Management of Intractable Nausea and Vomiting in Patients at the End of Life: Title and subTitle Break“I Was Feeling Nauseous All of the Time . . . Nothing Was Working”

Mr Q is a 50-year-old electronics designer with metastatic esophageal cancer treated with third-line palliative chemotherapy. Recently, he has spent more than half of his time in bed due to a general lack of energy, although he walks without assistance or dyspnea. He was admitted to a university hospital in May 2006 for intractable nausea and vomiting.

His medical history was remarkable for migraine headaches, depression, and ulcerative colitis during childhood. He was diagnosed with esophageal cancer by endoscopic biopsy in October 2005. Thoracic computed tomography (CT) scans at the time showed circumferential thickening of the distal esophagus and an enlarged gastrohepatic lymph node. In December 2005, he began presurgical chemotherapy with docetaxel and capecitabine. In February 2006, he underwent an exploratory laparotomy but the tumor was found to be unresectable. A 20 × 20-mm stent was inserted in the gastroesophageal junction for impending obstruction and a jejunostomy feeding tube (J-tube) was placed. In March 2006, CT scans showed evidence of liver metastases.

Mr Q had experienced intermittent nausea and vomiting throughout his course of chemotherapy and reported a painful burning sensation in the chest and epigastrium since the esophageal stenting. Ten days before admission he had begun palliative chemotherapy with capecitabine. Afterwards, his nausea and vomiting worsened considerably, with vomiting episodes occurring up to 10 times a day, consisting of both dry heaves and emesis of bilious fluid. There was no apparent temporal relation of these symptoms to oral intake or J-tube feedings. Normal daily bowel movements were noted and a trial of ondansetron was not effective. He and his wife became worried about his inability to keep down food or water so they came to the emergency department.

On admission to the hospital, Mr Q received intravenous fluids and nothing by mouth; however, his nausea and vomiting persisted. At that time, his antiemetic regimen consisted of 8 mg of ondansetron intravenously twice a day; a scopolamine patch, 1.5 mg topically; lorazepam, 1 mg intravenously every 4 to 6 hours as needed; and promethazine, 12.5 to 25 mg intravenously every 4 to 6 hours as needed. Additional medications included oral morphine elixir as needed, bupropion, docusate, potassium chloride, and transdermal and transmucosal fentanyl. Upon physical examination, his mucus membranes were moist, with no oral thrush. His abdominal examination revealed no tenderness or distention, no hepatosplenomegaly, and normoactive bowel sounds. Laboratory studies were unremarkable including a normal complete blood count, electrolyte panel, liver function tests, amylase, lipase, and urinalysis. An abdominal and pelvic CT scan showed no abnormally dilated bowel loops. A palliative care consultant, Dr O, was asked to assist with management of the patient's nausea and vomiting.

A Perspectives editor interviewed Mr Q and Dr O in May and June 2006.

MR Q: I was just feeling terrible. . . . I was nauseous all of the time and throwing up. My energy level was really low, and I was dropping weight. What prompted me to go into the hospital was . . . I really just couldn't eat or drink anything. Even feeding through a J-tube . . . was making me nauseous. My wife and I were afraid that I was starving. . . . [W]e went to the emergency department and did the long wait there. . . . They couldn't tell me to go home without figuring out how to give me food and liquids.

DR O [PALLIATIVECAREPHYSICIAN]: We were called to consult on [Mr Q] by the primary medical team for symptom management. . . . He [was] not eating much and feeling weaker as a result.

Nausea and vomiting are common symptoms at the end of life, occurring in 62% of terminally ill cancer patients with a prevalence of at least 40% during the last 6 weeks of life.¹ Although most extensively studied in the cancer setting, nausea and vomiting also occur frequently in other terminal illnesses such as congestive heart failure and AIDS.^{2 - 3} In a retrospective review of 100 consecutive patients with varying diagnoses admitted to a palliative care unit, 71% reported nausea during their stay.⁴ Nausea often presents with a cluster of symptoms⁵ ; in one study, 25% of cancer patients treated for pain also reported nausea.⁶ Nausea and vomiting cause substantial psychological distress for patients and families near the end of life,⁷ with poorly controlled symptoms contributing to fears about starvation, dehydration, and even disease progression.

