Glucose and Hemoglobin A1c Measurement Uncertainty at the Diagnostic Thresholds for Diabetes Mellitus: Implications for Patient Classification
Journal of Turkish Clinical Biochemistry · 06.11.2026
DOI: 10.65717/jtcb.2026.26054 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Turkish Clinical Biochemistry, online erschienen am 06.11.2026. Verfasst von Kaan Kuzu und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Themen: Diabetes mellitus
Abstract (Originalsprache)
Objective
Diagnosis of diabetes relies on fixed thresholds, although every result carries measurement uncertainty. The authors estimated the uncertainty of glucose and hemoglobin A1c (HbA1c) at their diagnostic limits and determined how many patients had a glucose result within the uncertainty interval around the threshold.
Methods
Uncertainty was estimated by the top-down Nordtest approach from internal quality control and external quality assessment, using the control level nearest each limit. Glucose and HbA1c results of adults between March 1 and June 1, 2023, were analyzed retrospectively at patient level. Patients whose glucose fell inside the interval around 126 mg/dL (7.0 mmol/L) were paired with the nearest HbA1c measured within 30 days before or after.
Results
Expanded uncertainty was 6.6% for glucose, or ± 8.3 mg/dL at the threshold, and 8.3% for HbA1c, or ± 0.54%. Of 46,639 adults, 4,114 (8.82%) had a glucose result between 117.7 and 134.3 mg/dL. An HbA1c paired within 30 days was available for 1,205 of them. In this selected subgroup, threshold-based classifications disagreed in 556 cases (46.1%; κ = 0.080), against 14.6% and κ = 0.662 among all 12,815 paired patients.
Conclusion
In this single-center cohort, 8.8% of the adults tested had a glucose result within the uncertainty interval around the diagnostic threshold, and in the subgroup with a paired HbA1c, the 2 tests disagreed almost half the time. The findings support reporting uncertainty with results near decision limits and confirming such results before a diagnosis is recorded, but they do not show that analytical uncertainty caused the disagreement.
Zitieren
APA
Kuzu, K., Asil, Ö., & Bozkaya, G. (2026). Glucose and Hemoglobin A1c Measurement Uncertainty at the Diagnostic Thresholds for Diabetes Mellitus: Implications for Patient Classification. Journal of Turkish Clinical Biochemistry. https://doi.org/10.65717/jtcb.2026.26054
Vancouver
Kuzu K, Asil Ö, Bozkaya G. Glucose and Hemoglobin A1c Measurement Uncertainty at the Diagnostic Thresholds for Diabetes Mellitus: Implications for Patient Classification. Journal of Turkish Clinical Biochemistry. 2026. doi:10.65717/jtcb.2026.26054
BibTeX
@article{Kuzu2026,
author = {Kuzu, Kaan and Asil, Öznur and Bozkaya, Giray},
title = {Glucose and Hemoglobin A1c Measurement Uncertainty at the Diagnostic Thresholds for Diabetes Mellitus: Implications for Patient Classification},
journal = {Journal of Turkish Clinical Biochemistry},
year = {2026},
doi = {10.65717/jtcb.2026.26054}
}
EUVIMED