Subject attrition from randomized Schizophrenia trials is a significant problem and has been found in meta-analysis to be as high as 76% (Robinowitz et al., 2005). High drop-out rates introduce selection bias, decrease the validity of intent to treat (ITT) analysis, can lead to a decrease in statistical power, and can ultimately cloud signal detection. The problem of drop-out is typically not addressed until the statistical analysis stage of the study which can decrease the validity of the results, increasing the likelihood of failed and negative trials.

Explore related content and topics