Nutritional support of critically ill patients
In recent years, great advances have been made in the field of critical care nutrition. While nutrition was once regarded as a low priority supportive measure, it is increasingly becoming recognized as an important therapeutic intervention in the care of critically ill patients. In human intensive care units, nutrition is not only a means of supportive therapy, but a means to modulate even severe diseases. While these developing applications are still years away from being standards in veterinary medicine, they highlight the possibilities of critical care nutrition in veterinary medicine. The current status of veterinary critical care nutrition and the major focus of this article involve careful selection of the patients most likely to benefit from nutritional support, deciding when to intervene, and optimizing nutritional support to individual patients.
Rationale for nutritional support in critical illness
Critical illness induces unique metabolic changes in animals that put them at high risk for malnutrition and its deleterious effects. An important distinction in the body’s response to inadequate nutritional intake occurs in disease (stressed starvation) compared to a healthy state (simple starvation). During fasting in the healthy state, utilization of glycogen stores is the primary source of energy. However, glycogen stores are quickly depleted, especially in strict carnivores such as the cat, and this leads to the initial mobilization of amino acids from muscle stores. Within days, there is a metabolic shift towards the preferential use of stored fat deposits, sparing catabolic effects on lean muscle tissue. In diseased states, the inflammatory response triggers alterations in cytokines and hormone concentrations and shifts metabolism towards a catabolic state. With a lack of food intake, the predominant energy source is derived from accelerated proteolysis, which in itself is an energy consuming process. Thus, these animals may preserve fat deposits in the face of lean muscle tissue loss.
These shifts in metabolism commonly result in a negative nitrogen balance, which sometimes provide a therapeutic target for intervention. The consequences of continued lean body mass loss include negative effects on wound healing, immune function, strength (both skeletal and respiratory), and ultimately, overall prognosis. While a definitive relationship between malnutrition and clinical outcome has not been demonstrated in companion animals, human beings afflicted with malnutrition and critical illness have been documented to have poorer outcomes (1). An important point in regards to nutritional support of hospitalized patients is that the immediate goal is not to achieve “weight gain,” per se, which most likely reflects shift in water balance, but rather to minimize further loss of lean body mass. Moreover, reversal of malnutrition hinges on resolution of the primary underlying disease, and provision of nutritional support is aimed at restoring nutrient deficiencies and minimizing the development of malnutrition in animals at risk.
As with any intervention in critically ill animals, nutritional support carries some risk of complications as well as having the potential to be beneficial. The risk of complications most likely increases with disease severity and the clinician must consider many factors in deciding to institute nutritional support. Of utmost importance for the clinician is to ensure that the patient is cardiovascularly stable before any nutritional support is initiated. In states of shock, perfusion of the gastrointestinal tract is often reduced in favor of maintaining adequate perfusion of the heart, brain, and lungs. With reduced perfusion, processes such as gastrointestinal motility, digestion, and nutrient assimilation are altered, and therefore feeding under such circumstances is likely to result in higher incidences of morbidity. In the setting of critical illness, an important goal of nutritional support is to minimize risk of complications. Other factors that should be addressed prior to nutritional intervention, include dehydration, electrolytes imbalances, and abnormalities in acid-base status.
In animals that have been stabilized, deciding on the appropriate time to start nutritional support also deserves careful consideration. The number of days an animal has not consumed adequate calories prior to hospitalization must be determined from the history and added to the number of days during hospitalization that the animal has not consumed any significant amounts of food. Documentation of total days without adequate nutrition should be listed in the patient’s daily progress notes to ensure that nutritional support remains an important therapeutic goal. The previously held notion that nutritional support is unnecessary until 10 days of inadequate nutritional support, is certainly outdated and unjustified. Commencing nutritional support within 3 days of hospitalization, even before determining the diagnosis of the underlying disease, is a more appropriate goal in most cases. However, other factors should also be considered and are discussed in nutritional assessment.
In order to design the most appropriate and effective nutritional strategy for a patient, a crucial first step involves making a systematic evaluation of the patient, and this is termed nutritional assessment. Nutritional ssessment identifies malnourished patients that require immediate nutritional support and also identifies patients at risk for developing malnutrition in which nutritional support will help to prevent malnutrition.
Since assessing nutritional status via objective measurements of body composition (e.g. anthropometry, bioelectrical impedance, dual energy x-ray absorptiometry, or serum indicators of malnutrition) continues to be of limited use in clinical veterinary medicine, subjective clinical assessment remains paramount in identifying malnourished patients that require nutritional support as well as those in which nutritional support will help prevent malnutrition. Indicators of overt malnutrition include recent unintentional weight loss of at least 10% of body weight, poor hair coat quality, muscle wasting, signs of poor wound healing, and hypoalbuminemia. However, these abnormalities are not specific to malnutrition and are not present early in the process. In addition, fluid shifts may mask weight loss in critically ill patients.
