Antimicrobial resistance in bacterial pathogens causing community-acquired and hospital-acquired lower respiratory tract infections (2010-24): a systematic review and meta-analysis.

BACKGROUND: Data on the prevalence of antimicrobial resistance (AMR) among the main bacterial pathogens causing community-acquired lower respiratory tract infections (CA-LRTIs) and hospital-acquired lower respiratory tract infections (HA-LRTIs) are scarce. This review estimated LRTI-specific AMR prevalence and time trends globally and for each WHO region. METHODS: In this systematic review and meta-analysis, we systematically searched Embase, Global Health, and MEDLINE for original peer-reviewed articles, published between Jan 1, 2010, and May 16, 2024, describing microbiologically confirmed LRTI caused by WHO's Global Antimicrobial Resistance and Use Surveillance System (GLASS) target bacteria, alongside antimicrobial susceptibility testing results. We included studies describing individuals seeking care at a health-care facility with a microbiologically confirmed LRTI, a culture-confirmed GLASS target bacterial pathogen from a lower respiratory tract or expectorated sputum specimen, and antimicrobial susceptibility testing results for one GLASS pathogen-antimicrobial combination or more, for 20 isolates or more. We excluded non-original or non-peer-reviewed articles. We used multilevel random-effect meta-analysis to estimate AMR prevalence for relevant pathogen-antimicrobial combinations. To assess article quality, we adopted the Microbiology Investigation Criteria for Reporting Objectively (MICRO) checklist. This review was registered with PROSPERO (CRD42024549753). FINDINGS: From 10 208 reports identified after deduplication, we included 506 reports and extracted data on 336 389 respiratory isolates from 51 countries. For CA-LRTIs, regional variations were observed in the prevalence of penicillin resistance in Streptococcus pneumoniae (32·4% globally [range: 26·5-38·7]; n=6475 isolates); ampicillin resistance in Haemophilus influenzae (51·6% [27·0-73·4]; n=4305); third-generation cephalosporin resistance (38·8% [21·3-84·0]; n=4198) and carbapenem resistance (7·2% [1·2-29·0]; n=3437) in Klebsiella pneumoniae; and meticillin resistance in Staphylococcus aureus (MRSA; 33·2% [11·0-43·2]; n=4698). For HA-LRTIs, regional prevalence of carbapenem resistance varied for Acinetobacter spp (86·0% [78·7-93·4]; n=17 843), Pseudomonas aeruginosa (42·1% [32·3-48·6]; n=22 341), K pneumoniae (37·0% [6·9-52·3]; n=8310), and Escherichia coli (13·2% [0·41-45·2]; n=6229); and for MRSA (55·2% [22·3-79·0]; n=6060). For HA-LRTIs, prevalence of third-generation cephalosporin resistance and carbapenem resistance in E coli was higher in low-income and middle-income countries (73·2% [65·5-80·3] vs 18·3% [5·2-37·0], p<0·001) than in high-income countries (23·2% [8·3-42·8] vs 0·30% [0·0-12·8], p<0·001). Between 2010 and 2021, overall prevalence of penicillin-resistant S pneumoniae declined and that of ampicillin-resistant H influenzae increased globally, both in high-income countries and low-income and middle-income countries, and in adults and children; carbapenem resistance in Gram-negative bacteria increased across multiple regions and globally; and MRSA increased globally. Among the 38 pathogen-antimicrobial combinations assessed, AMR prevalence was higher in HA-LRTI for 18 combinations (47·4%), in CA-LRTI for nine (23·7%), and did not differ for 11 (28·9%). INTERPRETATION: AMR in LRTIs is concerningly high, with pronounced regional and socioeconomic disparities. Systematic literature reviews can provide a complementary data source to national AMR surveillance, to better inform public health interventions. FUNDING: Fleming Fund, US Centers for Disease Control and Prevention, WHO.