Even if screening research shows a reduction in mortality, the “real” effect of screening may be diluted by factors such as pre-testing and contamination. In this paper, researchers aimed to describe the extent of pre-testing and contamination in the Göteborg-1 prostate cancer screening experiment.
Over the course of 10 years, beginning in 1994, a total of 20,000 men aged 50 to 64 were invited and randomly assigned to either a screening group (given prostate-specific antigen testing every 2 years) or a control group (not provided testing). The end of 2014 was the endpoint for the follow-up. The screening and control groups were subjected to comprehensive examinations to determine the results. Prostate-specific antigen levels more than or equal to 3 ng/ml were considered diagnostic of an enlarged prostate.
About 4.2% of men in the screening group and 4.6% of men in the control group were tested before the trial began. About 72% of men in the control group (contamination) took at least 1 prostate-specific antigen test during follow-up, compared to 87% of men in the screening group. Only 24% of the men in the screening group had prostate-specific antigens that were at or above the threshold, while 39% of those in the control group did. Within 12 months of a high PSA reading, 66% of men in the screening group and 28% of men in the control group had a prostate biopsy.
The screening and control groups had their prostate-specific antigen levels checked. However, only a small percentage of false-positive results resulted in a biopsy. Also, men in the screening group typically began the process earlier in life. Both of these possibilities could explain why investigators found that standardized testing for prostate cancer was more successful than informal screening. The benefits of structured screening are expected to be much bigger than those seen in the Göteborg screening study when tested against a non-screened population.