RESUMO
OBJECTIVE: To evaluate a screening questionnaire to identify individuals with undiagnosed diabetes. DESIGN AND METHODS: We conducted a community survey to detect people at increased risk for diabetes using a questionnaire proposed by Herman et al. (1995), that incorporates major risk factors as age, obesity, family history of diabetes, sedentary lifestyle and personal history of delivering a macrosomic infant. Blood glucose test was made by means of reagent strip and a reflectance meter system. We used the ADA recommendations and cut-points for screening programs, adjusted for fasting and random blood glucose. RESULTS: We included 360 participants older than 20 years of age. A total of 200 subjects (55.5%) were at risk for diabetes according to the questionnaire, of whom 31 (15.5%) had an abnormal glucose test compared to the 4.4% of the low-risk group (p < 0.001). The 1995 Herman et al. Questionnaire had sensitivity of 81.6%, specificity of 47.5, positive predictive value of 15.5 and negative predictive value of 95.6%. The high-risk group was older (44.9 vs. 34.6 y, p < 0.001) and heavier (30.5 vs. 24.4 kg/m2, p < 0.001) than the low-risk group. There were 38/360 (10.5%) abnormal glucose readings, of which 31 (81.5%) had a positive questionnaire (p < 0.01). The mean fasting glucose in the high-risk group was higher (90.6 vs. 84.2 mg/dL, p = 0.015) than in those with low risk by questionnaire, as well as for random blood glucose (116.1 vs. 100 mg/dL, p < 0.01). CONCLUSIONS: The questionnaire proposed by Herman et al. combined with capillary blood glucose testing performance good in mexican population to identify people at high risk for undiagnosed diabetes, and improved the detection rate.