Comparison of The World Health Organization (WHO) two-step strategy and OGTT for diabetes mellitus screening.

The new diagnostic criteria recommended by the American Diabetes Association (ADA) will only detect diabetic patients with fasting hyperglycemia, and leave patients with isolated post-challenge hyperglycemia (IPCH) and imparied glucose tolerance (IGT) unidentified. The WHO recommends that all those with abnormal fasting glucose should undergo the oral glucose tolerance test (OGTT) to exclude the diagnosis of diabetes (two-step strategy). This two-step strategy will leave out subjects with normal fasting glucose (<109 mg/dl). The aim of this study is to compare the WHO two-step strategy and the gold standard OGTT for all subjects. We re-analyzed the results of 907 high-risk patients who have been screened for diabetes mellitus and impaired glucose tolerance. All subjects were screened with an OGTT containing a 75-gram glucose load after fasting for 12 hours. The results were classified into three categories: the ADA criteria, the two-step strategy, and the OGTT. Using the ADA criteria, these 907 subjects can be classified has having normal fasting glucose (fasting plasma glucose - FPG < 109 mg/dl) in 715 subjects (78.9%), abnormal fasting glucose (FPG 110 - 125 mg/dl) in 107 subjects (11.8%), and diabetes mellitus (FPG > 126 mg/dl) in 85 subjects (9.4%). The WHO two-step strategy performed in 107 IFG subjects identified another 30 diabetic patients (FPG < 109 mg/dl and 2 hour post load > 200 mg/dl = IPCH) or 3.3%, and 49 patients with IGT, or 5.4% from all subjects. If the OGTT was performed on the 715 normal fasting glucose, it could identify another 40 diabetic patients or 4.4%, and another 178 IGT patients, or 19.6% of all subjects. This means that without OGTT to all subjects, 40 diabetic patients or 25.8% of all diabetic patients and 178 patients or 78.4% from all IGT subjects would have remained unidentified. From this study we can conclude that applying the WHO two-step strategy in subjects with IFG would fail to detect 25.8% of diabetic patients and 78.4% of IGT subjects. It is recommended that the old strategy of screening--the gold standard OGTT--should be used instead of the two-step strategy, at least in high-risk groups.