Using the case of Mr Q, this article reviews a general approach to caring for patients with nausea and vomiting near the end of life, relying on empirical evidence, and in its absence, our clinical experience. The approach involves: (1) careful evaluation to determine the etiology of the presenting symptoms; (2) using pathophysiology to determine the mechanism and, subsequently, receptors underlying the patient's nausea and vomiting; and (3) choosing an antiemetic to block the implicated receptors. Because of its importance at the end of life, this article places a special emphasis on how to approach intractable nausea, defined herein as nausea and vomiting that is not adequately controlled after multiple antiemetics are used in series and/or in combination. Although we believe a mechanism-based approach is applicable to any patient with nausea and vomiting, this article's focus may not be generalizable to populations with less limited life expectancies.

A history and physical examination represent essential first steps in the evaluation of nausea and vomiting, for they provide a measure of symptom severity⁸ and clues to the underlying etiology. Careful evaluation permitted physicians in one study to confidently establish the cause of nausea and vomiting for about 45 (75%) of 61 hospice patients.⁹ The most frequently cited etiologies were chemical abnormalities (metabolic, drugs, infection; 33%), impaired gastric emptying (44%), and visceral and serosal causes (bowel obstruction, gastric bleed, enteritis, constipation; 31%).⁹ A study of 40 patient-episodes of nausea, vomiting, or both on a palliative care unit identified 59 reversible etiologies, with medications (51%) and constipation (19%) presenting most commonly.¹⁰

The history should focus on characterizing the nausea and vomiting as well as any associated symptoms (Table 1).^{11 - 12} Special attention should be paid to complaints of anorexia because it may represent a constant low-grade nausea. Although Mr Q did not have a history of constipation, given its frequency near the end of life,¹⁰ constipation must be ruled out in every patient.^{11 ,13} This includes a detailed history of the frequency and consistency of stools because many patients with limited oral intake mistakenly believe it is normal to have infrequent bowel movements. Mr Q reported esophageal burning consistent with gastroesophageal reflux, a common complication after esophageal stent placement.¹⁴

Obtaining a complete medication history is essential, including a thorough evaluation of new and recently discontinued prescription and over-the-counter drugs. Chemotherapeutics, opioids, antidepressants, and antibiotics are frequent contributors to nausea and vomiting near the end of life.¹⁵ Recent and/or rapid discontinuation of corticosteroids or high-dose progesterones may cause nausea due to adrenal insufficiency.¹⁶

Nonpharmacological therapies must also be considered in the evaluation. Radiation therapy, especially to the abdomen or lumbosacral spine, can trigger nausea and vomiting.¹⁷ Any recent surgery, particularly abdominal surgery, can also produce symptoms.¹⁸ In the case of Mr Q, the esophageal stent placement, palliative capecitabine (though a low emetic risk agent), and opioid therapy could all be contributing to his nausea. Bupropion and potassium chloride can be emetogenic, but represent long-standing therapies for Mr Q and, as such, are less likely causes of his symptoms.

The past medical history provides additional critical clues. Peptic ulcer disease, gastroesophageal reflux, or both may explain symptoms. Diabetes mellitus, alcoholism, chronic renal failure, advanced cancer, autoimmune disorders, amyloidosis, and Parkinson disease are all associated with autonomic dysfunction and delayed gastric emptying.¹⁹ For cancer patients, the type of malignancy, its site of origin, and location of metastases are dispositive. For example, liver metastases, malignant bowel obstruction, and peritoneal carcinomatosis can all cause nausea and vomiting.¹² External compression of the stomach or duodenum by tumor or massive ascites is associated with nausea and vomiting through the “squashed-stomach syndrome.”¹² Primary or metastatic brain or leptomeningeal tumor can be emetogenic as well.¹² Finally, a patient's psychological state, particularly anxiety or depression, may be associated with nausea.²⁰ Mr Q's past medical history of migraines and ulcerative colitis can cause nausea but currently appear quiescent. Esophageal cancer, through direct extension, may irritate the esophageal or gastric mucosa, causing nausea and vomiting. Mr Q does not appear to have any distant contributory metastases.

The physical examination provides additional clues to the etiology of a patient's nausea and vomiting with important findings listed in Table 1. Mr Q, however, presented with a normal abdominal, rectal, and neurological examination.

Laboratory and radiology testing may provide diagnostic insights, but for patients in the home setting an exhaustive workup often distracts from minimizing symptom burden and optimizing management.¹¹ A laboratory evaluation may reveal renal failure, hyponatremia, liver failure, pancreatitis, or hypercalcemia, all of which may cause or contribute to nausea and vomiting. Drug toxicity from digoxin or anticonvulsants can precipitate symptoms and, if suspected, may warrant checking a serum level. A supine abdominal film helps identify constipation,¹³ and is especially useful in patients with delirium or dementia who are unable to give an accurate history of recent bowel movements. Finally, an upright abdominal film can identify air-fluid levels if gastrointestinal (GI) tract obstruction is suspected. Mr Q's laboratory studies were unremarkable, and a CT scan did not show evidence of bowel obstruction.

