GLI GLI Quality Tool
GLI Quality Tool — Version 2.0

A Practical Protocol For Validating New TB Laboratory Methods

Introducing a new tuberculosis test method can improve detection, drug-resistance testing, turnaround time, or access to diagnosis. It can also create avoidable risks if the laboratory begins routine use before confirming that the method performs as expected in its own setting. A written validation protocol turns that decision into a controlled, evidence-based process.

The protocol should explain why the method is being introduced, what claims must be tested, which specimens and controls will be used, who is responsible for each activity, and what evidence is required before release. It should apply to molecular assays, smear microscopy procedures, culture systems, line probe assays, antimicrobial susceptibility testing, and related changes to established workflows.

Validation is also part of a broader quality management system. The work should connect with document control, personnel competency, equipment management, biosafety, internal quality control, external assessment, and continual improvement. The GLI Quality Tool materials can help laboratories align this protocol with a phased approach to strengthening quality practices.

Define the method and its intended use

Begin by describing the test precisely. Record the manufacturer, assay version, platform, specimen types, target organism or gene, resistance markers, reagents, software, and reporting units. State whether the method is entirely new, transferred from another laboratory, modified locally, or being used for a new specimen type. These distinctions determine the amount and type of evidence required.

The intended use statement should identify the population, clinical purpose, testing algorithm, and decision supported by the result. For example, a molecular assay may be intended to detect Mycobacterium tuberculosis complex and rifampicin resistance in sputum from presumptive pulmonary TB patients. It may not automatically be suitable for extrapulmonary specimens, treatment monitoring, or use as a replacement for culture.

Separate validation from verification in the protocol. Validation establishes that a new or modified method is fit for its intended purpose, while verification confirms that an already validated method works correctly under local conditions. Both activities require planned evidence, but a locally modified procedure usually needs broader investigation than an unchanged commercial method adopted according to its instructions.

Build a risk-based validation plan

List the risks associated with the test before selecting samples. Consider false-negative results caused by low organism concentration, inhibitors, poor specimen quality, target mutations, contamination, reagent failure, or unsuitable storage. False-positive results may result from amplicon carryover, cross-contamination, non-specific reactions, or transcription errors. Operational risks include power interruptions, temperature excursions, instrument downtime, supply shortages, and insufficient staff competency.

The protocol should specify the performance characteristics relevant to the method. Depending on the assay, these may include accuracy, precision, analytical sensitivity, analytical specificity, limit of detection, reproducibility, invalid-result frequency, reportable range, carryover, interference, and turnaround time. For drug-resistance tests, the plan should address agreement with an accepted reference method and the interpretation of clinically important mutations.

Use a risk-based sample design rather than selecting only convenient specimens. Include positive samples across relevant concentration levels, negative samples, weak positives near the detection limit, and challenging specimens when available. If specimens are limited, document the rationale, use characterized archived material where appropriate, and identify which claims remain supported by external evidence rather than local testing.

Set acceptance criteria before testing

Acceptance criteria must be written before results are reviewed. They should be measurable, justified, and connected to the intended use. A protocol that states only that results must be “acceptable” leaves too much room for subjective decisions and makes an audit difficult.

For qualitative TB assays, criteria may include agreement with a reference result, a maximum invalid rate, correct detection of expected resistance markers, and acceptable reproducibility across operators or days. For quantitative or semi-quantitative procedures, define allowable bias, precision limits, calibration requirements, and the range over which results can be reported. Include rules for discordant results, indeterminate findings, repeat testing, and unresolved discrepancies.

The reference method should be identified and its limitations acknowledged. Culture, sequencing, microscopy, a previously established molecular assay, or a characterized proficiency panel may serve as a comparator depending on the claim being tested. If no perfect reference exists, use a composite reference approach or expert adjudication and explain how discordant results will be investigated.

