Lead-Time and Length-Time Biases

Two important concepts related to the reported benefits and risks of screening tests deserve consideration: lead-time and length-time biases.^[198] These two factors can indicate an apparent benefit of testing when there is none. Let us examine screening for lung cancer in smokers and screening for breast cancer. In a Czechoslovakian trial of screening of smokers, men at high risk were randomly assigned to either chest radiography twice a year or no screening.^[199] No difference in mortality from lung cancer occurred. However, the 5-year survival rate from the time of diagnosis was 23% for the screened group and 0% for the control group. Thus this apparent improvement was entirely owing to earlier diagnosis ("lead-time bias"). In fact, mortality from lung cancer was actually higher in the screened group, indicating that the real effect of screening and subsequent intervention was negative.

Black and Welch^[198] reviewed the only other trial of screening strategies using diagnostic radiology evaluated in a randomized trial. In the first (and to date, only) randomized trial of mammographic screening—the Malmö mammographic screening trial^[200] —the benefits of this test also seem to disappear. Women over 45 years of age were randomly assigned to either regular mammography or no screening. Although survival from the time of diagnosis was 80% higher for the screened group, no difference existed in mortality from the time of randomization. Thus, the apparent increase in the survival rate for screened patients was entirely owing to lead-time bias (earlier diagnosis) and length-time bias (less severe cases diagnosed) associated with screening test detection, and not to any benefit from the screening strategy itself.

Before one concludes that no tests should occur preoperatively, let us remember that detection of subclinical conditions in high-risk groups and optimization of therapy for clinical conditions can result in less perioperative morbidity, fewer changes in perioperative plans, and better informed discussions of risk with the patient and significant others. In fact, more recent tests of mammography have also shown this lead-time bias but have shown a survival benefit as well: about two thirds of the benefit of screening mammograms in the over-age-40 population was due to lead time and one third to a real survival advantage.^[201]