Introduction The demand for home parenteral nutrition (HPN) in advanced malignancy is increasing. This coincides with recent guidance recommending an explanation of the benefits, burden and risks of clinically assisted nutrition or hydration towards the end of life with patient and family.1 Furthermore, a 2nd expert opinion is required if clinically assisted nutrition is withdrawn or not commenced in end of life care. HPN related complications might be greater in malignant disease due to impaired immunity and prothrombotic tendency. The benefits and risks of HPN in malignant disease are currently unknown.

Methods To compare HPN and non-HPN related complications and outcomes in patients receiving HPN for malignant indications (MI) compared to non-malignant indications (non-MI). A retrospective case note review was undertaken from a comprehensive database from 1995 to 2010. Data collected included indication for HPN, time on HPN, cause and duration of hospital readmissions (HPN related defined as catheter-related sepsis, line blockage or thrombus) and reason for discontinuation. Data was standardised by presenting complication rates and days per year on HPN.

Results A total of 88 patients were identified. Fifty-one (58%) case notes were retrieved, of which 15(29%) had a MI and 36 (71%) non-MI. The median age of the cohort was 50 years (range 19–82); 23 (45%) male. The median time on HPN for MI was 164 days (range13–1089) which did not differ from non-MI; 156 days (range3–1117) (p=0.75).

The number of HPN related complications did not differ in MI compared to non-MI; 5 and 3.6 per HPN year respectively (p=0.33). There was a trend towards increased mean time in hospital due to HPN related complications per HPN year; MI 18 days (range 0–84) and non-MI 6 days (range 0–75) (p=0.06). There was no difference detected in mean time in hospital per HPN year due to non-HPN related complications; non-MI 26days (range 0–156) compared with non-MI 37 days (range 0–365).

HPN was discontinued in patients with MI due to; palliative care 9 (60%), death 3 (20%), enteral feed (13%). This did not differ significantly to non-MI; surgery 12 (33%), enteral feed 9 (25%) and death 5 (19%). Significantly fewer (5.6%) patients with non-MI were referred to palliative compared to MI (p=<0.001).

Conclusion Patients with advanced malignancy spend similar time on HPN and do not have greater complication rates compared to patients on HPN for non-MI. This suggests that HPN should be considered in end of life care as a result of advanced malignancy. A better understanding of the effects of HPN on quality of life in this setting is needed.