COMPLICATIONS - INFECTIONS

A. Eid¹, S. Waller¹, L. Clough¹, R. Gilroy^2
1Infectious Diseases, Kansas University Medical Center, Kansas City/KS/UNITED STATES OF AMERICA, ²Hepatology And Transplantation, Kansas University Medical Center, Kansas City/KS/UNITED STATES OF AMERICA

Body: Introduction: It is not uncommon that collections are drained in the earlier post-operative period related to concerns for untreated infection. The aim here is to identify predictive clinical or laboratory markers for the presence of an intra-abdominal abscess.
Methods: Between 2004 and 2009 thirty-one patients were identified for having undergone percutaneous drainage of a fluid collection during the first year following liver transplantation (LT). The reason for drainage of the collection was a clinical suspicion of infection or lack of response to empiric antibiotic therapy. Multiple clinical, laboratory, operative, donor and therapeutic variables were collected. Cases were patients in whom the aspirated fluid collection demonstrated positive cultures while controls were those with negative cultures. At the time of fluid aspiration, 88% of patients were on antibiotics. Results: The mean of age of the patients was 51.4 years and 54.8% were male. Hepatitis C (26.5%), cryptogenic cirrhosis (23.5 %) and primary sclerosing cholangitis (14.7%) were the most common indications for liver transplantation. Twelve out of 31 patients who underwent drainage procedure had positive culture. The most commonly cultured organism was Enterococcus (44%), all vancomycin resistant, followed by Klebsiella and Candida. Roux limb biliary reconstruction, reoperations were not more common among patients with proven infection. The site of the drained fluid collection, induction therapy, and the level of immunosuppression did not predict infection. Although not statistically significant, donor warm ischemia time tended to be longer in patients diagnosed with infected collection (73.5 min vs 59.2 min; p=0.15). Recipient clinical variables including MELD score failed to predict whether a fluid collection was infected or not. A temperature > 38.3 was present in 25% and 22% of patients with and without infected collection, respectively. An elevated white blood cell count (>11,000 cells/dL) was present in 92% of patients with positive culture (mean 19,880 cells/dL) versus 58% of those with negative culture (mean 12,520 cells/dL). CT image review shows that the average Hounsfield units (HU) did not predict whether an aspirated collection yielded a positive culture (12.98 HU vs 15.03 HU).Conclusions: We were not able to identify reliable markers topredict whether a fluid collection harboured an active infection. This study highlights that variables commonly associated with infections in non-immunosuppressed patients appear unreliable in livertransplant recipients on immunosuppression.

Disclosure: All authors have declared no conflicts of interest.