Microbial Identification

An unidentified isolate is regulatorily useless. The agency-required investigation cannot conclude (a) whether the organism is objectionable for the dosage form (e.g. Burkholderia cepacia in non-sterile aqueous oral product = critical; Bacillus subtilis in dry tablet = nominal); (b) whether the organism is resident in the facility (recurring same-species recoveries = facility-control failure); (c) whether the disinfectant rotation is effective against this specific organism; (d) whether the operator gowning programme adequately controls the source; (e) whether the contamination is intrinsic (process) or extrinsic (lab). Every meaningful investigation begins with species-level ID.

Any ID system depends on its reference database. MALDI-TOF libraries shipped from the vendor (Bruker MBT, Vitek MS Knowledge Base) are extensive but biased toward clinical organisms; environmental + pharmaceutical-relevant organisms (e.g. Ralstonia, Burkholderia, slow-growing fungi) may be underrepresented. USP <1113> requires that each ID method be verified for its intended use — for pharma QC: panel of typical site organisms tested with the chosen method; concordance with reference (sequencing); evaluation of false-positive / false-negative rates. A curated in-house library of site-isolate sequences extends the vendor library and dramatically improves species-level call rate.

USP <1111> + FDA's 2024 microbiological quality guidance establish the 'objectionable organism' framework. An organism is objectionable in a non-sterile drug product if it (a) is a known pathogen in the dosage form's intended administration route, (b) is capable of growth in the product matrix, (c) has caused historical adverse events in similar products, or (d) is on a regulatory specified-organism list (USP <62>: E. coli, Salmonella, P. aeruginosa, S. aureus, Candida albicans). Species-level ID is what enables the objectionable judgment. Same colony, different ID, different disposition.

A curated isolate library — every species-level ID from EM + product + water + utility recoveries, with date, location, source, and frozen stock — is the single most powerful tool for recurrence analysis. When a new positive comes in, the question 'have we seen this before, where, and when' is answered in seconds. Resident organisms recur from the same locations; transient organisms appear once. The library drives disinfectant-effectiveness re-validation (test the actual site organisms, not lab strains), HVAC + gowning investigations, and the trend reports that feed CCS reviews.

• Recovery — isolate from plate / broth at the time of count. Pure isolate is critical; mixed cultures give unreliable IDs.
• Sub-culture — single-colony pick to confirm purity before ID.
• Primary ID — MALDI-TOF preferred; cost-effective and fast. Genus + species call + score.
• Confirmatory ID — if MALDI score is below the species threshold OR the organism is consequential (specified-organism candidate, repeated environmental hit): 16S / ITS sequencing.
• Library check — does the site library have this species at this location? Recurrent? Outbreak?
• Disposition — objectionable judgement per the dosage form; release vs hold vs reject vs investigate further.
• CAPA — depending on disposition: clean + disinfect the location, re-validate disinfectant, re-train operators, investigate HVAC / water / supplier, escalate to CCS review.
• Frozen stock — archive the isolate (typical 5 – 10 years) for future outbreak typing.

• Genus-only IDs accepted as adequate — Annex 1 §9.34 violation in sterile facilities; investigation cannot conclude objectionable-organism question.
• MALDI library not extended with site isolates — recurring organisms keep coming back as 'no reliable ID' calls; species-level rate stuck at 70 – 80%.
• Investigation closed on low-score MALDI ID without sequencing confirmation — false-species call drives wrong CAPA.
• No frozen isolate archive — when a recurrent issue surfaces 18 months later, no stocks available for strain typing.
• Disinfectant rotation validated against ATCC reference strains only, never against site-resident isolates — disinfectant 'works' on the test panel but fails in the cleanroom.
• Specified-organism positive (USP <62>) handled as routine bioburden rather than triggering recall pre-flight.
• BCC / Ralstonia detected in aqueous oral product, not flagged as objectionable because investigation team didn't know FDA's published BCC-specific guidance.
• Method-verification per USP <1113> never done — system performance assumed from vendor brochure.
• Library curated by one person who leaves; institutional knowledge of site-organism ecology lost.

• Per-isolate record: source (EM point / product lot / water loop / utility), recovery date, plate / broth, count, two-person sampling e-sig.
• Primary ID capture: MALDI-TOF instrument feed (Bruker / Vitek MS); spectrum score; species + genus call.
• Confirmatory ID workflow: low-score / consequential isolates auto-routed for sequencing; result captured + linked.
• Site library: searchable by species, location, date range; recurrence flag fires automatically when a new isolate matches a recurrent species at the same location.
• Objectionable-organism rules: per-dosage-form rule set (sterile / aqueous oral / topical / dry oral / inhalation); auto-classification on ID assignment.
• Disposition gate: objectionable-classified positives block product release; investigation workflow auto-spawns; CAPA template attached.
• Frozen-stock register: per-isolate freezer / box / position; retrieval workflow with cold-chain audit trail.
• Disinfectant-effectiveness panel: maintained against the actual top-N site isolates from the library; re-challenge schedule + result trending.
• CCS feed: monthly + quarterly site-microbial-ecology summary auto-generated; ready for management review per Annex 1 §2.
• Outbreak typing workflow: when recurrence is suspected, WGS request workflow with batch linkage; comparison to historical isolates.