Alanine Aminotransferase (ALT) Precision Determination: Protocols, Statistics, and Acceptance Criter

Alanine aminotransferase (ALT) precision determination is a core analytical performance study that quantifies the random error of ALT measurement procedures in clinical laboratories and in vitro diagnostic (IVD) product development. This technical guide describes the sample requirements, experimental design, statistical analysis, and acceptance criteria for ALT precision studies based on CLSI EP05-A3, CLSI EP15-A3, and ISO 15189 requirements.

Technical Overview

Alanine aminotransferase (ALT, EC 2.6.1.2), formerly known as serum glutamic-pyruvic transaminase (SGPT), is a pyridoxal phosphate-dependent enzyme that catalyzes the reversible transfer of an amino group from L-alanine to alpha-ketoglutarate, producing pyruvate and L-glutamate. Because ALT is highly concentrated in hepatocytes, its serum activity is widely regarded as the most sensitive routine biomarker of hepatocellular injury, and it is measured in liver function panels, drug safety monitoring, blood donor screening, and occupational health surveillance using IFCC-recommended UV kinetic methods at 37 degrees C with NADH consumption monitored at 340 nm.

Precision determination is a formal performance evaluation that quantifies the random error, or imprecision, of an ALT measurement procedure under defined conditions. Precision is expressed as the closeness of agreement between independent replicate results and is statistically characterized by the standard deviation (SD) and the coefficient of variation (CV%). ALT precision studies are structured according to internationally recognized standards, including CLSI EP05-A3 for precision evaluation, CLSI EP15-A3 for user verification of precision claims, ISO 15189 for medical laboratory quality and competence, and applicable IVD regulatory frameworks such as EU IVDR 2017/746 and FDA 510(k) submissions.

Three levels of precision conditions are distinguished in ALT precision studies:

• Repeatability: measurements performed by the same operator, on the same instrument, with the same reagent lot, within a short time interval under identical conditions.
• Intermediate precision (within-laboratory precision): measurements performed within a single laboratory over multiple days, runs, recalibrations, and operator changes.
• Reproducibility: measurements performed across different laboratories, instruments, and reagent lots, typically evaluated by manufacturers in multi-site studies.

Testing Samples

Commercial Quality Control Materials

Liquid or lyophilized, serum-based quality control materials with assigned ALT activities are the most practical samples for precision studies. At least two, and preferably three, control levels should be used: one near the lower clinical range, one near the upper reference limit or medical decision level of approximately 40 to 50 U/L, and one in the clearly elevated range, for example 200 to 400 U/L. A single control lot should be used throughout the entire study to avoid confounding lot-to-lot variability with analytical imprecision, and controls should be reconstituted or handled exactly as instructed by the manufacturer.

Prepared Human Serum Pools

Pooled residual patient specimens provide superior commutability compared with processed commercial controls and are preferred when the study is intended to reflect performance on authentic clinical samples. Pools should be prepared under appropriate ethical and biosafety frameworks, screened for infectious agents where required, clarified by centrifugation, and mixed thoroughly to ensure homogeneity. Target activities can be adjusted by spiking with high-activity patient serum or by dilution with ALT-free serum matrix. The pool should be aliquoted into single-use portions, stored at minus 70 degrees C or below, and each aliquot thawed only once before use.

Concentration Level Selection

Because imprecision varies with analyte concentration, sample levels must be selected to span the measuring interval and the clinically important decision range. A low sample assesses performance near the limit of quantitation where absolute SD dominates, a medical decision level sample assesses performance near the reference limit where diagnostic classification occurs, and a high sample assesses performance under conditions of high enzymatic turnover where CV% is most informative.

Sample Integrity and Handling

All samples must be homogeneous, available in sufficient volume for the complete study, and stable under the storage and handling conditions used. Specimens should be free of clinically significant hemolysis, lipemia, and icterus, since these interfere with the 340 nm photometric reading. Pre-study stability verification and freeze-thaw control are essential elements of a defensible precision protocol.

Testing Items

An ALT precision determination generates the following measurable outputs, each reported per concentration level:

• Repeatability (within-run precision): SD and CV% of replicate measurements within a single analytical run.
• Between-run precision component: variability attributable to separate runs within the same operating day.
• Between-day precision component: variability attributable to different operating days.
• Within-laboratory (intermediate) precision: the combined estimate of imprecision across days, runs, and replicates, including routine recalibration events.
• Verification of manufacturer claims: comparison of observed SD and CV% against claimed performance and statistically derived verification limits.

The final report should include the number of results, mean activity in U/L, SD in U/L, and CV% for each precision condition and each concentration level, preferably with 95% confidence intervals for the imprecision estimates.

