Skip Navigation LinksHome > January 2008 - Volume 26 - Issue 1 > Artificial Nutrition and Hydration for the Terminally Ill: A...
Home Healthcare Nurse:
doi: 10.1097/01.NHH.0000305551.55467.24
Hospice & Palliative Care

Artificial Nutrition and Hydration for the Terminally Ill: A Reasoned Approach


Free Access
Article Outline
Collapse Box

Author Information

Paula M. Suter, MA, RN, is Disease Management/Telehealth Program Manager at Baptist Health Home Health and Hospice, Little Rock, AR.

Janice Rogers, RN, is Clinical Practice Specialist in Palliative Care at Baptist Health Home Health and Hospice, Little Rock, AR.

Carmen Strack, RD, LD, CNSD, is Clinical Dietitian in Oncology at Baptist Health Medical Center, Little Rock, AR.

Address for correspondence: Paula M. Suter, MA, RN, Baptist Health Home Health and Hospice, 11900 Colonel Glenn Road, Suite 2000, Little Rock, AR 72210 (e-mail:

Collapse Box


Knowledge about evidence-based research findings on the physiologic effects of artificial nutrition and hydration (ANH) can be used as a framework for assisting patients and families with decision making at the end of life. This article provides a review of the credible evidence related to ANH obtained through scientifically sound research, supplies a framework for discussion about ANH with patients and families, and lists basic educational recommendations and resources that both home health and hospice nurses can use to assist with patient and family supportive efforts. The charge to nurses caring for dying patients and their families is consistently to honor patient beliefs about life—their goals and values—in a supportive, unbiased, and informational manner, which will improve the quality of life—at the end of life.

Mr. Jones is a 64-year-old man with recurrent esophageal cancer. By his decision, he is no longer receiving chemotherapy and has been referred to hospice for care at home. He has a percutaneous endoscopic gastrostomy (PEG) tube in place and is receiving continuous tube feedings at 60 ml per hour. Mr. Jones is tolerating the feedings well and can move from bed to chair with minimal assistance. He asks his hospice nurse for advice as to whether he should continue with the feedings or request that they be discontinued to allow for greater mobility.

Mrs. Smith is 72-year-old woman with advanced pancreatic cancer. She has just been discharged from the hospital with home health services and is receiving total parenteral nutrition (TPN). Mrs. Smith is experiencing severe nausea and vomiting. The oncologist has told her family that she has only a few weeks to live. The family asks the home health nurse's opinion about what is causing the nausea and vomiting and wonders whether it is the TPN.

Mrs. Brown is 91-year-old woman with end-stage dementia. She is now refusing food and is sleeping 80% of the time. Her family feels that she has had a rich and full life and do not think she would want a feeding tube. They tell their nurse that they are feeling guilty about this and wonder whether this will lead to more discomfort and be the wrong decision.

The preceding patient scenarios demonstrate the complexity of decision making that patients and families must face and the wide spectrum of situations in which nurses may be called upon for guidance. This places an important responsibility on the shoulders of the caregiving professional to provide credible information, particularly in light of the fact that the literature does not categorically support the practice or nonpractice of ANH.

The high-profile cases of Terri Schiavo and Robert Wendland have clouded both public and professional understanding with regard to the indications and benefits of ANH. Two differing views represent those now commonly held by the 2 groups. One view is that the provision of ANH decreases suffering and discomfort by alleviating thirst and improving energy levels. The opposing view is that when ANH is introduced in advanced illness, it is poorly tolerated and results in increased discomfort.

The dichotomy of beliefs associated with ANH may lead to perpetuation of misinformation in its favor or opposition. The advice and information provided to the families in each of the introductory scenarios will vary depending on nurse familiarity with the ANH literature. At the core of this issue of providing ANH is the ability to differentiate patients in vegetative states (for which the literature clearly demonstrates a survival benefit with the use of ANH) from patients with a terminal illness deemed untreatable.

This article aims (1) to review the credible evidence related to the physiologic effects of ANH obtained through scientifically sound research, (2) to provide a framework for discussion about ANH with patients and families, and (3) to list basic educational recommendations and resources that both home health and hospice nurses can use to assist with patient and family supportive efforts.

Hospice and home health nurses have pivotal roles of teaching, collaborating, and supporting patients and families at the end of life. The provision of ANH for patients with a terminal illness is now commonplace in home settings. A critical mass of scientific evidence demonstrating the benefits or the risks associated with this therapy, as derived from randomized controlled trials, is lacking. Hospice and home health nurses may not be familiar with the few well-designed studies that currently exist.

