Screening: the information individuals need to support their decision: per protocol analysis is better than intention-to-treat analysis at quantifying potential benefits and harms of screening

Abstract

Doc number: 28

Abstract

Background: Providing individuals with the information necessary to make informed decisions is now considered an ethical standard for health systems and general practitioners.

Discussion: Results from intention-to-treat analysis have thus far been used to illustrate screening benefits and harms, but intention-to-treat analysis in most screening trials compares no intervention to invitation to screening. Therefore, the intervention arm includes everyone who was invited, regardless of actual participation. These results may be misleading for individual decision-making. We propose to use a per protocol analysis that includes all subjects who presented to screening and compares them to those in control arm, adjusting for self-selection bias. Such an analysis can give more accurate and useful information for individual decision-making.

Summary: Correct information for individual decision to participate in screening or not should consider the efficacy, benefits, and harms observed for subjects who actually participated at least once in screening compared to the control arm, adjusting for self-selection bias. Thus, per protocol analysis, even a very conservative one, should be used, not a full intention-to-treat analysis.