Introduction to nutritional dogma
Everyone has strong opinions about food. We all feel that we have some special, intuitive understanding of nutrition. Nonsense. Such intuitions have historically created a wide array of dogma regarding nutrition, complicating matters immensely.
Fortunately, the 2016 SCCM/ASPEN guidelines have stripped away much of the nonsense involved in nutritional support. This allows for a simple and effective approach to nutritional support in the ICU. Based on these guidelines, this post will attempt to sort out truth vs. fiction: which nutritional beliefs are valid, and which need to be discarded?
#1. Early enteral nutrition (within 24-48 hours) is helpful.
Benefits of early enteral nutrition may include maintence of gut integrity, reduction in bacterial translocation, and stress ulcer prophylaxis. Delayed initiation of nutrition may promote ileus, impeding later attempts to start feeding. Meta-analyses confirm benefits, including reduction of infectious complications.
Benefits of enteral nutrition largely relate to supporting bowel health, rather than necessarily providing nutritional fuel. This explains why enteral nutrition is generally superior to parenteral nutrition.
#2. The following are not contraindications to enteral nutrition.
- Lack of bowel sounds: This may simply indicate a lack of air in the bowel. The use of bowel sounds to drive any clinical decision has never been validated.
- Therapeutic paralysis: Paralytic infusions only affect the skeletal muscles, not smooth muscles (e.g. intestines, pupillary reflexes).
- Patient is required to lie flat: The evidence supporting feeding in a semi-recumbent position is somewhat weak. Semi-recumbent positioning is preferable but not mandatory (Marik 2014).
- Vasopressor use: Traditionally, it was feared that providing nutrition in the context of shock could cause intestinal ischemia. However, enteral nutrition actually appears to improve blood flow to the gut and preserve intestinal integrity. Once a patient has been fluid resuscitated and stabilized on vasopressors, enteral nutrition may be started at low rates (more on this below).
- Open abdomen: Enteral nutrition is recommended for patients with an open abdomen who don't have bowel injury.
- Pancreatitis: Evidence supports early enteral nutrition among patients with severe pancreatitis, similar to other critically ill patients.
Currently, the only major contraindications to enteral nutrition seem to be bowel obstruction, perforation, mesenteric ischemia, or major gastrointestinal bleeding. In short, the patient can be fed as long as they aren't having an intestinal catastrophe.
There is no known illness or disease that has been demonstrated to benefit from starvation. – Marik 2014.
#3. Starting feeds at 25-50% caloric goal and increasing to 100% over 3-7 days is reasonable.
The ideal initial rate for tube feeding is unknown.
- In previously well nourished patients, starting hypocaloric feeding appears safe (while targeting 100% protein goal). This was previously discussed here.
- In previously malnourished patients, starting at a lower rate may be wise to avoid refeeding syndrome.
- SCCM/ASPEN guidelines recommend starting with very low-level feeding (10-20 kCal/hr “trophic feeding”) for patients with septic shock or severe pancreatitis.
Hopefully this will eventually be clarified. For now, it seems reasonable to start feeding early, but at a reduced rate (e.g. 25-50% caloric goal). Subsequently, feeding can be advanced to 100% goal calories over 3-7 days, as tolerated. Efforts should be made to provide 100% protein requirements when possible. Determining a nutritional prescription that provides fewer calories yet adequate protein may be easily accomplished with the tube feed cheat sheet.
#4. Generally avoid fancy, designer tube-feed formulations.
Historically there has been much enthusiasm about a variety of specialized tube feeds. Evidence regarding these has generally been disappointing. Based on the SCCM/ASPEN guidelines, if you gave every patient a generic tube feed formulation, that would be fine.
These guidelines specifically recommend against the following designer tube feeds:
- High-fat formulations designed to help wean patients off ventilation.
- Anti-inflammatory formulations designed for patients with ARDS.
- Immune-modulating enteral formulations (e.g. with arginine) for patients in the medical ICU.
- Branched-chain amino acid formulations designed for patients with hepatic encephalopathy.
Specialty formulations may remain useful in the following situations:
- Fluid-restricted, energy-dense formulations may be useful in patients with volume overload (1).
- Renal formulations are recommended only for patients with renal failure who develop significant electrolyte abnormalities.
