Improving the reliability of the Reliable Change Index

Le résumé fourni par la source

The reliable change index (RCI) is a tool for evaluating change at the individual level. It compliments standard group-level change estimates and is central to evaluating clinical significance for interventions. In principle, the RCI provides common criteria for evaluating individual change. However, current practices use sample-specific estimates to create these criteria. Because sample estimates are subject to sampling error, these criteria are also subject to sampling error and therefore differ across studies. We illustrate how current practices can lead to differing criteria for reliable change and use simulations to identify the impact of sampling error on the RCI. Excessive error in the RCI began for the average sample when N < 30, and samples only comfortably avoided the risk of excessive error when N > 100. Finally, small errors in estimating a measure’s reliability sometimes had profound effects on the RCI. Recommendations on use of the RCI are provided.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.