Related Experiment Videos
Screening for prostate cancer: estimating the magnitude of overdetection
M McGregor1, J A Hanley, J F Boivin
1Department of Medicine, McGill University, Montreal, Que. mcre@musica.mcgill.ca
Background:
No randomized controlled trial of prostate cancer screening has been reported and none is likely to be completed in the near future. In the absence of direct evidence, the decision to screen must therefore be based on estimates of benefits and risks. The main risk of screening is overdetection--the detection of cancer that, if left untreated, would not cause death. In this study the authors estimate the level of overdetection that might result from annual screening of men aged 50-70.
Methods:
The annual rate of lethal screen-detectable cancer (detectable cancer that would prove fatal before age 85 if left untreated) was calculated from the observed prostate cancer mortality rate in Quebec; the annual rate of all cases of screen-detectable prostate cancer was calculated from 2 recent screening studies.
Results:
The annual rate of lethal screen-detectable prostate cancer was estimated to be 1.3 per 1000 men. The annual rate of all cases of screen-detectable prostate cancer was estimated to be 8.0 per 1000 men. The estimated case-fatality rate among men up to 85 years of age was 16% (1.3/8.0) (sensitivity analysis 13% to 22%).
Interpretation:
Of every 100 men with screen-detected prostate cancer, only 16 on average (13 to 22) could have their lives extended by surgery, since the prostate cancer would not cause death before age 85 in the remaining 84 (78 to 87).