Repeat Analysis and Replicate Agreement in Forensic Toxicology

Quick answer: Repeat analysis can show whether a result is reproducible under defined conditions, but replicate agreement is only one part of analytical reliability. Two closely agreeing measurements may support precision; they do not by themselves establish accuracy, analyte identity, absence of bias, or the broader interpretation of the result.

What does repeat analysis actually test?

A repeat measurement asks whether the analytical process produces sufficiently similar results when the same or a comparable specimen is tested again. The scientific meaning depends on what was repeated: the same extract, a new injection, a new aliquot, a fresh extraction, a second specimen, or a later analysis on another instrument or day. Those designs do not challenge the same sources of variability.

For that reason, repeat analysis should be interpreted together with the laboratory’s validated method, quality-control program, and predefined acceptance criteria. The Academy Standards Board’s ANSI/ASB Standard 036 addresses validation of analytical methods for forensic toxicology, while ANSI/ASB Standard 054 establishes minimum requirements for forensic toxicology quality-control programs.

Precision is not the same as accuracy

Closely agreeing replicate values are evidence about precision under the conditions tested. They do not prove that the measurements are close to the true or reference value. A systematic calibration error, matrix effect, preparation bias, incorrect reference material, integration problem, or shared interference can affect repeated measurements in the same direction.

This distinction is central to quantitative interpretation. See Accuracy and Precision in Forensic Toxicology for the difference between repeatability, reproducibility, bias, and closeness to a reference value.

Not all repeats are equivalent

A second injection from the same vial mainly challenges injection and instrumental repeatability. A new extraction from a separate aliquot reaches further upstream and can challenge specimen aliquoting, internal-standard addition, extraction, evaporation, reconstitution, and related preparation steps. A repeat on another day may also incorporate between-run variation, new calibration, different control results, environmental change, and other routine sources of variability.

Because the tested sources of variation differ, a laboratory record should make clear what was actually repeated. Describing every duplicate value as an independent confirmation can overstate the evidence when both values came from the same extract or the same preparation sequence.

What level of replicate agreement is acceptable?

There is no single universal percentage that applies to every analyte, matrix, concentration, and forensic toxicology method. Acceptance criteria should come from the validated procedure and should be appropriate for the intended use of the result. Analytical variability often changes across a calibration range, particularly near the lower end, so a fixed agreement rule must be interpreted in the context of the method’s demonstrated performance.

When replicate agreement is evaluated, review the unrounded results, the calculation used to compare them, the laboratory’s written acceptance criterion, and whether the criterion was established during validation or later quality review. A rounded pair of reported values can appear identical even when the underlying instrument results differ.

What does replicate disagreement mean?

Disagreement is a signal to investigate, not a diagnosis of one particular error. Possible causes include specimen heterogeneity, pipetting or aliquoting variability, extraction loss, matrix effects, carryover, integration differences, dilution error, calibration problems, instrument instability, or simple random analytical variation. The relevant explanation depends on the analytical workflow and the records from the affected run.

A scientifically useful review asks whether the laboratory documented the discrepancy, identified a plausible source, repeated an appropriate portion of the process, and applied its corrective-action or acceptance procedure consistently. Quality-control data should be examined alongside case-sample replicates because a case result can appear reproducible while the analytical run contains other evidence of a problem. See Quality Control in Toxicology and Method Validation in Forensic Toxicology.

Repeat testing near a cutoff or decision point

Repeated measurements near a reporting cutoff or other consequential boundary require particular care. A second value may help characterize short-term analytical variation, but it does not make measurement uncertainty disappear. The result remains an estimate produced by a validated measurement process.

ANSI/ASB Standard 056, first published in 2025, provides minimum requirements for evaluating measurement uncertainty in quantitative forensic toxicology testing. The relationship between uncertainty and a reported result is discussed further in Measurement Uncertainty in Forensic Toxicology.

Near a boundary, the reviewer should distinguish the laboratory’s analytical reporting rule from any separate legal, administrative, or clinical decision rule. A laboratory repeat may support the reliability of the measurement process, but it does not by itself answer the ultimate non-analytical question attached to the cutoff.

Repeat analysis versus a true independent confirmation

The phrase independent confirmation should be used carefully. A result repeated by the same analyst, on the same instrument, from the same extract, using the same calibration and reference materials shares many common sources of potential error. Greater analytical independence may come from a new aliquot, separate preparation, different run, alternative analytical method, or second specimen, depending on the scientific question.

Even then, independence is not absolute. Two methods can share the same specimen collection error, reference-material issue, chain-of-custody problem, or incorrect case assumption. The degree of independence should therefore be described rather than presumed.

Records that matter when reviewing repeated results

raw quantitative values before rounding;
sample identifiers and aliquot numbers;
sequence tables and injection order;
extraction batch and preparation records;
calibrators, blanks, and quality-control results;
internal-standard response and chromatographic data;
dilution calculations and reinjection records;
the laboratory’s replicate-agreement criterion;
technical-review notes and any corrective-action documentation;
the final report and any amended or supplemental report.

ANSI/ASB Standard 053 sets minimum requirements for report content in forensic toxicology. The analytical records behind the report remain important when evaluating how a repeated result was generated and accepted.

What replicate agreement does not prove

Agreement between repeated measurements does not, by itself, prove the exact dose, time of use, route of administration, impairment, tolerance, intent, toxicity, or causation. Those are separate interpretive questions. Replicate agreement is most defensible when described for what it is: evidence about reproducibility or precision under the conditions actually tested.

Bottom line

Repeat analysis is useful when the design of the repeat is understood. The strongest review identifies which steps were repeated, which sources of variability were challenged, what acceptance criterion applied, whether the result was within the validated range, how quality-control data performed, and whether measurement uncertainty matters to the conclusion. Reproducibility strengthens a measurement; it does not transform a measurement into proof of every conclusion that might later be drawn from it.

Primary standards and further reading

This technical resource is educational and should be read together with the laboratory’s own validated procedure, quality manual, case records, and the scientific question being evaluated.