Routinely collected general practice data aids identification of people with hyperglycaemia and metabolic syndrome

Abstract

To assess the performance of a risk score comprising data routinely available in general practice records (age, gender, body mass index, family history of diabetes, smoking habits and prescribed anti-hypertensive drugs or steroids) in detecting diabetes, impaired glucose tolerance and metabolic syndrome. In a population-based, cross-sectional study in a semi-rural general practice in Jutland, Denmark, Cambridge Risk Scores were calculated for 1355 patients without known diabetes (69% response rate) who completed questionnaires and underwent anthropometric measurement and an oral glucose tolerance test. Prevalences of diabetes, impaired glucose tolerance and metabolic syndrome were 2.29% (95% CI: 1.56-3.23), 6.64% (95% CI: 5.38-8.10) and 13.4% (95% CI: 11.5-15.2), respectively. Area under the ROC curve for the risk score and diabetes was 83.8% (75.9-91.7) and for metabolic syndrome [European Group for the Study of Insulin Resistance (EGIR)] was 78.1% (74.6-81.6). Twenty per cent of the population had a risk score above 0.246; at this threshold the sensitivity to detect diabetes was 71.0% (53.4-83.9), the specificity 81.2% (79.0-83.2), positive predictive value 8.1% (6.6-10.0) and likelihood ratio 3.77 (2.94-4.85). For metabolic syndrome (EGIR) corresponding values for sensitivity were 50.3% (43.1-57.5), specificity 84.7% (82.5-85.6), positive predictive value 33.6% (28.2-39.4), and likelihood ratio 3.28 (2.69-4.00). Undiagnosed hyperglycaemia and metabolic syndrome are common. The Cambridge Risk Score is a practical first step in a screening procedure to identify individuals with these disorders who might benefit from diagnostic testing or to direct preventive interventions.