Salmonella / Salmonella typhi / Salmonella typhimurium

Associated infections

• Non-typhoidal salmonellosis (acute gastroenteritis)
• Typhoid and paratyphoid (enteric) fever
• Bacteraemia and endovascular infection, including mycotic aneurysm
• Osteomyelitis and septic arthritis, particularly in sickle cell disease
• Meningitis, particularly in infants
• Reactive arthritis following gastroenteritis
• Chronic asymptomatic carriage, typically biliary (S. Typhi)

Transmission route

Salmonella is acquired by the faecal-oral route. Non-typhoidal serovars are usually foodborne, associated with poultry, eggs, meat and contaminated produce, with additional routes through direct animal contact, particularly reptiles and poultry, and contaminated water. S. Typhi and S. Paratyphi are human-restricted and are transmitted through water or food contaminated with the faeces of an acutely infected person, a convalescent case, or a chronic carrier. The infectious dose for non-typhoidal salmonellosis is generally higher than for Shigella, though it falls substantially in the presence of a fatty food vehicle or reduced gastric acidity.

Healthcare-associated salmonellosis is uncommon but documented, usually through contaminated food served within the facility or through person-to-person spread in settings with shared toileting and impaired continence. Historically, Salmonella is one of the few enteric pathogens with a documented record of transmission by gastrointestinal endoscopy, described in reports predating modern automated reprocessing and standardised high-level disinfection, in which instruments were inadequately cleaned or disinfected between patients. Those reports establish the biological plausibility of endoscopic transmission and are the reason the organism appears at all in reprocessing reference material; they do not describe a current pattern of risk where validated reprocessing is performed.

Salmonella is not part of the flora expected in an engineered, treated water system and is unusual in endoscope reprocessing final rinse water. It does not form the oligotrophic biofilms characteristic of Pseudomonas or Serratia in reverse osmosis plant and distribution loops, and it is readily inactivated by validated high-level disinfection. For these reasons a confirmed detection is serious and should not be dismissed as an incidental finding. The plausible explanations are ingress of untreated water or sewage into the treated water path through cross-connection or backflow, gross contamination of the sampling process, or carry-over of gastrointestinal soil from inadequate manual cleaning. Response should include immediate quarantine of the affected reprocessor and any endoscopes processed since the last satisfactory result, notification of infection prevention, review of the entire water path, and consideration of whether patient look-back is required.

Relevance in endoscopy and reprocessing

Salmonella occupies an unusual position in endoscopy infection prevention: it has documented historical transmission by flexible gastrointestinal endoscopy but essentially no role in the contemporary water-quality picture. The reported transmission events involved instruments that were inadequately cleaned or disinfected between patients, in an era before automated reprocessors, standardised high-level disinfection and validated cycle parameters were the norm. The mechanism was carry-over of gastrointestinal contents, in quantity, from one patient to the next. That mechanism remains theoretically available whenever manual cleaning fails or a reprocessing cycle is not completed, which is precisely why the organism retains a place in reprocessing reference material.

In the water system itself, Salmonella has no meaningful biofilm role. It does not compete in the low-carbon environment of a reverse osmosis distribution loop, it does not colonise reprocessor internal circuits the way Pseudomonas and Serratia do, and it will not be found persisting in a terminal connecting tube months after an ingress event. It is removed by reverse osmosis and by bacterial-retentive terminal filtration. This means the interpretation of a Salmonella result is fundamentally different from that of a non-fermenter result: the question is not where in the system the organism has established itself, but how faecal material entered a treated water path or a disinfected instrument.

Against high-level disinfection Salmonella has no relevant tolerance. Peracetic acid, glutaraldehyde and ortho-phthalaldehyde at validated concentrations and contact times inactivate it readily, as does thermal disinfection at washer-disinfector parameters. It is likewise readily inactivated by free chlorine at drinking water residuals. Its survival in a reprocessing context therefore depends entirely on physical protection: organic soil that the disinfectant cannot penetrate, a channel that was not brushed or perfused, a damaged lumen, or a cycle that did not run to completion. Drying and storage are secondary considerations for this organism compared with the non-fermenters, since it is not expected to be introduced from rinse water and does not proliferate in a clean wetted channel, though retained moisture will prolong the survival of any organisms already present. In CSD and dental water the same reasoning holds: a Salmonella detection points to faecal ingress or gross process failure, not to waterline colonisation.

