Role of the environment in transmission of Gram-negative bacteria in two consecutive outbreaks in a haematology-oncology department

Abstract

In 2019-2020, two subsequent outbreaks caused by phenotypically identical ESBL-producing Enterobacter cloacae and multi-drug-resistant (MDR) Pseudomonas putida were detected in respectively 15 and 9 patients of the haematology-oncology department. Both bacterial species were resistant to piperacillin-tazobactam, used empirically in (neutropenic) sepsis in our hospital, and ciprofloxacin, used prophylactically in selective digestive decontamination for haematology patients. The E. cloacae outbreak was identified in clinical cultures of blood and urine. Despite intensified infection control measures, new cases were found in weekly point-prevalence screening cultures. Environmental samples of sinks and shower drains appeared positive in 18.1%. To diminish the environmental contamination burden, all siphons of sinks were replaced, and disinfection of sinks and shower drains was intensified using chlorine and soda on a daily basis. Replacement of shower drains was not possible. The outbreak of P. putida remained limited to rectal cultures only, and disappeared spontaneously without interventions. During both outbreaks, multiple strains of the incriminated bacterium were found simultaneously (demonstrated by Amplified-Fragment Length Polymorphism and/or Whole-Genome Multi-locus Sequencing Typing) in patients as well as the environment. It was experimentally shown that a biofilm on the toilet edge may act as a source for nosocomial transmission of Gram-negative bacteria. In conclusion, the drainage system of the hospital is an important reservoir of MDR bacteria, threatening the admitted patients. In existing hospitals, biofilms in the drainage systems cannot be removed. Therefore, it is important that in (re)building plans for hospitals a plan for prevention of nosocomial transmission from environment to patients is incorporated.

Keywords: Gram-negative bacteria; Nosocomial infection; Water points.

Figure 1

Epidemic curve ESBL-positive E.cloacae outbreak. Chronologic schematic overview of the epidemic. Patient isolates. Colour or rectangle: red; blood culture, yellow; urine, grey: rectal swab. Colour of character: black; haematology patient, red; oncology patient. Environmental isolates: 321 samples from various water points were tested; 58 (18%) were positive for ESBL-positive E.cloacae with the epidemic phenotype. Genotyping results (patient and environmental isolates): Genotyping results were obtained by AFLP and/or wgMLST. A, C–E: genotype of ESBL-positive Enterobacter cloacae strain; italicized = genotyped by wgMLST. NT = not typed. NR = not related. Interventions: (1) contact-isolation in separate room for positive patients and all contact patients, (2) start of point prevalence screening of all patients on ward by rectal swab, (3) disinfection of sinks and showers with Na₂CO₃ on a weekly basis, (4) replacement of all syphons of sinks on the ward (replacement of shower drains not possible), (5) disinfection of shower drains with chlorine and Na₂CO₃ on a daily basis plus closure of culture positive shower drains in the corridor, (6) reopening of showers corridor, (7) termination of weekly point prevalence screening.

Figure 2

Overview environmental sampling E.cloacae. Situation in the haematology department in week 25. In this figure is indicated where patients and environmental samples tested positive for ESBL-positive E.cloacae. Inner rectangles: 1p/2p/4p = 1, 2, 4 persons patient room. Orange: environmental samples were positive for other Gram-negative bacteria. Red: environmental sample(s) positive for ESBL-positive E. cloacae and other Gram-negative bacteria. In the outer rectangles is indicated which patients had been admitted to the room before environmental sampling. Two shower drains were both positive at the corridor which were used by multiple patients were all positive for ESBL-positive E. cloacae.

Figure 3

Whole Genome Sequencing results ESBL-E.cloacae. Minimum spanning tree of wgMLST of ESBL-positive E. cloacae. Each circle represents one or multiple isolates, depending on the circle size and lines between circles indicate the allelic distance between the indicated isolates. Isolates are coloured by origin: patient isolates (green), environmental isolates (red) and reference non-related isolates (blue). Isolates from patients and the environment that cluster together within the cut-off value are depicted with a grey line zone and indicated as cluster A.

Figure 4

Epidemic curve multi-resistant Pseudomonas putida outbreak. Chronologic schematic overview of the epidemic. Patient isolates. Only positive cultures from rectal swabs for Selective Digestive Decontamination were involved. Environmental cultures: week 38; shower and sink in patient room frequently occupied by positive patients, week 40; ice cube machine; week 41; various water tap points (water from toilets, siphons of basins, buckets used by cleaning personnel), week 48; ridges of 7 toilets from patient rooms of positive patients (note: these cultures were negative for ESBL+ E.cloacae). Genotyping results (patient and environmental isolates): 3–7: genotype of multi-resistant Pseudomonas putida strain. NT = not typed. Interventions: no interventions were carried out.

Figure 5

Experimental study flushing toilet. Experiments were performed in a toilet room of a unused patient department. Experiment 1: 90 mL Visirub® concentrate (Hartmann, Nijmegen, The Netherlands) was mixed with the water present in the flushing cistern (estimated volume 9 L). Photographs were taken before and 5 minutes after flushing the toilet, with illumination of the toilet room using UV-light. UV-positive remainders of droplets were identified inside as well as outside the toilet bowl. Experiment 2: 24,25 mL Visirub® concentrate was mixed with 24,25 demineralized water and 1,5 mL gel-former (Jojoli, Barendrecht, The Netherlands). The resulting fluorescent mixture was applied underneath the toilet edge. Again, photographs were taken in a similar manner. Flushing water was able to remove the gel from the toilet.