Elizabethkingia meningoseptica

Associated infections

• Neonatal meningitis
• Bacteraemia and bloodstream infection in critical care
• Ventilator-associated and healthcare-associated pneumonia
• Catheter-related bloodstream infection
• Endocarditis
• Ocular infection, including keratitis and endophthalmitis
• Soft tissue and surgical site infection
• Peritonitis in patients on peritoneal dialysis

Transmission route

Acquisition in healthcare is predominantly environmental rather than person-to-person. A 22-month critical care investigation in London demonstrated, using epidemiological and molecular evidence, that patient acquisition was water-source associated, with matching isolates recovered from clinical sink units on the ward; automated flushing of clinical taps was followed by elimination of the organism from the water samples and cessation of new cases. That investigation is instructive because the intervention that worked was hydraulic rather than chemical, and because stagnation at infrequently used outlets was identified as the condition permitting the organism to establish.

Contaminated respiratory equipment, humidifiers, infusates and other fluids in contact with patients have also been implicated, as have contaminated solutions used for rinsing or flushing devices. The exposure routes that follow from these reservoirs are direct instillation, aspiration of contaminated fluid, and contact transfer from wetted equipment to a mucosal surface, indwelling device or wound. Hand transfer from a colonised sink or splash zone to a patient or device is a plausible intermediate step and is one reason why sink design, splash containment and the storage of clean items near sinks have received attention in critical care outbreak reports.

For reprocessing services, E. meningoseptica belongs to the same category as other non-fermenting waterborne opportunists: it is not a component of the flora expected in properly treated final rinse water, but it is well adapted to the low-nutrient, wetted conditions found in water treatment and distribution systems and in the internal surfaces of automated endoscope reprocessors. It is not among the organisms conventionally named as routine rinse-water indicators, so it is typically recovered from a total viable count plate and identified subsequently. Because of its intrinsic resistance profile and the documented healthcare risk, an identification of E. meningoseptica from rinse water or from a patient-ready endoscope warrants escalation, quarantine of the affected equipment, and a full review of the water path including filtration, stagnation points and reprocessor internal disinfection.

Relevance in endoscopy and reprocessing

There is no substantial body of literature linking Elizabethkingia specifically to flexible endoscope reprocessing, and it should not be presented as an established endoscopy-associated pathogen in the way that Pseudomonas aeruginosa or the non-tuberculous mycobacteria are. Its relevance is inferred from ecology rather than from documented endoscopy transmission: it is a waterborne opportunist that colonises the same premise-plumbing niches that supply reprocessing plant, and there is no reason to expect it to behave differently in a reprocessing water circuit than it does in a ward tap or a sink outlet. A detection in reprocessing water should therefore be understood as evidence about the water system rather than as evidence of a recognised endoscopy transmission pathway.

That inference is nonetheless operationally important. The organism's demonstrated ability to persist in chlorinated supplies means the incoming mains water cannot be assumed to exclude it, and its tolerance of oligotrophic conditions means the treated water side of the plant is a viable habitat. The locations to consider are those where water is warm, slow-moving or standing: terminal connecting tubes, point-of-use fittings, sample taps, reprocessor rinse reservoirs and internal tubing downstream of the terminal filter, and any branch of the loop serving a reprocessor used only intermittently. Where the organism participates in a mixed biofilm alongside better-characterised colonisers, planktonic counts in a spot sample will understate the burden and results will be intermittent.

Against validated high-level disinfection there is no evidence that Elizabethkingia possesses meaningful tolerance; its formidable resistance profile concerns antibiotics, not biocides, and the distinction should be made explicitly when communicating a result, because an isolate reported as resistant to carbapenems can be misread as resistant to disinfection. The practical risk is recontamination after the disinfection step from colonised rinse water or wetted reprocessor components, compounded by inadequate channel drying. Endoscopes stored with retained moisture allow any surviving or newly introduced waterborne organism to multiply during storage, and for an organism with this resistance profile the consequences of transferring even a small inoculum to a vulnerable patient are disproportionate to the count recovered. The same considerations apply to CSD final rinse circuits and to dental unit waterlines, where narrow bore and prolonged stagnation between uses create comparable conditions.

Interpreting a detection

An identification of Elizabethkingia meningoseptica from a water or final rinse sample should be treated as a real and significant finding. It is not a skin or airborne laboratory contaminant, and its recovery is consistent with an established water-side reservoir. It is worth confirming the identification, however, because routine biochemical systems misassign species within this genus and also confuse it with other non-fermenters; where the result will drive quarantine, escalation or reporting, confirmation by mass spectrometry or sequencing is justified before the investigation is built on the species name.

