GI Nutrition: GI Nutrition: Deciding to Feed and How

0:00Welcome to Board Pearls. This is episode one of four of the GI Nutrition chapter, in the Special Populations and Acute or Supportive Care module. This episode is deciding to feed and choosing how. The first patient is a starved man with alcohol use disorder, and his case shows why food becomes dangerous after starvation, why thiamine must come before glucose, who's at high refeeding risk, and how slowly to start. The second patient is a woman in septic shock, and her case shows why a low albumin isn't malnutrition, how malnutrition is diagnosed, and when nutrition support is indicated. The third patient is a man starting chemoradiation for oropharyngeal cancer, and his case shows why the gut is used whenever it works, which tube suits weeks and which suits months, why his cancer changes how his gastrostomy is placed, what a tube that won't turn means, and why the formula is usually innocent in tube-feed diarrhea. A woman with advanced dementia shows who shouldn't get a feeding tube at all.

0:58Here's the first patient. A fifty-four-year-old man with alcohol use disorder is admitted after two weeks of heavy drinking with almost no food. He weighs fifty kilograms, a body mass index, BMI, of fifteen and a half, and has lost ten kilograms over four months. His potassium is three point one, his phosphate one point nine, and his magnesium low. He's confused, with an unsteady gait. A dextrose infusion has been hung in the emergency department.

1:27He's at high risk of refeeding syndrome, the metabolic crisis that follows the return of carbohydrate after starvation, and his confusion and unsteady gait raise Wernicke encephalopathy. Preventing refeeding syndrome is the most important decision in inpatient nutrition, and the reason food is dangerous for him lies in what starvation has done. In starvation, glycogen runs out within twenty-four to forty-eight hours, and fatty acids and ketones become the main fuels. Intracellular phosphorus, potassium, and magnesium drift down because the kidneys' conservation lags behind ongoing losses. Total body stores fall even while serum levels can look normal, because serum is a small compartment kept topped up by the much larger intracellular pool. When carbohydrate returns, an insulin surge drives glucose, potassium, magnesium, and phosphorus into cells, and the serum compartment crashes within hours as the hidden deficit is exposed all at once.

2:23That crash is what makes refeeding lethal, and his starting levels predict how steep it would be. Phosphorus depletion causes diaphragmatic failure, because muscle needs phosphate to regenerate adenosine triphosphate, ATP. The crash causes arrhythmia, because cardiac muscle cells need intact magnesium and potassium gradients. And it causes acute encephalopathy. His potassium, phosphate, and magnesium are already low before any food, so his total deficit is large. In a man like him, normal serum levels wouldn't have meant full stores.

2:57His dextrose bag is stopped until he's had thiamine, because his confusion may already be Wernicke encephalopathy and glucose can make it worse. Thiamine is a cofactor for transketolase and pyruvate dehydrogenase. In a thiamine-depleted patient, a carbohydrate load drives glycolysis, but pyruvate can't enter the tricarboxylic acid cycle, so ATP fails in vulnerable thalamic and mammillary neurons and Wernicke encephalopathy appears. Glucose given to any thiamine-depleted patient, with alcohol use disorder, hyperemesis gravidarum, prior bariatric surgery, or prolonged fasting, can turn reversible Wernicke encephalopathy into permanent Korsakoff syndrome. So the rule that thiamine comes before glucose isn't ritual; it follows directly from glycolysis depending on thiamine.

3:45His dose depends on whether he's being protected or treated. For refeeding prevention alone, thiamine two hundred to three hundred milligrams by mouth or vein is given before any glucose and continued daily for five to seven days. Wernicke encephalopathy is a clinical diagnosis, treated empirically with five hundred milligrams intravenously three times daily for two to three days, then two hundred fifty milligrams daily for five to seven days. His confusion and ataxia earn him the treatment dose, and then his dextrose can resume.

4:17His risk is formally high under the twenty oh six National Institute for Health and Care Excellence criteria, which identify the patients in whom careful pacing prevents the syndrome. High risk is any one of these: a BMI below sixteen, weight loss over fifteen percent in three to six months, more than ten days of negligible intake, or low potassium, phosphate, or magnesium before feeding. Any two of a less stringent set also count: a BMI below eighteen and a half, weight loss over ten percent, more than five days of low intake, or a history of alcohol misuse or of insulin, chemotherapy, or diuretic use. He meets all four of the major criteria, though one would be enough: a BMI of fifteen and a half, about seventeen percent weight loss in four months, fourteen days of negligible intake, and low electrolytes. The high-risk groups are the ones expected to be both starved and thiamine-depleted: anorexia nervosa, prolonged fasting, often beyond seven days, chronic alcohol use disorder, bariatric surgery, cancer cachexia, and prolonged hyperemesis gravidarum.

