Objectives: The objective of this Review was to locate, appraise and summarise evidence from scientific studies on home parenteral nutrition (HPN) in order to answer specific research questions on the effectiveness of this technology. The following questions were asked. What patients have received HPN? What has been the experience of patients on HPN programmes? How have HPN programmes been organised, and what techniques and equipment have been used, and to what effect? What comparative information is available on effectiveness? What evidence exists for the cost-effectiveness of HPN? What questions about the provision of HPN could be answered with additional research, and what studies would be most suitable?
Data sources: A comprehensive list of studies was provided by an extensive search of electronic databases (including MEDLINE, Embase, Science Citation Index, Uncover, Cinahl, Caredata, Food Science and Technology Abstracts, NTIS, Pascal, Psychlit, and Economic Literature Index), relevant journals (including Journal of Parenteral and Enteral Nutrition, Clinical Nutrition, American Journal of Clinical Nutrition, Nutrition, Clinical Gastroenterology, Nutrition Reviews, Annals of Nutrition and Metabolism, Nutrition and Cancer, Nutrition and Health, and Journal of Paediatric Nutrition and Metabolism), and scanning of reference lists, as well as other search strategies outlined in the protocol.
Study selection: Studies relevant to the questions were selected. The inclusion criteria were fairly broad because of the quality of the studies located.
Data extraction: Data extraction forms were used to collect data from studies included in the review. The data was checked by a second researcher to reduce error.
Data synthesis: Quantitative analysis was difficult owing to the type of studies located. The data is discussed in a qualitative manner. Where complication rates have been given, we have attempted to combine the results in a quantitative manner.
Results: The age and sex of patients on HPN varies according to the underlying disease but, on the whole, patients are young (see Tables 4a and 4b). There are trends showing an increased use of the technology at the extremes of the age range. There are marked differences between countries on the underlying diseases for which HPN is indicated. For example, many more patients with an underlying malignancy are treated in Italy and the USA than in the UK (40-67% versus 8%). Morbidity rates for the majority of patients are acceptable (see Table 8), the complications tend to be related to the central venous catheter. It is fairly clear that a minority of patients are susceptible to recurrent problems and that many patients have very few complications. The mortality rate for HPN patients (see Table 10) was good for those patients with benign underlying disease (for example, 5% of Crohn's HPN patients die per year), and there are very few reports of patients dying from complications of the technology. The survival of those with malignant disease and AIDS is poor, almost all having died from the underlying disease at one year; despite this, most programme growth worldwide is due to an increase in the numbers of patients with these diagnoses (see Table 5). Quality of life is reasonable for patients with benign disease (see Table 9); no studies were found that examined the quality of life of HPN patients with malignant disease. Economic analysis shows that the cost of HPN treatment is cheaper than the alternative of in-patient care (see Table 18). There is a paucity of comparative studies examining different aspects of the technology, and this accounted for the majority of gaps in the evidence.
Conclusions: The use of HPN for benign intestinal failure is supported by evidence from the scientific studies located. There are, however, large gaps in the evidence, particularly relating to the use of HPN in malignant disease and AIDS. A programme of research is suggested at the end of this review.