[Skip to Content]
[Skip to Content Landing]
October 17, 1990

Benefits and Harms: The Balance Sheet

Author Affiliations

New Paltz, NY

New Paltz, NY

JAMA. 1990;264(15):1945. doi:10.1001/jama.1990.03450150043014

To the Editor.—  In his otherwise excellent article on the effectiveness of colorectal cancer screening, Dr David Eddy1 makes a common but serious arithmetic error.Without screening, a high-risk 50-year-old man (like me) has about a 10.3% probability of developing colorectal cancer sometime in the rest of his life and about a 5.3% probability of dying of colorectal cancer. If such a person is screened as Dr Eddy proposes, the probabilities of developing and dying of colorectal cancer are decreased to about 7.3% and 2.9%, respectively. Thus, the effect of this screening strategy is to decrease the actual probability of developing or dying of colorectal cancer by about 3.0% and 2.4%, respectively.Dr Eddy's calculations are erroneous, and the mistake recurs throughout the article. In actuality, a decrease in the likelihood of getting cancer from 10.3% to 7.3% is a 30% improvement. When cancer mortality drops, as a result