ACP American College of Physicians - Internal Medicine - Doctors for Adults

Effective Clinical Practice

Primer on Lead Time, Length and Overdiagnosis Bias

March/April 1999

The apparent effects of early diagnosis and intervention (measured in terms of how screen-detected cases compare to cases detected by signs and symptoms) are always more favorable than the real effects (measured in terms of how a population that is screened compares to a population that is not). The comparison between screen-detected cases and others overestimates benefit because the former consists of cases that were diagnosed earlier, progress more slowly and that may never become clinically relevant. This comparison, therefore, is said to be biased. There are, in fact, three biases that inflate the survival of screen-detected cases:

1. Lead time bias

Overestimation of survival duration among screen-detected cases (relative to those detected by signs and symptoms) when survival is measured from diagnosis. In the figure below (representing one patient) there is a 10 year survival at the point of the clinical diagnosis (old), but a 15 year survival from the early diagnosis (new). This is simply a reflection of earlier diagnosis, however, as the overall survival time of the patient is unchanged.

Schematic Representation

2. Length bias

Overestimation of survival duration among screen-detected cases due to the relative excess of slowly progressing cases. These are disproportionally identified by screening because the probability of detection is directly proportional to the length of time during which they are detectable (thus inversely proportional to the rate of progression). In the following figure (representing 12 patients) 2 of 6 rapidly progressive cases are detected, while 4 of 6 slowly progressive case are detected.

Schematic Representation

3. Overdiagnosis bias

Overestimation of survival duration among screen-detected cases due to the inclusion of pseudodisease - subclinical disease that would not become overt before the patient dies of other causes. Some researchers further divide pseudodisease into two categories: one in which the disease does not progress (Type I), the other in which the disease does progress -- but so slowly that it never becomes clinically evident to the patient (Type II). Inclusion of either type as being a "case" of disease improves apparent outcomes of screen-detected cases.