Rasch Analysis In The Development of A Specific Health-Related Quality of Life Questionnaire for Hypoglycemic Events In Diabetic Patients: The Qolhypo Questionnaire

To develop a specific questionnaire to assess the impact of hypoglycemia on the health-related quality of life (HRQoL) of patients with type 2 diabetes mellitus (T2DM). A first version of the QolHypo questionnaire was obtained from a literature review and focus groups with patients and physicians. QolHypo consisted of 37 specific items about the impact of hypoglycemia on HRQoL with 5 response categories. The final version was obtained by item reduction methods: floor and ceiling effects and Rasch analysis. Rasch methodology allowed obtaining a questionnaire which responses will be a trade-off between respondent’s abilities and item difficulty. Rasch assesses the appropriateness of item responses with probability curves that must appear ordered as the response categories. Overlapped categories suggested excess of response options and were reduced and items with disordered categories were eliminated. The goodness of fit of items in the Rasch model was calculated using the normalized Infit and Outfit statistics whose interval of good fit is (-2, 2). Items with statistics falling outside that interval were removed. Analyses were performed on a sample of 140 patients. 55.0% of patients were men, aged 63.0 (mean)(SD:9.6) years. Patients had 14.6 (mean)(SD:6.9) years of T2DM duration and 5.5 (mean)(SD:11.1) hypoglycemias in the previous 6 months. 11 items were deleted because of floor and ceiling effect. Most of probability curves showed overlapping response categories; therefore two pairs of categories were joined. Probability curves were represented again and 11 items that showed disordered categories were eliminated. Goodness of fit of the 15 remaining items showed only 2 items not adjusting to the model that were deleted. The final version of QolHypo consisted of 13 items with 3 response categories. Rasch analysis has led to obtain a reduced questionnaire for measuring how hypoglycemia affects the HRQoL in T2DM patients.