GLI GLI Quality Tool
GLI Quality Tool — Version 2.0

Writing a TB Laboratory Procedure for Critical Instrument Failures

A critical instrument failure can interrupt testing, compromise biosafety, delay treatment decisions, and make previously issued results difficult to defend. In a tuberculosis laboratory, the response must therefore be consistent, documented, and linked to the quality management system. A written procedure gives staff a reliable path from the first alarm or abnormal observation through containment, technical assessment, reporting, and recovery.

The procedure should apply to instruments that affect specimen integrity, analytical performance, environmental safety, or result traceability. Examples include molecular testing platforms, centrifuges, biosafety cabinets, incubators, refrigerators, autoclaves, microscopes, uninterrupted power supplies, and laboratory information systems. The level of response may differ by instrument, but the decision-making principles should remain clear.

A useful procedure is practical enough for a busy bench scientist to follow during an incident. It should state who takes immediate action, who authorizes testing to stop, which results require review, how service support is contacted, and what evidence must be retained. It should also connect with equipment management, incident reporting, document control, personnel competency, and continual improvement.

Define the failure and its impact

Begin by defining what the laboratory considers a critical instrument failure. The definition should focus on consequences rather than on a particular brand or error code. A failure is critical when it may expose staff to infectious material, invalidate a test run, alter specimen storage conditions, interrupt required monitoring, or prevent the laboratory from proving that a result is reliable.

Include clear examples and non-examples. A biosafety cabinet that loses airflow, a centrifuge with excessive vibration, an incubator outside its validated temperature range, or a molecular platform that stops during amplification may require immediate containment. A broken printer may be inconvenient but is usually not critical unless it prevents required records from being created or results from being controlled.

Describe how staff recognize a failure. Triggers may include an instrument alarm, failed internal quality control, calibration drift, unusual noise, visible damage, power instability, temperature excursion, software malfunction, or an operator’s observation that the instrument is behaving unexpectedly. Staff should be instructed to stop and report uncertainty rather than continue testing to meet a turnaround-time target.

Set immediate containment actions

The first response should protect people, specimens, records, and the surrounding work area. The procedure should tell the operator whether to stop the run, secure the instrument, close or isolate specimens, preserve error messages, and notify the responsible supervisor. Staff should never open electrical panels, bypass alarms, move contaminated components, or attempt an unapproved repair.

Safety instructions need to be instrument-specific. For a centrifuge, this may include waiting for the rotor to stop completely before opening the lid and managing any suspected tube breakage under the laboratory’s spill response procedure. For a failed biosafety cabinet, work should stop and potentially exposed material should remain contained until the area is assessed. For a refrigerator or freezer excursion, specimens should be identified, temperature data preserved, and transfer to a qualified backup unit considered.

The procedure should establish an escalation chain. The operator informs the section lead, who assesses whether the incident requires the laboratory manager, biosafety officer, quality manager, biomedical engineer, manufacturer, or external service provider. Contact details, service contract information, backup equipment locations, and after-hours arrangements should be available in a controlled location rather than held only in one person’s phone.

Create a decision path for testing and results

A strong procedure distinguishes between tests in progress, completed runs, stored specimens, and results already released. The instrument’s failure time should be recorded as accurately as possible, along with the last acceptable quality control, last verified operating condition, affected run identifiers, specimen numbers, and any result transmission status.

When analytical validity is uncertain, the default action should be to quarantine affected results until a qualified person completes a documented review. The review may consider internal controls, calibration status, environmental records, error codes, maintenance history, replicate testing, and comparison with a validated backup method. A result should not be released simply because the instrument produced a numeric or qualitative output.

The laboratory should also define reporting requirements. Internal reports normally go to the section supervisor and quality manager. Clinically significant delays, amended reports, recollection requests, or potentially incorrect results may require communication with clinicians, the referring facility, the national TB program, or another designated authority. The procedure should specify who communicates, what information can be shared, and how the communication is recorded.