Originally, ANH was believed to accomplish the following major goals: prevent aspiration pneumonia, provide comfort, prevent hunger and thirst, and prolong life. A true-false questionnaire based on scientific evidence for each of the aforementioned major goals and designed to elicit beliefs regarding goal attainment was administered to 30 hospice and home health nurses employed by Baptist Health Home Health and Hospice. The results showed that home health nurses were divided regarding their beliefs, whereas hospice nurses were completely unified in their belief that ANH leads to further discomfort. Although the survey has many limitations, a basic inference can be made: Beliefs do differ considerably within the field of nursing and may be due to a knowledge gap regarding ANH evidence.

Hospice nurses are continually involved with end-of-life support by nature of their specialty and are more likely to be familiar with ANH evidence. Some patients and families may not have access to hospice services due to service unavailability in their locale. Other patients and families elect not to have hospice services for various reasons. Many home health nurses, therefore, may care for patients at the end of life. Both nursing groups will benefit from familiarity with the ANH research evidence enabling them to provide accurate information and facilitate informed decision making. Guiding those for whom they provide care in an unbiased and informative manner is imperative during these times of high emotion and stress.

Back to Top | Article Outline

Artificial Nutrition and Hydration Defined

The Hospice and Palliative Nursing Association (HPNA, 2004) has defined ANH as described in the following sections.

Back to Top | Article Outline
Artificial Hydration

Artificial hydration involves administration of fluid through nonoral means. Hydration can be administered through intravenous or subcutaneous (also called hypodermoclysis), rectal (proctoclysis), and enteral routes.

Back to Top | Article Outline
Artificial Nutrition

Artificial nutrition involves nonoral, mechanical feeding either by the intravenous or the enteral route. Enteral feedings may be provided through nasogastric tubes, gastrostomy, esophagostromy, or jejunostomy tubes placed either endoscopically (often referred to as PEG tubes) or via open surgical procedures. Intravenous nutrition is administered through a central line and often is called TPN.

It stands to reason that conducting randomized, controlled, double-blind studies in a terminally ill population would pose significant ethical and humane considerations. Williams (2007) discusses these challenges and describes barriers associated with attempting research of this kind such as individual patients feeling too overwhelmed, end-of-life labeling, and family and healthcare professionals acting as gatekeepers. Owens (2007) conducted a review of the evidence on hydration among the terminally ill and concluded that scientific studies are scarce at best and that most literature on the subject was anecdotal in nature.

Many other experts in the field have used methods of evaluation such as retrospective correlational analyses. Some of these analyses have been conducted using small samples from which generalized inferences would be inappropriate. Therefore, these analyses are not included in the following brief review of the literature. Anecdotal accounts also are not included in this review.

Back to Top | Article Outline

Literature Review

Morita et al. (2005) conducted a multicenter, prospective observational study to explore the association between hydration volume and laboratory findings and to evaluate calculated fluid balance and changes in clinical signs of dehydration. The study enrolled 123 terminally ill cancer patients treated in various clinical settings in Japan. The researchers found that active hydration was significantly associated with hypoalbuminemia during the last 3 weeks of life, and that when artificial hydration was not performed, sodium and potassium levels were within normal limits during the last week of life. Calculated fluid balance was not strongly associated with changes in clinical signs of dehydration.

Bruera et al. (2005) randomly assigned cancer patients with clinical evidence of mild to moderate dehydration to groups that received either parenteral hydration with 1,000 ml or placebo with 100 ml of normal saline administered over 4 hours for 2 days. The patients then were evaluated for the target symptoms of hallucinations, myoclonus, fatigue, and sedation as well as global well-being. The researchers found that the treatment group experienced less symptom burden, myoclonus, and sedation and tolerated hydration well.

Burge (1993) conducted a cross-sectional survey of 52 patients in an attempt to estimate quantitatively the experiences of dehydration symptoms in patients with advanced cancer. Patients admitted to 2 palliative care units in Canadian hospitals were asked to assess symptom burden using a self-administered analog scale. Fluid intake for a 24-hour period was estimated by the nurses caring for the study patients. Although Burge (1993) found symptoms to be rated moderately severe, there was no association between thirst, the primary symptom outcome, and fluid intake. This researcher was surprised to find a virtual normal distribution in laboratory values for sodium, osmolality, and urea when dehydration was present.

