length bias; length-time bias

Introduction

One of 2 major biases that complicate observational studies including randomized-controlled trials on screening tests.

Length bias occurs as the result of the heterogeniety of a single disease. A disease with more aggressive course is more likely to be diagnosed when symptoms occur rather than with a screening program. A disease with a less aggressive course is more likely to be detected with a screening program. Early detection is obviously associated with improved survival, but this may be due to the greater likelihood of detecting less aggressive disease in screening programs.

More general terms

• bias