Escherichia coli / EHEC / ETEC / EPEC

Associated infections

• Watery diarrhoea and travellers' diarrhoea (ETEC)
• Infantile and paediatric diarrhoea (EPEC, EAEC)
• Haemorrhagic colitis and haemolytic uraemic syndrome (STEC/EHEC)
• Urinary tract infection, cystitis and pyelonephritis
• Bacteraemia and sepsis, frequently of urinary or biliary origin
• Neonatal meningitis
• Intra-abdominal and surgical site infection
• Cholangitis and hepatobiliary infection, including after biliary instrumentation
• Prostatitis and epididymo-orchitis

Transmission route

Diarrhoeagenic E. coli is acquired by the faecal-oral route through contaminated food or water, or by person-to-person contact; STEC has a notably low infectious dose. Waterborne transmission is well documented and has caused large community outbreaks where treated supplies were contaminated by animal or human faecal material. Extraintestinal disease more commonly arises from the patient's own gut flora, ascending the urinary tract or seeding the bloodstream from a urinary, biliary or intra-abdominal focus, and in healthcare from contaminated devices, hands and fluids.

In endoscope reprocessing, E. coli is best understood as a contamination indicator rather than a typical water-system coloniser. It is not a component of the normal oligotrophic biofilm flora of a treated-water distribution loop, so its recovery from final rinse water or from a patient-ready endoscope should be treated as a serious finding. Plausible explanations include ingress of untreated or partially treated mains water, backflow or cross-connection, a breached or bypassed filter, or carry-over of gastrointestinal soil due to inadequate manual cleaning before disinfection. Any confirmed detection warrants quarantine of the affected reprocessor and endoscopes, review of the complete water treatment and distribution path, and repeat testing only after a documented corrective action. E. coli itself is readily inactivated by validated high-level disinfection, which is precisely why its presence points to a process or water-integrity failure rather than to disinfectant resistance.

The distinction between water-side ingress and device-side carry-over is the central diagnostic question when the organism is recovered in a reprocessing department. Gastrointestinal endoscopes are, by the nature of the procedure, grossly contaminated with faecal flora at the point of removal from the patient, with bioburden in the working channel measured in high logarithms. Manual cleaning is the step that removes the bulk of that load, and if it is incomplete, organisms and organic material can persist into the disinfection stage, be shed within the reprocessor and appear in a subsequent sample. A finding of E. coli in an endoscope channel sample with clean water results throughout the distribution system points strongly in this direction. Conversely, recovery from supply water sampled before it reaches any reprocessor implicates the water path and excludes patient soil as the origin.

Relevance in endoscopy and reprocessing

Faecal Gram-negative organisms, E. coli among them, are the flora most directly relevant to gastrointestinal endoscope reprocessing because they are the organisms the process exists to remove. Prevalence studies of patient-ready duodenoscopes and gastroscopes have recovered gut-derived organisms from a substantial minority of instruments in routine use, and while much of that literature concerns organisms of low intrinsic virulence, the same failure mechanisms apply to E. coli. Documented endoscopy-associated transmission of Enterobacterales has generally involved duodenoscopes and other instruments with complex distal mechanisms, damaged or retained channels, or reprocessing shortcuts, rather than any failure of the disinfectant to kill the organism.

In the reprocessor circuit itself, E. coli does not behave like Pseudomonas or Serratia. It does not readily initiate biofilm on clean stainless steel or fluoropolymer under low-nutrient conditions, and it does not persist for months in a well-sanitised loop. Where it is recovered repeatedly, the usual explanation is either a continuing source of contamination, such as an unresolved cross-connection or a failed backflow prevention device, or incorporation into an established biofilm elsewhere in the system that is being periodically sloughed. A single terminal filter change will not resolve either situation. It is also worth noting that filters rated to remove bacteria are a barrier, not a treatment: a filter operating downstream of heavily contaminated water is being asked to perform outside its intended duty and will eventually fail or allow grow-through.

High-level disinfection inactivates E. coli comfortably at validated exposure conditions, and there is no recognised acquired tolerance of endoscope disinfectants in this species. Drying and storage remain relevant nonetheless. Any organism introduced into a channel after the disinfection step, whether from rinse water, from a contaminated air line or from handling, will multiply in retained moisture during storage. For dental unit waterlines and CSD washer-disinfector rinse supplies the same principle holds: the barrier is only as good as the integrity of everything downstream of it, and a faecal indicator organism appearing at the point of use means that barrier has been bypassed rather than overwhelmed.

