Abstract:
An epidemiological study was carried out to determine the prevalence of NIDDM and to
develop a screening tool to detect NIDDM in the community setting using a cross-sectional
descriptive study design. The study area was Kalutara district in Sri Lanka. A stratified
cluster samples of 3000 between the ages of 30 to 64 years was enrolled for the prevalence
study. The participation rate was 95.5 per cent. Important risk factors found were age more
than 40 years, dryness of the mouth, constipation, presence of a family history of diabetes
and nocturia. Screening tool was validated with the FPG value and the ROC analysis
showed that the screening tool carries 53.3 per cent sensitivity, 60.8 per cent specificity and
positive predictive value of 6 per cent. It is concluded that a questionnaire based on
symptoms and risk factors is not suitable to screen for diabetes in the field setting
