Palliative home parenteral nutrition: Clinical service evaluation and identifying potential prognostic factors to assist with patient selection.

BACKGROUND AND AIMS Home parenteral nutrition (HPN) for palliation has little evidence supporting existing guidance. Patient selection remains challenging. We aimed to evaluate use of palliative HPN in our service against ESPEN guidance, and to identify potential prognostic indicators. METHODS Palliative care patients commenced on HPN were identified. Medical notes, computer records and HPN database were accessed to identify patient demographics, primary diagnosis and aetiology of intestinal failure, blood test results potentially associated with prognosis (eGFR, albumin, CRP, Hb), presence of ascites, and PN duration. By dichotomising blood results Kaplan-Meier survival plots were derived to identify potential associations with survival. RESULTS From the HPN database of 111 patients, 20 (18%) were identified as palliative. Six were male (30%), median age (interquartile range (IQR)) 56.4 (51.5-66.8) years. Four patients commenced palliative HPN between 2000 and 2006, while 2007-2013 there were 16. The median number (IQR) of nights on HPN was 85 (19-352). The most common indication was gastro-intestinal obstruction (n = 13, 65%) and short bowel syndrome following palliative surgical resection (n = 4, 20%). Kaplan-Meier survival plots identified worse prognosis on HPN if the presenting albumin was ≤30 g/L p = 0.016. CONCLUSION The use of HPN in palliative care is increasing. Current patient selection meets with ESPEN guidance with respect to aetiology of intestinal failure and length of survival on PN. We suggest that a low albumin (not a marker of malnutrition) may help to predict those who are likely to survive less long on palliative HPN. A multi-centre prospective study, also examining quality of life would help define improved guidance