Corrective Actions for Failed Proficiency Testing in Microbiology

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-11-10

The most likely cause of a failed microbiology proficiency test is a break in specimen handling, test timing, or result transcription, and the first thing to check is the raw worksheet, the sample receipt time, and the exact result entered into the proficiency testing reporting form. Proficiency testing, or PT, is an external quality assessment in which a provider sends blind specimens to a laboratory and compares the laboratory's results with an assigned or consensus result. A PT failure is an unacceptable result on that external challenge. It is not a diagnosis of the laboratory, and it is not proof that every patient result is wrong. It is a signal that requires a documented investigation.

Corrective actions for failed proficiency testing in microbiology follow a disciplined sequence. Confirm the failure. Secure the raw data. Reconstruct the specimen timeline. Identify the immediate cause, the root cause, and contributing factors. Correct the defect. Verify that the correction worked. Change the process so the same failure does not return. A seven-year analysis of 19 food microbiology PT schemes involving 48 laboratories found that laboratories failed to detect or enumerate foodborne pathogens in 3.0% of tests. The same study identified causes in 74% of outliers, with preanalytical handling and timing, analytical method and counting errors, and postanalytical calculation or encoding errors among the main sources [1]. That pattern is the practical starting point for most microbiology PT investigations.

What a PT Failure Really Means

A PT failure means the laboratory's reported result fell outside the scheme's acceptance criteria. Those criteria vary by analyte, method, sample, and provider. In a SARS-CoV-2 whole-genome sequencing PT program, participants were assessed on genome coverage, Pango lineage, and sequence quality. Reassessing the 2021 results with the stricter 2022 metrics increased the number of participants who would have failed and increased the failure rate per sample [2]. The lesson is direct. A failure is defined by the scoring rules in force, so read the PT report and the provider instructions before drawing conclusions about the laboratory's performance.

A PT event can fail for a laboratory reason, a method reason, or a scheme reason. A laboratory reason is a defect in specimen handling, testing, interpretation, calculation, or reporting. A method reason is a limitation of the assay, such as a limit of detection above the panel concentration, a primer or probe mismatch, a narrow identification database, or a confirmation rule that does not fit the organism. A scheme reason is a problem with the panel, the assigned result, the homogeneity or stability of the material, or the scoring metric. Do not assume the panel is wrong because the result is inconvenient. Do not assume the laboratory is wrong before checking the panel and the reporting record.

A useful distinction is between a failed PT result and a failed PT investigation. The first is an unacceptable result on one challenge. The second is a corrective action that stops at retraining or a rewritten result without finding the process defect. The goal is to close the second gap. A PT failure is a free look at a process that usually remains invisible during routine work. The same study that identified causes in 74% of outliers concluded that regular PT participation, followed by feedback and appropriate corrective action, supports improvement [1].

First Response: The 60-Minute Triage

Treat the PT report like a clinical incident report. Do not alter the original worksheet, instrument file, photograph, or electronic record. Make a secure copy of the raw data. Pull the PT package, the provider instructions, the sample receipt log, the accession record, the worksheet, the instrument printout or file, the reagent lot records, the quality control records, the incubator and refrigerator temperature logs, the technologist's competency record, and the final report. If the result was entered into a laboratory information system or a PT portal, capture the exact entry and the date and time of entry.

Confirm the failure against the provider's criteria. Check whether the unacceptable result applies to one sample or multiple samples, one analyte or multiple analytes, one method or all methods, and one technologist or all technologists. A single-sample failure on a low-level challenge suggests a limit of detection or sample handling problem. A failure across multiple samples and methods suggests a reporting, calculation, or specimen mix-up problem. A failure limited to one method suggests method design, validation, or interpretation. A failure limited to one person suggests technique, reading, or data entry.

Safety note: handle PT materials as infectious clinical specimens. Use the laboratory's biosafety procedures, wear gloves and eye protection, work in a biosafety cabinet when aerosol or splash risks exist, and decontaminate benches after handling. Staining reagents, acids, alcohols, and molecular lysis buffers carry chemical hazards, so review the safety data sheet and follow local waste rules.

