A simple and objective method of estimating the longterm control of diabetes and derived from urine terms tests at routine clinic attendances has been investigated. The per cent of such urine tests showing 2 per cent glycosuria (glycosuria percentage) is shown to be significantly related to the mean of blood sugars determined at the same attendances. This relationship is unaffected by the number of clinic attendances, if these number five or more.

This content is only available via PDF.