- Patients with diarrhea may be managed with tube feeds that contain fiber or small-peptide formulations (2).
- Immune-modulating enteral formulations with arginine may be useful in post-operative patients and patients with severe trauma.
#5. Don't check gastric residual volumes (GRVs).
GRVs are a poor predictor of with pneumonia, aspiration, or regurgitation. Between ~2000-2010, several RCTs showed that using a lower GRV cutoff wasn't any safer than a higher cutoff. Consequently, the 2006 SCCM/ASPEN guideline increased the GRV target to <500 ml. More recently, one RCTs and one before/after implementation trial demonstrated that it was safe to stop checking GRVs entirely (Reignier 2013, Poulard 2010). This has led the SCCM/ASPEN to currently recommend eliminating GRVs entirely from routine care.
Measuring GRVs is a significant impediment to effective enteral nutrition, since it often leads to unnecessary interruption of feeding. Eliminating GRVs entirely is probably easier than trying to figure out what to do with them. Discontinuing GRVs may also reduce the burden on nurses, who are required to check these repeatedly.
#6. Don't interrupt feeding for procedures or possible extubation.
There is little rationale to interrupt tube feeds for most procedures (e.g. interventional radiology procedures). With adequate procedural sedation and analgesia, most procedures shouldn't provoke vomiting. If vomiting should occur, these patients are intubated with a protected airway.
A common practice is to hold tube feeding at midnight if the patient might possibly be extubated the following morning. In the absence of evidence, my preference is to continue tube feeds until a definite decision is reached to extubate the patient. Prior to extubation, the stomach should be suctioned empty (if the patient is being fed via a nasogastric or orogastric tube)(3).
#7. Difficulty tolerating tube feeds (e.g., vomiting) may be managed with post-pyloric tube placement or prokinetics.
- Post-pyloric feeding generally produces similar outcomes compared to gastric feeding (similar mortality, ICU length of stay, and duration of mechanical ventilation). The first-line approach for feeding intubated patients is typically gastric feeding, because this is easier. However, if there is difficulty tolerating gastric feeding, post-pyloric feeding is a rational approach. This may be a viable strategy within a hospital capable of placing post-pyloric feeding tubes rapidly (e.g. with Cortrak electromagnetic guidance)(4).
- Metoclopramide or erythromycin may improve gastric motility. These agents have been shown to reduce gastric residual volume and feeding intolerance.
#8. Patients with pancreatitis, septic shock, hepatic failure, or renal failure can all be fed the same way you would feed any other patient in the ICU.
Feeding patients with renal or hepatic failure has always been a source of confusion. Will protein worsen uremia in renal failure? Could it exacerbate hepatic encephalopathy in cirrhosis? What do we do about pancreatitis?
According to the SCCM/ASPEN guidelines, you can feed these patients the same way you would feed any patient in the ICU. Protein administration shouldn't be restricted as a strategy to avoid dialysis or hepatic encephalopathy. Some minor differences are:
- Patients with renal failure may be started on a regular tube feed formulation. However, if significant electrolyte abnormalities develop (e.g. hyperkalemia, hyperphosphatemia), then switch to a renal tube feed formulation.
- In critically ill patients on dialysis, the amount of protein administered should be increased up to a maximum of 2.5 grams/kg daily (5).
#9. Patients with morbid obesity can be fed similarly to other ICU patients, based on their ideal body weight.
The ideal number of calories for a patient with morbid obesity remains unknown. Theoretically, underfeeding could promote weight loss. However, underfeeding patients in the ICU tends to backfire, because patients burn their own protein for fuel and become protein-malnourished (6).
The perfect approach may be to provide fewer calories than the patient needs, but an ample amount of protein. The SCCM/ASPEN guidelines therefore recommend:
- For BMI 30-50, provide 11-14 kCal/kg (using the actual body weight) and 2 grams/kg protein (using the ideal body weight)
- For BMI>50, provide 22-25 kCal/kg and 2.5 grams/kg protein, both utilizing the ideal body weight.
This may seem complicated, but it doesn’t have to be. The high-protein tube feed cheat sheet may be used for patients with morbid obesity (example below; details about how to use this here). This sheet calculates tube feed formulations providing 25 kCal/kg ideal body weight and 2.25 grams protein/kg ideal body weight. This will produce a reasonable nutritional prescription for these patients (7).
