Abstract
The oral glucose tolerance test (OGTT) is poorly reproducible, time consuming and expensive. Manley et al. have developed a decision algorithm combining HbA1c and fasting plasma glucose (FPG) to identify the need for an OGTT using 500 UK patients who were referred for OGTT with impaired fasting glucose measurements. According to the proposed algorithm:
If FPG ≥7.0 mmol/L: suggestive of diabetes, confirm with second glucose measurement; If FPG <7.0 mmol/L and HbA1c <6.0%: diabetes not present; If FPG <7.0 mmol/L and HbA1c ≥ 6.0%: refer for OGTT.
On analysis of the 500 patients used to derive the algorithm, HbA1c ≥6.0% had poor sensitivity (83%) and specificity (46%) in isolation for the diagnosis of diabetes. However, combining the HbA1c and FPG to identify the need for an OGTT gave improved diagnostic sensitivity and specificity.
The algorithm was then validated in a further 500 UK patients and in 1175 unselected Australian patients who had been referred for an OGTT. The algorithm identified 234/241 patients with diabetes in the validation cohort with a sensitivity of 97% and a specificity of 100%. Of the seven patients not diagnosed, six had impaired fasting glycaemia and would therefore have continued to be monitored. In the Australian validation cohort, of whom 26% had impaired fasting glucose measurements, the algorithm identified 382/412 cases of diabetes with a sensitivity of 93% and a specificity of 100%. Isolated impaired glucose tolerance was identified in 5% of the UK validation cohort, one-third of whom had an HbA1c <6.0% and would not therefore have been referred using the algorithm.
The advantages of this algorithm to both the laboratory and the patient are clear. While HbA1c in isolation is unsuitable for the diagnosis of diabetes, combining it with fasting plasma glucose increases sensitivity to acceptable levels for a diagnostic test. The authors suggest that it could reduce the number of OGTTs required by one-third.