McCann et al. (1994) monitored mentally aware, competent patients with terminal illness in a long-term care facility from admission until death. The patients received no forced hydration, whereas symptoms of thirst, dry mouth, and hunger were recorded. The researchers reported that 63% never experienced hunger and that 34% did so only initially. They also found that 62% experienced either no thirst or thirst only initially. All the patients' symptoms of hunger, thirst, or dry mouth could be alleviated with small amounts of food, fluids, or ice chips.

Viola (1997) conducted a systematic review of the literature on the effects of fluid status and fluid administration on terminal patients. After a thorough review of pertinent ANH studies published in reputable journals, he concluded that only 6 met the goal of high methodologic quality. The studies conducted by Bruera et al. (2005), Burge (1993), and McCann et al. (1994) referenced earlier were included among the 6, but Viola (1997) listed methodologic concerns for these as well. The consistent lack of control groups and blinding, the small sample sizes, and detection bias (bias in assessment of outcomes) were cited as design flaws.

Finucane et al. (1999) conducted a MEDLINE search of clinical trials from 1966 to 1999, comparing tube feedings with oral feedings. The researchers sought to identify associations with the prevention of aspiration pneumonia, prolonged survival, improved function, or improved palliation between the 2 nutritional methods for patients with dementia. The researchers concluded that there was a lack of data to support tube feedings for patients with dementia. No studies were found that demonstrated a reduction in aspiration pneumonia, and no survival benefit was demonstrated. Finucane et al. (1999) in fact stated that the risks associated with tube feeding in this population were substantial and recommended that the practice be discouraged.

Mitchell et al. (1998) conducted a cohort study of nursing home residents living in Washington state who had chewing and swallowing problems. Clinical characteristics and survival data were compared for residents with and without feeding tubes. The strength of this study was the researcher's use of the minimum data set (MDS) to obtain clinical and functional data. The MDS allowed control to be used for a large number of potential confounding covariates, the major limitation of similar studies. The researchers concluded that residents with feeding tubes experienced poorer survival after 1 year than residents who were not tube fed. This study is of interest because the residents were not suffering from a terminal condition. Therefore, mortality could plausibly be explained by the adverse consequences of the feeding tube itself.

Findings have shown ANH to improve survival among patients in a permanent vegetative state. The Multi-Society Task Force on Persistent Vegetative States (1994) reported that patients may live up to 10 years or more while receiving ANH. When ANH was withdrawn, patients usually died within 10 to 14 days, with the immediate cause of death cited as dehydration or electrolyte imbalance.

Rapp et al. (1983) conducted a prospective randomized controlled clinical trial comparing the effects of TPN and traditional standard enteral nutrition (SEN) on the outcome for 38 patients with head injury during an 18-day period. The authors reported that the outcome for these 2 groups differed greatly, with 8 of the 18 SEN patients dying within 18 days of injury and no patient in the TPN group dying within this period. The authors suggest that the basis for the improved survival of the TPN-treated patients was improved nutritional status, because these patients had a more positive nitrogen balance, a higher serum albumin level, and higher total lymphocyte count. The improved lymphocyte count was cited as the plausible reason for decreased susceptibility to sepsis.

Lee et al. (1998) conducted a retrospective study of 88 patients with advanced head and neck cancer requiring an aggressive radiation therapy regimen. Of the 88 patients, 36 received PEG tubes due to the anticipated radiation toxic side effect of mucositis (inflammation of mucus membranes lining the digestive tract). The remaining 52 patients served as a control group. The authors found that the use of PEG tubes significantly reduced weight loss and hospitalizations due to nutritional or dehydration problems in the treatment group compared with the control group.

It can be concluded from this brief literature review that ANH did not meet the objectives of reducing hunger and thirst or the incidence of aspiration pneumonia, and that normal laboratory values did not correlate well with dehydration. Comfort was not increased, and in many studies, survival was not prolonged. However, a survival benefit was found in some specific patient populations studied, such as those with an acute medical problem or those receiving aggressive cancer therapy. The mixed findings from the literature review further validate the importance of beginning any discussion of ANH with the patient's primary goals as paramount.

Back to Top | Article Outline

Framework for Patient and Family Discussion

When a decision is made to withhold or withdraw ANH, clinicians may feel comfortable reassuring patients and families that discomfort, if present, can be managed fairly easily and effectively. Position statements by the HPNA (2004) and the American Academy of Hospice and Palliative Medicine (2006) are unified with regard to this viewpoint. Because a large body of empirical evidence is lacking, neither approach should be ruled out categorically. Some patients may attempt to prolong life to participate in a family milestone such as the birth of a grandchild or the marriage of an adult child. Additional nutrition may assist with the maintenance of independence and physical function. Others may wish to spend their remaining time and efforts focusing on matters other than nutrition and hydration, or they may have totally lost the desire to eat or drink.