A greater emphasis is now placed on evaluating overall body condition rather than simply noting body weight. The use of body condition scores (BCS) has been shown to be reproducible, reliable, and clinically useful measures in the process of nutritional assessment (2). In addition, as fluid shifts may significantly impact body weight, BCS may be particularly helpful in assessing critically ill animals. Various systems of BCS have been proposed and no system is necessarily better than another. The author extensively uses a 1-9 body condition scoring system, but perhaps more importantly, the consistent application of any particular scoring system by all members of the veterinary practice is preferable. It should be noted that body condition scoring schemes were designed and validated to assess body fat and do not incorporate loss of lean body tissue. A muscle condition score has been proposed and could enhance nutritional assessment, however, further studies are needed to demonstrate the clinical utility of this scoring scheme (3). Factors that predispose a patient to malnutrition include anorexia lasting longer than three days, serious underlying disease (e.g. trauma, sepsis, peritonitis, pancreatitis, and significant gastrointestinal surgery), and large protein losses (e.g. protracted vomiting, diarrhea, or draining wounds). Nutritional assessment also identifies factors that can impact the nutritional plan, such as cardiovascular instability, electrolyte abnormalities, hyperglycemia, and hypertriglyceridemia or concurrent conditions such as renal or hepatic disease that will impact the nutritional plan. Appropriate laboratory analysis should be performed in all patients to assess these parameters. Before implementation of any nutritional plan, the patient must be cardiovascularly stable, with major electrolyte, fluid, and acid-base abnormalities corrected.
Goals of nutritional support
Even in patients with severe malnutrition, the immediate goals of therapy should focus on fluid resuscitation, stabilization, and identification of the primary disease process. As steps are made to address the primary disease, formulation of a nutritional plan should strive to prevent (or correct) overt nutritional deficiencies and imbalances. Placement of feeding tubes (which require general anesthesia) should only be performed once the patient is deemed stable.
By providing adequate energy substrates, protein, essential fatty acids, and micronutrients, the body can support wound healing, immune function, and tissue repair. A major goal of nutritional support is to minimize metabolic derangements and catabolism of lean body tissue. During hospitalization, repletion of body weight is not a priority as this will only occur when the animal is recovering from a state of critical illness. Therefore body weight gain is not a goal whilst the animal is hospitalized for the majority of cases. However, continued weight loss during hospitalization is of particular concern and should be addressed. The ultimate goal of nutritional support is to have the patient eating adequate amounts of food in its own environment.
Proper diagnosis and treatment of the underlying disease is the key to the success of nutritional support. Based on the nutritional assessment, a plan is formulated to meet energy and other nutritional requirements of the patient and at the same time address any concurrent condition requiring adjustments to the nutritional plan.
The anticipated duration of nutritional support should be determined and factored into the plan. This will largely depend on clinical familiarity with the specific disease process and sound clinical judgment. For each patient, the best route of nutrition should be determined, i.e. enteral versus parenteral nutrition. This decision should be based on the underlying disease and the patient’s clinical signs. Whenever possible, the enteral route should be considered first and options for various feeding tubes are listed in Table 1. If enteral feedings are not tolerated or the gastrointestinal tract must be bypassed however, parenteral nutrition should be considered. Nutritional support should be introduced gradually and reach target levels in 48-72 hours.
With regards to appetite stimulants, it is the author’s opinion that they have no place in the nutritional management of hospitalized critically ill patients. The only means of ensuring adequate caloric intake is through nutritional support (i.e. tube feeding or parenteral nutrition). Appetite stimulants could be used once the patient is recovering from its disease and at home.
Calculating nutritional requirements
Ideally, the provision of nutritional support should provide ample substrates for gluconeogenesis, protein synthesis, and energy necessary to maintain homeostasis without causing complications. Ensuring that sufficient calories are being provided to sustain critical physiological processes such as immune function, wound repair, and cell division and growth, would necessitate actually measuring the patient’s total energy expenditure. However, precise measurements of energy expenditure in clinical veterinary patients remain largely impractical. While a few studies have used indirect calorimetry to estimate energy expenditure in select populations of clinical veterinary patients, the use of mathematical formulas remains the only practical means of estimating the patient’s energy requirement.
