Klebsiella pneumoniae

Associated infections

• Hospital-acquired and ventilator-associated pneumonia
• Bacteraemia and septicaemia
• Urinary tract infection, frequently catheter-associated
• Surgical site and intra-abdominal infection
• Pyogenic liver abscess associated with hypervirulent strains
• Neonatal sepsis and meningitis
• Cholangitis and biliary sepsis, including post-ERCP
• Endophthalmitis and meningitis complicating hypervirulent invasive syndrome

Transmission route

Transmission in healthcare occurs principally by the hands of healthcare workers and via contaminated equipment and environmental surfaces, with the gastrointestinal tract of colonised patients acting as the main reservoir. Colonised patients frequently outnumber infected patients by a substantial margin and are usually unrecognised without active screening, so the visible cases represent only part of the transmission burden.

Contaminated sink drains and wastewater plumbing constitute a secondary reservoir capable of sustaining transmission over prolonged periods, and migration of carbapenemase-producing Enterobacterales through connected pipework has been demonstrated in laboratory model systems. The route from drain to patient runs through splash and aerosol generated when water strikes a contaminated bowl or strainer, contaminating hands, gloves, adjacent surfaces and any clean items stored within the splash zone. Sink design, tap alignment relative to the drain, the practice of disposing of nutrient-rich waste into clinical handwash basins, and the storage of clean equipment near sinks are consequently all recognised contributory factors, and remediation of drain reservoirs has in several settings required replacement of sink units rather than disinfection alone.

With respect to flexible endoscope reprocessing, K. pneumoniae is relevant chiefly as an organism transmissible by inadequately reprocessed devices, most prominently duodenoscopes, where the complex elevator mechanism has been repeatedly implicated in transmission of carbapenemase-producing Enterobacterales. A German university hospital outbreak of OXA-48-producing carbapenem-resistant K. pneumoniae included six cases directly linked to endoscopic retrograde cholangiopancreatography performed with the same duodenoscope, with the outbreak resolving after manufacturer servicing of the instrument. Transmission in such events originates from a previously colonised patient rather than from the water supply, and it propagates because the device retains organic soil in a location that cleaning and high-level disinfection do not reliably reach.

In CSD and dental settings the organism is not a characteristic water-system coloniser, and its recovery would more plausibly indicate faecal, patient-derived or wastewater contamination of the sampled point than a treatment failure in the supply.

Relevance in endoscopy and reprocessing

K. pneumoniae occupies a distinctive position in endoscopy: it is one of the most significant organisms transmitted by endoscopes, but it is not primarily a rinse-water organism. The duodenoscope is the device of principal concern. Its distal elevator mechanism, forceps elevator recess and associated wire channel present crevices, moving parts and blind spaces that resist brushing and through which disinfectant flow is difficult to guarantee, and this geometry has been implicated in repeated international transmission events involving carbapenemase-producing Enterobacterales, K. pneumoniae prominently among them. Contributory factors identified in root-cause investigations have included instrument damage and wear, sealing failures within the distal assembly, deviation from the manufacturer's reprocessing instructions, and inadequate manual cleaning of the elevator recess. Notably, several such outbreaks continued despite reprocessing that appeared compliant on audit, which is why device-level surveillance culture and, in some jurisdictions, duodenoscope design change have been adopted.

Relevance to final rinse water is more limited than that of Pseudomonas aeruginosa. K. pneumoniae is not a classical low-nutrient water organism and is not among the indicator organisms named in the AS 5369:2023 final rinse water microbiological criteria. Nevertheless, its isolation from final rinse water is a significant finding precisely because it is out of character for that sample type. It indicates either faecal or patient-derived contamination entering the water circuit or reprocessing area, or biofilm establishment within the water system, and because rinse water contacts the endoscope after high-level disinfection is complete, any organism present at that point recontaminates a device intended to be patient-ready.

Drying and storage remain relevant but operate differently for this organism. Residual moisture in channels supports survival of Enterobacterales, and where organic soil remains in a channel or elevator recess, that soil provides the nutrient the organism requires. The combination of retained soil and retained moisture is therefore substantially more dangerous for K. pneumoniae than water quality alone. In CSD, the analogous concern is instruments with lumens or hinged assemblies where soil removal is incomplete; terminal steam sterilisation will kill the organism, but a wet or soiled load compromises that assurance.

