What organisms cause pyogenic liver abscess in Korea?

Published 2026-09-05 · Updated 2026-09-05 · Last checked against sources 2026-09-05

Before I looked it up

Liver abscess is, in fact, rarely something I diagnose myself. Most of the patients admitted to our ICU are ones who have had an abdominopelvic CT taken in the emergency room and come up with it; on the wards too, when clinical symptoms appear, a CT is taken, the diagnosis is made, and they come to the ICU. When the abscess is easy to approach and large in size, radiology performs a PCD as well and the patient is admitted afterwards; when it does not meet the indication for PCD, we end up giving antibiotics only. When the organism has not been identified at first, we usually have to cover anaerobes together as well. My job is to give the antibiotics and follow up the clinical symptoms, the labs and the imaging. I tried putting together the current knowledge on liver abscess.

What I found

The organism you should be treating for depends on where you are standing. In Korea the answer is Klebsiella pneumoniae, and it is not close.

• Korea and East Asia: K. pneumoniae first. In a two-centre Korean series of 402 liver abscesses, 61.2% were culture-positive; among those, K. pneumoniae (n = 133) outnumbered E. coli (n = 74) roughly two to one. A single-centre Korean series of 231 patients found K. pneumoniae in 69.9% of abscess isolates and 74.2% of blood isolates.
• The classical Western pattern is different — biliary-source, E. coli, often polymicrobial. This is background rather than a figure I can source precisely, so treat the contrast as a direction, not a percentage.
• Metastatic spread is the thing that makes K. pneumoniae different. Endophthalmitis, meningitis and necrotising fasciitis follow the abscess, and they are what cost the patient an eye.
• Cover anaerobes. In Korean population data this is not a theoretical point — see below.
• Drain it. Antibiotics alone are not the treatment.

How often the metastatic complications actually happen

Two numbers get quoted interchangeably and they are not the same number.

The meta-analysis is the one to quote when you are looking at a Klebsiella abscess — about 1 in 22. K1 capsular serotype was an independent risk factor. The heterogeneity was high (I² = 94.2%), so the confidence interval matters more than the point estimate.

The practical consequence does not change either way: ask about vision, and get ophthalmology in early if the answer is not clean. Endophthalmitis is diagnosed late by the patient’s account or not at all.

Why the abscess is Klebsiella in the first place

• Diabetes. In Korean national data, 37.24% of liver abscess patients had diabetes and 26.5% had a malignancy.
• Hypervirulent capsular types. K1 and K2 serotypes, hypermucoviscous phenotype, string test positive. This is the invasive syndrome described in Taiwan and Korea from the 1980s onward — abscess plus distant seeding in a patient with no biliary disease.
• Incidence in Korea is rising, not falling: 5.7 per 100,000 in 2007 to 14.4 per 100,000 in 2017, mean 10.9 per 100,000 across the period.

The standard approach, in one table

Expanded below. The right-hand column is the point — some of this is guideline, some is convention quoted as though it were.

Empiric therapy

A third-generation cephalosporin plus metronidazole covers the Korean epidemiology. Doses below are the adult IV doses from the 2010 SIS/IDSA complicated intra-abdominal infection guideline (Table 6) — see the caveat about that guideline’s status further down.

Two things worth flagging in that table:

• Piperacillin-tazobactam is 3.375 g q6h, not 4.5 g q6h. The 4.5 g every 6 h figure is the escalated dose the guideline reserves for Pseudomonas aeruginosa (footnote b), along with 3.375 g every 4 h. If you carry 4.5 g q6h in your head as the default, it came from the pseudomonal footnote.
• Ceftriaxone 2 g q24h and metronidazole 500 mg q8h — the doses most people actually write — sit inside the guideline ranges.

