The Re-Engineered Discharge for Diabetes Computer Adaptive Test (REDD-CAT): New patient-reported outcome measures that evaluate social determinants of health for persons with Type 2 Diabetes Mellitus

Abstract

<h3>Research Objectives</h3> To detail the development of patient-reported outcome measures (PROs) that capture social determinants of health (SDOH) among individuals with Type 2 diabetes mellitus (T2DM). <h3>Design</h3> Observational survey study. <h3>Setting</h3> Academic medical center. <h3>Participants</h3> 225 individuals with T2DM. <h3>Interventions</h3> N/A. <h3>Main Outcome Measures</h3> Participants completed five item pools: Healthcare Access (54 items), Medication Adherence (58 items), Illness Burden (47 items), Health-seeking Behavior (56 items), Housing Stability (22 items). Classical test and item response theory analyses were used to develop item banks, computer adaptive tests (CATs), and short forms (SFs). <h3>Results</h3> Eight new measures were established: Healthcare Access (44 items); Medication Adherence (27 items); Illness Burden (27 items); Housing Stability (20 items); and four Health-Seeking Behavior instruments: Primary Care Provider-Specific (6-item SF), General Beliefs (13 items), Family/Friends-Specific (5-item SF), and Internet-Specific (4-item SF). Exploratory and confirmatory factor analysis (EFA/CFA) supported unidimensionality for all measures. For EFAs, eigenvalue 1-to-2 ratio was >4 and proportion of variance accounted for by eigenvalue 1 was >0.40. Model fit criteria for CFAs (e.g., CFI ≥0.90, TLI ≥0.90, RMSEA < 0.15) were also met. Items had no differential item functioning (DIF) by gender, age, education, or socioeconomic status (i.e., no significant (p < 0.01) group-specific item parameter differences; ≤2% of DIF-corrected/uncorrected score differences exceeded uncorrected score standard errors). A constrained graded response model estimated item parameters, which enabled CAT simulations and SF construction informed by psychometric and clinical considerations. <h3>Conclusions</h3> These PROs are part of the REDD-CAT comprehensive measurement system for SDOH and are candidates for inclusion as standard outcomes measures for use with discharge planning to mitigate readmission risk in people with T2DM. <h3>Author(s) Disclosures</h3> The authors have no conflicts of interest.