Etiologic factors in carcinoma of the penis.

T h e main objective of this paper is to evaluate the significance of certain etiologic factors in carcinoma of the penis. The factors considered are venereal disease, circumcision, and the age and race of the patients. A second objective is to illustrate the use of control groups in a statistical study. The use of controls is routine in experimental work and every experimental group is checked by one or more controls. In statistical studies on cancer, however, control groups are not as frequently used. This paper exemplifies several types of control groups and considers the necessity and advantages in the use of controls in statistical work. Clinical and control grouu . -The first table summarizes some statistics on the clinical and control groups studied. The clinical group under investigation consisted of 139 men with carcinoma of the penis. These patients were admitted to Hines Veterans Hospital during the I4 year period from 1931 to 1944. A detailed report on these men was published by Lenowitz and Graham ( I ) . 1931 to 1944. For most purposes, this group is too large and cumbersome and has, therefore, limited value as a control. A second control group, B, was composed of 236 men admitted during 1931 to 1944 with carcinoma of the kidney. This group was useful in studying the incidence of venereal disease. It could not be used, however, to determine the incidence of circumcision because these patients had not been questioned in regard to this factor. A third group, C, was set up consisting of 209 patients who were admitted consecutively to the tumor clinic in 1944. These patients were interviewed by Mr. Philipp Zinkgraf for a study conducted by the U. S. Public Health Service. The last group included 4 men with carcinoma of the penis and 2 Jewish men. For theoretical reasons, it seemed advisable to exclude these 6 patients leaving a new control group, D, of 203 men with tumor. The number of colored men in the last control group