An improved approach to modelling patient reclassification: HbA_1c as an example.

Area of Science:

• Clinical Chemistry
• In Vitro Diagnostics
• Quality Assurance

Background:

• External quality assurance (EQA) is crucial for evaluating in vitro diagnostic (IVD) device analytical performance.
• Assessing device fitness for clinical use requires robust performance metrics.

Purpose of the Study:

• To develop and apply a novel method for assessing IVD device analytical performance using reclassification rates at clinical decision limits.
• To evaluate the performance of common haemoglobin A1c (HbA1c) IVD devices using this new method.

Main Methods:

• A commutable EQA program provided analytical performance data.
• Simulated 10,000 HbA1c measurements per subject using common IVD devices and the NHANES device.
• Calculated incremental reclassification rates (IRR) by comparing IVD device reclassification rates to the NHANES device, accounting for regression to the mean.

Main Results:

• Median IRRs were optimal for diabetes diagnosis (0.11–0.22%) and desirable for diabetes monitoring (-0.51–1.01%) across most tested HbA1c IVD devices.
• One device (Afinion) failed the minimal standard for prediabetes diagnosis (8.62%), while others met desirable standards (1.43–2.69%).

Conclusions:

• The incremental reclassification rate (IRR) method effectively identifies IVD device performance variations.
• EQA providers should consider reporting IRR for analytes to flag poorly performing IVD devices, especially for prediabetes screening.