The anaerobic half is not optional. In a Korean national analysis of 30,690 liver abscess patients, in-hospital mortality was 7.9% with anaerobe-covering antibiotics versus 15.6% without, and the adjusted odds ratio for death with anaerobic coverage was 0.42 (95% CI 0.38–0.46) after adjusting for age and comorbidity. This is observational claims data, not a trial, and confounding by indication runs in the obvious direction — but the effect size is large and it points the same way as the microbiology.

When to reach for a carbapenem. ESBL risk is the usual trigger. Worth knowing which abscess it attaches to: in the 402-patient Korean series, ESBL production, biliary disease, liver disease and malignancy were all more frequent in E. coli abscess than in Klebsiella abscess. The ESBL question is more an E. coli question than a Klebsiella one.

Source control

• Percutaneous catheter drainage beats needle aspiration. Pooled across 15 RCTs and 1,676 patients, catheter drainage had a higher success rate (RR 1.23, 95% CI 1.12–1.36), faster 50% cavity reduction and faster clinical improvement. Major complication rates were the same.
• The 5 cm threshold that everyone quotes is convention rather than a trial result — it is the entry criterion several of these studies used, not an outcome they tested. Treat it as a reasonable default, not a rule with evidence behind it.
• Multiloculated abscesses, failed percutaneous access, and ongoing sepsis are where surgical drainage enters.

Duration — where the note I started from was wrong

This is the part worth reading twice, because the number in my own note did not survive checking.

• The 2010 SIS/IDSA guideline says antimicrobial therapy of established intra-abdominal infection should be limited to 4–7 days, “unless it is difficult to achieve adequate source control,” and states that longer durations have not been associated with improved outcome (recommendation 76, B-III).
• 4–6 weeks is not an IDSA recommendation for intra-abdominal infection. It is what is conventionally done for liver abscess specifically, and it lives inside that source-control exception rather than contradicting it.
• What can be said with data: in a French cohort, observed antibiotic duration for pyogenic liver abscess was a median of about 6 weeks in both transplant and non-transplant patients. That is practice, described — not a duration that has been tested against a shorter arm.

So: long courses are defensible for liver abscess because source control is usually incomplete. They are not defensible by citing an intra-abdominal infection guideline that says 4–7 days.

Outcome

• In-hospital mortality in Korean national data: 9.6%, stable across 2007–2017. A single-centre Korean series reported 6.9%.
• Mortality rose with age and was driven by comorbidity — malignancy, diabetes, chronic kidney disease.

Amoebic liver abscess

Different organism, different drug. Entamoeba histolytica is rare in Korea but travel history changes the question, and the treatment is metronidazole plus a luminal agent rather than antibacterial therapy aimed at Klebsiella. Worth excluding before committing to a long antibacterial course in someone with the right exposure history. This paragraph is background; I have not sourced it here.

What I take from it

1. In Korea, treat for Klebsiella first. The Western teaching that liver abscess is a biliary E. coli disease does not describe the patients in front of me.
2. Cover anaerobes anyway — the Korean mortality data on this is more striking than I expected.
3. Ask about the eyes. About 1 in 22 with Klebsiella liver abscess, and the patient will not volunteer it.
4. Send anaerobic cultures. Nearly 40% of these come back with no organism at all, and missing the anaerobic bottle is a self-inflicted version of that.
5. The 4–6 week course is convention, not guideline. Say so honestly when handing over.

A note on the guideline I was citing

My original note cited “IDSA guidelines for intra-abdominal infections” with no edition. Checking it turned up something worth knowing:

• The current IDSA document is the 2024 update, Part 1 — risk assessment, diagnostic imaging and microbiological evaluation. It does not cover antimicrobial regimen selection.
• The 2010 SIS/IDSA guideline, which does carry the dosing table and the duration recommendation, is listed by IDSA as archived.
• So as of this note’s review date there is no current IDSA recommendation on which antibiotic to choose for complicated intra-abdominal infection. The 2010 numbers above are the most recent IDSA-published doses, and they are old enough that local susceptibility data should outrank them.