The first operational checks mirror those for other waterborne non-fermenters, with particular attention to stagnation. Establish the sampling point and stage, and extend sampling upstream to the reverse osmosis product water, the storage vessel, the loop return and each reprocessor on the same loop, so the extent of colonisation can be mapped. Identify infrequently used outlets, capped spurs and dead legs, and determine whether any part of the system stands idle for days at a time, since the London critical care experience indicates that restoring regular flow can itself eliminate the organism from affected outlets. Inspect and consider replacing terminal connecting tubing and point-of-use fittings. Review the terminal filter's type, integrity test result and change interval, and confirm there is no colonisable wetted component downstream of it. Review the reprocessor's internal self-disinfection cycle, its frequency, and whether it is actually run when the machine has been idle. Confirm sampling technique, including outlet disinfection, pre-flush volume, use of a sterile neutralising container and transport time, but do not settle on artefact as the explanation without positive evidence for it.

A single isolate at a low count from one outlet, with an identifiable period of stagnation and clean results elsewhere, may be managed by restoring flow, sanitising the affected segment and resampling with an expanded sample set. Given the organism's intrinsic resistance profile, however, the threshold for treating a first isolate seriously should be lower than for a common environmental Gram-positive, and the resample should include upstream points rather than only the original outlet. Repeated recovery, recovery from more than one outlet, or recovery persisting after sanitisation indicates an established biofilm reservoir and calls for physical replacement of the colonised components together with a review of loop design and circulation, rather than further chemical treatment.

Escalation to infection prevention is warranted at first confirmed identification when the organism is recovered from a patient-ready endoscope, when counts exceed the applicable action limit, when the finding persists after a documented corrective action, or where the facility serves neonatal, critical care, haematology or transplant populations for whom acquisition would carry high consequence. The microbiology laboratory should be asked whether any clinical isolates of Elizabethkingia have been reported from the facility in the relevant period, and environmental and clinical isolates should be retained for comparison, since establishing or excluding a link depends on typing that cannot be performed retrospectively if isolates have been discarded. Affected reprocessors and endoscopes should be quarantined while the investigation proceeds.

Antimicrobial resistance

Elizabethkingia species show extensive intrinsic resistance that complicates therapy. Isolates are typically resistant to most beta-lactams including carbapenems, owing to production of both extended-spectrum beta-lactamases and metallo-beta-lactamases, and are commonly resistant to aminoglycosides. The simultaneous presence of a metallo-beta-lactamase and a serine beta-lactamase in the same organism is unusual among Gram-negative pathogens and accounts for the breadth of beta-lactam resistance observed, including against agents that retain activity against many other multidrug-resistant non-fermenters.

Paradoxically, they are often susceptible in vitro to agents usually reserved for Gram-positive organisms, such as vancomycin, rifampicin and some fluoroquinolones, and trimethoprim-sulfamethoxazole, minocycline and piperacillin-tazobactam retain activity against some isolates. Reported susceptibility patterns vary considerably between series and between species within the genus, and in vitro activity does not translate reliably into clinical response, particularly for vancomycin. Combination therapy is often used in serious infection, though the evidence base is limited to case series.

Susceptibility testing is essential, and interpretive breakpoints for this genus are not well standardised. Laboratories generally apply breakpoints developed for other non-fermenting Gram-negative organisms or for Enterobacterales, neither of which is validated for Elizabethkingia, and method-dependent discrepancies between disc diffusion, gradient strip and broth microdilution have been described. Reported susceptibility results should therefore be interpreted in consultation with the laboratory and with infectious diseases input.

None of this bears on environmental control. Elizabethkingia is inactivated by chlorine, chlorine dioxide, peracetic acid and the aldehydes at the concentrations used in water treatment and device disinfection, and there is no established acquired biocide resistance in the genus. Its persistence in chlorinated municipal water reflects survival at the low residual concentrations found at the extremities of distribution systems, and protection within biofilm and by association with particulates, rather than any intrinsic capacity to withstand disinfection at working strength. Control in a reprocessing water system depends on eliminating stagnation, maintaining circulation, effective terminal filtration with no colonisable surface downstream, adequate sanitisation reaching every branch, and thorough drying of endoscope channels before storage.