Direct and Indirect Utilities of Severe Versus Non-Severe Carpal Tunnel Syndrome Patients

Background:

Carpal tunnel syndrome (CTS) is the most common entrapment neuropathy. To understand how patients experience CTS, quality of life (QoL) can be measured as a patient-reported outcome. One way to capture QoL is through health utility, which represents health states on a scale of 0 (death) to 1 (perfect health). Utilities can be measured using direct methods, where patients’ preferences are mapped directly onto the utility scale, and indirect methods, where patients complete a questionnaire that is converted to population norms. Utilities can be integrated into decision analysis models and combined with cost data to perform economic analysis, which is especially important in the face of rising healthcare costs. To date, few utility methods have been used to evaluate CTS, with direct methods especially lacking.

Purpose:

Our primary objective is to improve the understanding of QoL in CTS patients, through comparing direct and indirect utilities of patients with non-severe (mild or moderate) versus severe CTS. Our secondary objective is to determine inter-instrumental agreement of the various utility measurement methods.

Methods:

We conducted a prospective cohort study and recruited patients who were referred to a tertiary care center for CTS evaluation. Adults with a likely diagnosis of CTS based on a score of “Classic” or “Probable” on Katz Hand Diagram were included. Patients were stratified into two groups using the Boston CTS Symptom Severity Scale/Functional Severity Scale (SSS/FSS), with severe defined as SSS>33 and/or FSS>24. Utilities were measured indirectly by Short Form Sixth Dimension (SF-6D) and EuroQol 5D Questionnaire (EQ-5D-5L), and directly by Chained Gamble (CG) and Visual Analogue Scale (VAS). The Wilcoxon sum rank test was used to compare severe versus non-severe groups, and Pearson correlation coefficient was used to determine inter-instrumental agreement.

Results:

Twenty-nine patients (53%) met inclusion/exclusion criteria. Mean age was 57.8 years (SD=13.1), with 22 (75.9%) female patients, and 11 patients (37.9%) with severe CTS. Utilities for the entire cohort was 0.73 (SD=0.09) for SF-6D, 0.75 (SD=0.15) for EQ-5D-5L, 0.94 (SD=0.16) for CG, and 0.74 (SD=0.17) for VAS. Significant difference was found between the severe and non-severe groups for EQ-5D-5L (p=0.01), with utility of 0.64 (S=0.18) for severe and 0.81 (SD=0.07) for non-severe patients. However, this difference was non-significant for SF-6D (p=0.23), CG (p=0.53), and VAS (p=0.67). There was significant inter-instrumental agreement between SF-3D and EQ-5D-5L (Pearson=0.60, p<0.01), SF-3D and VAS (Pearson=0.67, p<0.01), and EQ-5D-5L and VAS (Pearson=0.55, p<0.01). Correlation of CG with the other methods, SF-3D (Pearson=0.18, p=0.34), EQ-5D-5L (Pearson=0.08, p=0.69), and VAS (Pearson=0.08, p=0.70), was not significant.

Conclusions:

Utility in this CTS patient population varied from 0.73 to 0.94 depending on the instrument. Variation in utilities between severe versus non-severe patients was significant for EQ-5D-5L only, demonstrating potential limitations in using utility tools to stratify patients according to disease severity. Given that utilities are incorporated as key measures in decision analysis and economic evaluation, the inter-instrument inconsistency found in this study may lead to challenges in constructing valid and reliable models. Future research is needed to determine how to reconcile such inconsistency.