Molecular epidemiology of a Pseudomonas aeruginosa hospital outbreak driven by a contaminated disinfectant-soap dispenser

Abstract

Background and objective: Pseudomonas aeruginosa infection represents a main cause of morbidity and mortality among immunocompromised patients. This study describes a fatal epidemic of P. aeruginosa that occurred in a hematology unit in Italy.

Methods: Retrospective cohort study, prospective surveillance, auditing, extensive testing on healthcare workers and environmental investigation were performed to define the dynamics and potential causes of transmission. RAPD, macrorestriction analyses and sequence typing were used to define relationships between P. aeruginosa isolates.

Results: Eighteen cases of infection were identified in the different phases of the investigation. Of these, five constitute a significant molecular cluster of infection. A P. aeruginosa strain with the same genetic fingerprint and sequence type (ST175) as clinical isolates strain was also isolated from a heavily contaminated triclosan soap dispenser.

Discussion and conclusions: Our results are consistent with the hypothesis that patients became indirectly infected, e.g., during central venous catheter handling through contaminated items, and that the triclosan soap dispenser acted as a common continuous source of P. aeruginosa infection. Since P. aeruginosa is intrinsically unsusceptible to triclosan, the use of triclosan-based disinfectant formulations should be avoided in those healthcare settings hosting patients at high risk of P. aeruginosa infection.

Figure 1. Hematology unit.

(A) Map of the unit. Areas for outpatients and inpatients are in yellow and blue, respectively. Crosses indicate the location of patients with at least one P. aeruginosa isolate according to the case definitions (red, incident case; green, prevalent case; see text for details). L1 to L4 indicate the sites from which Pseudomonas spp. were isolated: L1, soap dispenser; L2 soap dispenser; L3 and L4 water outlets. (B) Drugs deposit and preparation room where the contaminated soap dispenser (L2) was placed.

Figure 2. Epidemic curve.

The diagram shows the 10 incident cases (with respective codes) identified throughout the 16 time fractions (T1–T16) of the retrospective cohort study. Red and blue squares denote died and survived patients, respectively. The incidence rate per 1000 person-days with 95% CI and the total time at risk is reported for each time fraction.

Figure 3. Epidemiological typing of clinical and environmental P. aeruginosa isolates.

(A) RAPD analysis with primers 208, 228, 241, 272, 275, 277, 287 and ERIC 1, as indicated on top of electropherograms. (B) PFGE analysis of P. aeruginosa isolates. The dendrogram was generated with BioNumerics (Applied Maths) using the unweighted pair-group method with arithmetic averages (UPGMA) and the Dice coefficient. The similarity (%) between isolates is shown at each node of the dendrogram.M is the molecular weight marker (bp); 10 P. aeruginosa from sputum (patient 10); 11 P. aeruginosa from blood culture (patient 11); 13 P. aeruginosa from central venous catheter swab (patient 13); L2 P. aeruginosa from soap; L3 and L4 P. aeruginosa from water.