Hostname: page-component-669899f699-swprf Total loading time: 0 Render date: 2025-04-25T20:45:17.606Z Has data issue: false hasContentIssue false

A cluster of Burkholderia contaminans bloodstream infections in a rural hospital in Sierra Leone

Published online by Cambridge University Press:  15 April 2025

Ioana Diana Olaru*
Affiliation:
Institute of Medical Microbiology, University Hospital Münster, Münster, Germany
Laura C. Kalkman
Affiliation:
Masanga Medical Research Unit, Masanga Hospital, Masanga, Sierra Leone Centre for Tropical Medicine and Travel Medicine, Department of Infectious Diseases, Amsterdam University Medical Centers, Location AMC, Amsterdam Infection and Immunity, University of Amsterdam, Amsterdam, The Netherlands
Emmanuel Marx Kanu
Affiliation:
Masanga Medical Research Unit, Masanga Hospital, Masanga, Sierra Leone Centre for Tropical Medicine and Travel Medicine, Department of Infectious Diseases, Amsterdam University Medical Centers, Location AMC, Amsterdam Infection and Immunity, University of Amsterdam, Amsterdam, The Netherlands
Islam Mohamed Kargbo
Affiliation:
Masanga Medical Research Unit, Masanga Hospital, Masanga, Sierra Leone
Christian Böing
Affiliation:
Institute of Hygiene, University Hospital Münster, Münster, Germany
Stefan Bletz
Affiliation:
Institute of Hygiene, University Hospital Münster, Münster, Germany
Martin P. Grobusch
Affiliation:
Masanga Medical Research Unit, Masanga Hospital, Masanga, Sierra Leone Centre for Tropical Medicine and Travel Medicine, Department of Infectious Diseases, Amsterdam University Medical Centers, Location AMC, Amsterdam Infection and Immunity, University of Amsterdam, Amsterdam, The Netherlands Institute of Tropical Medicine & Deutsches Zentrum für Infektionsforschung, University of Tübingen, Tübingen, Germany Centre de Recherches Médicales, Lambaréné, Gabon Institute of Infectious Disease and Molecular Medicine, University of Cape Town, Cape Town, South Africa
Frieder Schaumburg
Affiliation:
Institute of Medical Microbiology, University Hospital Münster, Münster, Germany Masanga Medical Research Unit, Masanga Hospital, Masanga, Sierra Leone
*
Corresponding author: Ioana Diana Olaru; Email: [email protected]
Rights & Permissions [Opens in a new window]

Abstract

Type
Letter to the Editor
Copyright
© The Author(s), 2025. Published by Cambridge University Press on behalf of The Society for Healthcare Epidemiology of America

Dear Editor,

Burkholderia cepacia complex (BCC) are Gram-negative, non-fermenting organisms commonly found in soil and water, which are able to survive for months in humid environments. Reference Hafliger, Atkinson and Marschall1 BCC comprises more than 20 closely related species including B. cepacia, B. cenocepacia, and B. contaminans. Reference Ghafur, Balaguru and Ramanan2 BCC are recognized opportunistic pathogens in patients with cystic fibrosis, Reference Govan, Hughes and Vandamme3 but due to their low virulence, they are rare causes of infection among the general population. In healthcare settings, outbreaks of BCC were associated with contaminated surfaces, equipment, or liquid medicinal products and disinfectants. Reference Hafliger, Atkinson and Marschall1 B. cepacia is the most common species involved, with B. contaminans reported in only 4% (n = 5/121) of cases. Reference Hafliger, Atkinson and Marschall1 In this study we aimed to describe a cluster of B. contaminans bloodstream infections from a referral hospital in Sierra Leone.

Masanga Teaching Hospital, located in Tonkolili District in Sierra Leone, is a rural hospital with 120 beds providing health care for about 12,000 patients annually. Between April and October 2023, BCC was isolated from blood cultures of six patients. Four patients, three from the pediatric ward and one adult from the maternity ward had BCC bacteremia between April and August 2023.

