A screening tool to detect patients at increased risk of developing diabetes mellitus type 2
- Author:
Ramiro Fredjackson A.
;
Pineda Jr. Alejandro V.
- Publication Type:Journal Article
- MeSH:
Human;
Male;
Female;
DIABETES MELLITUS, TYPE 2
- From:
The Filipino Family Physician
2006;44(4):174-186
- CountryPhilippines
- Language:English
-
Abstract:
Objective: To modify a validated questionnaire to prospectively identify individuals at increased risk for undiagnosed diabetes.
Research Design and Methods: Each subject was asked to answer the ADA Diabetes Risk Test (ADADRT). The total score was tabulated to know if the patient was at increased risk or not for developing diabetes. All subjects submitted venous blood samples for a fasting blood sugar (FBS) determination. The ADADRT total scores were compared with the FBS values using chi square (chi2) tests and likelihood ratios (LR). The ADADRT of the study population was modified using the new Asian BMI. Scores were recomputed and compared with FBS using the same statistical measure. The ADADRT and the Modified Diabetes Risk Test (MDRT) were compared to determine if they were statistically different using Fisher's exact test and which one would have a better sensitivity, specificity and positive predictive value. The study population in the First Phase of the study included patients in a primary care clinic in a tertiary hospital, the Second Phase was the validation in a rural community and the Third Phase was the actual implementation of the validated screening tool among patients seen in a primary care clinic in a community set-up.
Results: In a representative sample of USTH patients (Phase 1), the sensitivity of the ADADRT was 41 percent, the specificity was 89 percent, x2 was at 10.59 which was significant at P> 0.05; likelihood ratio for a positive test LR (+) computed was 3.72; Likelihood ratio for negative test LR (-) computed was 0.66. In the same representative sample, the sensitivity of the MDRT was 81 percent, the specificity was 92 percent, x2 was 32.2 which was significant at P>0.05; LR (+) was 10.125 and the LR (-) was 0.21. Computed Fisher's exact test was 0.387, which was significant at P>0.05. In Phase II, ADADRT sensitivity was 64 percent, specificity was 83 percent LR (+) 3.11 and LR (-) 0.19 whereas the MDRT sensitivity was 86 percent, specificity was 72 percent, LR (+) 3.73 and LR (-) 0.43 and f-test 0.0063. In Phase III, the MDRT identified 18 as high risk where only 15 had elevated FBS.
Conclusion: The modified ADA Risk Test using the new Asian BMI performed significantly better than the existing ADADRT and should serve as a simple, noninvasive and potentially cost-effective add-on screening tool for detecting those at increased risk for diabetes mellitus type 2 in the local setting.
