Sink drains in healthcare facilities may provide an environment for antimicrobial-resistant microorganisms, including carbapenemase-producing Klebsiella pneumoniae (CPKP).

We investigated the colonization of a biofilm consortia by CPKP in a model system simulating a sink-drain P-trap. Centers for Disease Control (CDC) biofilm reactors (CBRs) were inoculated with microbial consortia originally recovered from 2 P-traps collected from separate patient rooms (designated rooms A and B) in a hospital. Biofilms were grown on stainless steel (SS) or polyvinyl chloride (PVC) coupons in autoclaved municipal drinking water (ATW) for 7 or 28 days.

Microbial communities in model systems (designated CBR-A or CBR-B) were less diverse than communities in respective P-traps A and B, and they were primarily composed of β and γ Proteobacteria, as determined using 16S rRNA community analysis. Following biofilm development CBRs were inoculated with either K. pneumoniae ST45 (ie, strain CAV1016) or K. pneumoniae ST258 KPC+ (ie, strain 258), and samples were collected over 21 days. Under most conditions tested (CBR-A: SS, 7-day biofilm; CBR-A: PVC, 28-day biofilm; CBR-B: SS, 7-day and 28-day biofilm; CBR-B: PVC, 28-day biofilm) significantly higher numbers of CAV1016 were observed compared to 258. CAV1016 showed no significant difference in quantity or persistence based on biofilm age (7 days vs 28 days) or substratum type (SS vs PVC). However, counts of 258 were significantly higher on 28-day biofilms and on SS.

These results suggest that CPKP persistence in P-trap biofilms may be strain specific or may be related to the type of P-trap material or age of the biofilm.

Describe the epidemiological and molecular characteristics of an outbreak of Klebsiella pneumoniae carbapenemase (KPC)–producing organisms and the novel use of a cohorting unit for its control.

Observational study.

A 566-room academic teaching facility in Milwaukee, Wisconsin.

Solid-organ transplant recipients.

Infection control bundles were used throughout the time of observation. All KPC cases were intermittently housed in a cohorting unit with dedicated nurses and nursing aids. The rooms used in the cohorting unit had anterooms where clean supplies and linens were placed. Spread of KPC-producing organisms was determined using rectal surveillance cultures on admission and weekly thereafter among all consecutive patients admitted to the involved units. KPC-positive strains underwent pulsed-field gel electrophoresis and whole-genome sequencing.

A total of 8 KPC cases (5 identified by surveillance) were identified from April 2016 to April 2017. After the index patient, 3 patients acquired KPC-producing organisms despite implementation of an infection control bundle. This prompted the use of a cohorting unit, which immediately halted transmission, and the single remaining KPC case was transferred out of the cohorting unit. However, additional KPC cases were identified within 2 months. Once the cohorting unit was reopened, no additional KPC cases occurred. The KPC-positive species identified during this outbreak included Klebsiella pneumoniae, Enterobacter cloacae complex, and Escherichia coli. blaKPC was identified on at least 2 plasmid backbones.

A complex KPC outbreak involving both clonal and plasmid-mediated dissemination was controlled using weekly surveillances and a cohorting unit.