Managing out-of-specification results in TB DST testing

A clear procedure for managing out-of-specification results from tuberculosis drug susceptibility testing (DST) assays protects patients, laboratory staff and public health programmes. It gives the team a consistent response when a result is invalid, internally inconsistent, technically impossible, outside an accepted control range or unexpected in relation to the patient’s previous results.

The procedure should support both phenotypic DST and molecular methods, including liquid culture systems, solid-media testing, line probe assays and rapid molecular tests that detect resistance markers. It must be practical for the laboratory’s equipment, staffing, referral arrangements and biosafety controls. A well-designed process also creates objective evidence for accreditation, incident review and continual improvement.

Define what counts as out of specification

Begin by describing the results covered by the procedure. An out-of-specification result may include a failed internal control, contamination, a mixed or indeterminate pattern, an instrument flag, an unacceptable growth control, a result outside validated performance limits or a resistance profile that conflicts with quality-control expectations. A result can also require review when it is technically valid but clinically implausible, such as a sudden change from a previously stable susceptibility pattern.

The procedure should state which tests are included, who may authorise a result, and when the issue must be escalated to a senior scientist, microbiologist or reference laboratory. Include the relevant test method, specimen type, critical reagents, control strains, reporting system and turnaround-time target. For laboratories in Australia, the document should align with the local quality management system and the requirements applied by NATA-accredited facilities, while allowing state and territory referral pathways to be specified.

Use a risk-based format rather than copying generic instructions. The GLI Quality Tool’s quality checklists can help the laboratory connect the procedure with document control, equipment management, assessment, personnel competency and continual improvement. The final document should identify the current version, approval date, review date, author and approving authority.

Protect the specimen, result and patient

The first response is containment. A staff member who detects an unusual DST result should stop final reporting, preserve the original specimen or culture where possible, and record the finding in the laboratory information system or quality incident register. The procedure should prevent an unchecked result from being released automatically, especially when it could change treatment for rifampicin-resistant or multidrug-resistant TB.

Specify how to label and segregate material awaiting investigation. Cultures and plates must remain secure under the laboratory’s biosafety arrangements, with access limited to authorised personnel. Staff should not repeat a test casually by opening additional containers or transferring material without a documented plan. For potentially infectious Mycobacterium tuberculosis complex material, the procedure should refer to the facility’s approved biosafety, exposure response and waste management instructions.

The case record should include the accession number, patient identifiers according to privacy policy, specimen type, collection date, assay, instrument, operator, run number, control results and the precise reason for the hold. Record whether the result has already been communicated to a clinician or public health unit. In a remote service supporting communities in the Northern Territory or Western Australia, the procedure should include a reliable after-hours contact and a defined route for urgent referral to a larger centre.

Investigate the technical cause

Investigation should follow a logical sequence. First review specimen integrity, transport conditions, storage time, culture purity and the possibility of mixed infection or cross-contamination. Then examine the run controls, reagent lot, expiry dates, preparation records, instrument maintenance, temperature logs and environmental conditions. For molecular assays, check extraction controls, inhibition indicators, amplification curves and software interpretation. For phenotypic DST, examine inoculum preparation, media, drug concentration, incubation conditions and reading criteria.

The procedure should distinguish between an assay failure and a true resistant isolate. A failed control does not establish susceptibility or resistance, while a reproducible resistant result supported by valid controls may be reportable even if it is unexpected. Repeat testing should be performed only when the method and sample quantity support it. Where the result is critical, confirmation by a second validated method or referral to an accredited reference laboratory may be necessary.

Include decision points for common causes. A contaminated culture may require decontamination or a new specimen, subject to the laboratory’s validated process. A result linked to a single reagent lot may require quarantine of that lot and review of other patient runs. An instrument malfunction may require service intervention and examination of results produced since the last acceptable performance check. When testing is referred to a laboratory in Sydney, Melbourne, Brisbane or another state centre, document the referral date, packaging, courier conditions and result reconciliation.

Report with appropriate clinical urgency

The procedure must explain what can be reported, what must be withheld and what wording should be used. Avoid converting an unresolved finding into a definitive “susceptible” or “resistant” interpretation. Suitable interim language may state that testing is inconclusive, that repeat or confirmatory testing is in progress, or that the result requires correlation with another method. The approved report comment should be written in advance and reviewed by the laboratory’s clinical governance lead.

Critical findings need prompt communication through an agreed channel. A suspected rifampicin resistance result, for example, may require direct contact with the treating team and the relevant public health service rather than waiting for routine electronic reporting. Record the name and role of the person contacted, date and time, information provided, advice received and any agreed follow-up. Australian laboratories should account for local public health arrangements, which can differ between jurisdictions and between metropolitan hospitals and regional services.

Communication should be factual and proportionate. Explain whether the problem concerns specimen quality, assay validity, control performance or the biological interpretation. Do not describe a technical repeat as confirmation unless the defined acceptance criteria have been met. If an amended report is issued, retain the original report, record the reason for amendment and ensure the change is visible to authorised users of the laboratory information system.

Close the incident and prevent recurrence

Once the result has been resolved, document the evidence supporting the final decision. This may include an acceptable repeat run, a second method, a reference laboratory report, a new specimen or an investigation showing that the original result was invalid. The authorised reviewer should sign or electronically approve the record. Any affected patient results from the same run, reagent lot or instrument period should be assessed before the incident is closed.

Corrective and preventive action should address the cause rather than simply require staff to “take more care”. Actions might include revising a work instruction, retraining staff in inoculum preparation, adding a second check for drug concentrations, changing stock control, improving courier temperature monitoring or scheduling instrument maintenance earlier. Assign each action an owner and due date, then verify that it was completed and effective.

Trend review is valuable when unusual results recur. Monitor invalid rates, contamination, control failures, repeat testing, amended reports, referral turnaround times and errors by assay or operator. A small laboratory may review these measures monthly, while a larger service may use a dashboard. The findings should feed into management review, internal audits and competency assessments. Staff in a busy Australian laboratory, including a private provider supplying rural hospitals, should be able to see that incident reporting leads to practical changes rather than blame.

Keep the approved procedure accessible at the point of work and remove obsolete copies. The GLI Quality Tool user instructions provide useful guidance for navigating phase-based quality improvement resources and applying them to local laboratory priorities. Review the procedure after a significant incident, method change, new instrument, revised resistance breakpoint or change in referral arrangements, as well as at the scheduled document review date.

Adopt the procedure through controlled document management, train every relevant staff member and test it with a simulated out-of-specification case. A short exercise can reveal missing contacts, unclear authority, weak specimen tracking or delays in critical communication. With clear criteria, disciplined investigation and documented follow-up, TB laboratories can manage uncertain DST findings safely while strengthening the wider quality management system.