a1 Mailman School of Public Health, Columbia University, New York, NY, USA
a2 New York State Psychiatric Institute, New York, NY, USA
Background Studies comparing cases with controls to uncover the causes of psychiatric disorders are common in biological research. The validity of these studies depends upon adherence to the methodological principles underlying the case-control design. However, these principles are often violated. One common practice that violates these principles is the use of well controls. In this paper we describe the bias that it can cause and discuss why the use of well controls leads to invalidity in case-control studies.
Method Using hypothetical numerical examples we illustrate the consequences of using well controls.
Results The results illustrate that the use of well controls can cause substantial bias. In no instance does the use of well controls improve validity.
Conclusions We conclude that the use of well controls is an unhealthy practice in psychiatric research.
(Received January 04 2010)
(Revised June 09 2010)
(Accepted July 17 2010)
(Online publication September 01 2010)