Validation element Evidence to collect Example decision rule
Accuracy or agreement Results from characterized positive and negative specimens compared with a defined reference Meets the preapproved agreement target, with all discordances reviewed
Analytical sensitivity Dilution series or low-concentration specimens tested across repeated runs Detection is consistent at the intended clinical threshold
Precision and reproducibility Replicates across operators, days, runs, instruments, or reagent lots Results remain within stated qualitative or quantitative limits
Specificity and interference Non-target organisms, common specimen interferents, and challenging matrices No clinically significant false-positive or false-negative pattern
Operational performance Invalid rate, turnaround time, hands-on time, errors, and material use Workflow meets service requirements without compromising safety
Report verification Sample reports, flags, interpretive comments, and result transmission Reports are accurate, clear, authorized, and traceable

Control specimens, equipment, and safety

The protocol should define the controls used in every run and during the validation study. Positive controls demonstrate that target detection and reagents are functioning. Negative or no-template controls help identify contamination, while extraction controls can reveal failures during specimen processing. Control acceptance rules must state when a run is valid, when it must be repeated, and when affected patient results must be withheld.

Equipment readiness is part of method validation, not an administrative afterthought. Instruments should be installed, qualified where applicable, calibrated or checked, and linked to maintenance records. Temperature-controlled storage, biosafety cabinets, centrifuges, pipettes, and backup power should be assessed against the method’s requirements. For laboratories using centrifugation during TB specimen processing, the centrifuge maintenance guide provides practical support for preventive maintenance planning.

Biosafety controls should be reviewed before any validation specimen is handled. Define the containment area, personal protective equipment, aerosol-minimizing practices, spill response, waste treatment, and exposure reporting process. Validation should never create a separate, less controlled pathway for working with infectious material. The same safe practices expected during routine testing must apply from the first study run.

Assign responsibilities and manage records

Name the protocol author, technical lead, testing personnel, quality officer, biosafety representative, and approving authority. Each person should understand the part they play in specimen selection, testing, data review, deviation management, and final authorization. Staff should demonstrate competency on the new method before producing validation data, with training and assessment records retained.

Use controlled forms or electronic records to capture specimen identifiers, storage history, run dates, operators, reagent lots, instrument status, control results, raw data, calculations, deviations, and corrective actions. Protect confidentiality by using coded identifiers where possible. Ensure that changes to the protocol, worksheets, acceptance criteria, or final report are version controlled and approved.

A deviation does not automatically invalidate a study, but it must be documented and assessed. For example, a delayed run, control failure, temperature excursion, or incorrect incubation time may affect some or all results. The technical lead should record the event, determine its impact, define repeat testing when necessary, and ensure that the final report explains how the deviation was resolved.

Review evidence and authorize implementation

At the end of testing, prepare a validation report that summarizes the method, study design, specimen set, results, deviations, limitations, and decision. Include raw-data references or attachments so that another qualified reviewer can trace the conclusions. The report should state clearly whether the method is approved, approved with restrictions, or not approved for the proposed use.

A multidisciplinary review is particularly valuable when results are discordant or the assay affects treatment decisions. The laboratory manager, quality officer, clinician or program representative, and technical experts may need to assess whether an unexpected result reflects the assay, the comparator, the specimen, or an interpretation issue. Restrictions might include limiting the method to certain specimen types, requiring referral of resistant or inconclusive results, or using confirmatory testing.

Implementation should begin only after the SOP, reporting templates, quality control plan, referral rules, inventory system, maintenance schedule, and staff competency records are ready. Establish a defined go-live date and a period of enhanced monitoring. Review early invalid rates, rejected specimens, turnaround time, control failures, complaints, and discordant results to confirm that routine performance matches the validation findings.

Practical safeguards for a defensible study

A concise checklist can keep the work focused while preserving essential evidence:

The protocol should also include a post-implementation review date. Early monitoring may reveal problems that were not visible in a controlled validation study, such as reagent delivery delays, unusual local specimen types, software errors, or a higher invalid rate during busy periods. These findings should feed into corrective action, staff retraining, SOP revision, and further verification when the method or its operating conditions change.

A carefully documented validation protocol gives the laboratory more than permission to start a test. It creates a defensible link between scientific evidence, local capability, patient safety, and ongoing quality improvement. Put the approved plan into controlled use, retain the evidence for review, and monitor the method as a living part of the TB diagnostic service.