The hospital infection prevention and control team presumed that contaminated intravenous solutions could be the source. Reference Held, Begier and Beardsley4 Contamination probably occurred extrinsically, from environmental sources, because of repeated use. These solutions were used to flush peripheral intravenous catheters over several days after being opened for multiple patients. The solutions may have been contaminated either through leaving syringe needles attached to an opened bag and/or through lacking disinfection of the rubber top when accessing a previously opened bag.

On August 27, 2023, nurses on the ward were requested not to use intravenous solutions packages for longer than two days after opening. Additionally, hand hygiene practices and adherence to sterile procedures were reinforced. Two additional patients, both from the pediatric ward, had positive blood cultures with BCC in October 2023 likely due to continuing practices. Following this, infection prevention and control practices were reinforced, and no other cases occurred.

Available BCC isolates from blood cultures from this time period (n = 5, as one isolate could not be recovered for sequencing) and an additional isolate from a wound swab from a patient who presented to the Masanga Teaching Hospital in 2019 (outlier isolate) underwent whole genome sequencing (WGS). WGS was conducted at the University Hospital Münster in 2024; sequences were submitted to BioProject (PRJNA1196134).

Four isolates, all from the pediatric ward were confirmed by WGS as B. contaminans belonging to sequence type ST2327 (Supplement). They differed by only ≤2 alleles of the 1,857 target genes in the BCC cgMLST typing scheme and were considered identical. Of the four patients, two had overlapping stays on the same ward in April and the other two in October 2023. The other isolates, which were from an adult from the maternity unit, and another one from a patient presenting with osteomyelitis in 2019 were B. cepacia (ST2325 and ST2328, Figure 1). Without any known related cases identified, the infection of the adult patient could also have been due to contaminated solutions.

Figure 1. Genomic relatedness of the Burkholderia cepacia complex isolates from Masanga Hospital and clinical characteristics of patients. The minimum spanning tree (A) was constructed using the 1,857 genes of the BCC core genome multilocus sequence typing (cgMLST) scheme. Nodes are labeled with the patient numbers (P1 to P7). Numbers between the nodes indicate the numbers of differing alleles between isolates. The neighbor-joining tree (B) was constructed using the concatenated sequence of the cgMLST typing scheme. Sex, wards, diagnosis sample, species, sequence type (ST), and outcome were color-coded. The phylogenetic analysis was visualized using R version 4.4.1 (packages ggtree, pheatmap and treeio.

Microbiological testing of the intravenous solutions and environmental sampling, which might have enabled us to establish the origin of the cluster, were not conducted because of resource constraints.

A systematic review of outbreaks due to contaminated substances found that multiuse solutions were linked to outbreaks in almost half of cases, with BCC as a leading causative organism. Reference Vonberg and Gastmeier5 However, few studies from low- and middle-income countries Reference Vonberg and Gastmeier5 were included likely due to considerable underreporting. The lack of identification of outbreaks in low-resource settings is largely due to limited access to microbiology testing, the need for patients to pay out of pocket for health care including diagnostics, and the difficulties in conducting environmental sampling within hospital facilities. Recognition of outbreaks due to BCC may be further challenged by difficulties in identification of these organisms using manual methods and the cost and availability of laboratory consumables needed for identification.

Overall, few studies from Africa report outbreaks due to contamination of solutions for parenteral use. Reference van Nierop, Duse, Stewart, Bilgeri and Koornhof6Reference Moodley, Coovadia and Sturm8 A study in Gambian neonates identified contamination of intravenous fluids and the multiple use of single-dose antibiotics as sources for outbreaks with Klebsiella pneumoniae and B. cepacia. The study included 49 neonates with B. cepacia bacteremia and was associated with a considerable mortality. Reference Okomo, Senghore and Darboe7 Generally, the mortality reported for BCC outbreaks is low (1%). Reference Hafliger, Atkinson and Marschall1 In our study, none of the four children with related BCC isolates died. It is unclear whether the adult patient died because of the BCC infection or due to complications following delivery. Despite the low mortality, outbreaks can be associated with substantial costs due to prolonged hospital stays and the need for conducting outbreak investigations. Reference Hafliger, Atkinson and Marschall1 Furthermore, BCC is intrinsically resistant to many antimicrobials which importantly limits treatment options, particularly in low-resource settings.