Over time, critical care nutrition is becoming progressively simpler and easier. Take-home messages from the 2016 SCCM/ASPEN guidelines include:
- Early enteral nutrition should be provided to nearly all intubated patients.
- The only strong contraindication to enteral nutrition is an intestinal catastrophe. The following are not contraindications: lack of bowel sounds, therapeutic paralysis, vasopressors, pancreatitis, or open abdomen.
- Don't check gastric residual volumes.
- Patients with pancreatitis, septic shock, hepatic failure, or renal failure can all be fed in essentially the same way that you would feed any patient in the ICU.
- Fancy, designer tube-feed formulations are rarely needed.
Related information:
- SCCM/ASPEN 2016: Guidelines for the provision and assessment of nutrition support therapy in the adult critically ill patient.
- Related posts:
Notes
- Please note that in order for this to make a difference with edema and fluid balance, the key factor is the sodium content of the tube feeds (not the actual volume). In some cases, concentrated tube feed formulations actually have less sodium per kCal, but in other cases they don't (you have to look up the sodium content of the specific tube feed brand). Using a concentrated tube feed formulation that has less volume but lots of sodium doesn't actually get you anywhere – the patient will wind up being hypernatremic… and then you will be forced to give them free water to normalize their sodium.
- It is OK to use tube feed formulations that contain fiber. However, the best source of fiber appears to be soluble fiber products, which may be metabolized into short-chain fatty acids that nourish the colon (as opposed to insoluble fiber like Metamucil). Thus, the best approach to adding fiber may be add a specific supplement that contains soluble fiber (e.g. guar gum).
- For patients being fed via a small-bore nasal post-pyloric feeding tube, it is impossible to aspirate tube feeds. However, aspiration risk in this context may be lower because the feeding is being delivered distal to the stomach.
- At Genius General Hospital, the Cortrak system generally allows our nurses to place post-pyloric feeding tubes at the bedside in the ICU. Alternatively, if you're depending on interventional radiology or gastroenterology to place these, then this can be logistically challenging. (I have no conflicts of interest with Cortrak or any other company.)
- This may be achieved using the increased-protein version of the tube feed cheat sheet.
- Morbid obesity provides zero protection against protein-calorie malnutrition. It's not uncommon in the ICU to encounter patients who are hundreds of pounds overweight yet profoundly protein-depleted.
- For patients with BMI on the lower end (e.g. BMI 30-35), this table may provide a few hundred more calories than recommended by SCCM/ASPEN. It's doubtful that this will have a significant clinical impact, since the evidence surrounding this topic isn't very precise to begin with. Furthermore, most studies show that patients rarely receive 100% target calories.
- Pulmcrit wee: The cutoff razor - April 15, 2024
- PulmCrit Blogitorial – Use of ECGs for management of (sub)massive PE - March 24, 2024
- PulmCrit Wee: Propofol induced eyelid opening apraxia – the struggle is real - March 20, 2024
Thank you for this wonderful update. I do have a few quick questions (all regarding enteral nutrition).
1. Do you feed patients who are on non invasive CPAP/BIPAP?
2. Do you feed patients with ARDS who are in prone position?
3. When do YOU usually start feeding a patient who is in florid sepsis (say, from pneumonia or soft tissue infection)? My feeling (yes, I know…) would be to start after 24 hours with 10mL/h, but I’d be hesitant if pressors where running in high/escalating rates.
with best regards
C.
