When 'Verification Bias' Removed, PSA of 2.6 Now Upper Limit of Normal
For Screening

August 21, 2003

Abstracted from:
New York Times and New England Journal of Medicine

NY Times Editorializes on PSA Test

New York, July 30, 2003 - It's not everyday that the New York Times takes a stand on a blood test. Yet on today's editorial page of that great newspaper---alongside opinions on the U.S. role in Israel, the proposed terrorism futures' market, and plans for nuclear waste disposal---we find "A Fallible Prostate Cancer Test." Such is the importance of prostate cancer and PSA screening in today's world.

New England Journal of Medicine

Stimulus for the Times editorial was an article in the July 24 issue of the New England Journal of Medicine, "Effect of Verification Bias on Screening for Prostate Cancer by Measurement of Prostate-Specific Antigen." In their premise for this article, the Harvard authors state that previous studies have failed to take into account the fact that only a fraction of men screened have actually had their disease status adequately verified. Thus, the true value of PSA as a screening test may be different from the numbers usually quoted.

The authors applied a mathematical model that estimates probability of prostate cancer in a population as a function of the following clinical variables: digital rectal exam, race, family history, PSA level, and age. A fitted logistic regression model was taken from the subgroup undergoing biopsy to predict the probability of prostate cancer in the entire group. Adjusted ROC curves were then derived by calculating the probability of being in a particular PSA category for men in the entire sample with and without prostate cancer. The study showed that after correction for verification bias, improvement in both sensitivity and specificity of PSA screening is observed. The key figure is shown here.

From analysis of the ROC curves, the authors found "that lowering the threshold for biopsy from 4.1 to 2.6 ng per milliliter in men younger than 60 years would double the cancer-detection rate from 18 percent to 36 percent, whereas the specificity would fall
only from 0.98 to 0.94." The authors concluded, "Both of these findings support the use of a lower threshold PSA value for a recommendation of biopsy."

In a NEJM editorial accompanying the article, Fritz Shroder from Rotterdam cautioned that "lowering the PSA threshold for performing a biopsy will increase the rate of overdiagnosis and, potentially, overtreatment. New recommendations for screening should arise from ongoing, randomized studies that are designed to show whether screening indeed reduces mortality from prostate cancer without unacceptably reducing the quality of life."

The New York Times editorial concluded that, "The deeper problem is that no one has yet demonstrated that finding prostate tumors early saves lives or improves health."