Resumen de Adding questions about cardiovascular risk factors improve the ability of the ADA questionnaire to identify unknown diabetic patients in Spain
Fernando Álvarez Guisasola, Ignacio Conget Donlo, Josep Franch Nadal, M. Mata, José Javier Mediavilla Bravo, A. Sarria, J. Soler, E. Ramírez, C. García, L. Clerch
Introduction:
Currently, there are not specific questionnaires for Spanish population to identify people at risk of undiagnosed diabetes. When American Diabetes Association (ADA) test is validated in the Spanish population, the sensitivity and specificity values obtained are lower than those found in the USA.
Objectives:
To develop a screening tool based on the ADA questionnaire, to prospectively identify undiagnosed type 2 diabetes in Spanish ambulatory patients.
Methods:
Epidemiological, transversal, multicentre study, including 2,662 ambulatory patients of Primary Care centres, mean age (SD) 61.7 (10.2) years (53% women), needing a blood test and attending follow-up protocols for chronic pathologies or periodic screening programs. Classification tree construction was achieved through classical and Artificial Intelligence (AI) methods. The sensitivity, specificity, and the positive and negative predictive values were described and compared with the ADA questionnaire.
Results:
The final selected classification tree included the following variables: previous impaired fasting glucose or glucose intolerance; recent weight gain; parents, siblings or children with diabetes; smoking habit and pharmacologic treatment for lipid disorders (sensitivity: 80.7%; specificity: 70.9%; positive predictive value: 45.3%; negative predictive value: 92.5%). This tree showed a better Receiver Operating Characteristic curve than that of the ADA test (sensitivity: 84.3%; specificity: 20.9%).
Conclusions:
The inclusion of questions regarding lipid disorders, smoking habit and weight gain increase the specificity of the ADA test to identify undiagnosed type 2 diabetes in Spanish patients older than 45 years.