Do not immediately retest the PT sample unless the provider permits it and the material is stable. A retest can answer a narrow technical question, such as whether a result was a one-time pipetting error. It does not replace an investigation. Do not send the PT sample to another laboratory as the primary corrective action. A reference result can help with identification, but the laboratory still needs to find why its own process failed.

Troubleshooting Table for Failed Microbiology PT

Work from the Most Common Cause to the Rarest

Preanalytical Handling and Timing

Preanalytical errors are the first place to look because they are common, easy to miss, and often invisible in the final result. The PT sample can be delayed at receipt, stored at the wrong temperature, opened too early, exposed to light, thawed repeatedly, or tested after the provider's stability window. A sample can be mislabeled or mixed up with another sample. A worksheet can carry the wrong accession number. The result then looks like an analytical failure when the specimen itself was compromised.

The food microbiology PT analysis found handling and timing among the main preanalytical causes of outliers [1]. A centralized malaria service monitored preanalytic and analytic turnaround times as part of its performance review and reported that more than 90% of community blood specimens arrived within 2 hours of collection [3]. That does not set a universal transport limit for every microbiology specimen, but it shows the principle. Time and temperature are test-specific requirements, and the laboratory must be able to prove that the PT sample met them.

The diagnostic questions are simple. When did the sample arrive? What was the temperature at receipt? Was the package intact? When was it stored, and at what temperature? When was it opened? When did testing start? Was the sample tested with the same workflow as a patient sample? If the PT sample was handled as a special project, the investigation must say so. A special handling path is a process defect because it does not represent routine performance.

Analytical Method and Technique

Analytical causes include an unsuitable method, a method used outside its validated scope, a confirmation step that was skipped, a counting rule applied incorrectly, a culture read too early, a selective medium that suppressed the target, an identification database that did not include the organism, or a susceptibility panel read with the wrong inoculum. The same PT analysis grouped analytical causes under unsuitable methods, sample confusion, colony counting errors, and confirmation errors [1]. That list is a practical checklist.

Method suitability is not a vague concept. A method has a validated limit of detection, a validated specificity, a validated range, and a validated set of acceptable specimen types. If the PT challenge falls below the limit of detection, a negative result can be a correct result for that method and still be unacceptable to the scheme. The corrective action is not always retraining. It may be a method change, a more sensitive confirmation step, a referral pathway, or a documented limitation with an alternative test.

Postanalytical and Reporting Errors

Postanalytical errors include calculation errors, unit errors, encoding errors, transcription errors, wrong sample entry, and result entry into the wrong field. These are common because they occur after the bench work looks correct. A technologist can perform the test perfectly and still fail PT by entering a result in the wrong units or by transposing digits. The food microbiology PT analysis identified calculation and encoding of results among the postanalytical mistakes [1].

The first check is a direct comparison of the raw worksheet, the instrument output, the calculated result, the laboratory information system entry, and the PT portal entry. Photograph or copy each step. If the result is correct on the worksheet but wrong in the portal, the root cause is in reporting. If the result is wrong on the worksheet, the root cause is upstream. If the calculation is wrong, check the formula, the dilution factor, the conversion factor, and the significant figures.

Quality Control, Reagents, and Instruments

Quality control failures can cause PT failures even when the technologist follows the procedure. An expired reagent, a contaminated lot, a drifted instrument, a miscalibrated pipette, a failed temperature controller, or a compromised media batch can change the result. Internal quality control is daily evidence that the test system is performing as expected. PT is external evidence that the laboratory's result agrees with other laboratories or with a reference result.

Review the quality control records for the PT test date and the surrounding days. Check positive and negative controls, sterility checks, media performance checks, instrument calibration, temperature logs, and maintenance records. If the control was out of range and the test was repeated, the review must show what was done and whether the PT result came from a valid run. If controls were acceptable but the PT sample failed, the defect may be sample-specific, method-specific, or interpretation-specific rather than a general system failure.

Personnel Competency and Reading Errors

Competency problems are real, but they are often overused as an explanation. A reading error can come from a difficult colony, a mixed plate, a low-positive slide, a faint band, or a subtle biochemical reaction. In a Mexican tuberculosis microscopy network, 196 of 430 technicians scored below 80% in 1998. After intensive training, their mean score rose from 65% to 90%, and the mean was 83% two years later [4]. The same study found that the number of low-positive slides in the test affected scores and that laboratories whose work was rechecked had better scores. Training helped, but the challenge design and the rechecking system also mattered.

Use competency review when the failure is linked to interpretation, reading, or technique. Observe the technologist performing the same task. Review a blinded set of previously validated slides, plates, or sequences. Check whether the technologist used the current procedure and the current reading guide. If the failure is a knowledge gap, retraining and documented competency assessment are appropriate. If the failure is a process gap, retraining alone will not fix it.

Molecular and Whole-Genome Sequencing Failures

Molecular methods add failure modes that culture-based laboratories may not monitor well. Extraction can fail. Inhibitors can suppress amplification. A primer or probe can have mismatches against a variant. A low viral load can fall below the limit of detection. A sequence can have low coverage, poor quality scores, contamination, or a bioinformatics pipeline threshold that removes the true signal. A reference database can be outdated. A lineage assignment can be wrong because the software version or the Pango nomenclature changed.

An HIV-1 genotyping external quality assessment found high concordance for single-variant samples, but three laboratories failed to detect minor variants that made up one-half of the sample. Those three laboratories used a common primer that had four internal mismatches to the minor template [5]. That is a method design failure, not a careless bench error. The corrective action is to review primer and probe design, verify the limit of detection for minority variants, and use an alternate assay when the clinical question requires it.

Whole-genome sequencing in routine microbiology requires managed validation, quality control, and PT. The practical implementation review emphasized that laboratories must adhere to accreditation requirements and that PT is part of quality control for WGS [6]. A SARS-CoV-2 WGS PT program used genome coverage, Pango lineage, and sequence quality as assessment metrics and showed that stricter metrics changed the apparent failure rate [2]. For a sequencing failure, review the library preparation, run metrics, coverage, quality scores, contamination controls, database version, pipeline version, and the exact lineage assignment. A retest without a pipeline review may reproduce the same error.

Panel, Scheme, and Scoring Problems

Panel and scheme problems are rare, but they must be considered. A panel member can be nonviable, contaminated, mislabeled, or unstable. The assigned result can be wrong. The acceptance criteria can be too tight for the method or the sample. A scoring change can make a previously acceptable result look unacceptable. In a platelet bacterial screening PT program, apheresis platelets were spiked with 10^1 and 10^2 colony forming units per milliliter of Staphylococcus epidermidis, Pseudomonas aeruginosa, and Streptococcus pneumoniae. Streptococcus pneumoniae did not grow consistently and was not selected for the implemented program [7]. That is an example of panel validation mattering. If a PT organism does not behave reliably in the panel matrix, the laboratory's result may be a true reflection of a panel problem.

Do not use a panel problem as a default explanation. Use it only when the evidence supports it. Ask the provider for the panel validation data, homogeneity data, stability data, and assigned-result rationale. Compare your result with the provider's expected result and with the results of peer laboratories using the same method. If the provider confirms a panel issue, document the communication and the resolution. If the provider confirms the panel was valid, return to the laboratory process.

A Step-by-Step Diagnostic Sequence

Step 1. Confirm the failure and define its scope. Read the PT report and the provider instructions. Record the sample, analyte, expected result, reported result, acceptance criteria, and score. Determine whether the failure is isolated or part of a pattern.

Step 2. Secure the raw data. Copy the worksheet, instrument output, photographs, quality control records, reagent lot records, temperature logs, and reporting records. Do not overwrite or discard original data.

Step 3. Reconstruct the specimen timeline. Record receipt, storage, opening, testing, reading, calculation, and reporting times. Compare each step with the procedure and the provider's instructions.

Step 4. Review the test system. Check media, reagents, controls, calibration, instrument maintenance, incubation, atmosphere, and pipette calibration. Confirm that the method was used within its validated scope.

Step 5. Review the interpretation and reporting steps. Compare the raw result with the calculated result, the laboratory information system entry, and the PT portal entry. Check units, significant figures, dilution factors, and sample identifiers.

Step 6. Compare with peer methods and reference information. Ask the provider for method-specific performance summaries when available. Use a reference laboratory or an orthogonal method for identification or confirmation when the result is uncertain.

Step 7. Identify the immediate cause, the root cause, and contributing factors. The immediate cause is what went wrong at the bench or in the report. The root cause is why the process allowed it. Contributing factors are conditions that made it more likely, such as understaffing, a temporary instrument, or an outdated procedure.

Step 8. Implement containment and correction. If patient results may be affected, perform a risk assessment and review results since the last acceptable PT event. Correct the immediate defect. Replace reagents, repair instruments, retrain staff, or revise the procedure as indicated.

Step 9. Verify effectiveness. Use a repeat PT challenge if available, a blinded internal challenge panel, a method comparison, a competency observation, or an audit. Define what evidence will prove the action worked.

Step 10. Document and close. Record the problem, investigation, root cause, action, owner, due date, verification evidence, and closure date. Link the record to the affected procedure, training record, and quality control data.

Corrective Action Plan: Contain, Correct, Prevent, Verify

A corrective action plan has four parts. Containment limits the immediate harm. Correction fixes the defect. Prevention removes the root cause. Verification proves the fix worked.

Containment for a microbiology PT failure may include holding patient reports that depend on the affected method, reviewing patient results since the last acceptable PT event, quarantining a reagent lot, or stopping a test until the investigation is complete. The laboratory director should decide whether patient impact assessment is needed. Not every PT failure has patient impact, but the decision must be documented.

Correction is the immediate fix. Examples include recalculating a result, repeating a failed control, replacing a contaminated reagent, recalibrating a pipette, correcting a reporting rule, or repeating identification with a pure culture.

Prevention is the process change. Examples include a second-person review for calculations, a mandatory purity check before identification, a media lot verification step, a primer design review before a new molecular assay is released, a sequencing run quality gate, a competency schedule, or a revised sample receipt checklist. Prevention should target the root cause, not the person.

Verification is the evidence that the action worked. A repeat PT result, an internal blinded panel, a proficiency challenge, a chart audit, or a direct observation can serve as verification. If the verification fails, reopen the investigation. Do not close the action because the paperwork is complete.

Corrective Actions by Failure Pattern

False-Negative Culture Results

A false negative means the laboratory did not detect an organism that was present. Start with specimen viability and timing. Review transport, storage, and testing delay. Check the medium, incubation temperature, atmosphere, and incubation time. Review the reading schedule. If the organism is fastidious or slow-growing, check whether the method's limit of detection and recovery are adequate. Subculture residual material if it is available and safe. Use an orthogonal method for confirmation when the organism is difficult to recover. The corrective action may be a method change, a longer incubation, a different medium, or a referral pathway.

False-Positive and Contaminated Results

A false positive means the laboratory reported an organism or signal that was not present. Look for contamination before looking for a panel problem. Review negative controls, reagent sterility, work flow, environmental cleaning, pipetting technique, and instrument carryover. Check whether the result came from a mixed culture, a contaminated slide, or a non-specific signal. Map the contamination by sampling surfaces, reagents, and equipment. Correct the work flow, replace the affected reagent, clean the instrument, and retrain aseptic technique. Repeat the challenge with negative controls.

Quantitative Count Errors

Count errors can come from the dilution scheme, the plating volume, the counting rule, or the calculation. Recalculate from the raw counts. Check the dilution factor, the volume plated, and the conversion to colony forming units per milliliter or per gram. Review the countable range and the rule for merged colonies. Compare duplicate plates. Verify pipette calibration. If an automated counter is used, verify it against manual counts. Retrain counting and calculation, then use a blinded set of plates for competency assessment.

Identification and Susceptibility Errors

Identification errors often begin with an impure culture. Review the Gram stain, the purity plate, and the colony morphology. Check the identification method and the database version. Use an orthogonal method when the score is low or the organism is unusual. For susceptibility testing, review the inoculum preparation, panel lot, incubation, reading, and breakpoint table. A wrong breakpoint can turn a susceptible result into a resistant result without any bench error. Update the breakpoint table and retrain reading.

Molecular Detection and Sequencing Errors

For a molecular detection failure, review the extraction control, internal control, amplification curve, cycle threshold, primer and probe design, and the validated limit of detection. Check for inhibitors and for a low target level. Repeat the extraction if material remains. Use an alternate target or a different sample type when appropriate. For sequencing, review coverage, quality scores, read counts, contamination controls, reference database version, and pipeline thresholds. Validate the pipeline before use and revalidate after a major update. A lineage assignment error may require a review of the nomenclature version and the assignment software.

Calculation, Encoding, and Reporting Errors

These errors are often the easiest to fix and the easiest to prevent. Compare the raw result with the calculated result and the final report. Check units, significant figures, dilution factors, and sample identifiers. Add a second-person review for manual calculations. Build auto-verification rules with delta checks and range checks. Use barcode scanning for sample entry. Audit the PT portal entry against the final report before submission.

Root Cause Analysis That Finds the Real Defect

Root cause analysis is a structured method for finding why a process failed. The five-why method asks why repeatedly until the answer reaches a process condition. A fishbone diagram groups possible causes into specimen, method, reagent, instrument, personnel, environment, and reporting categories. Both methods are useful, and neither requires blaming an individual.

Start with the event statement. For example, the laboratory reported a negative result for an organism that was present in the PT sample. Ask why. The answer may be that the organism was not recovered on the selected medium. Ask why. The medium may have been inhibitory to a stressed organism. Ask why. The medium was selected for a different specimen type and was not validated for the PT matrix. Ask why. The method validation did not include that matrix. The root cause is a validation gap, not a careless technologist.

Use evidence, not memory. Worksheets, timestamps, lot records, temperature logs, quality control records, and competency records are stronger than recollections. Interview staff without blame. Ask what made the task difficult. Look for work-arounds, ambiguous instructions, and competing priorities. A corrective action that ignores the work system will fail.

Microbiology Quality Control and Laboratory Quality Management

Microbiology quality control is the daily internal monitoring of the test system. It includes positive and negative controls, media sterility and performance checks, organism viability checks, instrument calibration, temperature monitoring, and molecular extraction and inhibition controls. PT is the external check that compares the laboratory's performance with other laboratories or with a reference result. Both belong to laboratory quality management. The practical implementation review of whole-genome sequencing in diagnostic microbiology placed PT within the quality control framework needed for accreditation and safe routine use [6].

Quality management connects PT results to procedure review, training, competency assessment, equipment maintenance, supplier qualification, incident reporting, and management review. A PT failure should generate a record that is visible to the quality manager and the laboratory director. The record should show the investigation, the root cause, the action, the verification, and the closure. If the same failure appears in a later PT event, the earlier record should be reopened. Repeat failures are a sign that the corrective action did not address the root cause.

Prevention: Build a Workflow That Withstands PT

Prevention begins with treating PT samples as routine samples. Do not create a special PT path that only one expert uses. Do not test PT samples at a different time of day, with a different method, or with a different review process. Use the same accessioning, handling, testing, reading, calculation, and reporting steps that apply to patient specimens.

Assign a PT coordinator who tracks schedules, receives the package, logs the samples, and confirms the testing plan. Rotate staff through PT testing so the competency check is meaningful. Review the provider instructions before testing. Confirm the sample type, storage, stability, test method, reporting units, and deadline. Check the PT portal entry against the final report before submission.

Maintain method validation and verification records. Know the limit of detection, the specificity, the range, and the acceptable specimen types for each method. Review primer and probe design before introducing a molecular assay. Validate the bioinformatics pipeline and the reference database. Use quality gates for sequencing runs. Verify media, reagents, and controls before use. Keep temperature logs and equipment maintenance records current.

Trend PT results over time. Track failures by method, sample type, technologist, reagent lot, and failure category. A single failure is an event. A cluster is a system problem. Use the trend to prioritize training, method review, or equipment replacement.

Common Mistakes and Limitations

The first mistake is blaming the PT provider before checking the laboratory's own process. The second is retesting the PT sample and treating a passing retest as a closed investigation. The third is correcting only the reported result without correcting the process that produced it. The fourth is retraining a technologist when the root cause is a bad reagent, an unsuitable method, or a reporting rule. The fifth is assuming internal quality control will catch every PT problem. Quality control samples are not identical to PT samples, and a method can pass daily controls while failing a specific challenge matrix.

Another mistake is ignoring the PT instructions. Storage, stability, and reporting requirements matter. A correct result reported in the wrong units can fail. A correct result entered into the wrong field can fail. A result reported after the deadline can fail. Read the instructions and verify the entry.

A common limitation is that PT samples are not patient samples. They may be spiked, freeze-dried, or stabilized in a matrix that behaves differently from a clinical specimen. A PT failure can reveal a real process defect, but a PT pass does not prove that every patient result is correct. PT is a sample of performance, not a complete validation. A laboratory director should assess patient impact when a failure suggests a systematic error. Individual patient decisions need clinical review.

Frequently Asked Questions

What is the first corrective action after a failed microbiology PT?

Secure the raw data and confirm the failure before changing the method. Pull the PT report, worksheet, instrument log, reagent lot, and reporting record, then compare what the scheme expected with what the laboratory actually did.

Does a failed PT always mean the laboratory made an error?

No. A failure can reflect a laboratory error, a method limitation, or a PT panel or scoring problem. Confirm the result, review panel viability and instructions, and investigate the laboratory process before assigning blame.

How do I investigate a false-negative PT result?

Start with specimen timing, storage, and viability, then check media, incubation, atmosphere, and reading. Subculture any residual material, review confirmation criteria, and compare the result with the method's validated limit of detection.

How do I investigate a false-positive PT result?

Look for contamination, carryover, and cross-reaction before assuming the panel was wrong. Review negative controls, work flow, reagent lots, instrument cleaning, and colony or signal confirmation.

Should I retest the PT sample after a failure?

Retest only when the PT provider permits it and the original material is stable. A retest can clarify a technical problem, but it cannot replace a root cause analysis or an effectiveness check.

How long should corrective action records be kept?

Follow the laboratory's accreditation, legal, and institutional record retention rules. Keep the investigation, root cause, action, verification, and closure records together so an assessor can reconstruct the event.

What is the difference between correction and corrective action?

Correction fixes the immediate problem, such as a calculation or a mislabeled report. Corrective action removes the underlying cause so the same failure does not recur.

Can training alone fix a PT failure?

Training alone works only when the root cause is a knowledge or technique gap. If the cause is a bad reagent, an unsuitable method, poor sample handling, or a reporting rule, training will not prevent recurrence.

Sources

1. Analytical performances of food microbiology laboratories - critical analysis of 7 years of proficiency testing results.
2. Continued improvement in the development of the SARS-CoV-2 whole genome sequencing proficiency testing program.
3. A regional centralized microbiology service in Calgary for the rapid diagnosis of malaria.
4. Implementation of proficiency testing in conjunction with a rechecking system for external quality assurance in tuberculosis laboratories in Mexico.
5. Japanese external quality assessment program to standardize HIV-1 drug-resistance testing (JEQS2010 program) using in vitro transcribed RNA as reference material.
6. Practical issues in implementing whole-genome-sequencing in routine diagnostic microbiology.
7. Implementation of a proficiency testing programme for bacterial screening in platelet preparations in Canada.