The part I found interesting

The common causative organisms differ between the East and the West (KP, E. coli), and it is said that this is because in the West there are many cases of ascending biliary infection — that is, an abscess in a person who has hepatobiliary disease — whereas in the East Asian type there are many cases that arise on their own without a biliary lesion (the route presumed to be entry from the bowel through the portal vein..). Interesting.

Sources

Doses, thresholds and rates above are traceable to the following. Everything not tied to one of these is background and is written as background.

1. Solomkin JS, Mazuski JE, Bradley JS, et al. Diagnosis and management of complicated intra-abdominal infection in adults and children: guidelines by the Surgical Infection Society and the Infectious Diseases Society of America. Clin Infect Dis. 2010;50(2):133–164. doi:10.1086/649554 — Table 6 (adult IV doses), recommendation 76 (4–7 days). Archived by IDSA.
2. 2024 Clinical Practice Guideline Update by the Infectious Diseases Society of America on Complicated Intra-abdominal Infections: Risk Assessment, Diagnostic Imaging, and Microbiological Evaluation in Adults, Children, and Pregnant People. Clin Infect Dis. 2024. doi:10.1093/cid/ciae346 — Part 1; does not address antimicrobial selection.
3. Yoo JJ, Lee TK, Kyoung DS, et al. A population-based study of pyogenic liver abscess in Korea: incidence, mortality and temporal trends during 2007–2017. Liver Int. 2021;41(11):2747–2758. doi:10.1111/liv.15034 — incidence, diabetes 37.24%, malignancy 26.5%, metastatic infection 1.74%, mortality 9.6%.
4. Myeong JH, Kyoung DS, Park MA, et al. Anaerobe coverage is important for the prognosis of pyogenic liver abscess: a population-based study in Korea. J Infect Public Health. 2022;15(4):425–432. doi:10.1016/j.jiph.2022.03.003 — n = 30,690; mortality 7.9% vs 15.6%; aOR 0.42 (0.38–0.46).
5. Yoon JH, Kim YJ, Kim SI. Prognosis of liver abscess with no identified organism. BMC Infect Dis. 2019;19:488. doi:10.1186/s12879-019-4131-z — n = 402; 61.2% culture-positive; K. pneumoniae 133 vs E. coli 74; ESBL pattern.
6. Sohn SH, Kim KH, Park JH, Kim TN. Predictors of mortality in Korean patients with pyogenic liver abscess: a single center, retrospective study. Korean J Gastroenterol. 2016;67(5):238–244. doi:10.4166/kjg.2016.67.5.238 — K. pneumoniae 69.9% of abscess isolates, 74.2% of blood isolates; mortality 6.9%.
7. Hussain I, Ishrat S, Ho DCW, et al. Endogenous endophthalmitis in Klebsiella pneumoniae pyogenic liver abscess: systematic review and meta-analysis. Int J Infect Dis. 2020;101:259–268. doi:10.1016/j.ijid.2020.09.1485 — pooled incidence 4.5% (95% CI 2.4–8.2), 11,889 patients, K1 as risk factor.
8. Siu LK, Yeh KM, Lin JC, Fung CP, Chang FY. Klebsiella pneumoniae liver abscess: a new invasive syndrome. Lancet Infect Dis. 2012;12(11):881–887. doi:10.1016/S1473-3099(12)70205-0 — invasive syndrome, capsular types, metastatic complications.
9. Al-Sayaghi KM, Alhujaily M, Zaky MK, et al. Percutaneous needle aspiration versus catheter drainage in the management of liver abscess: an updated systematic review and meta-analysis. ANZ J Surg. 2023;93(4):840–850. doi:10.1111/ans.18129 — 15 RCTs, 1,676 patients; success RR 1.23 (1.12–1.36).
10. Lafont E, Roux O, de Lastours V, et al. Pyogenic liver abscess in liver transplant recipient. Transpl Infect Dis. 2020;22(6):e13360. doi:10.1111/tid.13360 — observed antibiotic duration median ~6 weeks.