Thank you, tough questions. (1) BiPAP/CPAP: I generally don’t feed patients who are on BiPAP/CPAP for acute respiratory failure. However, I’m not dogmatic about this. There are patients out in the community who wear BiPAP/CPAP every night, and I’m sure that they sometimes wear these after dinner with a full stomach. For a patient who is stable and gradually improving, it’s probably OK to remove the mask intermittently to have some food (although I wouldn’t encourage this). (2) There isn’t great evidence about feeding in prone position, but some studies have been done on it and it seems OK. I… Read more »
Hi Dr. Farkas, I am a dietitian who works critical care exclusively (Burns, medical, neuro, surgical, and trauma units). I listen to your podcast as well as the emcrit one to help understand a bit more about what is going on around me, and with my patients. As far as the above- mentioned questions, we generally don’t feed on full face masks like Bipap or NIV. However, when we get CF patients in who they are trying to prevent from being intubated, we have/do feed them, as long as they are also hemodynamically stable. Granted, the best nurse in the… Read more »
most cardiologists caution against full feeding in pts with severe CHF or cardiogenic shock. The idea being that the bowel will steal blood flow and that in low flow states other organs would be deprived ie kidney, liver etc. Thibault R, Pichard C, Wernerman J, et al. Cardiogenic shock and nutrition: safe? Intensive Care Med. 2011;37:35–45. Berger MM, Revelly JP, Cayeux MC, et al. Enteral nutrition in critically ill patients with severe hemodynamic failure after cardiopulmonary bypass. Clin Nutr. 2005;24:124–132. Zaloga GP, Roberts PR, Marik P. Feeding the hemodynamically unstable patient: a critical evaluation of the evidence. Nutr Clin Pract.… Read more »
I’m in agreement with these papers, specifically: (1) When approaching a patient in shock (septic, cardiogenic, or otherwise), the first step is hemodynamic stabilization (e.g. fluid, pressor, inotropes). This should be achieved within a 6-12 hour period ideally. (2) Following initial fluid and vasopressor/inotrope stabilization, feeding may be started at a reduced rate (25%-50%, depending on how sick the patient is). This should be monitored carefully and advanced as tolerated. (3) I agree that abruptly initiating full feeding in a patient with cardiogenic shock isn’t a good idea. (4) Would note that the second paper you cited (Berger 2005) supports… Read more »
This pre-dates current recommendations
Thank you for your excellent summary and key points. I have a question in regards to point 6 – we’ve moved away from midnight feed cessation to 4 – 6 hours prior to extubation. You mentioned that your practice was to delay until a definite decision was made and then ensure the stomach was suctioned prior to extubation – I read this as indicating that there was essentially no cessation of NGT feeds prior to suctioning and extubation. I’d like to implement this as I’ve seen this documented as recommended practice elsewhere, however haven’t been able to find any formal… Read more »
Thank you for the awesome summary!
I’ve been trying to figure out the best way to feed ICU patients i.e. bolus vs continuous. So far, many papers didn’t find any differences between the two approaches (For instance, Evans et al conducted a good study that didn’t show any difference between the two approaches, see link below). The bolus feeding seems more physiological though, and should simulate the normal gastric hormonal variations. What are your thoughts about this?
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4795366/
Hi there,
There is research about this topic. While bolus may be “more physiologic,” this method of feeding is generally reserved for more stable/floor patients. While in the ICU, these patients should generally be receiving continuous feeding, as there is less chance for aspiration. The
aforementioned study only compared the following measures: This study examines if there is a difference in glycemic variability, insulin usage, EN volume, and caloric delivery. This study did not include increased risk for aspiration as one of their stated measures, which is the primary reason feedings are generally run continuously.
I’m a little confused – You have here that the guidelines say it’s okay to feed patients EN on vasopressor support. However, when I actually read the guidelines, it states only to trial EN if the patient is on stable and LOW doses of pressors. Am I missing something? I Currently only initiate EN if they are on the appropriate level of pressors as outlined in Enteral Feeding and Vasoactive Agents – Practical Gastroenterology 2009.
Great article! I’m a bit late to the party, but this seems like a good place to leave this. What is your opinion on post-pyloric feeding on NIPPV?
do you have a tube cheat sheet for pediatric formula ?
Any protein requirement that uses nitrogen balance for critically ill, bed-bound, sedated, and/or paralyzed patient is based on lack of understanding of protein needs and muscle synthesis. Anyone who was previously active and is now lying in bed for 24 hours a day, will lose muscle. Nitrogen balance does equal muscle balance. It may be a proxy in growing infants, but not in adults who are no longer in homeostasis. Giving more protein does not mean muscle protein synthesis will increase to match muscle breakdown. Critically ill patients lose muscle mostly from decreased activity. YOU DON’T USE IT, YOU LOSE… Read more »
Nitrogen balance does not equal muscle balance. I could not edit sentence in prior comment, so I am correcting it here.