GLI GLI Quality Tool
GLI Quality Tool — Version 2.0

Controlling Non-Conforming TB Results And Retesting

Reliable tuberculosis testing depends on more than accurate instruments and skilled scientists. Laboratories also need a defined response when a result, control, specimen, process, or report does not meet an approved requirement. A non-conforming result may arise from contamination, an unsuitable specimen, a transcription error, a failed internal quality control, instrument malfunction, or an unexpected external quality assessment finding.

A practical system protects patients, staff, referring clinicians, and public health programmes. It ensures that questionable results are contained, investigated, corrected, and communicated consistently. For Australian laboratories, the process should fit NATA accreditation expectations, ISO 15189 principles, state and territory reporting duties, privacy obligations, and the realities of testing across metropolitan and remote services.

Event Immediate response Retesting decision Required evidence
Failed internal control Stop or hold affected runs Repeat after resolving the control failure Control chart, investigation, corrective action
Suspected contamination Quarantine batch and review workflow Reprocess or recollect according to risk Contamination assessment and authorisation
Unsuitable specimen Notify the requesting service Request a new specimen where clinically safe Rejection reason and communication record
EQA failure Withhold affected interpretation if necessary Repeat retained material or test a known panel Root-cause analysis and follow-up review
Incorrect report Correct the record and notify recipients Retest only when the analytical result is uncertain Amended report, notification log, approval

Define What Counts As Non-Conforming

A laboratory should define non-conformity broadly enough to capture failures before they affect a patient. Examples include a sputum specimen without adequate identification, leakage during transport, delayed receipt beyond stability limits, an invalid molecular assay, an interrupted culture incubation, a failed positive or negative control, and a result released by an unauthorised person.

The definition should cover pre-examination, examination, and post-examination stages. It should also distinguish a technical failure from a clinically unexpected result. An unusual result is not automatically wrong, while a technically valid result can still require urgent review if it conflicts with specimen history, earlier testing, or an established contamination pattern.

A controlled procedure should state who can place a test on hold, who may authorise a retest, and when the laboratory director or senior microbiologist must be involved. The quality roadmap from the GLI Quality Tool can help teams align this process with documentation, personnel, equipment, assessment, and continual improvement requirements.

Contain The Risk Before Retesting

The first action is containment, not repetition. Staff should identify the affected patient, specimen, assay, instrument, reagent lot, run, and reporting status. Results that may be unreliable should be held in the laboratory information system when possible. If a report has already been issued, the responsible clinician or public health contact should receive a prompt, documented notification.

A retest cannot repair a compromised specimen or erase the impact of a delayed notification. For suspected cross-contamination, the laboratory may need to quarantine related specimens, pause a work area, inspect extraction steps, review staff movements, and check environmental cleaning. In a busy Melbourne or Sydney service, this may involve several benches and shifts; in Darwin or a regional hub, it may also involve longer transport times and fewer replacement resources.

The laboratory should record the time of detection, immediate risk assessment, people notified, tests placed on hold, and the person approving the containment plan. A simple incident number can link the original request, worksheets, instrument files, amended reports, and investigation findings.

Make Retesting A Controlled Decision

Retesting should answer a defined question. The laboratory may need to establish whether the first result was analytically valid, whether a specimen was compromised, whether a new specimen is required, or whether a reporting error occurred. Repeating a test without identifying the question can produce conflicting results and create false confidence.

The procedure should specify when to repeat from the original specimen, when to use a second aliquot, and when to request a fresh specimen. For example, a failed control may justify repeating the run after resolving the reagent or instrument problem. An inadequate or contaminated specimen may require recollection. A suspected clerical error may require record verification rather than another analytical run.

Retesting should use an appropriate method and, where relevant, an independent operator, reagent lot, instrument, or extraction process. The original result must remain visible in the audit trail, with its status clearly marked. Staff should never overwrite a non-conforming result or silently replace it with a later value.

Investigate Root Causes And Trends

Each event should receive a proportionate investigation. Minor issues may need a brief documented review, while repeated invalid runs, contamination events, or EQA failures require a formal root-cause analysis. Useful evidence includes control charts, maintenance records, temperature logs, reagent lot details, specimen acceptance records, staff training files, environmental records, and instrument middleware logs.

The investigation should look beyond individual blame. A recurring error may reflect unclear instructions, a crowded workflow, insufficient competency assessment, poor shift handover, inadequate barcode scanning, or a courier arrangement that does not protect specimens. Australian laboratories should consider state and territory public health pathways, NATA assessment findings, and local work health and safety requirements when reviewing the system.

For external quality assessment failures, laboratories can use the out-of-control results guide to support a structured response. The review should identify whether the failure is isolated or systemic, whether patient results may be affected, and what evidence will demonstrate that corrective action worked.

Document Corrective Action And Communication

Corrective action should be specific, assigned to a person, and given a due date. It may include retraining, a revised worksheet, equipment servicing, reagent replacement, environmental cleaning, workflow separation, a new specimen transport instruction, or additional review of released results. “Staff reminded” is rarely enough evidence unless the reminder is linked to a clear competency or process change.

The laboratory should define how effectiveness will be checked. Examples include successful repeat controls over a specified number of runs, review of contamination rates, observation of a revised procedure, a targeted competency assessment, or a follow-up EQA sample. Closure should be approved by someone with appropriate authority and should include evidence rather than simply a completion date.

Communication must be timely and traceable. Corrected reports should identify the amended information without obscuring the original record. Notifications should go to the requesting clinician, relevant hospital team, and public health authority when required. Patient information should be handled under applicable Australian privacy requirements, with access limited to staff who need it for care, surveillance, quality, or legal purposes.

Build Retesting Into Routine Quality Management

A robust system becomes easier to use when it is built into everyday work. The laboratory can include non-conformity fields in the laboratory information system, create standard reason codes, use decision trees at reporting stations, and review open incidents during quality meetings. Staff working weekend shifts or rotating through high-volume molecular benches should have the same access to current instructions as weekday teams.

Managers should monitor indicators such as invalid-test frequency, specimen rejection, contamination, turnaround time after retesting, amended reports, repeat EQA failures, and overdue corrective actions. Results should be reviewed by assay, site, specimen type, instrument, shift, and staff group where appropriate. A rising rejection rate at a remote collection point may indicate packaging or transport problems rather than a laboratory testing defect.

A short set of practical controls can keep the process consistent:

The system should be reviewed after major method changes, new instruments, revised reporting rules, or a serious incident. Australian services may need to coordinate across hospital networks, private providers, state reference laboratories, and public health units, so contact details and escalation pathways should be checked regularly rather than assumed to remain current.

Put The Process Into Practice

A useful first step is to map the current pathway from specimen receipt to final report and mark every point where a result can become unreliable. Compare that pathway with the laboratory’s quality manual, local reporting obligations, instrument instructions, and accreditation evidence. Then run a short scenario exercise involving a failed control, a suspected contamination event, and a corrected report.

Staff feedback is essential because the people handling specimens and entering results often see practical weaknesses before managers do. Laboratories can share feedback on tools and guidance that support quality improvement in varied resource settings. Use those observations to refine forms, escalation rules, training, and retesting criteria.

Adopt the procedure, train every relevant shift, audit its use, and review the results at the next quality meeting. A clear record of containment, investigation, retesting, communication, and effectiveness checking turns an unexpected TB laboratory failure into a controlled quality event and strengthens confidence in every result that follows.