Interpreting a detection

Isolation of K. pneumoniae from a final rinse water or reprocessing water sample is an abnormal result and should be escalated. Unlike the non-fermenters that populate treated water systems, this organism is not an expected constituent of purified water, so its presence is not adequately explained by ordinary biofilm shedding from a distribution loop. Two explanations should be considered first, and they lead in different directions. The first is contamination of the sample or the sampling point by a person or by the immediate environment: hands, gloves, a splashed outlet, a sink in the vicinity of the sampling point, or a container handled without aseptic technique. The second is genuine ingress of wastewater, faecal material or patient-derived contamination into the water path or the reprocessing area. Both must be excluded; neither can be assumed.

The practical checks differ from those for a water-system organism. Alongside the standard review of filtration, storage, loop temperature, stagnation, dead legs and sanitisation, the investigation should specifically examine backflow prevention and air gaps, any cross-connection between clean water and drainage, the position and design of sinks and drains relative to the clean workflow and the sampling point, splash exposure of stored clean items, the drainage arrangements of the washer-disinfector itself, and whether the sampling point is subject to contamination from the surrounding environment. Sampling technique should be reviewed in detail, since a plausible handling artefact is a common explanation for a single enteric isolate from an otherwise well-controlled circuit. A repeat sample, taken by a different operator with strict technique, is a reasonable early step, but it should run in parallel with, not instead of, the investigation.

A single isolate is sufficient to justify escalation because the alternative explanations are both serious. The affected reprocessor should be withdrawn from use for high-risk devices pending investigation, and infection prevention should be notified. Where the isolate carries a carbapenemase or another resistance mechanism of significance, notification under local multidrug-resistant organism protocols is required, the isolate should be retained for typing against any concurrent clinical isolates, and the investigation should extend to whether a colonised patient has been in the department and whether any device, particularly a duodenoscope, may have been the source rather than the recipient. Look-back should identify devices reprocessed through the affected pathway since the last satisfactory result. Verification of remediation should include repeat water sampling at more than one time point and, where a device is implicated, device surveillance culture rather than water testing alone.

Antimicrobial resistance

K. pneumoniae is intrinsically resistant to ampicillin through a chromosomal SHV-type beta-lactamase and has an exceptional capacity to acquire further resistance determinants on plasmids and other mobile elements. Extended-spectrum beta-lactamases, plasmid-mediated AmpC enzymes and carbapenemases including KPC, NDM, OXA-48-like and VIM types are widely disseminated, and resistance to fluoroquinolones, aminoglycosides and colistin frequently accompanies them, producing extensively drug-resistant phenotypes with very limited therapeutic options. The species is also an efficient recipient and donor of resistance plasmids, and functions as a hub for the movement of carbapenemase genes between Enterobacterales, which is why a single carbapenemase-producing isolate in a facility has implications extending beyond the patient from whom it was recovered.

Carbapenem-resistant K. pneumoniae was the highest-ranked organism in the WHO bacterial priority pathogens list 2024 and sits within the critical priority group. Convergence of carbapenem resistance with hypervirulence determinants in some lineages is an additional emerging concern, producing strains that combine the invasive capability associated with community-acquired liver abscess syndromes with an extensively drug-resistant phenotype.

Resistance to chemical disinfectants is a separate question and is not a defining feature of the species. K. pneumoniae in suspension is readily inactivated by the biocides used in high-level disinfection and in water treatment, and multidrug resistance confers no meaningful protection against them. Its survival through a reprocessing cycle is attributable to protection by residual organic soil and by biofilm within channels, recesses and drain systems, not to biocide tolerance. This has a direct operational implication: where the organism survives a validated disinfection process, the fault lies in cleaning, device condition or process delivery rather than in the choice of disinfectant, and escalating the chemistry will not correct it. Thorough manual cleaning, verified removal of organic soil, attention to device damage and wear, and elimination of the drain and biofilm reservoirs are the effective controls.