It is possible that bacteremia episodes due to BCC also occurred in other hospitalized patients without being identified as blood cultures are not usually collected after the initial presentation.

These findings from a rural hospital in Sierra Leone highlight the importance of the availability of microbiology diagnosis for outbreak identification and for the implementation of control measures. Although multiuse intravenous solutions have repeatedly been linked to outbreaks, these practices remain common in limited resource settings.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/ice.2025.63

Data availability statement

The data for this work has been made available in the manuscript. The whole genome sequences of the bacterial isolates were submitted to BioProject (PRJNA1196134).

Financial support

This work was supported by funds from the ‘Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ)/Klinikpartnerschaften” (project: 81281918). We acknowledge support from the Open Access Publication Fund of the University of Münster.

Competing interests

The authors declare none.

Ethical standard

Hospital policies relating to data privacy were followed.

References

Hafliger, E, Atkinson, A, Marschall, J. Systematic review of healthcare-associated Burkholderia cepacia complex outbreaks: presentation, causes and outbreak control. Infect Prev Pract 2020;2:100082.CrossRefGoogle ScholarPubMed
Ghafur, A, Balaguru, P, Ramanan, SG, et al. Diving deep for the needle in the haystack: an outbreak investigation of Burkholderia cenocepacia bacteremia. Infect Control Hosp Epidemiol 2024;45:677680.CrossRefGoogle ScholarPubMed
Govan, JR, Hughes, JE, Vandamme, P. Burkholderia cepacia: medical, taxonomic and ecological issues. J Med Microbiol 1996;45:395407.CrossRefGoogle ScholarPubMed
Held, MR, Begier, EM, Beardsley, DS, et al. Life-threatening sepsis caused by Burkholderia cepacia from contaminated intravenous flush solutions prepared by a compounding pharmacy in another state. Pediatr 2006;118:e2125.CrossRefGoogle ScholarPubMed
Vonberg, RP, Gastmeier, P. Hospital-acquired infections related to contaminated substances. J Hosp Infect 2007;65:1523.CrossRefGoogle ScholarPubMed
van Nierop, WH, Duse, AG, Stewart, RG, Bilgeri, YR, Koornhof, HJ. Molecular epidemiology of an outbreak of Enterobacter cloacae in the neonatal intensive care unit of a provincial hospital in Gauteng, South Africa. J Clin Microbiol 1998;36:30853087.CrossRefGoogle ScholarPubMed
Okomo, U, Senghore, M, Darboe, S, et al. Investigation of sequential outbreaks of Burkholderia cepacia and multidrug-resistant extended spectrum beta-lactamase producing Klebsiella species in a West African tertiary hospital neonatal unit: a retrospective genomic analysis. Lancet Microbe 2020;1:e119e29.CrossRefGoogle Scholar
Moodley, P, Coovadia, YM, Sturm, AW. Intravenous glucose preparation as the source of an outbreak of extended-spectrum beta-lactamase-producing Klebsiella pneumoniae infections in the neonatal unit of a regional hospital in KwaZulu-Natal. S Afr Med J 2005;95:861864.Google Scholar
Figure 0

Figure 1. Genomic relatedness of the Burkholderia cepacia complex isolates from Masanga Hospital and clinical characteristics of patients. The minimum spanning tree (A) was constructed using the 1,857 genes of the BCC core genome multilocus sequence typing (cgMLST) scheme. Nodes are labeled with the patient numbers (P1 to P7). Numbers between the nodes indicate the numbers of differing alleles between isolates. The neighbor-joining tree (B) was constructed using the concatenated sequence of the cgMLST typing scheme. Sex, wards, diagnosis sample, species, sequence type (ST), and outcome were color-coded. The phylogenetic analysis was visualized using R version 4.4.1 (packages ggtree, pheatmap and treeio.

Supplementary material: File

Olaru et al. supplementary material

Olaru et al. supplementary material
Download Olaru et al. supplementary material(File)
File 23 KB