DR O: I went down a lengthy list of the . . . causes of intractable nausea and vomiting. . . . It's important to have an etiologic diagnosis, so you know which treatments are going to be most helpful.

After elucidating the most likely etiology of nausea and vomiting, the next step is to determine which mechanism is triggering symptoms to guide therapy. Nausea and vomiting are caused by the stimulation of at least 1 of the 4 pathways. Each of these provides input into the vomiting center in the brainstem, which produce nausea or vomiting when the minimum thresholds are reached (Figure). The 4 pathways are^{12 ,21 - 24}

Opioid-Induced Nausea and Vomiting. Up to 40% of opioid-treated patients experience nausea and vomiting,²⁵ triggered by constipation, stimulation of the CTZ, gastroparesis, and sensitization of the labyrinth.²⁶ The effects in the CTZ are largely mediated through central dopamine type 2 (D₂) receptors, whereas the gastroparesis is mediated through peripheral D₂ receptors. Although early studies attributed opioid-induced nausea and vomiting to the accumulation of metabolites, particularly morphine-6 glucuronide,²⁷ more recent studies do not support this theory.²⁸

Chemotherapy-Induced Nausea and Vomiting.Chemotherapy causes nausea and vomiting by a complex set of mechanisms.²⁹ First, chemotherapy is thought to directly stimulate the CTZ. This effect appears to be mediated by 5-hydroxytryptamine type 3 (5HT₃) and neurokinin type 1 (NK₁) receptors. Second, chemotherapy is thought to damage the GI mucosa and cause release of neurotransmitters including 5HT₃. This stimulates nausea and vomiting via peripheral pathways mediated by vagal and splanchnic nerves. Third, there appears to be some neurohormonal etiology to these symptoms via alteration in arginine vasopressin and prostaglandin levels.²⁹ Finally, chemotherapy-induced nausea and vomiting may be mediated by anxiety, which can trigger symptoms via central pathways.^{30 - 31}

Malignant Bowel Obstruction. Malignant bowel obstruction can occur with any malignancy but is most commonly associated with advanced ovarian and colorectal cancer.³² Peripheral pathways are stimulated because of the stretch of bowel wall, pain, and colic associated with accumulating food and fluids proximal to the obstruction. Additionally, the CTZ is likely triggered by inflammatory mediators and bacterial toxins.²⁴

Impaired GI Tract Motility of Advanced Cancer. Autonomic dysfunction may play a central role in chronic nausea and vomiting in patients with advanced cancer as a result of gastroparesis and constipation.³³ Symptoms are likely triggered by activation of peripheral pathways due to stretch of the gastric or esophageal wall from this poor motility. The etiology of autonomic failure in patients with advanced cancer is multifactorial, including malnutrition and cachexia, chemotherapy and other drugs, radiation therapy, paraneoplastic phenomena, nerve invasion by tumor, and comorbidities such as diabetes mellitus.¹²

Mr Q's esophageal irritation due to tumor burden and poststent reflux is likely triggering nausea via vagal input into the vomiting center. The opioids he is receiving may be activating central D₂ receptors in the CTZ, and the capecitabine chemotherapy may be activating NK₁ receptors in the CTZ and 5HT₃ receptors in the GI tract and the CTZ.

DR O: We generated a list of possible etiologies and tried to rank them. . . . We recommended adding [prochlorperazine] to cover the possibility that the opiates were producing the nausea. Because of the possibility that he was having esophageal candidiasis, we recommended nystatin. Because of the stent and the possibility that reflux was creating irritation in his esophagus and upper GI tract, we thought about [adding sucralfate].

MR Q: They started me on different [combinations of] pain and antinausea medications. The thing that made the most difference, I think, is when they put me on an antacid called[lansoprazole]. . . . [t]he acid reflux got better 2 or 3 days later. It was in conjunction with other antinausea medications. . . . By the second day, I wasn't taking in anything orally, but I wasn't throwing up.

Thoughtful evaluation to determine both the etiology of the symptoms and the pathophysiological mechanism by which they are triggered allows directed therapy to begin. Therapy should not only include antiemetics, but also measures to alleviate the cause of the symptoms, such as the proton pump inhibitor for Mr Q.

