Global genomic surveillance and temporal trends of carbapenemase-encoding β-lactamase genes in Klebsiella pneumoniae using NCBI pathogen detection data

The global spread of carbapenem-resistant Klebsiella pneumoniae is weakening the effectiveness of last-resort antibiotics and making severe infections increasingly difficult to treat. This study examined the geographical distribution, temporal patterns and co-occurrence of carbapenemase genes among K. pneumoniae genome assemblies deposited in the National Center for Biotechnology Information Pathogen Detection database. Genome metadata and antimicrobial resistance genotype information were retrieved between January and February 2026. After data cleaning, 67,057 unique genome assemblies from 123 countries contributed 77,975 carbapenemase-gene detections. Gene frequencies were summarised by continent, country, year and sample source. Poisson and negative binomial regression models were used to examine temporal patterns from 2013 to 2025. Genome-level co-occurrence was assessed using phi correlation coefficients, Jaccard similarity, odds ratios and Fisher’s exact tests. bla KPC was the most frequently detected gene, with 39,832 detections (51.08%), followed by bla NDM with 24,412 (31.31%). bla NDM was the dominant gene in Africa, accounting for 1,101 detections (52.68%), whereas bla KPC was dominant in the Americas, Asia, Europe and Oceania, accounting for 59.29%, 56.50%, 38.37% and 38.77% of detections, respectively. Urine and blood were the main sample sources, representing approximately 30% and 22% of the analysed genomes, respectively. Yearly trend analysis showed increases in bla NDM, bla OXA-232 and bla OXA-181. The strongest negative association was found between bla KPC and bla NDM (φ = −0.69), although genomes carrying either gene frequently contained additional antimicrobial resistance genes. These findings demonstrate clear geographical and temporal differences in carbapenemase genes among genome assemblies deposited in NCBI pathogen detection. However, differences in sequencing capacity and database submissions across countries and years should be considered when interpreting the findings. Continued genomic surveillance, antimicrobial stewardship and stronger infection prevention and control measures are needed, particularly in underrepresented regions.