5:30His high risk sets how slowly he's fed. High-risk patients start at ten to twenty kilocalories per kilogram per day for the first twenty-four to forty-eight hours and advance over four to seven days to the full target. The restriction is the treatment, not an oversight, because it limits the insulin surge and the electrolyte shift that follows. Phosphate, potassium, and magnesium are checked twice daily for the first seventy-two hours and repleted aggressively to stay ahead of the shift, and calories aren't increased until the electrolytes are stable on replacement. At fifty kilograms, he starts at five hundred to one thousand kilocalories a day, at the low end, with his electrolytes replaced before feeding and rechecked that evening. If his phosphate fell to zero point eight on day two, his calories would be held while it's replaced. Severe refeeding syndrome carries a mortality approaching thirty percent without electrolyte support, and with proper pacing it's largely preventable.

6:29The second patient shows the opposite mistake, treating a lab value as starvation. A seventy-one-year-old woman is in the intensive care unit, ICU, on day four with pneumonia and septic shock. She ate normally until a week ago and has a BMI of twenty-six. Her albumin is two point one and her prealbumin is low, and her team asks about parenteral nutrition to correct her albumin.

6:53She isn't malnourished; her low albumin reflects inflammation. Albumin and prealbumin are negative acute-phase reactants, suppressed by interleukin-six and tumor necrosis factor alpha as the liver is reprogrammed to make C-reactive protein, fibrinogen, and ferritin, so a falling albumin marks the inflammatory response, not poor protein intake. Prealbumin, also called transthyretin, has a half-life of two to three days and was once promoted as a nutrition marker, but it's suppressed by the same cytokines and behaves just like albumin in an inflamed patient. Aggressive calorie or protein supplementation doesn't normalize albumin in inflamed patients or improve survival. Randomized trials in oncology, sepsis, and critical illness have shown that the early weight gain in fed catabolic patients is fluid, not muscle. Feeding won't raise her albumin, and a central line would add infection risk for no benefit. What decides her nutrition is how long she'll go without eating.

7:51Because albumin can't diagnose malnutrition, a defined set of criteria does. The Global Leadership Initiative on Malnutrition, GLIM, criteria require one phenotypic criterion plus one etiologic criterion. The phenotypic part identifies the deficit, and the etiologic part identifies the cause that nutrition support is meant to fix. The phenotypic criteria are unintentional weight loss over five percent within six months or over ten percent beyond six months; a BMI below twenty under age seventy, or below twenty-two over seventy; or reduced muscle mass on a validated measure, such as mid-upper-arm circumference, calf circumference, or imaging. The etiologic criteria are reduced intake or assimilation, meaning less than half of requirements for more than a week, any reduction for more than two weeks, or chronic malabsorption, or inflammation from acute illness or chronic disease. GLIM explicitly excludes albumin and prealbumin, because they don't respond to feeding in inflamed patients.

8:53Applied to the two patients so far, the criteria separate them. She has inflammation, an etiologic criterion, but no weight loss and a BMI of twenty-six, so she has no phenotypic criterion and isn't malnourished. He has seventeen percent weight loss and a BMI of fifteen and a half, two phenotypic criteria, plus more than two weeks of reduced intake, so he meets GLIM. Screening comes before diagnosis. Inpatients are screened with the Nutritional Risk Screening tool, NRS-2002, which combines BMI, weight loss, recent intake, and illness severity, and patients in the community with the Malnutrition Universal Screening Tool, which uses BMI, weight loss, and the effect of acute illness. Both flag patients for a full GLIM evaluation, and neither one is the diagnosis.

9:41What decides her feeding is whether nutrition support is indicated at all. The American Society for Parenteral and Enteral Nutrition and the European Society for Clinical Nutrition and Metabolism indicate it when oral intake will be inadequate for more than seven days, when malabsorption causes ongoing weight loss, or when hypermetabolism from burns, polytrauma, or severe pancreatitis outpaces what the patient can eat. It isn't indicated to chase albumin or to reverse the cachexia of metastatic cancer. If she's still intubated at seven days, she'll need support, through her gut if it works, at about twenty-five to thirty kilocalories per kilogram and one point two to one point five grams of protein per kilogram, the targets in critical illness.

