Rasch analysis of the carpal tunnel syndrome instrument

Background

The carpal tunnel syndrome instrument (CTSI) is the most widely used patient-reported outcome measure (PROM) in carpal tunnel syndrome (CTS). However, CTSI is an ordinal-type questionnaire and might have caused misinterpretations of the PROM between surgical outcomes of CTS (Camitz and extra/open carpal tunnel release).

Purpose

This study aims to convert the CTSI to an interval scale using Rasch analysis (RA) and evaluate the outcome differences between the original and transformed scales.

Study Design

Prospective control study.

Methods

Four hundred twenty-four patients with 567 CTSs had been interviewed for CTSI perioperatively and treated with either endoscopic/open carpal tunnel release or Camitz tendon transfer. Each CTSI was analyzed for dimensionality, fit statistics, and a transformation of the ordinal-to-interval scale by RA. We compared the two groups perioperative scores of three CTSI versions (original 11-item, modified 8-item, and transformed 8-item).

Results

Based on the RA, the original CTSI was not unidimensional. We identified two dimensions. After removing misfit items, the perioperative course of each score by three versions of each dimension was compared (Repeated 2-factor analysis of variance). The transformed interval scales of CTSI provided different assessments of score changes from the ordinal scale of CTSI analyses.

Conclusions

Original CTSI consisted of ordinal scale items that yielded different conclusions than scores converted to interval scale by Rasch analysis. CTSI should convert into an interval scale after reclassifying dimensionality by Latent Factor Analysis and removing misfit items