|
Objectives:To develope the Type 2 Diabetes Screening Questionnaire for Rural Residents (DSQRR), and furtherly to provide scientific suggestions for intervention programs in rural areas.Methods:Sample villages were selected by multi-stage cluster random sampling method, and were randomly divided into training group and test group of DSQRR respectively. Permanent residents in training group aged 15 and over took a peripheral blood test by fast blood glucose meter, and were interviewed by self-developed questionnaire of type 2 diabetes risk factors, which included 30 variables as demographic data, general data, family history of type 2 diabetes, suspected diabetes symptoms, chronic disease history, lifestyle, diet, life stress and related physiological indices. DSQRR was developed by analysis results of training group. Then DSQRR was applied and to be evaluated in test group. Permanent residents in test group aged 15 and over took a peripheral blood test by the same way of residents in training group.Results:683 valid questionnaires were finally obtained in training group. Age, family history of diabetes, number of suspected clinical symptoms of diabetes and central obesity were selected as items of DSQRR finally, and the OR values were 1.480 (95% CI,1.050～2.085)、 2.800 (95%CI,1.026～7.646)、2.759 (95% CI,2.022～3.764)、2.873 (95% CI,1.156～7.145) respectively, corresponding risk score of type 2 diabetes was 1,3,3,3 respectively, whereby DSQRR was developed, and the total score ranged from 0 to 21.919 valid questionnaires were finally obtained in test group. DSQRR were applied in test group, total score varied from 0 to 21, quartiles were 3,4,6 respectively. Chi-square test results showed that DSQRR scores were significantly associated with prevalence of type 2 diabetes (P<0.001). A UC of DSQRR was0.748 (95% CI,0.684～0.811), the best cutpoint was≥7, then the Youden’s index was 0.387, and sensitivity, specificity, PPV, NPV was 58.3% ,80.4%,20.2%,95.8% respectively.Conclusions:DSQRR performs well for type 2 diabetes screening in rural areas, which includes data of age, family history of diabetes, number of suspected clinical symptoms of diabetes and central obesity, and corresponding risk score of type 2 diabetes is 1,3,3,3 respectively, and the total score ranges from 0 to 21.
|