Diagnostic accuracy of ADA 2020 criteria for undiagnosed diabetes in a Peruvian population

Leonardo Albitres-Flores, Antonio Bernabe-Ortiz

Research output: Contribution to journalArticlepeer-review

Abstract

Aims: To assess the prevalence of undiagnosed diabetes (UDD) and the diagnostic accuracy of the American Diabetes Association (ADA) criteria to detect UDD cases in a Peruvian population. Methods: Population-based cross-sectional study. UDD was defined using fasting plasma glucose (FPG), 2-hour post-prandial plasma glucose (2 h-PPG), and glycated hemoglobin (HbA1c) traditional cut-offs. Diagnostic accuracy was estimated using areas under the receiver-operating characteristic (ROC) curve, compared with the combination of oral glucose tolerance test (FPG plus 2 h-PPG) plus HbA1c as gold standard. Results: 1609 subjects were evaluated; mean age 48.2 (SD: 10.6) years, 50.3% were women. A total of 179 (11.3%) subjects were classified as having diabetes, 41.3% of them had UDD. Area under the curve for FPG, 2 h-PPG and HbA1c was 86.5% (95% CI: 81.4–91.6%); 87.2% (95% CI: 82.2–92.2%) and 80.4% (95% CI: 74.8–86.0%), respectively. FPG sensitivity was 73.0%, whereas this value was 74.3% for 2 h-PPG and 60.8% for HbA1c. Of 74 UDD cases, 45 were positive for HbA1c, 54 for FPG and 55 for 2 h-PPG. Conclusions: 41.3% of people with diabetes do not know their diagnosis. Diagnostic accuracy of FPG and 2 h-PPG was higher than HbA1c. The most sensitive combination of two tests to detect UDD cases was FPG plus 2 h-PPG.

Original languageEnglish
Article number108475
JournalDiabetes Research and Clinical Practice
Volume169
DOIs
StatePublished - Nov 2020

Keywords

  • Diabetes mellitus
  • Epidemiology
  • ROC curve
  • Sensitivity and specificity

Fingerprint

Dive into the research topics of 'Diagnostic accuracy of ADA 2020 criteria for undiagnosed diabetes in a Peruvian population'. Together they form a unique fingerprint.

Cite this