Nonpharmacological therapy is an important first consideration in the management of intractable nausea. Simple recommendations like avoiding strong smells or other nausea triggers, eating small, frequent meals, and limiting oral intake during periods of extreme emesis are helpful.^{34 - 35} Psychological techniques, especially those that promote relaxation, can be helpful.^{36 - 37} Acupuncture and acupressure may provide some benefit in the setting of chemotherapy or surgery. A systematic review found benefit to P6 stimulation (just above the wrist) in 11 of 12 randomized placebo-controlled trials.³⁸ Acupressure wrist bands, however, have not been shown to be effective.³⁹ Medical devices including gastric electrical stimulation⁴⁰ and transcutaneous electrical nerve stimulation units⁴¹ are currently under investigation, but a lack of convincing evidence and substantial cost currently limit their use.

A mechanism-based treatment scheme administering the most potent antagonist to the implicated receptors has been shown to be effective in up to 80% to 90% of patients near the end of life.^{9 - 10 ,42} It should be noted that some practitioners recommend starting an empirical antiemetic regimen, typically with a D₂ antagonist, regardless of the presumed etiology.⁴³ To date, no head-to-head comparisons between mechanism-based and empirical therapy exist.⁴⁴ We advocate and practice a mechanism-based management paradigm because it facilitates a systematic approach to caring for the patient, identifies all potential symptomatic contributors, directs therapy, and minimizes the risk of overmedicating a vulnerable population.

In practice, multiple etiologies are often at play and patients are acutely symptomatic on presentation, requiring empirical treatment and numerous interventions while evaluation is ongoing. All potential underlying causes, such as constipation, opioids, and electrolyte abnormalities should be addressed simultaneously to provide the greatest chance of rapidly resolving symptoms. When choosing antiemetics for these patients, we favor initiating medications that target the D₂ receptor, such as metoclopramide, prochlorperazine, or haloperidol, which are the foundation of many of the empirical schemes.^{43 ,45 - 47} Choosing one of these agents makes mechanistic sense because D₂ antagonists block CTZ-mediated nausea, a common cause of symptoms in patients near the end of life.

Another important consideration when selecting an antiemetic is the medication's adverse-effect profile. For example, a patient with nausea due to stimulation of the CTZ may benefit from either a 5HT₃ or D₂ antagonist. If the patient is concerned about excessive sedation, the clinician might avoid the D₂ antagonist, whereas, if constipation has been particularly problematic, the D₂ antagonist might be the better choice.

A recent development is the incorporation of 5HT₃ antagonists such as ondansetron. Evidence supports the use of these agents for chemotherapy-induced nausea and vomiting,⁴⁸ radiation therapy-induced nausea,⁴⁹ and postoperative nausea.⁵⁰ Smaller studies suggest efficacy of 5HT₃ antagonists in nausea and vomiting due to opioids⁵¹ and uremia.⁵² However, the literature does not support using these agents empirically outside of the noted clinical scenarios. Moreover, for the most common etiologies of nausea and vomiting at the end of life, 5HT₃ antagonists are no more effective than the less expensive D₂ antagonists.^{53 - 54}

Despite evidence supporting its use, a mechanism-based monotherapy approach may not reduce nausea and vomiting to an acceptable level.⁹ Before changing regimens, practitioners should ensure that the prescribed therapy was properly administered. A common management pitfall is that first-line antiemetics are prescribed on an as-needed basis instead of scheduled around-the-clock.¹¹ If nausea and vomiting continue despite effective blocking of the targeted pathway, a second agent that antagonizes other implicated neurotransmitters should be added. Adding a second agent is preferred to switching agents because nausea is often multifactorial and several neurotransmitters are active at each receptor site. This approach has proved effective in chemotherapy⁴⁸ and for patients at the end of life.^{43 ,55 - 57}

Prophylactic dosing prior to known emetogenic triggers has value particularly with chemotherapy,⁴⁸ radiation therapy,¹⁷ in the postoperative setting,⁵⁸ or in patients with known prior adverse reactions to, eg, opioids.⁵⁹ Prevention of nausea is particularly important if the stimulus is likely to be repeated, such as with chemotherapy, because of the high potential for developing learned responses.³⁰

In the case of Mr Q, a careful evaluation revealed several possible contributory etiologies. As such, Dr O recommended prochlorperazine to block D₂ receptors in the CTZ to counteract nausea and vomiting due to opioids. In addition, Dr O recommended lansoprazole and sucralfate to treat Mr Q's gastroesophageal reflux.

