Introduction Atlantaclassification stratifies acute pancreatitis (AP) into mild and severe. Severe acute pancreatitis (SAP) is best managed in HDU or ITU setting and associated with high mortality and morbidity despite best efforts at attaining early diagnosis and timely intervention.

Aim To compare management strategies and mortality of patients admitted to ITU with SAP against national standards and study the group who succumbed to their disease in detail in an attempt to define the circumstances that lead to this event and identify the most accurate prognostic indicators in this group of patients.

Methods Retrospective audit of management and outcome of consecutive patients admitted to ITU with SAP during the period of 2007–2010. The development of necrosis, infected necrosis (IN) or organ failure (OF) was recorded. Patients were classified into group I (No necrosis or OF), group II (sterile necrosis or transient OF), group III (IN or persistent OF) and group IV (infected necrosis and persistent OF). The four groups were compared regarding the clinical course, radiological/surgical intervention, any post-intervention complications, use of antibiotics/antifungal and nutritional support.

Results 51 patients were admitted to ITU with SAP (APACHE II >8, modified Glasgow score >3) during the period of 2007–2010. All cases fulfilled the Atlanta criteria of SAP. Median age: 66±17.5. The pancreatitis was alcohol induced in 12% and due to gallstones in 59% of patients; no cause was found in 25% of patients. Median ITU stay was 3.23 days. The overall mortality rate during the study period (3 years) was 38% (n−19) above national standard of 30%. All seven patients in group IV died, five of them underwent necrosectomy and one had CT guided drainage of infected acute fluid collection. The Abstract PWE-135 table 1 shows the total number of patients and respective mortality of SAP in all four groups. Forty-one patients (80%) received antibiotics and 35 patients (69%) had nutritional support but neither of them seems to have a significant impact on survival (p=0.6 and 0.06 respectively). Outcome (death) correlated with organ dysfunction criteria (Atlanta criteria and APACHE II score).

Conclusion While the presence of infected necrosis or persistent organ failure in SAP (group III) is associated with high mortality, the combination of “infected necrosis and persistent organ failure” (group IV) is uniformly fatal. Further research is necessary to confirm the findings in our study and to explore ways of optimising patients in group III to improve survival.

Competing interests None declared.