There is one guiding principle that can be applied to all 3 opening scenarios, in fact to all situations, and this principle should be the foundation for nursing support irrespective of one's knowledge or beliefs: Discussion around ANH must have the patient's own goals of care and values at the core of all decision making, and the nurse must facilitate open discussions relating to these in a nondirective manner. This is consistent with the HPNA's calls for a “middle road” approach to foster effective decision making around ANH (HPNA Position Statement, 2004).

Facilitating discussions around life-sustaining treatments is an essential skill for nurses who care for patients at the end of life. Engaging families and patients in meaningful dialogue while providing compassionate support is a hallmark of sound nursing practice. A framework for discussion can ensure that the nurse addresses most of the common misconceptions associated with ANH. Providing information will help families cope with feelings of helplessness. The following framework is offered as a guideline for discussion.

1. Determine the patient's primary goal of care. Is it for palliation of symptoms or for maximal prolongation of life? A discussion based on the evidence about the potential for ANH to achieve these goals should ensue.

2. Clearly define ANH. Provide education about the technical aspects required and the benefits as well as the risks. Delineate the difference between comfort experienced with regular food and liquids as compared with ANH.

3. Discuss myths associated with ANH and the dying process. Misconceptions are commonplace. Educate the family about how nutritional needs lessen as illness advances, and how hunger and thirst may be greatly affected as body systems shut down. They should have an understanding that this is the normal part of the dying process and not the cause. In particular, patients and families should be aware of literature that fails to show an improvement of symptoms with ANH, and should know that a reduction in distress often occurs with ANH withdrawal.

4. Clarify the cause of fatigue or somnolence. An explanation about how underlying medical disorders are at the root of the disability and about ANH's lack of impact on this root cause will help redirect families to patient goals.

5. Explore alternate ways to demonstrate care and concern. Providing favorite foods and liquids is a symbol of caring in our society. As thirst and appetite decrease, nurses can help families identify new ways to show love. For example, family members can read books by the patient's favorite authors, bring favorite music, read scripture, or take turns holding vigil while reminiscing about past events they shared together.

6. Provide specific interventions to assist with comfort measures. Key points may include the following:

a. Put the patient in the driver seat. You can gently encourage intake but should never pressure patients to eat. Patients should not feel guilty for a lack of desire to eat. The dying process may shut down many normal functions, resulting in a lack of hunger or thirst.

b. Do not place undue emphasis on weight loss.

c. Small meals on small plates usually are received better than normal plates and portions.

d. Offer patients a large straw with beverages. Straws encourage more fluid intake and require less energy than drinking from a glass.

e. Alcoholic beverages before a meal may stimulate appetite.

f. High-calorie, high-protein liquid supplements are sometimes easier to consume than a regular meal.

g. Forget set mealtimes. Patients should be given food whenever they express a desire for it.

h. Expect taste changes. Favorite foods may no longer be well received and might even become distressing. Offer a variety of choices.

i. For problems with dry mouth, add gravies and sauces to meat, rice, or pasta. Saliva substitutes may be useful. If a sore mouth is the problem, avoid spicy, acidic, and crunchy foods. Avoid very hot or very cold foods.

j. Offer sips of water, ice chips, a mouth rinse, or moistened swabs often if the patient exhibits thirst. Never force large quantities of fluid if not desired.

Back to Top | Article Outline


The most difficult aspect of witnessing the decline of a loved one is the feeling of helplessness. When a patient can no longer eat or drink, nurses can help family members clarify patient goals and gently redirect care efforts in a more meaningful way. Families may need a better understanding of patient rights and should be encouraged to help their loved one complete advance directives, including the patient's preferences for ANH.

Credible resources to assist nurses in better engaging patients and families in meaningful discussions are listed in Table 1. Ethical and legal information regarding ANH is beyond the scope of this article but is included in this resource list.

Table 1
Table 1
Image Tools

Collaboration with nutritionists, speech and physical therapists, counselors, and palliative care nurse specialists can facilitate the identification of additional strategies to enhance comfort. The charge to nurses caring for dying patients and their families is that they consistently honor patient beliefs about life—their goals and values—in a supportive, unbiased, and informational manner. By ensuring familiarity with ANH evidence and using a discussion framework ensuring that salient issues are addressed, both hospice and home health nurses will improve the quality of life—at the end of life.

