In the 1920s, continuous infusion of i.v. glucose was introduced in humans. It was not until the 1960s that parenteral nutrition was used, first in seriously ill adult surgical patients and then in children and adults with short bowel syndrome (3). These children, who before these therapies died of starvation, were able to live for years, sustained with artificial nutrition. Parenteral nutrition use then expanded to many other patient populations, often without clear or well-established indications. Only in the last decade has some light been shed in critical care settings as to when TPN is beneficial and when it may be more harmful (4,5).

In the late 1970s, gastrostomies began to be performed and were often used for swallowing problems in children (6). Their use became widely generalized to adults such that EN is now commonly used among patients with stroke, neurologic disease, and cancer (7). Between 1988 and 1995, the number of tubes placed in the United States doubled; in 2000, more than 216,000 tubes were placed. Recent data from Veterans Administration (VA) and Medicare database reviews suggest a stabilization of this trend in some settings (Fig. 60.1).

Whereas i.v. hydration is a well-established part of medical practice, its role in end-of-life care remains less clear. Intravenous hydration is often used for the treatment of terminal delirium and agitation. Whether it is beneficial, and if so, for what outcomes and at what rate of infusion, remains an area of controversy.

A number of retrospective studies and a few prospective studies have been performed in patients to ascertain survival benefit in patients receiving artificial nutrition. The largest study to date was a retrospective review by Grant et al. of 81,105 Medicare beneficiaries who received gastrostomies in 1991. No comparison group was identified for this study. Cerebrovascular disease, neoplasms, fluid and electrolyte disorders, and aspiration pneumonia were the most common primary diagnoses. The mortality rate at 1 and 3 years was 63.0% to 81.3%. The median survival was 28.9 weeks for women and 17.6 weeks for men. At 30 days, primary diagnoses of malnutrition and fluid and electrolyte disorders and secondary diagnoses of swallowing disorders, dementia, or cerebrovascular disease were characterized by the lowest mortality rates. Thirty-day mortality rates were highest among those with primary diagnoses of nonaspiration pneumonia or influenza and secondary diagnoses of congestive heart failure or any neoplasm (12). Rabeneck et al. studied 7,369 patients receiving PEG tubes at VA facilities between 1990 and 1992. In this retrospective cohort study, 23.5% died during their index hospitalization. The median survival of the full cohort was 7.5 months from the time of tube placement. The overall mortality rates at 1, 2, and 3 years were 59%, 71%, and 77%, respectively. The highest mortality rates were observed for patients with lung or pleural cancer (46.4%), followed by esophageal cancer (20.8%) and head and neck cancer (18.8%) (17). Survival decreased with increasing age. The median survival across clinical diagnostic categories was 13.9 months for cerebrovascular disease, 13.4 months for other organic neurologic diseases, 9.6 months for nutritional deficiency, 8.0 months for head and neck cancer, and 4 or fewer months for all other cancers. Among 674 older
(age >50 yr) adults referred to a community gastroenterology group for PEG insertion over a 10-year period, mortality rates at 1, 2, and 3 years were 54.3%, 73.2%, and 84.5%, respectively. Like Rabeneck’s study, the overall median survival was between 6 months and 1 year; however, those receiving tube feeding in this cohort were much more likely to be patients with stroke or other neurologic conditions. Very few of the PEGs were placed in patients with cancer. Risk factors for mortality in this cohort were being male, having feeding difficulty, having diabetes, being referred from a hospital, and being 80 years of age or greater (18). Similar to Grant’s cohort, dementia was not an independent risk factor for decreased survival. Because these studies used large databases, identification of truly comparable control groups would have been challenging. Nevertheless, because there was no comparable control group, the impact of gastrostomies on survival could not be ascertained.

The small body of literature evaluating the patient symptoms at the end of life suggests a relatively low prevalence of hunger (19). In McCann’s study of 32 patients in a comfort care unit, 63% denied hunger entirely, while 34% reported
hunger during the first quarter of their course in the unit. In all patients reporting either hunger or thirst, these symptoms were consistently and completely relieved by oral care or the ingestion of small amounts of food and fluid. In a case study of a patient refusing nutrition and hydration, the patient experienced no discomfort and died peacefully (19). A survey of Oregon hospice nurses found that among those who had cared for patients who declined nutrition and hydration, the majority reported the ensuing death to be peaceful (20).

Studies of healthy volunteers engaged in fasting report resolution of hunger in <24 hours. The resulting ketosis is associated with the relief of hunger and a mild euphoria. Animal studies suggest that ketosis may also have a mild analgesic effect. When ketosis is minimized by small feedings, hunger may persist (21).

There is some evidence that malnutrition is positively correlated with pressure ulcer incidence and severity (22,23). However, only two nutrition intervention studies for pressure ulcer prevention or treatment included artificial nutrition. Hartgrink et al. performed a randomized controlled trial (RCT) of 140 patients with fracture of the hip and an increased pressure ulcer risk. The intervention group was treated with a standard hospital diet and an additional NGT feeding administered overnight. The comparison group received the standard hospital diet alone (24). No significant difference was found between the groups. Ohura et al. utilized artificial nutrition in a trial that enrolled 60 tube-fed patients. The treatment comparison, however, was protein and micronutrient intake, not tube feeding (25). Thus, while it is not possible to draw any firm conclusions on the effect of enteral and parenteral nutrition on the prevention and treatment of pressure ulcers, there is growing evidence that disease-specific nutrition support may offer benefit with regard to pressure ulcer incidence and healing (26).