Carbapenem-resistant Acinetobacter baumannii at a hospital in Botswana: Detecting a protracted outbreak using whole genome sequencing

Carbapenem-resistant Acinetobacter baumannii (CRAb) has emerged as a major and often fatal cause of bloodstream infections among hospitalized patients in low- and middle-income countries (LMICs). CRAb outbreaks are hypothesized to arise from reservoirs in the hospital environment, but outbreak investigations in LMICs are seldom able to incorporate whole genome sequencing (WGS) due to resource limitations. We performed WGS at the National Institute for Communicable Diseases (Johannesburg, South Africa) on stored A. baumannii isolates (n=43) collected during 2021-2022 from a 530-bed referral hospital in Gaborone, Botswana where CRAb infection incidence was noted to be rising. This included blood culture isolates from patients (aged 2 days - 69 years), and environmental isolates collected at the hospitals 33-bed neonatal unit. Multilocus sequence typing (MLST), antimicrobial/biocide resistance gene identification, and phylogenetic analyses were performed using publicly accessible analysis pipelines. Single nucleotide polymorphism (SNP) matrices were used to assess clonal lineage. MLST revealed 79% of isolates were sequence type 1 (ST1), including all 19 healthcare-associated blood isolates and three out of five environmental isolates. Genes encoding for carbapenemases (blaNDM-1, blaOXA-23) and biocide resistance (qacE) were present in all 22 ST1 isolates. Phylogenetic analysis of the ST1 clade demonstrated spatial clustering by hospital unit. Nearly identical isolates spanned wide ranges in time (>1 year), suggesting ongoing transmission from environmental sources. One highly similar clade (average difference of 2.3 SNPs) contained all eight neonatal blood isolates and three environmental isolates from the neonatal unit. These results were critical in identifying environmental reservoirs (e.g. sinks) and developing remediation strategies. Using a phylogenetically informed approach, we also identified diagnostic genes useful for future tracking of outbreak clones without the need for WGS. This work highlights the power of South-South and South-North partnerships in building public health laboratory capacity in LMICs to detect and contain the spread of antimicrobial resistance.