Interpreting a detection

E. coli in a final rinse or supply water sample is one of the few findings in reprocessing water surveillance that should be treated as significant on a single detection, at any count. It is the internationally accepted indicator of recent faecal contamination, it is not a normal inhabitant of treated water, and it is not a plausible incidental laboratory contaminant in the way that some environmental Gram-positives or moulds can be. The correct initial posture is that the result is real and that the water path or the process has been compromised.

Sampling artefact must still be excluded, but by evidence rather than by assumption. Establish exactly where and how the sample was taken: whether the outlet was disinfected and flushed, whether a sterile neutralising container was used, whether the sampler had been handling soiled endoscopes or working in the decontamination area immediately beforehand, and whether the sample could have contacted a drain, a sink surface or a glove used for dirty-side work. Confirm the transport interval and temperature, and check whether the laboratory processed other faecally contaminated samples in the same batch. Where a single positive is accompanied by an otherwise normal total viable count, no confirmatory organism from other outlets, and an identifiable breach of sampling technique, artefact becomes a defensible conclusion, but that conclusion should be documented with its reasoning rather than asserted.

If artefact is not clearly demonstrated, the investigation should proceed on the assumption of faecal ingress. The immediate operational actions are to quarantine the affected reprocessor and the endoscopes processed on it since the last satisfactory result, to notify infection prevention, and to stop using the affected water outlet. The technical review should cover backflow prevention devices and any cross-connection between treated water and mains, drain or waste lines; the integrity and change history of the terminal filter and its housing; recent plumbing work, temporary hoses or bypass arrangements; the condition of the reverse osmosis plant and whether product water has been blended or supplemented; and the possibility of drain or waste ingress into the reprocessor chamber. In parallel, manual cleaning practice should be audited directly, including brush condition and size, channel access, detergent contact time and dilution, and leak testing, because inadequate cleaning is the most common non-water explanation for a faecal organism in a reprocessing sample.

A single isolate warrants full investigation. A trend, meaning recovery on more than one occasion or from more than one outlet, indicates either an unresolved structural fault in the water system or a systemic failure of manual cleaning, and should trigger escalation beyond the department to facilities engineering and to the executive responsible for the service. Escalation is also warranted at first detection if the organism is recovered from a patient-ready endoscope, if the isolate is a Shiga toxin-producing strain, if it carries an extended-spectrum beta-lactamase or carbapenemase, or if any patient who underwent a procedure in the relevant period has developed a compatible infection. Isolates should be retained. Resumption of service should follow a documented corrective action and confirmatory sampling from an expanded set of points, not a single clean repeat from the original outlet.

Antimicrobial resistance

Resistance in E. coli is a substantial and growing global concern. Extended-spectrum beta-lactamase-producing strains, notably those carrying CTX-M enzymes, are widespread in both hospital and community settings, and fluoroquinolone and trimethoprim-sulfamethoxazole resistance is common. Community-onset urinary infection caused by an ESBL-producing strain is now routine in many regions, which has shifted empirical prescribing and increased carbapenem consumption, with predictable secondary effects.

Carbapenemase-producing E. coli, while less frequent than in Klebsiella pneumoniae, is established in many regions and severely limits treatment options. The relevant enzymes, including NDM, OXA-48-like, KPC and IMP types, are carried on mobile genetic elements that also disseminate among other Enterobacterales, so recovery of a carbapenemase-producing E. coli from any healthcare source has implications wider than the single isolate. Plasmid-mediated colistin resistance has also been described in this species.

Antimicrobial therapy is not recommended in STEC infection because it may increase the risk of haemolytic uraemic syndrome, and management is supportive. This is one of the few settings in which the identification of the pathotype directly changes clinical management, and it is a reason to characterise an isolate fully where the clinical context warrants it.

None of this resistance affects the organism's susceptibility to physical and chemical decontamination. E. coli remains readily inactivated by free chlorine at drinking water residuals, by peracetic acid, glutaraldehyde and ortho-phthalaldehyde at high-level disinfection concentrations, by thermal disinfection at washer-disinfector parameters and by steam sterilisation. Antimicrobial resistance genes confer no protection against these processes. The operational implication is that a resistant isolate is not harder to remove than a susceptible one, but the consequences of failing to remove it are greater, and the threshold for escalation and for patient look-back should be correspondingly lower.