To estimate the prevalence of known diabetes in a Swedish county (Skaraborg) and to evaluate capture-recapture methods for validation of the diabetes prevalence.
Four sources were used to find known diabetic patients: 1) the Skaraborg Diabetes Registry (SDR), a case-finding inventory 2) an administrative hospital registry 3) a registry for the retinopathy screening program, and 4) a pharmacy prescription inventory. The capture-recapture method was applied to various combinations of sources to determine the prevalence of diabetes.
The completeness of the SDR was 88.4 ± 1.3%, indicating a diabetes prevalence of 3.20 ± 0.08%. When the SDR was not used as one of the sources, the estimated prevalence varied between 2.2 and 4.5%.
A case-finding registry is important in reliably estimating the prevalence of diabetes.