The following decision matrix can make the response easier to apply during a stressful event:

Failure situation Immediate control Result decision Escalation
Safety function is lost Stop work, secure the area, notify the supervisor Hold all affected work Biosafety and laboratory management
Quality control fails with an instrument alarm Stop reporting, retain run data and controls Quarantine the run and review validity Section lead and quality manager
Temperature moves outside an accepted range Identify and protect specimens, record excursion Assess specimen stability and affected tests Equipment lead and supervisor
Power or software interruption occurs mid-run Preserve instrument status and electronic records Review run completeness before release Technical support and quality manager
Failure is repaired and verified Complete maintenance and performance checks Resume only after authorization Laboratory manager or designee

Record the incident and notify stakeholders

The incident record should capture facts, decisions, and evidence in chronological order. A useful form includes the date and time, instrument identification, location, operator, failure description, alarm or error message, affected specimens and batches, immediate actions, people notified, service ticket number, temporary controls, and final disposition.

The report should avoid speculation and blame. “Amplification stopped during cycle 32 and error code X appeared” is more useful than “the operator mishandled the machine.” Objective language supports root-cause analysis and helps distinguish equipment failure, environmental conditions, user error, supply problems, software issues, and weaknesses in the procedure itself.

Result communication deserves special attention. If a report was issued before the failure was recognized, the laboratory needs a controlled process for correction, withdrawal, amendment, and notification. The process should preserve the original record according to applicable requirements while clearly identifying the corrected interpretation, reason for the change, authorizing person, date, and recipients.

Laboratories can use the GLI Quality Tool’s phase-specific checklists to compare their procedure with broader quality system activities. Checklists can help identify missing controls for equipment, documentation, assessment, personnel responsibilities, and records without replacing local technical judgment.

Restore service and verify performance

Repair does not automatically mean that an instrument is fit for use. The procedure should define release criteria after corrective maintenance. Depending on the device, this may include cleaning and decontamination, inspection, calibration, preventive maintenance, temperature or airflow checks, software verification, internal quality control, external quality assessment considerations, and comparison with an approved reference method.

A competent person should authorize return to service. The authorization should identify what was repaired, who performed the work, which checks were completed, acceptance limits, supporting records, and the date testing may resume. If a manufacturer or service engineer provides a report, the laboratory should review it rather than file it without assessing whether the work addressed the original risk.

Specimens and pending work require a separate recovery plan. Staff should identify material that can still be tested, material requiring recollection, and material that must remain quarantined. If a backup instrument or referral laboratory is used, the procedure should address transport conditions, chain of custody, data transfer, result reconciliation, and communication of any revised turnaround time.

Relevant downloadable materials can support controlled forms, worksheets, or other records used to standardize this recovery process. Any adapted document should carry the laboratory’s document number, version, approval date, and review date.

Review causes and strengthen the system

After the immediate event is controlled, the laboratory should determine whether a formal root-cause analysis is needed. Repeated alarms, missed maintenance, unstable power, inadequate environmental monitoring, insufficient training, unavailable spare parts, or unclear escalation instructions may indicate a system problem rather than an isolated equipment defect.

The review should examine trends across incidents, downtime, rejected runs, delayed reports, service response times, and repeat failures. Corrective and preventive actions should have an owner, due date, verification method, and effectiveness review. For example, replacing a damaged cable may fix the immediate issue, while adding a voltage stabilizer and reviewing preventive maintenance may reduce recurrence.

The procedure itself should be tested. A short simulation can reveal whether staff know where to find contact details, how to quarantine results, how to preserve electronic evidence, and who may authorize restart. The quality manager should revise the document when equipment changes, regulations change, a serious incident occurs, or the review identifies unclear instructions. Guidance on continual improvement can help place incident learning within a broader quality improvement cycle.

Practical recommendations for a usable procedure

Keep the controlled procedure concise, while placing detailed instrument-specific instructions in approved work instructions or manufacturer-supported documents. Make sure printed copies at the bench match the current electronic version and that obsolete copies are removed.

A TB laboratory procedure becomes valuable when it converts an unexpected failure into controlled action. Approve it through the laboratory’s document-control process, train all relevant personnel, and place the response flow where staff can reach it without entering an unsafe area. Then review the first incidents against the procedure and refine the controls using documented evidence. Building this discipline into routine laboratory practice protects patients, staff, specimens, and confidence in TB test results.