Aims: the main aims of this study were to investigate the feasibility of screening for noninsulin dependent diabetes mellitus (NIDDM) and impaired glucose tolerance (IGT) in general practice and to compare random blood glucose measurements with the two hour oral glucose tolerance test as screening methods.
Methods: one thousand, one hundred and eighty-four people aged 39-79 years who were registered with two general practitioners were invited to be screened using a random blood glucose test. Subjects who had a random test and were aged 39-69 years were subsequently invited to have a two hour oral glucose tolerance test (OGTT). Those subjects who had a raised random blood glucose level or a raised fasting and/or two hour plasma glucose level according to WHO criteria, were invited back for two additional OGTTs.
Results: the overall response rate for the random blood glucose testing was 67%. Forty-seven percent of those aged 50-69 years and 35% of those aged 39-49 years completed one OGTT. A number of people with high random blood glucose levels had normal OGTTs which suggests that the risk of a false positive result from the random test is high. In addition, a number of false negative results from the random test were identified. A total of 20 people were found to have IGT, although only three of these were identified from the random test which confirms the necessity of administering an OGTT for the identification of IGT. Only 44% of those identified with IGT from a single test remained in the IGT category after repeat testing. The prevalence of NIDDM and IGT for people aged 39-69 years was 4.4%.
Conclusions: the random glucose test did identify people with NIDDM however this test was found to be an insensitive technique for identifying IGT. A number of cases of NIDDM and IGT did not persist after repeat testing. The small number of new diabetics diagnosed from this screening study suggests that screening for diabetes in a predominantly European general practice is not cost effective in terms of the resources required.