Interpreting a detection

A confirmed Salmonella isolate from a water, final rinse or endoscope channel sample is an abnormal result and should be treated as a potential serious incident from the moment it is reported. There is no benign routine explanation. The organism is not part of treated water flora, is not a recognised environmental coloniser of reprocessing plant, and is not a plausible laboratory contaminant, since it is not something a laboratory encounters incidentally and its identification requires positive biochemical or molecular confirmation. The first action is to confirm with the laboratory that the identification is genuine and, where the isolate will drive an incident response, to arrange referral for serotyping, since distinguishing a typhoidal from a non-typhoidal serovar materially changes the public health implications.

Sampling and handling artefact must be considered but requires a specific and identifiable mechanism, not a general assumption. The realistic scenarios are contamination of the sample bottle or the sampler's gloves in the decontamination area, contact between the sample and a drain, sink or soiled item, or a laboratory cross-contamination event from a concurrently processed faecal specimen. Each of these can be investigated concretely: review who took the sample and what they had handled, whether outlet disinfection and pre-flush were performed, whether a sterile neutralising container was used, the transport interval, and what else the laboratory processed in the same batch. If none of these mechanisms is present, the finding must be treated as real.

The operational response runs on two tracks in parallel. On the water track, review backflow prevention and any possible cross-connection between treated water and mains, drain, waste or non-potable services; review recent plumbing work, temporary hoses and bypass arrangements; check the integrity and change record of the terminal filter and whether the housing has been disturbed; verify reverse osmosis plant performance and whether product water has been blended or supplemented; and consider whether waste or drain water could enter the reprocessor chamber. On the process track, audit manual cleaning directly rather than by record review, covering brush type and condition, channel access and perfusion, detergent dilution and contact time, leak testing, the interval between procedure and cleaning, and whether any cycle has been aborted or interrupted. Where the organism was recovered from an endoscope channel rather than from water, the process track is the priority.

The single-isolate versus trend distinction does not apply in the usual way. One confirmed isolate is sufficient to warrant a full investigation and immediate quarantine of the affected reprocessor and of endoscopes processed on it since the last satisfactory result. Waiting for a second result to establish a trend is not appropriate for an organism of this type. Escalation to infection prevention should be immediate at first confirmed detection, and should extend to the facility executive and to the relevant state or territory health authority where the finding is confirmed and a transmission pathway to patients cannot be excluded; salmonellosis is a notifiable disease in all Australian jurisdictions. Isolates must be retained for typing. Infection prevention, not the reprocessing service, should determine whether a patient look-back exercise is required, and that determination should take account of the interval since the last satisfactory result, the number of patients processed, and whether any of them has presented with a compatible illness.

Antimicrobial resistance

Antimicrobial resistance in Salmonella is a recognised global public health problem. Resistance to older first-line agents such as ampicillin, chloramphenicol and trimethoprim-sulfamethoxazole is widespread, and decreased susceptibility or frank resistance to fluoroquinolones has emerged in both typhoidal and non-typhoidal serovars. Reduced ciprofloxacin susceptibility in S. Typhi is now the norm in much of South Asia, and detecting it requires appropriate testing methods, since isolates may appear susceptible by disc diffusion using older breakpoints while failing clinically.

Extensively drug-resistant S. Typhi, additionally resistant to third-generation cephalosporins, has been reported and constrains treatment to azithromycin and carbapenems. A large and sustained outbreak of this phenotype originating in Pakistan has produced travel-associated cases in Australia and elsewhere, and any S. Typhi isolate with a relevant travel history warrants full susceptibility testing and infectious diseases input. In non-typhoidal serovars, multidrug resistance is driven substantially by antimicrobial use in food animal production, and specific resistant clones, including certain S. Typhimurium and monophasic variants, have disseminated internationally through the food chain.

Most uncomplicated non-typhoidal gastroenteritis does not require antimicrobial therapy, and treatment may prolong faecal carriage without shortening illness. Antimicrobials are reserved for invasive disease, for patients at high risk of invasive disease including infants, older adults and the immunocompromised, and for enteric fever, which always requires treatment.

As with other Enterobacterales, none of this resistance affects susceptibility to physical or chemical decontamination. Salmonella is inactivated by chlorine at drinking water residuals, by peracetic acid, glutaraldehyde and ortho-phthalaldehyde at high-level disinfection concentrations, by thermal disinfection and by steam sterilisation. A resistant isolate is no harder to remove from a water system or an endoscope than a susceptible one. What resistance changes is the consequence of failure, and an extensively drug-resistant isolate recovered from any point in a reprocessing pathway should lower the threshold for escalation, for external notification and for patient look-back.