Results of indirect calorimetry studies in dogs support the recent trend of formulating nutritional support to meet resting energy requirements (RER), rather than more generous illness energy requirements (4, 5). Until recently, many clinicians multiplied the RER by an illness factor between 1.1 – 2.0 to account for increases in metabolism associated with different diseases and injuries. However, less emphasis is now being placed on such subjective and extrapolated factors and the current recommendation is to use more conservative energy estimates, i.e. start with the animal’s RER, to avoid overfeeding and its associated complications (6).
While there are several formulas proposed to calculate the RER, a widely used allometric formula can be applied to both dogs and cats of all weights. The formula most commonly used by the author is:
RER = 70 (BW in kg)0.75
Alternatively, for animals weighing between 2 and 45 kg the following may be used:
RER = 30 (BW in kg) + 70
Currently, it is generally accepted that hospitalized dogs should be supported with 4-6 grams of protein/100 kcal (15-25% of total energy requirements), while cats are usually supported with 6 or more grams of protein/100 kcal (25-35% of total energy requirements) (7). Patients with protein intolerance (e.g. hepatic encephalopathy, severe azotemia) should receive reduced amounts of protein (approximately 3g of protein/100 kcal). Similarly, patients with hyperglycemia or hyperlipidemia may also require decreased amounts of carbohydrates and lipids, respectively. Other nutritional requirements will depend upon the patient’s underlying disease, clinical signs, and laboratory parameters.
In animals with a functional gastrointestinal tract, the use of feeding tubes is the standard mode of nutritional support in critically ill animals. (Editor’s note: within Europe, Convalescence Support is ideal for tube feeding). As discussed previously, a key decision is determining whether the patient can undergo general anesthesia for placement of feeding tubes. In animals with surgical diseases requiring laparotomy, placement of gastrostomy or jejunostomy feeding tubes should receive particular consideration. Feeding tubes commonly used in critically ill animals include nasoesophageal, esophagostomy, gastrostomy, and jejunostomy feeding tubes (Table 1). The decision to choose one tube over another is based on the anticipated duration of nutritional support (e.g. days versus months), the need to circumvent certain segments of the gastrointestinal tract (e.g. oropharynx, esophagus, pancreas), clinician experience, and suitability of patient to withstand anesthesia (very critically ill animals may only tolerate placement of nasoesophageal feeding tubes). In-depth instructions for the placement of feeding tubes are beyond the scope of this article, and the reader is referred elsewhere (8-11).
The major advantages of nasoesophageal feeding tubes is that placement is relatively simple, requires minimal if any, sedation, and no special equipment is necessary. Since this is a largely blind procedure, verification of placement within the esophagus with radiography or an end-tidal CO2 monitor is recommended (12). Tubes placed with the gastrointestinal tract should yield no CO2 when checked with an end-tidal CO2 monitor (12). Disadvantages of nasoesophageal tubes include patient discomfort, and the need to use an exclusively liquid diet, because tubes are typically 3.5 to 5 Fr in size.
Esophageal feeding tubes are an excellent choice in many critically ill animals and have completely supplanted the need for pharyngostomy tubes. They are also relatively easy to place, require only brief anesthesia and can accommodate more calorically-dense diets (i.e. > 1 kcal/mL), making them ideal for patients that are fed limited volumes. Tubes ranging from 12 to 19 Fr are commonly used and patient discomfort is usually not an issue. The most common problems associated with this tube are tube obstruction and cellulitis at the stoma site. Feeding through esophageal feeding tubes is usually performed via intermittent bolus feeding, but low rate continuous infusions can be used in animals that cannot tolerate bolus feeding.
Surgically placed and percutaneous endoscopy-guided gastrostomy tubes are good options for patients undergoing laparotomy and endoscopy, respectively. These tubes can be used for long term nutritional support (i.e. months) and feeding with these tubes are usually achieved via bolus feeding. Gastrostomy feeding tubes are generally larger (16 – 32 Fr) and can accommodate almost any diet with little further processing. These tubes do require special equipment and considerable experience, but can be quite effective. Complications associated with these tubes range from mild cellulitis around the stoma site, to more serious life-threatening peritonitis. Patients with premature tube dislodgement (prior to 14 days) should be immediately evaluated for the need for possible surgery.
Animals requiring laparotomy and deemed to be in need of a delivery system which bypasses the stomach or pancreas (e.g. significant stomach wall resection, severe pancreatitis, pancreatectomy) should have a jejunostomy feeding tube placed. These tubes are similar in size to nasoesophageal tubes and therefore can only accommodate liquid diets. Feeding through these tubes should also be performed via continuous infusions (e.g. 1 mL/kg/hr initially and slowly increased) rather than with bolus feeding. Complications associated with these feeding tubes include tube occlusion, diarrhea, and tube dislodgement, which results in peritonitis. Calculation of amounts of enteral feeding that should be administered to patients with feeding tubes are found in Table 2.