In the following section, we apply the mechanistic approach to the management of some of the most common etiologies of nausea and vomiting in patients near the end of life (Table 2). Table 3 provides a list of frequently used antiemetics, their mechanism of action, dosage, common adverse effects, and cost. Table 4 reviews selected studies supporting the use of these agents in patients near the end of life.

Generally, opioid-induced nausea and vomiting occurs with initiation of opioids or with dose escalation and resolves within 3 to 5 days of continued use. If nausea develops, antiemetics targeting D₂ receptors should be prescribed around-the-clock for several days and then tapered as tolerated.^{24 ,70} Haloperidol, droperidol,^{46 ,59 ,71} and metoclopramide^{59 ,72} all have demonstrated efficacy. Limited evidence suggests that promethazine may potentiate the effects of opioids.⁷³ Although some clinicians see this interaction with opioids as a therapeutic advantage, others avoid promethazine due to sedation and the increased risk of respiratory depression.⁷⁴

A small number of patients develop persistent nausea that may improve with an opioid dose-reduction or rotation. A 10% to 20% reduction in daily opioid dose often alleviates nausea without a loss in analgesia.⁷⁵ However, if dose reduction is not feasible or is ineffective, opioid rotation demonstrates efficacy in both prospective and retrospective studies.^{76 - 77}

The patient's goals of care are paramount when considering the use of chemotherapeutic agents near the end of life. Management of chemotherapy-induced nausea and vomiting is preventive and based on the emetogenicity of the prescribed agent (Table 5).⁴⁸

Some of the nausea associated with chemotherapy may also be anxiety-related or “anticipatory” because patients associate receiving chemotherapy with becoming nauseated.³⁷ This may partially explain the observed decreasing efficacy of antiemetics in patients undergoing multiple cycles of chemotherapy.⁷⁸ Although not strictly classifiable as antiemetics, benzodiazepines such as lorazepam are effective in preventing anticipatory nausea.^{79 - 80} Outside of this setting, however, the use of benzodiazepines for nausea is generally discouraged.⁸¹

Management of malignant bowel obstruction often involves both pharmacologic and nonpharmacologic interventions. Surgery is generally not recommended for persons with a life expectancy of less than 2 months^{82 - 83} because it does not improve survival, rarely palliates symptoms, and is associated with a high complication rate.⁸⁴ Gastrointestinal tract stents may have a role, depending on the location of the obstruction, but have been associated with complications.⁸⁵ Nasogastric tubes can relieve symptoms but should only be used temporarily while other treatment is pursued given the complications and discomfort associated with their long-term use.³²

Fortunately, medical management provides very effective symptom control.⁸⁶ Recommended pharmacologic therapy includes analgesics, antisecretory agents, and antiemetics.³² Opioids are used for pain control. Anticholinergics such as hyoscyamine and a somatostatin analogue (octreotide) diminish secretions and potentially reduce pain and nausea by decreasing mucosal distention and peristalsis. Octreotide can be administered subcutaneously beginning at 50 to 100 μg 3 times daily (to a maximum of 900 μg per day). Some palliative care units will administer octreotide via continuous infusion at much higher doses, although evidence to support this practice is scarce. Metoclopramide is recommended for patients with nausea and a partial obstruction without colic. In patients with complete obstruction, metoclopramide can induce colic through its peripheral D₂ receptor stimulation of GI motility, although this concern may be overstated.⁸⁷ For these patients, the recommended agents are central D₂ antagonists, such as haloperidol, which work primarily at the CTZ. Antihistamines that work through peripheral pathways and the vomiting center may also be effective. Corticosteroids, such as dexamethasone, are generally included in most antiemetic regimens for their potential effect on tumor-associated inflammation. A recent Cochrane review found a nonsignificant (P>.05) trend suggesting that corticosteroids may be effective in helping resolve the obstruction.⁸⁸

If medical therapy provides insufficient relief, a venting gastrostomy tube may be placed. With this, gastrointestinal and oral secretions are removed without a nasogastric tube, and the patient may continue liquid oral intake as desired.⁸⁹

MR Q: We tried these little dots [ondansetron ODT] for nausea. But nothing was working. It wasn't until we went into the hospital and just started experimenting that I really got some relief.

