Cover: Evaluation of the Patient-Reported Outcomes Information System (PROMIS®) Spanish-language Physical Functioning Items

Evaluation of the Patient-Reported Outcomes Information System (PROMIS®) Spanish-language Physical Functioning Items

Published in: Quality of Life Research, v. 22, no. 7, Sep. 2013, p. 1819-1830

Posted on Sep 1, 2013

by Sylvia H. Paz, Karen Spritzer, Leo S. Morales, Ron D. Hays

PURPOSE: To evaluate the equivalence of the PROMIS® physical functioning item bank by language of administration (English versus Spanish). METHODS: The PROMIS® wave 1 English-language physical functioning bank consists of 124 items, and 114 of these were translated into Spanish. ANALYSIS: Item frequencies, means and standard deviations, item-scale correlations, and internal consistency reliability were calculated. The IRT assumption of unidimensionality was evaluated by fitting a single-factor confirmatory factor analytic model. IRT threshold and discrimination parameters were estimated using Samejima's Graded Response Model. DIF by language of administration was evaluated. RESULTS: Item means ranged from 2.53 (SD = 1.36) to 4.62 (SD = 0.82). Coefficient alpha was 0.99, and item-rest correlations ranged from 0.41 to 0.89. A one-factor model fits the data well (CFI = 0.971, TLI = 0.970, and RMSEA = 0.052). The slope parameters ranged from 0.45 ("Are you able to run 10 miles?") to 4.50 ("Are you able to put on a shirt or blouse?"). The threshold parameters ranged from −1.92 ("How much do physical health problems now limit your usual physical activities (such as walking or climbing stairs)?") to 6.06 ("Are you able to run 10 miles?"). Fifty of the 114 items were flagged for DIF based on an R 2 of 0.02 or above criterion. The expected total score was higher for Spanish- than English-language respondents. CONCLUSIONS: English- and Spanish-speaking subjects with the same level of underlying physical function responded differently to 50 of 114 items. This study has important implications in the study of physical functioning among diverse populations.

This report is part of the RAND external publication series. Many RAND studies are published in peer-reviewed scholarly journals, as chapters in commercial books, or as documents published by other organizations.

RAND is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.