Building A Reliable Root Cause Process For Recurring TB Laboratory Problems
Recurring non-conformities in a tuberculosis laboratory are rarely caused by a single careless action. A delayed culture, incorrect specimen label, failed contamination control check or incomplete temperature record usually reflects weaknesses across people, procedures, equipment and workload. Treating each event as an isolated mistake can leave the underlying risk untouched.
A structured root cause analysis (RCA) helps the laboratory move from correction to prevention. The purpose is to understand why the failure was possible, identify contributing conditions, and introduce controls that remain effective during staff changes, busy periods and supply interruptions.
For Australian laboratories, the method should fit an accredited quality management system and the realities of a geographically large health service. State and territory networks may send specimens between metropolitan centres and remote communities, while work health and safety duties, NATA expectations and local operating procedures all influence how non-conformities are investigated.
Define The Non-Conformity Precisely
Begin with a factual problem statement. It should describe what happened, where it happened, when it was detected and which requirement was not met. “Poor labelling” is too broad; “three respiratory specimens received without a second patient identifier during the night shift in June” gives the investigation a usable starting point.
Separate the event from its impact. A specimen may be rejected, a result may be delayed, or a biosafety risk may be created, but the consequence does not explain the cause. Record the process stage, test method, instrument, staff role, specimen type and relevant environmental conditions. Include whether the issue was found through internal quality control, an external assessment, a complaint, a safety report or routine review.
A recurring problem should be defined using a consistent threshold. The same non-conformity may be considered recurrent when it happens twice in a quarter, appears across two sites, returns after a corrective action, or shows a statistically meaningful trend. Establishing this rule prevents teams from relying on memory or waiting for a serious incident before escalating the investigation.
Gather Evidence Before Assigning Blame
An RCA is stronger when evidence is collected before explanations are discussed. Review bench sheets, electronic records, instrument logs, maintenance reports, temperature charts, reagent lot details, training records and shift rosters. Preserve relevant records rather than editing them retrospectively, and note gaps as findings in their own right.
Interview people who perform and supervise the work, preferably using open prompts. Ask what normally happens, what happened on the day, what makes the task difficult and what workaround staff use when the intended process cannot be followed. A technologist may reveal that a form is unavailable at the point of use, while a courier coordinator may explain why specimens from a rural clinic arrive outside the expected transport window.
Avoid framing the conversation around individual fault. In a public laboratory, workload can change sharply during an outbreak investigation, staff may rotate between platforms, and a replacement reagent may arrive through a different supplier. These conditions are part of the system. The investigation should still document any failure to follow a procedure, but it should examine why the procedure was not followed or was impractical.
Map The Process And Its Control Points
Create a simple process map from request and collection through transport, accessioning, preparation, testing, result review, reporting and storage or disposal. Mark each point where an error could be detected or prevented. This visual record often exposes handovers that are absent from the formal procedure.
For recurring bacterial contamination, map cleaning, media preparation, cabinet use, workflow direction, environmental monitoring and waste handling. The practical guidance on contamination monitoring can help a TB service consider how monitoring data should be collected and interpreted rather than treating contamination as an unexplained laboratory event.
Then compare the written procedure with actual practice. Look for duplicate data entry, unclear acceptance criteria, shared equipment, cramped workspace, interruptions and steps that depend on memory. In Australian services, a metropolitan reference laboratory may process material from several hospitals, while a regional site may rely on scheduled transport. The same procedure may therefore have different failure points at each location.
Analyse Contributing Factors Systematically
Use a small set of complementary tools rather than selecting a fashionable technique. A “five whys” discussion is useful for a narrow event, provided each answer is supported by evidence. A fishbone diagram can group causes under personnel, methods, equipment, materials, environment, measurement and management. A timeline is valuable when delays or handovers are involved.
For example, repeated missed temperature checks may initially appear to be a staff compliance issue. Further analysis could show that the monitoring sheet is kept in another room, the alarm is not visible from the work area, the relief staff member was not trained, and the procedure does not state who checks the unit after a public holiday. The root causes are then access, role clarity, training and process design.
Rank causes by their relationship to the event, their frequency and the strength of available evidence. Distinguish a root cause from a contributing factor and from a symptom. “The result was released late” is an outcome; “the review queue had no escalation rule for unverified results” is a potential process cause. When evidence is insufficient, record the uncertainty and plan how it will be tested.
Connect Findings To Quality Systems Essentials
Root cause analysis should sit within the laboratory’s broader quality management system, not operate as a separate investigation file. Link each finding to the relevant Quality Systems Essential, such as organisation, personnel, equipment, purchasing and inventory, process management, documents and records, assessment or continual improvement.
This connection makes patterns easier to see. Repeated failed instrument checks may involve equipment maintenance, competency assessment, inventory control and document management at the same time. A cluster of specimen rejections may point to collection instructions, transport contracts, communication with requesting services and the laboratory’s acceptance policy.
The GLI Quality Tool provides phase-based checklists and practical materials that can support this review; laboratories can access the quality management resources while aligning investigations with their current level of system maturity. A small regional service may begin with a paper-based RCA record, while a large state network may integrate findings into an electronic corrective action register.
Design Corrective And Preventive Actions
Actions should address the verified cause and include an owner, due date, required resources and success measure. “Retrain staff” is rarely sufficient on its own. A stronger action could revise the specimen acceptance form, place the form beside the receiving bench, introduce barcode verification, observe competency during the next two shifts and review rejected specimens weekly for three months.
Use the hierarchy of controls where safety or contamination is involved. Change the process or workspace before relying on reminders. Examples include separating clean and contaminated traffic, introducing closed waste containers, using an automated temperature alarm, standardising reagent storage or limiting access to an authorised testing area. Personal protective equipment remains important, but it should not be the only response to a preventable system weakness.
Consider Australian operational constraints when selecting actions. Work health and safety requirements apply under the relevant state or territory legislation, and a control must be practical for the local facility. A remote laboratory may need a validated contingency process for delayed deliveries, while a Sydney or Melbourne service may need stronger controls for high-volume accessioning and multiple courier routes. Procurement decisions should account for local supplier availability, service contracts and replacement lead times.
Verify That The Problem Has Stayed Fixed
Closing an action because a procedure was amended does not demonstrate effectiveness. Define a verification plan before implementation. Choose a measure tied directly to the failure, such as the rate of rejected specimens, contamination episodes per batch, overdue maintenance tasks, transcription errors or days to final result.
Review the measure over a suitable period and compare it with the baseline. Include enough work cycles to cover different shifts, staff groups and predictable pressure points. If an intervention is introduced in winter, assess it during a busier period as well; Australian laboratories may experience changing respiratory demand, annual leave patterns and transport disruptions across the year.
Use audits, direct observation, trend charts, proficiency testing, incident reports and staff feedback together. If the same non-conformity returns, reopen the analysis rather than simply extending the due date. The first corrective action may have addressed a visible symptom, or a new cause may have appeared through a change in instrument, supplier, staffing or workflow.
Record lessons in a way that supports continual improvement. Share the outcome with affected staff, update controlled documents, revise training where necessary and feed significant trends into management review. The objective is a more reliable process that can withstand routine variation, not a file that merely demonstrates that an investigation occurred.
A repeatable RCA method gives TB laboratories a disciplined way to turn incidents into system knowledge. Start with one recurring non-conformity, document the facts, map the process, test the causes and measure whether the control works. Build the approach into incident management, internal audit and corrective action review so that learning becomes part of everyday laboratory practice.