Diagnostic accuracy of two confirmatory tests for diabetes mellitus in adult Ugandans with recently diagnosed tuberculosis

Abstract

Objective: The optimal confirmatory tests for diabetes mellitus (DM) in patients with tuberculosis (TB) vary across populations. This study aimed to evaluate the performance of two confirmatory tests for DM against the oral glucose tolerance test (OGTT) as the reference test in adult Ugandans with recently diagnosed TB. Methods: A total of 232 adult participants receiving TB treatment underwent initial screening for DM with random blood glucose (RBG) measurement. Participants with a RBG level ⩾6.1 mmol/l received additional screening with fasting blood glucose (FBG), laboratory-measured glycated haemoglobin (HbA1c) and an OGTT. Using the latter as the gold standard and reference test, we evaluated the diagnostic accuracy of laboratory-measured HbA1c and FBG. Results: Of the 232 participants initially screened for DM using RBG measurement, 117 participants (50.4%) had RBG level ⩾6.1 mmol/l and were scheduled to return for additional blood glucose testing. Of these, 75 (64.1%) participants returned for FBG and HbA1c measurements. A diagnosis of DM was made in 32 participants, corresponding to a prevalence of 13.8% [95% CI 9.9–18.9]. The areas under the curve (AUC) for FBG and laboratory-measured HbA1c were 0.69 [95% CI 0.47–0.90] and 0.65 [95% CI 0.43–0.87], respectively. The sensitivity and specificity of a FBG level of ⩾7 mmol/l were 57.1% [95% CI 18.4–90.1] and 74.6% [95% CI 62.5–84.5], respectively, whereas the sensitivity and specificity for laboratory-measured HbA1c of ⩾6.5 mmol/l (48 mmol/mol) were 14.3% [95% CI 0.40–57.9] and 95.3% (86.9–99.0%), respectively. Conclusion: FBG may be better than laboratory-measured HbA1c in confirming DM in adult Ugandans with recently diagnosed TB. However, because of the small study sample size, larger studies evaluating the diagnostic utility of these diabetes screening tests in adult Ugandans with TB are needed to confirm these findings.