Minggui Wang, Michelle Earley, Liang Chen, Blake M Hanson, Yunsong Yu, Zhengyin Liu, Soraya Salcedo, Eric Cober, Lanjuan Li, Souha S Kanj, Hainv Gao, Jose M Munita, Karen Ordoñez, Greg Weston, Michael J Satlin, Sandra L Valderrama-Beltrán, Kalisvar Marimuthu, Martin E Stryjewski, Lauren Komarow, Courtney Luterbach, Steve H Marshall, Susan D Rudin, Claudia Manca, David L Paterson, Jinnethe Reyes, Maria V Villegas, Scott Evans, Carol Hill, Rebekka Arias, Keri Baum, Bettina C Fries, Yohei Doi, Robin Patel, Barry N Kreiswirth, Robert A Bonomo, Henry F Chambers, Vance G Fowler, Cesar A Arias, David van Duin, Lilian M Abbo, Deverick J Anderson, Kean Lee Chew, Heather R Cross, Partha Pratim De, Samit Desai, Sorabh Dhar, Valentina Di Castelnuovo, Lorena Diaz, AN Q Dinh, Brandon Eilertson, Beth Evans, Vance G Fowler Jr, Julia Garcia-Diaz, Omai B Garner, Kerryl Greenwood-Quaintance, Blake Hanson, Erica Herc, Jesse T Jacob, Jianping Jiang, Robert C Kalayjian, Keith S Kaye, Angela Kim, Courtney Lauterbach, Steven H Marshall, Todd McCarty, Jose Munita, Oon Tek Ng, Jose Millan Oñate Gutierrez, Anton Peleg, Robert A Salata, Suzannah Schmidt-Malan, Nares Smitasin, Maria Spencer, Martin Stryjewski, Jiachun Su, Paul Ananth Tambyah, Sandra Valderrama, Maria Virginia Villegas Botero, Mary Waters, Darren Wong, Glenn Wortmann, Yang Yang, and Fujie Zhang
Summary Background Carbapenem-resistant Klebsiella pneumoniae (CRKP) is a global threat. We therefore analysed the bacterial characteristics of CRKP infections and the clinical outcomes of patients with CRKP infections across different countries. Methods In this prospective, multicentre, cohort study (CRACKLE-2), hospitalised patients with cultures positive for CRKP were recruited from 71 hospitals in Argentina, Australia, Chile, China, Colombia, Lebanon, Singapore, and the USA. The first culture positive for CRKP was included for each unique patient. Clinical data on post-hospitalisation death and readmission were collected from health records, and whole genome sequencing was done on all isolates. The primary outcome was a desirability of outcome ranking at 30 days after the index culture, and, along with bacterial characteristics and 30-day all-cause mortality (a key secondary outcome), was compared between patients from China, South America, and the USA. The desirability of outcome ranking was adjusted for location before admission, Charlson comorbidity index, age at culture, Pitt bacteremia score, and anatomical culture source through inverse probability weighting; mortality was adjusted for the same confounders, plus region where relevant, through multivariable logistic regression. This study is registered at ClinicalTrials.gov , NCT03646227 , and is complete. Findings Between June 13, 2017, and Nov 30, 2018, 991 patients were enrolled, of whom 502 (51%) met the criteria for CRKP infection and 489 (49%) had positive cultures that were considered colonisation. We observed little intra-country genetic variation in CRKP. Infected patients from the USA were more acutely ill than were patients from China or South America (median Pitt bacteremia score 3 [IQR 2–6] vs 2 [0–4] vs 2 [0–4]) and had more comorbidities (median Charlson comorbidity index 3 [IQR 2–5] vs 1 [0–3] vs 1 [0–2]). Adjusted desirability of outcome ranking outcomes were similar in infected patients from China (n=246), South America (n=109), and the USA (n=130); the estimates were 53% (95% CI 42–65) for China versus South America, 50% (41–61) for the USA versus China, and 53% (41–66) for the USA versus South America. In patients with CRKP infections, unadjusted 30-day mortality was lower in China (12%, 95% CI 8–16; 29 of 246) than in the USA (23%, 16–30; 30 of 130) and South America (28%, 20–37; 31 of 109). Adjusted 30-day all-cause mortality was higher in South America than in China (adjusted odds ratio [aOR] 4·82, 95% CI 2·22–10·50) and the USA (aOR 3·34, 1·50–7·47), with the mortality difference between the USA and China no longer being significant (aOR 1·44, 0·70–2·96). Interpretation Global CRKP epidemics have important regional differences in patients’ baseline characteristics and clinical outcomes, and in bacterial characteristics. Research findings from one region might not be generalisable to other regions. Funding The National Institutes of Health.