Testing Methods

Experimental Design According to CLSI EP05-A3

The classical protocol for establishing ALT precision is the 20 x 2 x 2 design: twenty operating days, two runs per day separated by at least two hours, and two replicate determinations per run, yielding 80 results per sample level. The same instrument, reagent lot, calibrator lot, and control lot should be maintained throughout the study. Calibration is performed according to the routine laboratory or system protocol, and any deviation, recalibration, or maintenance event must be documented. Samples should be positioned to reflect routine workflow, and run timing should vary realistically across the day.

Data Screening and Outlier Handling

Before statistical analysis, data are inspected for outliers using predefined criteria. A within-run outlier is flagged when the range of a duplicate pair exceeds 5.5 times the mean within-run duplicate range, and overall outliers are identified using Grubbs' test at the 1% significance level. Results may be excluded only when a documented analytical error is identified; unexplained exclusions must be reported transparently.

Statistical Analysis by Nested ANOVA

Precision components are estimated using a two-stage nested random-effects analysis of variance. With D days, R runs per day, and n replicates per run, the variance estimates are calculated as follows:

• Repeatability: Sr = square root of MSwithin.
• Between-run component: Srun = square root of ((MSrun minus MSwithin) / n), where n = 2 replicates per run.
• Between-day component: Sday = square root of ((MSday minus MSrun) / (n x R)), where n x R = 4.
• Within-laboratory precision: Swl = square root of (Sday squared + Srun squared + Sr squared), with negative variance components truncated at zero.

The coefficient of variation is then calculated as CV% = (SD / mean) x 100 for each precision condition and sample level.

Verification Protocol According to CLSI EP15-A3

When the objective is to verify a manufacturer's precision claim rather than establish new performance data, a shortened verification experiment is performed, commonly five operating days with three replicate measurements per day in a single daily run. Verification limits are derived from the claimed SD and the appropriate statistical degrees of freedom; the claim is verified if the observed SD does not exceed the verification limit. If verification fails, the laboratory may extend the study toward the full EP05-A3 protocol or contact the manufacturer for corrective action.

Acceptance Criteria

ALT precision results are judged against one or more of the following criteria:

• Manufacturer performance claims, verified when observed SD is below the manufacturer's stated limits or the EP15-A3 verification limit.
• Analytical performance goals based on biological variation, for which the desirable imprecision for ALT corresponds to a CV of approximately 12% or lower at medically relevant concentrations.
• Regulatory proficiency testing criteria, such as the CLIA acceptable performance specification of within plus or minus 20% of the target value for ALT.
• State-of-the-art comparisons against peer-reviewed method data or interlaboratory quality assessment schemes.

Applications

ALT precision determination supports multiple stakeholders across the in vitro diagnostic and clinical laboratory community:

• IVD manufacturers: generation of precision performance evaluation reports required for product registration, FDA 510(k) submissions, EU CE technical documentation under IVDR, and other national market authorization files.
• Clinical laboratories: verification of a new ALT assay, instrument, or reagent system before release for patient testing, as required by ISO 15189 method verification clauses.
• Accreditation bodies: objective evidence of analytical quality during laboratory inspections and assessments.
• Laboratory quality management: evaluation after instrument relocation, major maintenance, software updates, or reagent lot changes, and support of ongoing quality planning and sigma-metric analysis.
• Research and harmonization: characterization of method performance in method comparison, reference interval transfer, and standardization studies.

FAQ

What is the difference between repeatability and intermediate precision for ALT?

Repeatability describes the agreement of results obtained under identical conditions within one run, operator, and instrument, while intermediate precision describes agreement across different days, runs, calibrations, and operators within the same laboratory. Intermediate precision is always equal to or greater than repeatability and better reflects routine long-term performance.

How many measurements are required for an ALT precision study?

For establishing precision under CLSI EP05-A3, the standard design produces 80 results per level over 20 days. For verifying manufacturer claims under CLSI EP15-A3, a shortened experiment of approximately 15 results over 5 days is typically sufficient when the claim is met.

What CV is considered acceptable for ALT measurement?

Acceptance depends on the applied criterion. Goals based on biological variation suggest a desirable CV of approximately 12% or lower, while most modern automated ALT assays achieve within-laboratory CVs of 3% to 6% at normal and elevated concentrations. The acceptable limit should always be defined prospectively in the study protocol.

What should be done if observed imprecision exceeds the acceptance limit?

The laboratory should investigate technical causes such as reagent integrity, calibration stability, pipetting accuracy, temperature control, and sample handling, then repeat the study after corrective action. Persistent failure indicates that the system does not meet the intended performance specification and should not be released for clinical use.

Can frozen serum pools be used instead of commercial controls?

Yes. Frozen aliquoted human serum pools are highly suitable because of their commutability, provided homogeneity and stability are verified and each aliquot is thawed only once. Many registration studies combine commercial controls for regulatory traceability with serum pools for real-matrix evidence.