Reliability and factor analysis of a questionnaire about diabetic patients’ access to consultations with endocrinologists

Abstract

Introduction: Access to health services has been addressed in discussions about health care systems; the literature describes several dimensions of this construct. Objective: Evaluating the reliability and configural structure of a questionnaire about the access of diabetic patients to consultations with endocrinologists at SUS, based on availability, financial viability, and acceptability dimensions. Methods: A questionnaire was developed to assess the access of diabetic patients to consultations with endocrinologists at SUS, based on availability, financial viability, and acceptability dimensions. Telephone interviews were conducted with 472 diabetic patients; interviews were scheduled by the Regulation System at SUS, in Espírito Santo State. Test-retest reproducibility analyses, and the assessment of internal consistency by Cronbach’s alpha, were performed to determine the reliability of the questionnaire. Configural structure and theoretical-empirical congruence were determined based on principal component analysis with orthogonal varimax rotation. Results: Most responses have shown substantial and almost perfect agreement in the three assessed dimensions. One item was excluded from the availability dimension because it presented moderate stability (k = 0.41, 95% CI = 0.08 - 0.73). In the end, the initial instrument, which comprised 25 questions, was reduced to 15 questions covering the following dimensions: availability (Cronbach’s α 0.699), financial viability (Cronbach’s 0.526), and acceptability (Cronbach’s 0.568). Resulting components have shown high factor loads (> 0.3) and acceptable reliability (> 0.5) for most items. Conclusion: Reliability and configural structure have indicated the feasibility of measuring the measurement properties of the questionnaire and suggested its applicability in populations similar to the herein investigated one.