Dysphagia-Specific Instrument Based on Item Response Theory and International Classification of Functioning, Disability and Health.

Abstract

This study aimed to identify functional challenges faced by individuals with non-esophageal dysphagia and to offer a tool for quantitatively evaluating the person abilities within the framework of the International Classification of Functioning, Disability and Health (ICF). Additionally, this study attempted to differentiate the personal abilities of individuals with dysphagia and hierarchize item difficulties using the ICF-based Item Response Theory (IRT) modeling approach. This cross-sectional study enrolled a cohort of 150 patients with dysphagia (105 male and 45 female) from a tertiary hospital in China. Participants were assigned to evaluate the 114-item ICF dysphagia questionnaire. To further assess their swallowing capabilities, eating patterns, quality of life, and nutritional status, participants underwent a battery of five additional scales. The ICF qualifiers underwent data shaping including dichotomization and missing value imputation, Mokken scale analysis (MSA) for checking unidimensionality, local independence, monotonicity, and invariant item ordering (IIO), and parametric IRT modeling for identifying an optimal model from the 1-parametric logistic model (1PLM), 2PLM, 3PLM, and 4PLM. Finally, we tested the robustness of the optimal model via Monte Carlo simulation and illustrated the usefulness of the model by its person-item map. The 1PLM emerged as the optimal model with a total of 50 ICF items (12 'd-Activities and Participation', 33 'b-Body Functions', and 5 'e-environmental' items). The final scale presented strong reliability with Cronbach's Alpha = 0.967. Furthermore, the scale showed good validity with a significant positive correlation (p < 0.001, ${\hat{r}}_{\mathrm{Winsorized}}$ = 0.60) between model-estimated person abilities and swallowing-quality of life (SWAL-QoL) scores. The findings also demonstrated measurement equivalence of the final model for individuals with different genders or across various age groups. The utilization of the person-item map can effectively compare the difficulty levels of items with the abilities of patients, thereby facilitating the delivery of tailored care and precise rehabilitation strategies that match the individual competency of those suffering from dysphagia. This study developed a parsimonious dysphagia-specific ICF outcomes tool derived from the IRT, named iSWAL-Performance Scale. The findings complement quantitative information on the psychometric characteristics of this 50-item scale.