10:27Her targets would shift if she had a chronic gastrointestinal or liver disease, because each one adds its own deficits. In cirrhosis, sarcopenia is common, and protein at one point two to one point five grams per kilogram with a late-evening snack limits overnight gluconeogenesis. Inflammatory bowel disease adds inflammatory cachexia and specific micronutrient deficits from ileal disease and steroid use. Celiac disease causes iron, folate, calcium, and vitamin D deficiencies that resolve on a gluten-free diet. Chronic pancreatitis causes fat malabsorption and fat-soluble vitamin deficiency that respond to pancreatic enzyme replacement.

11:08Once support is indicated, the next decision is how to deliver it, and the third patient shows that choice. A sixty-eight-year-old man with oropharyngeal squamous cell carcinoma is starting seven weeks of chemoradiation and is expected to be unable to swallow enough for several months. He's lost six kilograms, and his stomach and bowel work normally.

11:28He needs nutrition support, and because his gut works, it should be used. Enteral nutrition preserves the gut lining, maintains the microbiome, supports gut-associated lymphoid tissue, and avoids the central line complications of parenteral nutrition; in acute pancreatitis, parenteral feeding doubles infections. The expected duration chooses the tube. For less than four to six weeks, a nasogastric or nasojejunal tube is enough, because the need is brief and a percutaneous procedure isn't justified. Beyond that, a percutaneous endoscopic gastrostomy, PEG, is placed, with surgical or radiologic placement when endoscopy isn't feasible. His treatment plus months of swallowing recovery puts him well beyond six weeks, so he needs a PEG. A woman after a stroke expected to swallow again within three weeks would need only a nasogastric tube.

12:18His cancer changes how his PEG is placed. In the pull technique, the tube is drawn through the mouth and out through the abdominal wall with an endoscopic snare. In the introducer technique, it's inserted directly through the abdominal wall under endoscopic view. The pull technique is the default, but the introducer technique is preferred in oropharyngeal cancer, because a pulled tube passes through the tumor and can carry tumor cells to the gastrostomy tract. So he has an introducer PEG, with the reason written on the request. A PEG is a high bleeding-risk procedure and one of the few endoscopic procedures that gets antibiotic prophylaxis: a single dose of cefazolin, one gram intravenously thirty minutes before.

12:59He also meets the conditions a PEG requires: a functional stomach without prohibitive gastric residuals, the capacity to take part in his care or a designated caregiver, and an expected need beyond four to six weeks. Some patients need feed delivered past the stomach, and a PEG with a jejunal extension, PEG-J, is reserved for post-pyloric feeding in severe gastroparesis, recurrent aspiration, or a proximal obstruction. His stomach empties normally, but if he later had recurrent aspiration or his stomach failed, an extension through his existing PEG could deliver feed beyond the pylorus without a new tract.

13:32The conditions for a PEG also show who shouldn't get one. An eighty-six-year-old woman with advanced dementia in a nursing home no longer recognizes her family, is bedbound, has had two aspiration pneumonias this year, and eats less and less, and her family asks for a tube so she won't aspirate again. In advanced dementia, a PEG doesn't prevent aspiration, because aspiration in dementia comes largely from oropharyngeal dysphagia rather than gastric contents. It doesn't extend survival, and it isn't standard of care. Her family is counseled against it, clearly and kindly, with the explanation that the tube can't stop the swallowing problem that causes her aspiration, and that declining it reflects what the tube can and can't do, not giving up on her.

14:16The man's PEG works for months, and then it stops turning. Five months later he has pain with feeds, his tube resists flushing, and his nurse can't rotate it. That triad is buried bumper syndrome. Excessive external traction on the tube pulls the internal bumper into the gastric wall, where the lining grows over it until it's buried in the wall and blocks the tube. Feeds aren't forced through it. The bumper is removed endoscopically or surgically and the tube replaced. Prevention is rotating the tube weekly, to break any early seal of tissue over the bumper, and avoiding excessive traction. His external bolster had been pulled tight to stop a leak, which is how his bumper was drawn in.

