Interpreting Patient-reported Outcome Scores for Clinical Research and Practice: Definition, Determination, and Application of Cutpoints

Abstract

Objectives: Cutpoints are specific numeric values used to create discrete categories for patient-reported outcome (PRO) items or scales. Cutpoints are widely used in both clinical research and practice. This article offers a definition for cutpoints, describes strategies for determining actionable cutpoints, and discusses considerations related to interpreting cutpoints in clinical applications.

Methods: We clarify the definition of cutpoints for PRO measures and summarize the major statistical approaches for identifying cutpoints, including multivariate analysis of variance and receiver operating characteristic and regression modeling.

Discussion: We review issues related to cutpoint determination and interpretation that should be considered when integrating PROs into clinical research and practice, including the selection of anchors, variability of cutpoints, and clinical burden that may be generated when a cutpoint is used as a threshold for further clinical action.

Key points: Cutpoints are widely used to categorize PRO responses in both clinical research and practice. Cutpoints can be derived for PRO measures regardless of the response scale used; however, the mild, moderate, and severe categories generated from numeric cutpoints are distinct from the mild, moderate, and severe categories found in some PRO measures that use verbal descriptors as response options. Bootstrap analysis is recommended to quantify the variability of cutpoints. The application of cutpoints is limited by how well the anchors are chosen and how cutpoints developed using group-level data are applied at the individual level.