ROBERT S. HOLZMAN, M.D.
This content is PDF only. Please click on the PDF icon to access.
To the editor: It has become a common practice, in the Annals and elsewhere, to describe the duration of a clinical study in patient-years (the sum of the number of years each patient was followed). Thus, if 3 patients were followed for 6 years and 12 for 1 year, we would have 30 patient-years of experience.
The inflationary nature of this procedure is obvious in the example. More important, no adequate estimate of the usual duration of follow-up and its variability can be obtained from the number of patient-years. Dividing by the number of patients does give the mean duration,
HOLZMAN RS. Proper Statistics for Clinical Studies. Ann Intern Med. 1971;75:649. doi: 10.7326/0003-4819-75-4-649_2
Download citation file:
Published: Ann Intern Med. 1971;75(4):649.
Copyright © 2019 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use