Article Text

Download PDFPDF

PTH-122 Home Parenteral Nutrition In Palliative Care: A Retrospective Study To Identify Potential Prognostic Indicators
  1. B Drinkwater,
  2. B Clarke,
  3. S Cooper
  1. Russell’s Hall Hospital, Dudley, UK


Introduction The use of home parenteral nutrition (HPN) in the palliative setting is less common in the UK than in Europe and the USA. It remains controversial with little evidence base supporting the existing guidance. Patient selection remains challenging. We aimed to identify potential prognostic indicators in palliative care patients receiving HPN using a retrospective analysis.

Methods Palliative care patients commenced on HPN were identified. Medical notes, computer records and the HPN database were accessed to identify patient demographics, primary diagnosis and aetiology of intestinal failure, blood test results potentially associated with prognosis (eGFR, albumin (not employed as a marker of malnutrition), CRP, Hb), presence of ascites prior to starting HPN, and total number of nights on HPN. By dichotomising the blood results into CKD stage 3 or worse (eGFR <60), albumin ≤30, CRP >50, Hb <10, we were able to derive Kaplan-Meier survival plots to identify statistical significance associated with survival.

Results From the HPN database of 111 patients, 20 (18%) were identified as palliative. Six were male (30%), and the median age (interquartile range (IQR)) was 56.4 (51.5–66.8) years; range 25.2–81.6. Between 2000 and 2006 four patients commenced palliative HPN, while 2007–2013 there were 16. At the time of data collection 80% (16/20) had died. The median number (IQR) of nights spent on HPN was 85 (19–352). The most common indication for HPN (aetiology of intestinal failure) was gastro-intestinal obstruction (n = 13, 65%) and short bowel syndrome following palliative surgical resection (n = 4, 20%).

The median (IQR) eGFR was 90 (53.5–90), with eGFR <60 n = 4, median (IQR) albumin 30 (24.5–35); low albumin ≤30, n = 10.; median Hb 10.4 (9.3–10.9);, Hb <10 n = 7; median (IQR) CRP 10.5 (3.8–89.3), high CRP >50 n = 4. Of the 14 available case notes reviewed, there was evidence of ascites prior to starting HPN in 35.7% (5/14).

Kaplan-Meier survival plots identified a statistically significant shorter prognosis on HPN if the presenting albumin was ≤30 p = 0.016, median 66 (5–115) days, and similarly if the CRP >50 p = 0.04, median (IQR) 98 (56–122). No significance was found with low Hb, CKD stage 3 or worse or presence of ascites when compared with prognosis by Kaplan-Meier survival plots.

Conclusion We conclude that a low albumin (not a marker of malnutrition) and high CRP may help to predict those who are likely to survive less long on palliative HPN. When examining median days this appears most pertinent with a low albumin. There does not appear to be any significant association between ‘presence of ascites’, a low Hb or a low eGFR and survival time on HPN. While this study does not examine quality of life, the use of albumin and CRP may help direct whether HPN may be of benefit in the palliative setting.

Disclosure of Interest None Declared.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.