Precision and Trueness Verification of Routine Biochemistry Tests using CLSI EP15-A3

Method verification is a fundamental requirement to ensure reliable test results in diagnostic and therapeutic processes. This study aimed to verify the precision and trueness of routine biochemistry assays on the Beckman Coulter AU680 analyzer according to the Clinical and Laboratory Standards Institute (CLSI) EP15-A3 and to evaluate analytical performance based on biological variation (BV)-derived specifications.

Twelve routine biochemical analytes, including albumin, alkaline phosphatase (ALP), alanine aminotransferase, aspartate aminotransferase, lactate dehydrogenase (LDH), amylase, glucose, total cholesterol, total protein, triglycerides, urea, and uric acid, were evaluated. Two levels of Bio-Rad Lyphochek Assayed Chemistry Controls (lots 89701 and 89702) were analyzed using a 5 × 5 experimental design. Repeatability and within-laboratory precision were estimated using one-way ANOVA and compared with the manufacturer's claims and CLSI EP15-A3 upper verification limits (UVL). Trueness was assessed using Unity Real Time peer group means as the target values. The results falling outside the verification intervals were further evaluated against the BV-based allowable bias limits. The total analytical error (TAE) was calculated by combining the bias and within-laboratory imprecision estimates derived from the CLSI EP15-A3 verification protocol, and was subsequently evaluated against BV-based specifications.

Repeatability CV values ranged from 0.37% to 1.51%, and within-laboratory CV values ranged from 0.56% to 3.70%. Although several observed CV values exceeded the manufacturer's claims, all remained below the CLSI EP15-A3 UVL, confirming acceptable analytical precision. Trueness was verified by compliance with verification intervals or by bias values remaining within the allowable BV-based limits. Optimal bias performance was achieved for ALP, LDH, total protein, triglycerides, and urea at both control levels, whereas the remaining analytes met the desirable bias specifications. All analytes fulfilled either the optimal or desirable BV-based total error specifications, and none exceeded the desirable allowable limits.

All evaluated assays demonstrated satisfactory analytical performance and successfully met the CLSI EP15-A3 verification criteria. Furthermore, this integrated verification approach may assist laboratories in generating objective evidence to support quality assurance activities and comply with ISO 15189:2022 accreditation requirements.