Back to Top | Article Outline


American Academy of Hospice and Palliative Medicine. (2006). Statement on the use of nutrition and hydration. Retrieved August 20, 2007 from

Bruera, E., Sala, R., Rico, M. R., Moyano, J., Centeno, C., Willey, J., et al. (2005). Effects of parenteral hydration in terminally ill cancer patients: A preliminary study. Journal of Clinical Oncology, 23(10), 2366–2371.

Burge, F. I. (1993). Dehydration symptoms of palliative care cancer patients. Journal of Pain and Symptom Management, 8(7), 454–464.

Finucane, T. E., Christmas, C., & Travis, K. (1999). Tube feeding in patients with advanced dementia: A review of the evidence. Journal of the American Medical Association, 282(14), 1365–1370.

Hospice and Palliative Care Nurses Association. (2004). Position statement on artificial nutrition and hydration. Retrieved August 20, 2007 from

Lee, J. H., Machtay, M., Unger, L. D., Weinstein, G. S., Weber, R. S., Chalian, A. A., et al. (1998). Prophylatic gastrostomy tubes in patients undergoing intensive irradiation for cancer of the head and neck. Archives of Otolaryngol Head and Neck Surgery, 124(4), 871–875.

McCann, R. M., Hall, W. J., & Grother-Juncker, A. (1994). Comfort care for terminally ill patients: The appropriate use of nutrition and hydration. Journal of the American Medical Association, 272(16), 1263–1266.

Mitchell, S. L., Kiely, D. K., & Lipsitz, L. A. (1998). Does artificial enteral nutrition prolong the survival of institutionalized elders with chewing and swallowing problems? Journal of Gerontology: Medical Sciences, 53A(3), M207–M213.

Morita, T., Hyodo, I., Yoshimi, T., Ikenaga, M., Tamura, Y., Yoshizawa, A., et al. (2005). Artificial hydration therapy, laboratory findings, and fluid balance in terminally ill patients with abdominal malignancies. Journal of Pain and Symptom Management, 31(2), 130–139.

Owens, D. A. (2007). Hydration in the terminally ill: A review of the evidence. Journal of Hospice and Palliative Nursing, 9(3), 122–123.

Rapp, R. P., Young, B., Twyman, D., Bivins, B. A., Haack, D., Tibbs, P. A., et al. (1983). The favorable effect of early parenteral feeding on survival in head-injured patients. Journal of Neurosurgery, 58, 906–912.

The Multi-Society Task Force on Persistent Vegetative State. (1994). Medical aspects of persistent vegetative state. New England Journal of Medicine, 330(22), 1572–1579.

Viola, R. A. (1997). The effects of fluid status and fluid therapy on the dying: A systematic review. Journal of Palliative Care, 13(4), 41–52.

Williams, A. (2007). Recruitment challenges for end-of-life research. Journal of Hospice and Palliative Nursing, 9(2), 79–85.

Back to Top | Article Outline
Medicare Announces Plans for Home Health Pay-for-Performance Demonstration

The Centers for Medicare & Medicaid Services (CMS) announced plans for a home health pay-for-performance demonstration, an important new step in its drive to become a more effective purchaser of quality healthcare.

Figure. No caption a...
Image Tools

CMS will begin soliciting home health agencies for the project this fall with the actual demonstration performance period to begin January 1, 2008. The demonstration will operate for 2 years in 7 states. Under the demonstration, home health agencies (HHAs) will be eligible to receive incentive payments if their quality improvement efforts result in the highest performance levels or significant improvements in patient outcomes. The availability of incentive payments will depend on whether or not the demonstration results in improvements in the quality of care and the actual savings to the Medicare program overall—not just for home health services provided to the patients served under the demonstration.

System-wide savings can be achieved when a home health agency prevents a re-hospitalization of the Medicare beneficiary or a further complication stemming from their illness. As the payments will be funded out of Medicare savings, none of the participating organizations will face payment reductions as a result of their participation in the demonstration.

CMS has selected the following states to provide a nationally representative sample of both HHAs and Medicare beneficiaries who utilize home health services:

* Northeast region: Connecticut, Massachusetts

* Southern region: Alabama, Georgia, Tennessee

* Midwestern region: Illinois

* Western region: California

CMS will be soliciting HHAs in the seven states to participate in the demonstration.

SOURCE: Department of Health & Human Services, CMS Public Affairs Office. Available at

© 2008 Lippincott Williams & Wilkins, Inc.