Disclaimer: Early release articles are not considered as final versions. Any changes will be reflected in the online version in the month the article is officially released.
Volume 32, Number 11—November 2026
Research
Five-Biomarker Definition for Hypervirulent Klebsiella pneumoniae Bloodstream Infection and Abscess Complications, Japan, 2022–2024
, Tomohisa Watari, Yoshihito Otsuka, and Tomoh Matsumiya
Table 2
Clinical outcomes and adjusted associations among Klebsiella pneumoniae bloodstream infection patients in a study of 5-biomarker definition for hvKp bloodstream infection and abscess complications, Japan, 2022–2024*
| Clinical outcome | hvKp, n = 28 | cKp, n = 179 | Total no. | Estimate (95% CI)† |
|
|---|---|---|---|---|---|
| Crude | Adjusted | ||||
| Any abscess complication | 17 (61) | 23 (13) | 207 (100) | 10.5 (4.4–25.2) | 10.7 (4.4–26.2) |
| Liver abscess | 9 (32) | 12 (7) | 207 (100) | – | – |
| Other abscess sites‡ | 11 (39) | 12 (7) | 207 (100) | – | – |
| Metastatic infection |
4 (14) |
1 (<1) |
207 (100) |
– |
– |
| Median postonset hospitalization, d (IQR)§# | 28 (14–42) | 14 (11–18) | 206 (99.5) | 1.6 (1.2–2.2) | Model A, 1.6 (1.2–2.2) |
| Model B, 1.1 (0.8–1.5) |
|||||
| Median antibiotic duration, d (IQR)#†† | 43 (16–67) | 14 (14–18) | 168 (81) | 2.1 (1.6–2.7) | Model A, 2.1 (1.6–2.8) |
| Model B, 1.2 (1.0–1.5) |
|||||
| 30-d mortality rate, deaths/no. | 4/27 (15) | 34/170 (20) | 197 (95) | 0.8 (0.2–2.1) | 0.9 (0.3–2.4) |
*Data are no. (%) or median (IQR) unless otherwise indicated. Adjusted estimate is shown only for prespecified primary, secondary, and supportive outcomes. Liver abscess, other abscess sites, and metastatic infection are presented descriptively and were not modeled separately. cKp, classical K. pneumoniae; hvKp, hypervirulent K. pneumoniae; IQR, interquartile range. †Estimates are odds ratios for abscess complications and 30-day mortality, and ratios for length of hospitalization and antibiotic duration. Adjusted estimates for abscess complication were derived from multivariable logistic regression adjusted for age, sex, and Charlson comorbidity index (CCI); for post–disease onset length of stay, from multivariable linear regression of ln(length of stay + 1) adjusted for age, sex, CCI, and infection acquisition category; for total antibiotic duration, from multivariable linear regression of log-transformed duration among treatment completers adjusted for age, sex, CCI, infection acquisition category, and primary infection source; and for 30-d mortality, from Firth penalized logistic regression adjusted for age and CCI. ‡Other abscess sites included pelvic or intra-abdominal abscess, lung abscess or necrotizing pneumonia, prostatic abscess, endophthalmitis, central nervous system abscess, and musculoskeletal or soft tissue infection; episodes may have had >1 site. §Post–disease onset length of stay was defined as days from index blood culture collection to hospital discharge or interhospital transfer. If hospitalization continued for reasons judged unrelated to the bloodstream infection, follow-up was censored at the end of the bloodstream infection-directed therapy. Episodes ending in in-hospital death were included, with length of stay calculated up to the date of death. #Model A indicates the base-adjusted model and model B indicates the model additionally adjusted for the development of an abscess complication. ††Total antibiotic duration was analyzed among treatment completers.