Is Clinical Relevance Sometimes Lost in Summative Scores?

Abstract

It is possible to have a reliable and valid measure of a clinical phenomenon but to score the measure in a way that inaccurately represents the clinical meaning of the measured phenomenon. The purpose of this article is to demonstrate the risk to research and clinical assessments caused by a discrepancy in the clinical meaning of scores obtained by total scale scoring and those obtained by individual item scoring. The clinical phenomenon that is used in this demonstration is symptom distress as measured by the Symptom Distress Scale (SDS). In this example, the discrepancy in meaning between total scale scores and the individual item scores of the SDS has potential clinical implications as more than 30% of participating adolescent patients at each of four time points spanning a 6-month time period would have been inaccurately assessed when the assessment was based on the total scale score.