15-Year Followup of a Population Based Prostate Cancer Screening Study

Abstract

We evaluated long-term survival in attendees and nonattendees of a 1-time screening for prostate cancer. A total of 2,400 men 55 to 70 years old in 1988 were randomly selected and invited to a screening for prostate cancer. Of the invited men 1,782 (74%) attended. Screening attendees were examined with digital rectal examination, transrectal ultrasound and prostate specific antigen analysis. When cancer was suspected, prostate biopsies were taken. A total of 65 men with prostate cancer were detected by this procedure. The entire source population comprising 27,204 men, including 618 nonattendees (26%), was followed for prostate cancer diagnosis and survival for 15 years. Incidence rate ratios were calculated using Poisson regression models. We found no effect of this screening procedure on the risk of death from prostate cancer and other causes of death (incidence rate ratio 1.10, 95% CI 0.83-1.46 and 0.98, 95% CI 0.92-1.05, respectively) when comparing all invited men with the source population. However, attending the screening program was associated with a significantly decreased risk of death from causes other than prostate cancer (vs source population incidence rate ratio 0.82, 95% CI 0.76-0.90). In contrast, the corresponding incidence rate ratio in nonattendees was 1.53 (95% CI 1.37-1.71). We found no evidence of a beneficial effect of this specific screening procedure but strong evidence of a difference in overall survival in screening attendees and nonattendees. These findings should be considered when interpreting previous and upcoming studies of the effect of screening programs.