Introduction Untreated, pyogenic liver abscesses have a mortality approaching 100%. Three admissions in a week sparked interest in the best management of this condition. Although common, no comprehensive management guidelines could be found, prompting further review into how well this condition was managed locally

Methods Retrospective analysis of all patients admitted to Watford Hospital between 2006 and 2011 with a diagnosis of pyogenic liver abscess. Data was collected to evaluate use of cultures, radiological intervention (aspiration or drain insertion), source of infection, investigation for cause, follow up and outcome

Results Fourty four admissions were identified: 39 patients with 5 re-admissions. Mean age was 62 yrs, 59% male, 41% female. Eleven patients were managed by the Gastro team.

Assumption of source was made on CT imaging results: 46% presumed portal translocation (most diverticular disease), 36% biliary, 18% unidentified. Blood cultures were taken in 24 patients (42% positive). Abscess aspirates were taken in 33 cases, sent for culture in 30 (50% positive). Presumed biliary or unidentified sources grew gramme negative organisms in 12/13 cases. Presumed portal sources grew gramme positives in 7/8 and anaerobes 1/8.

Abscess size was < 3 cm in 5 cases (incl. 2 readmissions). Four received anitbiotics (Abc) alone:resolution in 3/4, 1/4 no follow up. One was managed to resolution with Abc and aspiration. Mean length of stay was 11 days.

In 6 patients the abscess was 3–5 cm. In this group, 1 patient with malignancy died, 1 treated successfully with Abc alone. The remaining 4 were treated with Abc and aspiration: 1/4 resolution, 1/4 readmitted, 2/4 no follow up. Mean length of stay 15 days.

Thirty three patients had abscesses > 5 cm (incl. 3 readmissions). Nineteen were treated with Abc and drainage: 2/19 had underlying malignancy and died, 6/19 resolution, 3/19 readmitted, 7/19 no follow up, 1/19 referred to surgery. Of the remaining fourteen, 3/14 had Abc alone (2 resolved, 1 patient with two readmissions no follow up), 1/14 a readmission referred for surgery and 10/14 Abc and aspiration. Outcome in these ten: 1/10 mutiple aspiration, 2/10 drain insertion, 1/10 surgical referral, 1/10 readmitted, 1/10 partial response, 3/10 no follow up, 1/10 resolution. Mean length of stay in > 5 cm group : 27 days.

Overall in all patients: investigation for cause limited to 3 colonoscopies and 1 ERCP, follow up inadequate in 15 (38%), mortality 8%.

Conclusion Management of this serious condition was remarkably poor, with limited use of cultures, inconsistant radiologic intervention, no search for cause, and scanty follow up. Despite this confirmed mortality was just 8%.Guidelines for future management have been drawn up including recommendation that all patients are looked after by the Gastro team.

Disclosure of Interest None Declared