How to Validate a Bioanalytical Method
Prove a method measures what it claims in the matrix it will actually see.
What a how-to is not
A how-to is SPEQ’s practitioner method, not a procedure. It does not replace your own SOP, it is not a validated approach, and the judgement calls in it belong to your quality unit.
Bioanalytical method validation demonstrates that a method reliably quantifies an analyte in a biological matrix, and it is the point where nonclinical and clinical work meet: the same harmonised expectations apply to method validation and study sample analysis for both. That is worth stating plainly, because teams often assume the laboratory standard follows the study type, and here it does not.
- 1
Define the analyte, the matrix and the intended range
Validation is specific to the analyte in a matrix over a range. A method validated in one species or one anticoagulant is not validated in another, and the substitution is easy to make and hard to spot afterwards.
- 2
Establish selectivity and specificity against real interference
Demonstrate that the method distinguishes the analyte from matrix components, metabolites, concomitant medication and degradation products. Selectivity shown only in blank matrix from healthy donors is not selectivity in the population that will be sampled.
- 3
Characterise the calibration curve and quality control samples
Establish the calibration model, its range, and quality control samples across it. The lower limit of quantification must be supported by demonstrated accuracy and precision, not asserted from the lowest calibrator that produced a signal.
- 4
Demonstrate accuracy, precision, and matrix effect
Within-run and between-run performance across the range, plus recovery and matrix effect where the technique is susceptible. Matrix effect assessed in one lot of matrix says nothing about lot-to-lot variability, which is where it usually bites.
- 5
Establish stability under every condition a sample will meet
Bench-top, freeze-thaw, long-term storage, processed-sample and stock stability, covering the real conditions and durations from collection to analysis. Sample stability is where validated methods most often fail in practice, because the study exposes samples to a timeline the validation never tested.
- 6
Run study samples with the controls that let a run be rejected
Acceptance criteria for a run, incurred sample reanalysis where required, and rules for repeats fixed in advance. Repeat rules decided after seeing a result are the mechanism by which an inconvenient value becomes an analytical error.
- !A method validated in one matrix, species or anticoagulant used in another without bridging.
- !The lower limit of quantification asserted from the lowest calibrator rather than demonstrated.
- !Matrix effect assessed in a single lot, missing the lot-to-lot variability that actually causes failures.
- !Repeat and acceptance rules decided after results are seen, turning an inconvenient value into an error.
How to Validate a Bioanalytical Method: frequently asked questions
Common questions on validate a bioanalytical method.
Do nonclinical and clinical samples follow different bioanalytical expectations?
No — the harmonised guidance covers method validation and study sample analysis for both. This surprises teams who assume the laboratory standard follows the study type. What does differ is the surrounding quality system: a nonclinical safety study sits under good laboratory practice, while clinical trial sample analysis sits under clinical laboratory expectations.
When does a validated method need revalidation?
When something it was validated against changes — the matrix, species, anticoagulant, range, a significant instrument or reagent change, or a change in sample handling. Partial validation may suffice for a limited change, but the judgement is documented rather than assumed, and "it is the same method" is not the test.
Why is stability the most common practical failure?
Because validation tests the conditions someone anticipated and the study exposes samples to the conditions that actually occurred — a shipment delayed, an extra freeze-thaw, a longer interval before analysis. Establishing stability across the real timeline from collection to analysis is what closes that gap.