke cigarettes have a higher incidence j25 oflung cancer and certain other diseases than do non-smokers, but correlation does not prove causal connection. Condensates oftobacco smoke can cause tumors ofthe skin ofmice, but it has not been possible to produce human-type lung cancers in laboratory animals upon demand. Tobacco smoke has been found to suppress ciliary motility in epithelium isolated from the upper respiratory tract ofseveral species oflaboratory animals. A number of pathologists who have reported on pathologic changes in the bronchi ofsmokers consider these lesions to be premalignant. The correlation between the amount of cigarette smoking and the incidence of disease and life expectancy is impressive. The evidence that it is salubrious for a smoker to stop the use ofcigarettes is impressive, but there is still the factor of non-random self-selection, which may invalidate the evidence. Not all alternative hypotheses have been excluded. The hypothesis that the association between smoking and disease is due to poorer health habits and to constitutional differences in susceptibility to disease by those who choose to smoke as compared to those who choose not to smoke is supported by some evidence and has not been disproved. It is not possible to do the crucial experiment ofproducing lung cancer at will in randomly selected human subjects. Attempts to produce human-type lung cancer at will in laboratory animals have failed. Although some medical scientists claim that a causal connection between lung cancer and cigarette smoking has been proven beyond all doubt, the evidence does not meet all requirements for proofofa proposition , and further study is needed. However, the hypothesis is plausible, and different lines of evidence—although indirect—converge in support ofit. There is no evidence that non-smokers would benefit by adopting the habit of smoking cigarettes. It seems reasonable to assume that cigarette smoking is a serious health hazard and to teach that this is probably true until there is impressive new evidence to the contrary. The risk that the hypothesis is incorrect seems small and is not greater than is commonly assumed in the practice of medicine or in the making of other decisions affecting human welfare. D.J.I. 526 Editorial Perspectives in Biology and Medicine · Summer 1968 ...