I wish to report the results of a screening project for type 2 diabetes performed over a 10-month period in our practice of 9700 patients. Patients presenting to the practice were screened opportunistically for type 2 diabetes if they showed two or more of the following risk factors, using a protocol based on an Australian screening study: aged >40 years, obesity, family history of type 2 diabetes, hypertension, atherosclerosis (coronary heart disease, cerebrovascular disease, or peripheral vascular disease), previous abnormality of glucose tolerance including gestational diabetes, ethnic predisposition, or polycystic ovarian syndrome.1
Patients were screened by random or fasting venous glucose, and those with blood glucose of ≥5.6mmol/l were intended to have an oral glucose tolerance test. Of the 271 patients, 130 (48%) had an initial blood sugar of ≤5.5mmol/l. Of the remainder, 91 patients (33.6%) proceeded to have oral glucose tolerance tests; 44 (16.2%) had normal results, 27 (10.0%) had impaired fasting glycaemia or impaired glucose tolerance, and 20 (7.4%) had results within the diabetic range. A further three patients were diagnosed as diabetic on high random blood sugars alone.
Some 47 (17.3%) patients did not proceed to have glucose tolerance tests; in 24 cases this was due to administrative failure. Three patients declined to have the test due to the inconvenience, nine patients were tested for diabetes using an alternative method, and in 11 cases the GP decided not to proceed with further investigation for clinical reasons.
Overall 8.5% of patients screened had type 2 diabetes and a further 9.6% had impaired glucose tolerance, or impaired fasting glycaemia. The clinical workload and demand for appointments was manageable: on average only nine oral glucose tolerance tests were performed every month, and many of the random blood sugars were taken simultaneously with other recommended annual screening blood samples.
Further modification of the protocol is suggested, so that patients in some clinical situations are tested by fasting glucose and haemoglobin A1c(HbA1c) measurement rather than by oral glucose tolerance test.2 Allowing some flexibility in interpretation of random blood sugar results according to the prandial state of the patient may also restrict unnecessary glucose tolerance tests.
This project demonstrates a feasible method for screening for type 2 diabetes and its precursors.
- © British Journal of General Practice, 2004.
