Sensitivity and Specificity

 

 

Part I

Imagine if we could prevent cancer the way vaccination has eradicated polio and smallpox.

In the absence of a definitive prevention strategy, medical science’s best efforts have been applied to finding the disease early enough so that it is curable.

The logic is simple: Cancer evolves from a pre-cancerous state to a noninvasive state to an invasive state. The greater the number of cancer cells, the greater the risk of mutation to an invasive state. The greater the number of invasive cells, the greater the risk of those invading cells escaping into the circulation. The greater the number of cells in the circulation, the greater the likelihood that some cells will anchor and nest in the blood vessels of another ­organ, creating metastases (spread). The greater the probability of metastases, the lower the probability of cure.

Following this logic, the key to success in cancer management is to find cancer before it is invasive, or if invasive, as soon as possible before it metastasizes. Unfortunately, some cancer screening tests that attempt to find the cancer early, before it produces symptoms and when it is curable, have had mixed success. Unlike the unquestioned success of the pap smear in allowing the early detection and cure of cervical cancer and colonoscopy for colon cancer, other tests like PSA (prostate specific antigen) testing for prostate cancer, mammogram screenings for breast cancer, and chest imaging for lung cancers remain controversial.

The controversy surrounding prostate, breast, and lung cancer screening relates to the concept that a screening tool must be both Sensitive and Specific.  A sensitive test clearly distinguishes diseased from normal structures so there are no false negative results, if disease is there it will be detected.  A specific test means that any abnormality detected is the disease we’re looking for and not something else so there are no false positive results.

Numerous observations challenge the specificity and sensitivity of prostate, breast, and lung cancer screenings.  Furthermore, even when the test is sensitive and specific it must be proven to save lives.  In the coming weeks I’ll explain why each of these cancer screenings: prostate breast and lung each remain controversial. We’ll be talking more about sensitivity and specificity, false positives and false negatives, and clinical validity and clinical utility.

What has been your experience with cancer screening?

Be Well,

Dr. Bruce Feinberg