Aportaciones al análisis de datos de calidad de vida relacionada con la salud, desde una perspectiva multivariante

Résumé

The aim of this thesis is to provide a critical review of the questionnaire QUALEFFO (Badía, 1999). Traditionally, it has been used in a hospital environment to evaluate the quality of life in osteoporotic patients; its validation in Primary Health Care using Item Response Theory (IRT) is studied. Traditionally it has been accepted that Osteoporosis affects mainly women, nevertheless men are progressively more affected. The second objective of this work is studying if the functioning of the questionnaire is the same for men and women; for that, DFIT (Differential Functioning of Items and Test) as described in Raju, van der Linden, & Fleer's (1995), and Ordinal Logistic Regression have been used. Some of the most relevant results are: The B8 item (Sleep disturbance due to pain) of the dimension �Activities of Daily Living�, C13 (Can you lift a heavy object?) of the dimension �Jobs around the House�, D19 (Changes in the figure due to osteoporosis) of the dimension �Mobility�, G34 and G35 (Afraid of fractures) of the dimension �Mental Function�, do not add any information to the questionnaire. They could therefore be removed. The C13 item (Lift a heavy object) of the dimension �Jobs around the House�, G29 (Feeling downhearted), G30 (Are you hopeful about your future?), G33, G34 and G35 (Afraid of fractures) of the dimension �Mental Function�, have differential item functioning. Taking into account the main results we propose a reduced version of QUALEFFO, with 23 rather than 35 items, and with less response categories in 11 of the items. The reduced version has high factorial validity and reliability, and provides important time and expenses savings for the physicians and the patients.