In some cases, nausea and vomiting may persist despite a mechanism-based approach using several medications at appropriate dosages taken around-the-clock targeting multiple pathways. In these situations, less traditional agents can be considered, but evidence supporting their use remains limited. For instance, dexamethasone, is widely used for its antiemetic effects in palliative care, even though a recent study demonstrated no greater effect than placebo when added to metoclopramide for patients with chronic nausea of advanced cancer.⁹⁰ Despite this study's results, corticosteroids have well-described antiemetic properties,⁹¹ and in our experience are extremely effective at decreasing symptom severity. Mirtazapine, an antidepressant that antagonizes the 5HT₃ receptor, is also frequently used to alleviate intractable symptoms. To date, evidence supporting its use is limited to small trials and case reports.^{68 ,92} Cannabinoid agents, such as dronabinol, can be effective antiemetics in patients with AIDS^{93 - 94} and cancer^{95 - 96} but should be used with caution in older adults or cannabinoid-naive patients because adverse effects, including confusion and hallucinations, may be pronounced. Olanzapine, an atypical antipsychotic, blocks several receptors associated with nausea and vomiting including dopamine, acetylcholine, histamine, and serotonin receptors. Larger studies are needed to better define its role.^{97 - 99} Megestrol acetate and thalidomide decreased nausea in patients enrolled in clinical trials for appetite stimulation,^{100 - 101} but they are rarely used solely for their antiemetic properties. The ABHR suppository, a combination preparation of lorazepam (Ativan), diphenhydramine (Benadryl), haloperidol (Haldol), and metoclopramide (Reglan), is often used for home hospice patients, although there are no data to support its benefit. It is well tolerated,¹⁰² but, in our experience, exerts its effect mainly through sedation. Herbal medicines have been used to treat chemotherapy-induced¹⁰³ and pregnancy-induced¹⁰⁴ nausea and vomiting, but little evidence exists to support their use in end-of-life populations.¹⁰⁵ Finally, 5HT₃ antagonists are sometimes used to treat intractable nausea and vomiting,^{106 - 108} but, as noted above, there is little justification for their use outside of circumscribed clinical scenarios.

Refractory nausea and vomiting may make oral administration of medication unfeasible so alternate routes must be considered. Many of the most common antiemetics are available in several preparations, such as rectal suppositories, subcutaneous infusions,¹⁰⁹ and orally dissolvable tablets (Table 3), allowing patients to be treated at home.

DR O: Ordinarily, I like to do things one at a time. If you do a bunch of things at once, you never know what the useful things were. . . . I was a little nervous that the medical team was using such a variety of antinausea medicines.

Avoiding polypharmacy is a critical aspect of nausea and vomiting management for the reasons Dr O observes. If patients are taking multiple medications, it may be difficult to identify the effective agent, and the patient is at increased risk for adverse effects as well as for drug-drug interactions.¹¹⁰ Precipitating delirium in patients near the end of life is of particular concern as they exhibit diminished cognitive reserve, and most antiemetic agents are centrally acting.^{111 - 113} Standardized tools such as the Confusion Assessment Method¹¹⁴ are effective and should routinely be incorporated into clinical practice to screen for delirium in patients with advanced life-limiting diseases.

One common misstep in the management of nausea and vomiting is the coadministration of multiple antiemetics that antagonize the same receptor, resulting in adverse effects at lower than expected doses. For example, if a patient is taking prochlorperazine and haloperidol, both of which work on the D₂ receptor, the risk of a dystonic reaction or akathisia increases. A mechanism-based approach helps avoid this pitfall and facilitates a step-wise introduction of medications that exert their effects at different receptor sites.

If nausea and vomiting remain intractable despite aggressive, multimodal attempts at control, palliative sedation may be considered for patients with a limited life expectancy.^{115 - 116} Although symptoms of nausea and vomiting are rarely the primary indication for palliative sedation,¹¹⁷ they are commonly noted secondary symptoms of patients choosing palliative sedation for other reasons (36%-44% of cases).¹¹⁵ No standard regimen exists for sedation of patients with intractable nausea; however, propofol has been proposed as an ideal agent because it blocks 5HT₃ receptors, resulting in an antiemetic effect in addition to its sedative effects.¹¹⁸

A step-wise, mechanism-based approach to treatment of nausea and vomiting has proved effective for a majority of patients experiencing these symptoms toward the end of life. A thorough assessment to ascertain potential etiologies, pathways, and respective transmitters and receptors allows the clinician to prescribe the most appropriate antagonist to the offending receptor. If monotherapy is ineffective, a trial combining several therapies to block multiple emetic pathways is recommended. Further research will refine palliative care management strategies that minimize adverse effects and maximize control of these highly distressing symptoms.