14:57His leak points to the other PEG problems, each with its own fix. Peristomal cellulitis gets topical or oral antibiotics and wound care. Peristomal leakage, from an enlarged stoma, a malpositioned bumper, or excessive traction like his, is managed by downsizing the tube, repositioning it, or removing it temporarily so the tract can contract, rather than by pulling the bolster tighter. A blocked tube is cleared by flushing with warm water and pancreatic enzyme solution, not carbonated drinks, which don't work and may corrode the lumen. A dislodged tube is replaced within twenty-four hours through the mature tract, because a longer delay risks tract closure and may mean a new radiologic or endoscopic placement.

15:38His feeds also bring diarrhea. A week after starting PEG feeds he has six watery stools a day, and his dietitian suggests a peptide-based formula. Tube-feed diarrhea is overwhelmingly caused by medications rather than formula, so his medication list comes first. Sorbitol-containing liquid medicines, metronidazole suspension, acetaminophen elixir, theophylline syrup, and many others, deliver an osmotic load straight into the gut, and because the colon can't absorb sorbitol, it draws water into the lumen. Broad-spectrum antibiotics cause dysbiosis or unmask Clostridioides difficile infection. And contaminated formula or open delivery systems can introduce bacteria that cause enteritis.

16:21His history fits the first two causes. He takes liquid acetaminophen elixir through his tube four times a day for mucositis pain, a sorbitol vehicle, and he had a week of antibiotics for aspiration pneumonia. So his acetaminophen is changed to a sorbitol-free form and his stool is tested for C. difficile before any formula change. The test is negative, and his diarrhea settles within three days on the new acetaminophen; switching formula first would have left the cause in his medication cup. A concentrated hypertonic formula would have been the wrong change too, because it's designed for fluid restriction in heart failure or cirrhosis with ascites and can worsen diarrhea by raising the osmotic load. Peptide formulas are reserved for documented digestive insufficiency, such as severe pancreatic exocrine failure or short bowel syndrome.

17:11Here are the traps the boards set in refeeding and malnutrition. Glucose is never given before thiamine in an at-risk patient. A high-risk patient starts at ten to twenty kilocalories per kilogram, not full calories, and calories aren't advanced while phosphate is falling. Electrolytes are checked twice daily for the first seventy-two hours in high-risk patients. Normal serum electrolytes in a starved patient don't mean normal stores. A low albumin isn't malnutrition, and parenteral nutrition isn't started to raise it. Prealbumin falls with inflammation just as albumin does. Malnutrition needs both a phenotypic and an etiologic criterion. And feeding doesn't reverse metastatic cancer cachexia.

17:52For enteral access, the traps are these. Parenteral nutrition isn't chosen when the gut works. A PEG isn't placed for a need shorter than four to six weeks. The pull technique is avoided in oropharyngeal cancer. A PEG doesn't prevent aspiration in advanced dementia. A tube that hurts, won't flush, and won't rotate is a buried bumper. A blocked tube isn't cleared with cola. A dislodged tube is replaced within a day. Sorbitol medications, antibiotics, and C. difficile are checked before the formula is changed. And a concentrated formula isn't a treatment for diarrhea.

18:29So, start to finish. Identify refeeding risk before feeding: a BMI under sixteen, more than fifteen percent weight loss, more than ten days of little intake, or low potassium, phosphate, or magnesium, or any two of the lesser criteria. Give thiamine before any glucose, at the treatment dose if Wernicke is suspected. Start at ten to twenty kilocalories per kilogram a day, advance over four to seven days, and check and replace electrolytes twice daily for three days. Diagnose malnutrition with GLIM, not albumin or prealbumin, screen inpatients with NRS-2002, and start support when intake will be inadequate for more than seven days or malabsorption causes weight loss.

19:11Feed through the gut whenever it works. Use a nasal tube for less than four to six weeks and a PEG beyond that, placed by pull by default and by introducer in oropharyngeal cancer, with post-pyloric feeding for aspiration, gastroparesis, or outlet obstruction. Don't place a PEG in advanced dementia. Pain, a tube that won't rotate, and resistance to flushing mean buried bumper, prevented by weekly rotation and avoiding traction. Unblock with warm water and enzymes, replace a dislodged tube within twenty-four hours, and check drugs, antibiotics, and C. difficile before changing formula.

19:47Episode two is parenteral nutrition and short bowel syndrome.

19:52For the full chapter, the practice vignettes, and the topic-tagged question bank, head to board pearls dot com. You'll find the rest of the series on Apple Podcasts, Spotify, or wherever you listen to podcasts. That brings us to the end of episode one of four of chapter thirty-two, and I'll see you in the next one.