• Rhys Thomas, Rana Madani, Tim Worthington, and Nariman Karanjia.
• Department of Hepatopancreaticobiliary Surgery, Royal Surrey County Hospital, Guildford, Surrey, UK.
• Hepato Gastroenterol. 2012 Nov 1; 59 (120): 2421-7.

Background/AimsP-POSSUM predicts operative risk and mortality, although few reports describe its use in elective liver resection. We analysed P-POSSUM scores in patients undergoing resection for colorectal liver metastases, non-colorectal cancer and benign liver dis-ease.MethodologyData for 664 elective liver resections were included between 1998 and 2009; 480 for colorectal metastases; 96 for benign and 88 for non-colorectal disease. Peri-operative management was standardised according to unit protocol.ResultsP-POSSUM predicted mortality was 3.73%, 6.27% and 9.99% for benign, colorectal and non-colorectal can-cer patients. 60-day mortality was 1.04%, 1.88% and 6.81%, p<0.001. There was correlation between predicted and actual mortality in non-colorectal patients, p=0.029 and weak correlation between physiological score and mortality in each group. There were high-er physiological scores in patients with morbidity and higher predicted mortality in colorectal patients experiencing morbidity, p=0.014. There was correlation be-tween observed and expected operative mortality in benign and non-colorectal patients. Long-term survival was unaffected by various aspects of the P-POSSUM system.ConclusionsPredicted mortality is over estimated in patients undergoing liver resection. Pre-operative,physiological score and predicted mortality may have a role predicting morbidity. P-POSSUM is ineffective in predicting prognosis following colorectal cancer metastases resection, although may be used to stratify risk in high risk cancer patients.

29 keywords...

hide…