Gastrointestinal Carriage Is a Major Reservoir of Klebsiella pneumoniae Infection in Intensive Care Patients

Abstract

Background: Klebsiella pneumoniae is an opportunistic pathogen and leading cause of hospital-associated infections. Intensive care unit (ICU) patients are particularly at risk. Klebsiella pneumoniae is part of the healthy human microbiome, providing a potential reservoir for infection. However, the frequency of gut colonization and its contribution to infections are not well characterized.

Methods: We conducted a 1-year prospective cohort study in which 498 ICU patients were screened for rectal and throat carriage of K. pneumoniae shortly after admission. Klebsiella pneumoniae isolated from screening swabs and clinical diagnostic samples were characterized using whole genome sequencing and combined with epidemiological data to identify likely transmission events.

Results: Klebsiella pneumoniae carriage frequencies were estimated at 6% (95% confidence interval [CI], 3%-8%) among ICU patients admitted direct from the community, and 19% (95% CI, 14%-51%) among those with recent healthcare contact. Gut colonization on admission was significantly associated with subsequent infection (infection risk 16% vs 3%, odds ratio [OR] = 6.9, P < .001), and genome data indicated matching carriage and infection isolates in 80% of isolate pairs. Five likely transmission chains were identified, responsible for 12% of K. pneumoniae infections in ICU. In sum, 49% of K. pneumoniae infections were caused by the patients' own unique strain, and 48% of screened patients with infections were positive for prior colonization.

Conclusions: These data confirm K. pneumoniae colonization is a significant risk factor for infection in ICU, and indicate ~50% of K. pneumoniae infections result from patients' own microbiota. Screening for colonization on admission could limit risk of infection in the colonized patient and others.

Keywords: Klebsiella pneumoniae; gastrointestinal colonization; genomic epidemiology; hospital acquired infection; intensive care.

Figure 1.

Flowchart outlining number of patients included in each part of carriage rates analyses.

Figure 2.

Genome diversity of isolates from ICU patients identified as Klebsiella pneumoniae. All trees are maximum likelihood trees inferred from core genome SNP alignments. Scale bars indicate average number of substitutions per site across the genome. Tip colours indicate isolate source as per inset legend. *Possible mixed isolate (0.02–0.1 het/hom SNP ratio, excluded from pairwise SNP analysis in Figure 3). **Clinical isolate from sputum (KC0048), may not represent an infection. Phylogenetic lineages to which more than one ICU isolate belongs are highlighted and labeled with their corresponding multi-locus sequence type (ST) and the total number of SNPs identified between isolates in the lineage; darker shading indicates multiple patients contributed isolates in that cluster, as per inset legend. (A), Unrooted tree of all isolates, revealing three distinct species that are typically identified as K. pneumoniae in diagnostic laboratories. (B), Midpoint rooted species tree for K. variicola isolates. (C), Midpoint rooted species tree for K. pneumoniae sensu stricto isolates. Abbreviations: ICU, intensive care unit; SNP, single-nucleotide polymorphism.

Figure 3.

Pairwise genetic distances between isolates belonging to the same lineage, expressed as SNPs per 5 Mbp of genome in order to normalise for differences in shared gene content between strain pairs. Violin plots showing distribution of pairwise genetic distances intra- and inter-patient; black bars indicate the median value. Note the log₁₀ scale which excludes display of 1 strain pair that was separated by 0 SNPs. Abbreviation: SNP, single-nucleotide polymorphism.

Figure 4.

Timelines for all lineages detected in multiple patients that show any inter-patient pairwise genetic distance between isolates of ≤25 SNPs per 5 Mbp. Lineages are boxed and labeled with their multi-locus sequence type (ST). Each horizontal dashed line indicates the time line for a patient, labelled to the left (crosses indicate date of death where applicable). Periods of Alfred Hospital admission are indicated as white boxes, periods in ICU as pink shading. Circles indicate K. pneumoniae infection isolates (red, belonging to the lineage; black, other lineage); triangles indicate rectal screening swabs (red, K. pneumoniae belonging to the lineage; black, K. pneumoniae of another lineage; unfilled, negative for K. pneumoniae). Orange boxes indicate groups of isolates for which all patients have at least one pairwise genetic distance of ≤10 SNPs per 5 Mbp with another in the group; similarly for yellow (≤25 SNPs) and blue (≤100 SNPs) boxes. Abbreviations: ICU, intensive care unit; SNP, single-nucleotide polymorphism.