Screening Tests For Illness: How Often Is a Positive Result an “Error”?

I have continuously stressed that you should be screened for serious illnesses. This includes heart disease, bone disease, diabetes, as well as cancer. Cancer screening tests are very well defined, and you probably have already had some of them if you are over the age of 40 to 50. Screening for serious chronic disorders can allow your doctors to diagnose these diseases earlier when there has been less damage to organs, and this can preserve your quality of life.

But how often are the results of these tests erroneous? How often does a test suggest that you may have a serious illness, but it is not really cancer or is just an error in the testing process?

To help to answer this, Dr. J. Croswell and Dr. B. Kramer and co-authors published an article to look at the accumulative incidence of false/positive results in cancer screening tests over a period of time (Annals of Family Medicine, Volume 7, page 212, 2009). The authors from the National Institutes of Health and participating centers throughout the country looked at the data from the ongoing PLCO (Prostate, Lung, Colorectal, and Ovarian) cancer screening trial. This trial consisted of over 68,000 men and women between the ages of 55 and 74. Half of these patients randomly received usual care, and the other half received screening tests for cancer which included flexible sigmoidoscopy (insertion of a tube in the rectum to detect cancers of the colon and rectum), chest x-rays, digital rectal examinations and PSA (prostate specific antigen) testing in men for prostate cancer, transvaginal sonograms and CA-125 blood tests in women to detect endometrial and ovarian cancer. Over a three-year screening period, 14 cancer screening tests were conducted for men and 14 tests were conducted for women.

In the report, the risk of at least one screening test positivity that was not associated with cancer was 60% for men and 49% for women! The frequency of having an additional diagnostic procedure as a result of the positive tests were 29% for men and 22% for women.

These results are very impressive. It means that whenever you get a positive test in a screening process, it is quite possible that the test is indicating a benign condition, not cancer (!) or has just shown a positive result erroneously.

In looking at the results in more detail, the most common cause of a false positive result was a finding on flexible sigmoidoscopy that showed a condition, but not cancer. Very often this was finding a polyp, evidence of colitis (inflammation of the colon) or evidence of some bleeding or diverticulosis (small out-pocketings from the colon). I must emphasize that each of these conditions is important to diagnosis even though it is not cancer. This allowed these men and women with abnormalities to be more appropriately treated. Removal of the polyps prevented many cases of cancer in the future. Although these were considered false positives in the article which was published, they are more likely significant results that show illness other than cancer.

In women, the next most common findings of false positive results were abnormalities on the chest x-ray or abnormalities on transvaginal ultrasound. Again, the chest x-ray may have indicated chronic conditions that require follow up, such as emphysema or lung nodules from previous infection or abnormal cardiovascular findings, and these were helpful for the individual even though they were not cancer. Similarly, transvaginal ultrasounds may have found ovarian cysts which required observation, but were not cancer, or may have found that uterine masses such as leiomyoma or polyps or fibroids which require follow up, and which are important.

In men, abnormalities were found not only in flexible sigmoidoscopy, but false positives were also found on chest X-rays, and on digital rectal examination and PSA. In addition to the abnormalities on chest X-ray mentioned above for women, abnormalities in the prostate may have included prostatic enlargement due to benign conditions (benign prostatic hypertrophy or BPH) resulting in symptoms that can be treated with medications, or prostatitis which can again be treated with antibiotics to reduce abnormalities.

The conclusions of the authors are subject to severe misinterpretation. The conclusions which an individual might come up with is that since 50% of people undergoing cancer screening have false positive tests, I should not have cancer screening! This is entirely wrong. The findings on cancer screening result in diagnosing cancer at earlier stages when it is more curable. The false positives identified in this article showed other illnesses that require treatment or observation, which is helpful to the individual.

The authors of this article appropriately concluded that individuals need more education about the possible false positives and resulting possible surgeries to correct these abnormalities, before they begin cancer screening. I agree completely with this!

When you are having cancer screening, you should know that it is likely that benign conditions will be found and that many of these benign conditions may require treatment, (medical or even surgical). This is helpful to you. However, physicians should also emphasize that it is important to undergo appropriate cancer screening so that cancer will not be found when it is incurable, but instead be found when it is at a very early stage and easily treated.