Twiss J, Whalley D, Doward L, Balp MM, Brass C, Lopez P, Loeffler J, Cai J, Cryer D, Sanyal AJ, Anstee QM. Psychometric evaluation of NASH-CHECK, a patient-reported outcome measure for Nonalcoholic Steatohepatitis. Poster to be given at the ISPOR 2019 European Conference; November 6, 2019. Copenhagen, Denmark. [abstract] Value Health. 2019 Dec; 22(S3).


OBJECTIVES: Nonalcoholic steatohepatitis (NASH) is the most advanced form of nonalcoholic fatty liver disease characterised by liver fat accumulation, inflammation, cell injury, and fibrosis. This study reports on the preliminary psychometric evaluation of NASH-CHECK, a new NASH-specific patient-reported outcome measure (PROM) comprising 31 items measuring symptoms and health-related quality of life (HRQOL).

METHODS: Data from a randomised, double-blind, placebo-controlled, phase 2 trial in adult NASH-patients were analysed. Item-level analyses (correlations, exploratory factor analysis [EFA], and Rasch analysis) were conducted to determine the preliminary dimensional structure of NASH-CHECK and identify potential items for removal.

RESULTS: The analysis sample included 104 patients (54.8% female; mean age=51.2 [standard deviation=12.7]; age range 19.0-79.0; NASH fibrosis grades 1-3; 62[59.2%] obese [BMI≥30]; type 2 diabetes=60[65.4%]). EFA supported a 4-item cognitive symptoms scale (focusing, thinking clearly, following conversation, forgetfulness), with other symptoms items as individual items. For HRQOL, EFA supported a distinction between activity limitations and emotions/lifestyle issues. EFA supported one- and two-factor models for both activity limitations items and emotions/lifestyle items (comparative-fit index ≥0.95, Tucker-Lewis index ≥0.95, root mean square error of approximation ≤0.06, standardised root mean-square residual ≤0.08). For activity limitations, an overall activity limitations summary score, and daily activities and ambulation domains were supported. For emotions/lifestyle, EFA provided evidence for an overall psychosocial summary score, and emotions and social domains. Rasch analysis showed preliminary support for EFA-identified scales. High interitem and item-total correlations (>0.80) suggested some potential redundancy, notably among items assessing aspects of fatigue and activity limitations. One item (food restriction) had low correlations with other items suggesting a lack of association.

CONCLUSIONS: The psychometric evaluation provides evidence for a preliminary dimensional structure for NASH-CHECK. The PROM properties will be further evaluated via confirmatory analyses using additional datasets to identify the final items for removal and to determine the final dimensional structure.

Share on: