# Building a QC Data Management System A quality control (QC) [data management](/blog/guides/data-management-basics-principles-processes-and-best-practices) system is the organized structure a microbiology laboratory uses to record, store, review, and act on every quality control result it generates. It covers media QC, control organism performance, antimicrobial susceptibility testing (AST) controls, staining reagents, rapid biochemical identification panels, and molecular assay controls. The system turns isolated QC observations into a continuous, reviewable record that shows whether the laboratory's methods performed as intended on every day that patient results were reported. This guide walks through building such a system from a blank page. The work suits any clinical, public health, veterinary, or research microbiology laboratory that reports patient, animal, or environmental results and must demonstrate that those results are reliable. A first-time build takes roughly 20 to 30 hours of hands-on work spread over 4 to 6 weeks of elapsed time. Daily operation afterward takes 15 to 30 minutes of data entry and review per bench. The biggest single determinant of success is not software. It is deciding, before the first result is entered, exactly what "acceptable" means for every control the laboratory runs, and then recording every deviation against that definition. The payoff is concrete. When an inspector, an accreditor, or a clinician asks how you know last Tuesday's blood cultures were valid, the system answers with a record instead of a memory. When a lot of media underperforms, the system shows when the drift started and which patient results are affected. When a new molecular workflow is introduced, the system proves its controls worked before results were released. Laboratories that skip this structure accumulate exactly the gaps documented in a survey of 124 Ontario microbiology laboratories, where media QC records showed marked variability and inadequacy despite a mandatory reporting program [1]. ## The Principle Behind QC Data Management Quality control in microbiology rests on a simple premise. You cannot verify a patient result directly, so you verify the process instead by running known materials through the same steps the patient specimen will take. If the known materials behave within defined limits, the process performed as designed, and the patient results from that run are trustworthy. If they do not, the process is out of control, and results must be held until the problem is found and fixed. A QC data management system exists because single QC results say very little on their own. A control organism growing on chocolate agar today tells you the medium worked today. A twelve-month record of that same control tells you whether performance is drifting, whether a new lot behaves differently from the old one, and whether a failure on one bench is isolated or part of a pattern across the laboratory. The system is therefore built around three functions: capture, which records every QC event in a uniform structure, evaluation, which compares each result to pre-set acceptance criteria, and trend review, which examines results over time rather than one at a time. The same principle applies across method types. In a metagenomic next-generation sequencing (mNGS) workflow for respiratory viruses, three controls run through the entire process: a no-template control to detect contamination introduced during processing, an internal control added to each sample to check extraction efficiency and inhibitors, and an external control to verify reagent and equipment integrity. In that published workflow, 34 of 37 clinical respiratory samples met the QC validation criteria, and viral genotypes were correctly determined for all 36 viruses detected by comparator PCR, which shows how well-defined controls let a laboratory defend every reported result [2]. One caution shapes everything that follows. The Ontario survey found that even laboratories participating in a formal external quality program produced QC records that were incomplete, inconsistent, and in many cases absent for media that needed them most [1]. Building the system is not the hard part. Keeping it complete every day, including on busy days, is. ## Materials and Software The table below lists what you need before the first QC result is entered. Concentrations and specifications shown are standard bench practice for [clinical microbiology](/knowledge/diagnostics/microbiology/clinical-microbiology-from-specimen-collection-to-pathogen-identification). | Item | Specification or working concentration | Purpose | |--|--|--| | Data platform | Spreadsheet program (Microsoft Excel or equivalent) or a laboratory information system (LIMS) module | Primary repository for all QC records | | QC organisms | Reference strains such as *Escherichia coli* ATCC 25922, *Staphylococcus aureus* ATCC 25923, and *Pseudomonas aeruginosa* ATCC 27853 | Verify growth, selectivity, and AST performance | | Specialized QC strains | *Haemophilus influenzae* ATCC 10211 for chocolate agar, *Neisseria gonorrhoeae* ATCC 49226 for selective gonococcal media, *Campylobacter jejuni* ATCC 33291 for Campylobacter media | Verify fastidious growth and selectivity of enriched and selective media | | Inoculum standard | 0.5 McFarland standard (approximately 1 x 10^8 CFU/mL for *E. coli*) | Standardize inoculum density for AST and media QC | | Maintenance medium | Tryptic soy broth or similar with 10 to 15 percent glycerol for frozen stocks | Preserve QC organism working stocks | | Lot documentation | Certificates of performance or analysis from media and reagent manufacturers | Anchor each QC result to a specific lot number | | Out-of-range log | Structured form (paper or electronic) for documenting deviations and corrective action | Close the loop when QC fails | | Review calendar | Defined weekly, monthly, and annual review dates with named reviewers | Ensure trends are actually examined | | Backup storage | Secondary location for the QC database, checked monthly | Prevent loss of the record itself | Choose the data platform deliberately. A spreadsheet is adequate for a laboratory running a few dozen QC events per week. A LIMS becomes necessary when multiple benches enter data concurrently, when results must be linked automatically to patient result release, or when the laboratory must demonstrate an audit trail showing who changed what and when. Whatever you choose, the platform must support three non-negotiable features: a fixed field structure so records are comparable across time, restricted editing so historical entries cannot be silently altered, and regular backup. ## Building the System Step by Step ![Flow of steps for building a microbiology QC data system, from inventorying tests to training every user.](/article-images/illustrations/microbiology-qc-data-management-build-steps.webp) The following procedure takes a laboratory from nothing to a functioning system. Each step lists the reason it matters, because skipping steps is how incomplete records like those in the Ontario survey arise [1]. **Step 1. Inventory every test and every QC event.** List each test the laboratory performs and, for each, every control it requires: media lots, staining reagents, biochemical reagents, antisera, AST control organisms, molecular controls, and instrument checks. Record how often each control runs. The reason for this step is scope. A system built around blood agar but blind to the syphilis serology reagent on the serology bench is not a system, it is a folder. **Step 2. Define acceptance criteria for every QC item before collecting data.** For each control, write down what "acceptable" means. For qualitative controls, this is a defined growth or reaction pattern, for example *S. aureus* ATCC 25923 showing beta-hemolysis on blood agar or *H. influenzae* ATCC 10211 growing on chocolate agar after the specified incubation time and temperature. For quantitative controls such as AST zone diameters, use the published ranges for the reference strain and method. For molecular controls, define both a detection threshold and a maximum background signal for the no-template control, mirroring the validation logic used in published mNGS workflows where the internal control spiking level was set high enough to be detected but low enough to avoid consuming sequencing reads [2]. The reason for this step is simple. A result without a pre-defined criterion cannot fail, and a QC program in which nothing can fail is decoration. **Step 3. Choose and configure the data platform.** Set up one table (or LIMS module) per bench or test category, with identical column structure across tables: date, test or medium name, lot number, expiration date, control organism and lot, incubation time and temperature, observed result, acceptance criterion, pass or fail code, analyst initials, and reviewer initials. The reason for identical structure across benches is aggregation. When every record has the same fields, the laboratory director can review all media QC in one pass, and a software transition years later does not orphan old records. **Step 4. Build lot and expiration tracking.** Create a register of every consumable lot in use, with receipt date, opening date, expiration date, and the date QC was completed on that lot. New lots of media and reagents must be qualified with control organisms before the old lot is retired, and the parallel-testing results belong in the same record. The reason is traceability. If a lot of MacConkey agar later proves defective, the register tells you within minutes which plates were poured from it and which specimens were processed on them. **Step 5. Set up the QC organism inventory.** Record each control strain, its source, the lot or passage in use, storage conditions, and the date of the last subculture. Maintain a working stock and a reserve stock so a contaminated working stock never leaves the laboratory without a control organism. Standard practice stores frozen stocks in glycerol broth at 10 to 15 percent glycerol at minus 20 degrees Celsius or colder, with working agar slants replaced on a defined schedule. The reason is biological stability. Repeated subculture selects for variants, and a QC strain that has drifted no longer verifies anything. **Step 6. Enter results daily, with standardized codes.** Use a fixed vocabulary for results. A practical three-level code is S (satisfactory), A (acceptable with comment), and F (failed). Record actual observations, not just checkmarks, for example "growth at 24 hours, oxidase positive, zone 32 mm" rather than "OK." The reason is that trends hide inside unrecorded detail. A zone diameter drifting from 30 mm to 26 mm over six weeks is visible only if the numbers were recorded. **Step 7. Apply automated checks.** In a spreadsheet, use conditional formatting to highlight any entry outside its acceptance range and any lot past expiration. In a LIMS, configure the same logic as system rules that block patient result release for the affected test until QC passes. The reason is human attention. Reviewers reliably catch highlighted cells and reliably miss a normal-looking row that is quietly out of range. **Step 8. Define the QC failure response before the first failure.** Write a one-page procedure that answers three questions. Which patient results are held when this control fails. Who is authorized to troubleshoot and release. What documentation the corrective action requires. The reason is that decisions made during a failure are rushed decisions. A pre-defined pathway keeps the bench from reporting results on an unverified process under time pressure. **Step 9. Build the trend review layer.** Set up a weekly review in which the senior technologist scans the previous week's entries for anything marked A or F and for visually obvious drift. Set up a monthly review with a chart for each quantitative control, plotting zone diameters or quantitative molecular control values across time, in the manner of a Levey-Jennings chart used throughout laboratory medicine. Standard practice applies rule-based checks such as rejecting a run when a single point falls beyond three standard deviations or when two consecutive points fall beyond two standard deviations on the same side of the mean. The reason is that control charts answer a different question than pass-or-fail does. A control can pass every single day and still be drifting steadily toward failure, and only the chart shows it. **Step 10. Create the deviation and corrective action log.** Every A and F entry gets a corresponding log entry with a unique number, the investigation, the root cause when identified, the corrective action taken, and the date results were released or repeated. The reason is closure. An open deviation is an unresolved question about the validity of patient results, and it should sit in a visible queue until it is answered. **Step 11. Set the review and sign-off hierarchy.** Define who reviews daily, who reviews monthly, and who performs the annual QC program audit, with signature or electronic sign-off required at each level. The reason is accountability and continuity. When a reviewer leaves, the signature record shows exactly where the next reviewer must pick up. **Step 12. Control access and preserve history.** Restrict editing of completed entries to the QC supervisor, require a reason field for any post-entry correction, and keep the original entry visible rather than overwritten. The reason is that a QC record that can be silently edited is not evidence of anything, and accreditors treat unexplained corrections as red flags. **Step 13. Back up and test the backup.** Schedule automatic backup to a second location and restore the database from backup once a month on a test copy. The reason is the obvious one, and the test matters more than the schedule, because a backup that has never been restored is a hypothesis. **Step 14. Validate the system and train every user.** Before go-live, enter a month of simulated data, run the automated checks, generate the trend charts, and confirm the failure pathway works. Train all bench staff on entry conventions and require a supervised first week of entries. The reason is consistency. A system with three different entry styles per bench cannot be trended, and inconsistent records were among the defining defects found in the Ontario laboratory survey [1]. ## Core Controls and What to Record ### Media quality control Media QC verifies that each medium supports the organisms it must recover and inhibits those it must suppress. Record the control organism, the lot number, the incubation time and temperature, the inoculum preparation, and the observed growth or inhibition. Typical acceptance criteria for common media, which should be confirmed against the current CLSI media QC standard that descends from the NCCLS M22-A2 document used in the Ontario survey [1], include the following. | Medium | Typical control strain(s) | What to verify | |--|--|--| | Blood agar | *S. aureus* ATCC 25923 | Growth and expected hemolysis | | Chocolate agar | *H. influenzae* ATCC 10211 | Growth of fastidious organisms | | MacConkey agar | *E. coli* ATCC 25922, *S. aureus* ATCC 25923 | Growth with lactose fermentation, and inhibition of staphylococci | | Sorbitol MacConkey agar | *E. coli* O157:H7 sorbitol-negative strain | Colorless sorbitol-negative colonies and selectivity | | Campylobacter selective agar | *C. jejuni* ATCC 33291 | Growth under microaerobic conditions | | Selective *Neisseria* media | *N. gonorrhoeae* ATCC 49226 | Growth and selectivity against ordinary flora | Inoculum preparation is part of the record, not an afterthought. In the Ontario survey, only 65 of 117 laboratories testing selective media for pathogenic *Neisseria* used a correct inoculum, and incubation times and temperatures of QC plates were not always appropriate [1]. A medium QC result generated with the wrong inoculum or the wrong incubation verifies nothing, yet it fills a row in the record and looks identical to a valid result. Recording the inoculum method beside the result makes that failure visible at review. Physical inspection belongs in the same record. Hemolysis pattern, agar color, consistency, and the presence of contamination or desiccation are part of the QC observation, and the Ontario survey documented that these physical characteristics were inadequately recorded across participating laboratories [1]. ### Molecular and sequencing QC Molecular workflows add controls that the classical bench never needed. In the published mNGS workflow for respiratory viruses, the no-template control caught contamination introduced anywhere in the extraction and library process, the internal control, spiked into each sample, detected inhibition and extraction failure, and the external control verified reagent and equipment integrity [2]. For a QC data management system, this means the molecular bench needs fields the classical benches do not: the internal control read count or cycle threshold in each sample, the background signal in the no-template control, and a validity call per sample that combines both. The system should require that per-sample validity call before a result can be released, exactly as the published workflow validated results sample by sample against pre-set QC criteria [2]. ### Antimicrobial susceptibility testing QC AST QC uses reference strains run with each new lot of cards, panels, or disks and at the defined routine frequency. Record the strain, the drug, the zone diameter or minimum inhibitory concentration (MIC), and compare it to the published range for that strain and method. This is the bench where quantitative trend review earns its keep, because zone and MIC data are continuously distributed and reveal drift long before a result falls outside its range. ## Expected Results and How to Read Them A healthy system produces three things, and each is readable at a glance. First, the daily record. Every control scheduled for that day has an entry with an S code, actual observations, lot numbers, and two sets of initials. A missing entry is itself a finding, because an unrecorded control is an unverified day of testing. Second, the exception list. Entries marked A or F, plus any automated flags, appear in a short list that the weekly reviewer works through to closure. A healthy week produces a list of zero to a few items. A list that is always empty means the criteria are too loose or the entries are not honest, and a list that is always long means the criteria are set wrong or a process is chronically failing. Third, the trend charts. For quantitative controls, points should scatter randomly around the mean with no direction. Read the chart for four patterns. A shift is a run of points on one side of the mean, which usually indicates a new lot, a new analyst, or a changed incubator. A trend is a steady slope, which indicates gradual degradation such as reagent aging. Increased scatter indicates inconsistent technique, most often inoculum preparation. A single point beyond the action limits calls for immediate investigation and a repeat. For qualitative media controls, the analog of the chart is the lot history table. When a new lot is qualified, record its results beside the outgoing lot's results. Equivalent performance across lots is the expected pattern, and a new lot that merely squeaks past its criteria deserves a comment in the record even when it passes. ## Practical Tips for Daily Operation Enter QC results at the bench, at the time of observation. Transcribed results acquire transcription errors, and the Ontario survey's findings of incomplete and variable records grew from exactly this kind of gap between observation and recording [1]. Keep one person accountable for the database even when many people enter into it. Shared responsibility for data hygiene reliably becomes no responsibility. Resist adding fields. Every column must have a defined use in review or corrective action. Fields that exist because they might be useful someday become the blank cells that make records look complete while saying nothing. When a control produces an A code, write the comment as if it will be read by someone who was not there, because in an audit or an inspection, it will be. ## Troubleshooting | Symptom | Likely cause | Fix | |--|--|--| | QC entries missing on busy days | Entry depends on one person's memory rather than a scheduled task | Assign entry to the close-of-shift checklist and verify at daily review | | Records complete but identical day after day | Copied entries or checkmark-only recording | Require actual observations, audit a week of raw entries against bench worksheets | | Control strain grows weakly on all media | QC stock has drifted from repeated subculture | Replace working stock from the frozen reserve and requalify | | New medium lot fails qualification repeatedly | Storage or shipment problem, or genuinely defective lot | Check storage temperature records, quarantine the lot, contact the manufacturer, and document parallel testing | | AST zones drifting steadily toward one limit | Aging reagent, degraded disks, or incubator temperature drift | Check disk storage and incubator logs, replace reagents, then repeat the control series | | No-template control shows background signal in molecular assays | Contamination introduced during extraction or library preparation | Trace with the lot and analyst fields, clean work surfaces, replace reagent aliquots, and revalidate with fresh controls [2] | | Internal molecular control absent or weak | Extraction failure or inhibitors in the specimen | Repeat extraction before releasing any result from that sample, consistent with per-sample validity logic [2] | | Trend charts unreadable because entry styles differ | No entry standardization enforced at training | Reissue the field definitions, retrain, and clean the backlog in a documented batch | | Reviewers approve without reading | Review is a signature ritual, not an analysis | Change the review task from signing to answering defined questions, and record the answers | ## Variations for Different Laboratory Sizes A small laboratory with one to five benches can run the entire system in a single spreadsheet workbook with one table per bench, a weekly review by the supervisor, and a monthly chart review by the director. The critical discipline is the fixed field structure and the correction rules, which preserve the record's value even in a simple format. A high-volume laboratory should run QC inside the LIMS, linked to patient result release. The defining advantage is enforcement. When the system physically withholds a result until its control passes, the QC failure response in Step 8 stops depending on human discipline. This linkage matters most where per-sample validity is required, as in molecular and mNGS workflows, where each specimen's result depends on its own internal and external control performance and cannot be validated by a run-level control alone [2]. A public health or reference laboratory handling surge events adds a variation on Step 1. Preparedness exercises show that laboratory processes, not just reagents, need documented readiness. In a 2024 survey of Canadian hospital laboratories, 32 of 33 microbiology laboratories responded to a preparedness assessment covering their laboratory processes for a difficult-to-detect pathogen, and 81 percent identified all clinically relevant characteristics in testing, demonstrating how defined, documented laboratory processes support consistent performance [3]. A QC data system should therefore include competency records for each analyst on each method, so that performance differences between analysts can be detected in the data rather than in an outbreak investigation. ## Storage, Backup, and Record Retention Store the QC database with the same care as any patient record. Keep at least one backup copy on separate storage, in a separate location, and verify monthly by performing a real restore on a test copy. Restrict write access to defined staff, and keep an access log if the platform supports one. For record retention, follow the requirement of your accreditation program and jurisdiction, which commonly exceed two years for laboratory QC records. Retain lot-level records for at least the life of the lot plus the retention period, because a lot defect may surface long after the lot is consumed. Electronic records should be exported in a non-proprietary format, such as comma-separated values, at defined intervals so the record survives any future software change. Store QC organisms in two tiers. Working stocks live on agar slants or plates replaced on a fixed schedule. Reserve stocks live frozen in glycerol broth at 10 to 15 percent glycerol at minus 20 degrees Celsius or colder, and a new reserve is prepared from a verified original source, never from a working stock. Record every subculture in the organism inventory from Step 5. **Safety note.** QC organisms are handled at the containment level appropriate to each strain, and standard biosafety practices apply to every manipulation. Prepare inocula from stock cultures in a biosafety cabinet when the procedure can create aerosols, decontaminate work surfaces after QC setup, and autoclave or otherwise inactivate spent control cultures before disposal. Glycerol stocks and the low-temperature freezer introduce cryogenic and cold-contact hazards, so use insulated gloves and face protection when retrieving frozen stocks. ## Common Mistakes and Limitations The mistakes below are the ones that appear repeatedly in real laboratories, several of them documented at scale. **Treating media QC as a formality.** In the Ontario survey of 124 laboratories, in-house QC was not performed by 9 percent of laboratories for chocolate agar, 87 percent for MacConkey agar without crystal violet, and 53 percent for sorbitol MacConkey agar, and fewer than half used the recommended ATCC strains [1]. The lesson is not that these laboratories were careless. It is that purchased media is often assumed to be qualified, and selectivity media that no one expects to fail quietly stops being tested. A system must explicitly require QC for every medium in the inventory, including the ones that always pass. **Recording pass without recording how.** Time, temperature, and inoculum determine whether a media QC result means anything. The survey documented inappropriate incubation times and temperatures and, for pathogenic *Neisseria* media, incorrect inoculum in more than a third of responding laboratories [1]. Record the conditions beside the result. **Lot tracking without patient linkage.** A failed lot is only actionable if the system can name the specimens tested on it. Build the lot register and the date linkage in Step 4, not after the first recall. **Criteria set after the fact.** Acceptance criteria defined after results arrive turn QC into narrative rather than measurement. Define them in Step 2 and treat any later change as a documented, justified revision. **Data without review.** The most common failure mode in otherwise complete systems is the reviewer who signs without analyzing. Convert review from a signature into the defined questions in Step 9. **Trending only failures.** A control that passes every day tells you nothing until you look at the direction of its numbers. Quantitative controls need charts, not just checkmarks. **Assuming molecular controls transfer from other methods.** Control design for molecular assays must be validated in the actual workflow. The published mNGS work determined the optimal internal control spiking level empirically, balancing detectability against consumption of sequencing reads, before applying the controls to clinical samples [2]. Adopt the logic, not just the control names, and validate the levels in your own workflow. **Limitations of the system itself.** A QC data management system verifies process performance, not clinical truth. It cannot detect an error shared by both the test and the control, it depends on the honesty and consistency of entry, and it cannot substitute for method validation or proficiency testing. Individual patient cases always require interpretation by a qualified professional in the context of the whole clinical picture. One broader point about why this record matters. A mixed methods study of public trust during the COVID-19 pandemic, based on a nationwide survey of 5,607 Canadian residents and 41 qualitative interviews, examined how trust in health and scientific institutions shapes adherence to public health guidance [4]. Laboratory quality data is part of the substrate of that trust. Results that trace back to a complete, honest QC record support the institutional credibility that the pandemic showed is fragile and consequential. ## Frequently Asked Questions ### What is a QC data management system in microbiology? It is the structured record a laboratory keeps of every quality control result, its acceptance criteria, its trends over time, and the corrective actions taken when criteria are not met. It spans media, control organisms, staining reagents, AST controls, and molecular assay controls. ### Can spreadsheets handle microbiology QC data, or do we need a LIMS? ![Side by side view of spreadsheet and LIMS across concurrent entry, result blocking, audit trail and best fit.](/article-images/illustrations/microbiology-qc-data-management-spreadsheet-vs-lims.webp) Spreadsheets handle QC data well for small and medium laboratories if the structure is fixed, editing is restricted, and backups are tested. A LIMS becomes necessary when you need concurrent multi-user entry, automatic blocking of patient result release on QC failure, or a full audit trail. ### How often should QC data be reviewed? Review daily entries weekly for exceptions and drift, review quantitative control charts monthly, and audit the whole program annually. The weekly review is the one that catches problems early, so protect it on the calendar. ### What QC strains are used for media quality control? Common choices include *S. aureus* ATCC 25923 for blood agar, *H. influenzae* ATCC 10211 for chocolate agar, *E. coli* ATCC 25922 for MacConkey agar, and *N. gonorrhoeae* ATCC 49226 for selective gonococcal media. Confirm the specific strains and criteria for each medium against the current CLSI media QC standard. ### What are the most common QC data management mistakes? The most common are missing entries on busy days, pass-only recording without observations or conditions, acceptance criteria defined after results, and reviews that consist of signing rather than analyzing. Documented surveys also show selectivity media being skipped entirely, often because purchased media is assumed to be pre-qualified. ### How long should QC records be kept? Follow your accreditation program's retention requirement, which commonly exceeds two years, and keep lot-level records for at least the life of the lot plus the retention period. Export electronic records in a non-proprietary format so they survive software changes. ### What is an acceptance range and how do we set one? An acceptance range is the pre-defined interval of results that indicates the process performed as designed. For quantitative controls such as AST, use the published ranges for the reference strain and method. For qualitative controls, define the expected growth or reaction pattern and the required incubation conditions before data collection begins. ### How does QC differ for molecular tests such as mNGS? Molecular workflows require a no-template control for contamination, an internal control in each sample for extraction and inhibition, and an external control for reagent and equipment integrity. Each sample's result is validated against these per-sample controls, and the internal control level must be validated in your own workflow so it is detected without overwhelming the assay. ## Further Reading - [Using high frequency GPS data to assess wintering goose proximity to commercial poultry facilities on the Delmarva peninsula for avian influenza risk management and surveillance.](https://pubmed.ncbi.nlm.nih.gov/42678946/) - [Efficacy of isoniazid in paediatric tuberculosis: an individual participant data meta-analysis.](https://pubmed.ncbi.nlm.nih.gov/41412716/) ## Sources 1. [Compliance with NCCLS approved standard M22-A2 for bacteriologic media quality assurance: a survey of 124 Ontario microbiology laboratories.](https://pubmed.ncbi.nlm.nih.gov/11821168/) 2. [Quality control implementation for universal characterization of DNA and RNA viruses in clinical respiratory samples using single metagenomic next-generation sequencing workflow.](https://pubmed.ncbi.nlm.nih.gov/30373528/) 3. [Preparedness for Candida auris in Canadian Nosocomial Infection Surveillance Program (CNISP) hospitals, 2024.](https://pubmed.ncbi.nlm.nih.gov/41165332/) 4. [Trust Dynamics and Equity in Public Health in Canada: Protocol for a Mixed Methods Project in the Postpandemic Era.](https://pubmed.ncbi.nlm.nih.gov/41313631/)