No Differential Item Functioning of Patient-Reported Outcomes Measurement Information System Global Health Items by Neighborhood Socioeconomic Deprivation

Yadi Li; Irene L Katzan; Nicolas R Thompson; Brittany Lapin

doi:10.1016/j.jval.2025.12.012

No Differential Item Functioning of Patient-Reported Outcomes Measurement Information System Global Health Items by Neighborhood Socioeconomic Deprivation

Value Health. 2026 Jan 13:S1098-3015(26)00001-X. doi: 10.1016/j.jval.2025.12.012. Online ahead of print.

Authors

Yadi Li¹, Irene L Katzan², Nicolas R Thompson¹, Brittany Lapin³

Affiliations

¹ Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, OH, USA; Center for Outcomes Research and Evaluation, Neurological Institute, Cleveland Clinic, Cleveland, OH, USA.
² Center for Outcomes Research and Evaluation, Neurological Institute, Cleveland Clinic, Cleveland, OH, USA; Cerebrovascular Center, Neurological Institute, Cleveland Clinic, Cleveland, OH, USA.
³ Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, OH, USA; Center for Outcomes Research and Evaluation, Neurological Institute, Cleveland Clinic, Cleveland, OH, USA. Electronic address: lapinB@ccf.org.

PMID: 41539559
DOI: 10.1016/j.jval.2025.12.012

Abstract

Objectives: We aimed to evaluate differential item functioning (DIF) of Patient-Reported Outcomes Measurement Information System (PROMIS) Global Health (GH) items by neighborhood socioeconomic deprivation. In addition, we also sought to examine DIF by age, sex, race, marital status, insurance, and comorbidities.

Methods: This retrospective study included patients who had a primary care appointment within Cleveland Clinic health system, lived within the county, and completed PROMIS-GH in 2022. DIF analyses were conducted within the logistic regression framework across groups based on area deprivation index (ADI), age, sex, race, marital status, insurance, and comorbidities. Latent trait was estimated using the graded response model and DIF magnitude was evaluated using McFadden's R square.

Results: A total of 263 150 adult patients had a primary care appointment in 2022, and 158 685 (60.3%) completed PROMIS-GH (mean age 51.9 ± 17.6 years, 60.7% female, 73.5% White race). Patients in the higher disadvantage group were, on average, 3.6 T-score points worse in global mental and physical health compared with the lower disadvantage group. The largest item-level differences were in physical health, quality of life, and social satisfaction. DIF was not detected for any item across ADI, age, sex, race, marital status, insurance, or comorbidities.

Conclusions: No DIF of PROMIS-GH items were detected, supporting the comparison of PROMIS-GH across levels of socioeconomic deprivation, age, sex, race, marital status, insurance, and comorbidities. Patients living in more disadvantaged neighborhoods have significantly worse health-related quality of life compared with those who live in less disadvantaged neighborhoods. Future research should investigate the factors contributing to health inequities by ADI.

Keywords: PROMIS Global Health; area deprivation index; differential item functioning.