A common misconception is that animals fed via feeding tubes will not eat voluntarily and so feedings are withheld to evaluate the animal’s appetite. It is important to remember that the main purpose of nutritional support is to provide nutrients and calories that the animal needs and less emphasis should be placed on appetite per se. As discussed earlier, reversal of anorexia should be corrected once the primary disease is addressed. Weaning animals from tube feedings whilst they are still hospitalized is discouraged and perhaps it should be more appropriately done once the animal is discharged from the hospital and recovering in its own environment. As feeding regimens by the time of hospital discharge should have been reduced to 3 or 4 times daily, owners can be instructed to offer oral feeding before each tube feeding to monitor for return of adequate spontaneous feeding. Based on reevaluation and reassessment by the clinician, tube feedings can then be reduced or discontinued depending on the progress made.
In the author’s opinion, mastering the placement of esophagostomy feeding tubes is essential in the management of critically ill animals and this technique should be adopted in almost all practices. A step-by-step description of this technique is outlined in Table 3. A volume of 5-10 mL/kg per individual feeding is generally well-tolerated but this may vary with the individual patient. In patients that are generally healthy but cannot consume food orally, e.g. jaw fracture, larger volumes of food per feeding (15-20 mL/kg) may be tolerated. As enteral diets are mostly composed of water (most canned foods are already >75% water) the amounts of fluids administered parenterally should be adjusted accordingly to avoid volume overload. Care should be taken to avoid wrapping the feeding tube too tightly as this could lead to patient discomfort and even compromise proper ventilation.
Parenteral nutrition (PN) is more expensive than enteral nutrition and requires vigilant and intensive monitoring. Indications for PN include protracted vomiting, acute pancreatitis, severe malabsorptive disorders, and severe ileus. While terminology of PN can be confusing there are two major types:
• Total parenteral nutrition (TPN) is typically delivered via a central venous (jugular) catheter and provides all of the energy requirements of the patient.
• Partial parenteral nutrition (PPN) delivers only a portion of the animal’s energy requirements (40-70%) but because of the lower osmolarity of the solution, it can usually be administered through a large peripheral vein such as the lateral saphenous in dogs and femoral vein in cats (hence PPN is sometimes referred to as Peripheral Parenteral Nutrition). Because PPN only provides a portion of the patient’s requirements, it is only intended for short-term use in a non-debilitated patient with average nutritional requirements. Regardless of the exact form of PN, intravenous nutrition requires a dedicated catheter that is placed using absolute aseptic technique. Multi-lumen catheters are often recommended for PN because they can remain in place for longer periods of time as compared to normal jugular catheters and provide other ports for blood sampling and administration of additional fluids and IV medications. Most PN solutions are composed of a carbohydrate source (dextrose), a protein source (amino acids), and a fat source (lipids). Vitamins and trace metals can also be added.
Formulation of TPN and PPN solutions can be individualized to each patient and the reader is referred to other references for further details (6, 13). In most cases, it is easiest to have a local human hospital formulate TPN and PPN. However, PN therapy should be reserved for practices that can provide a high level of critical care e.g. 24 hr nursing, and dedicated intensive care units.
Monitoring and reassessment
Body weight should be monitored daily with both enteral or parenteral nutrition. However, the clinician should take into account fluid shifts in evaluating changes in body weight. For this reason, body condition scores are important as well. The use of the RER as the patient’s caloric requirement is merely a starting point. The number of calories provided may need to be increased to keep up with the patient’s changing needs, typically by 25% if well tolerated. In patients unable to tolerate the prescribed amounts, the clinician should consider reducing the amounts of the enteral feedings and supplementing the feeding with PPN.
Possible complications of enteral nutrition include mechanical complications such as obstruction of the tube or early tube removal. Metabolic complications include electrolyte disturbances, hyperglycemia, volume overload, and gastrointestinal signs (e.g. vomiting, diarrhea, cramping, bloating). In critically ill patients receiving enteral nutritional support, the clinician must also be vigilant for the development of aspiration pneumonia. Monitoring parameters for patients receiving enteral nutrition include body weight, serum electrolytes, tube patency, appearance of tube exit site, gastrointestinal signs (e.g. vomiting, regurgitation, diarrhea), and signs of volume overload or pulmonary aspiration.
Possible complications with PN include sepsis (low risk), thrombophlebitis, and metabolic disturbances such as hyperglycemia, electrolyte shifts, hyperammonemia, and hypertriglyceridemia. Avoiding serious consequences of complications associated with PN requires early identification of problems and prompt action. Frequent monitoring of vital signs, catheter-exit sites, and routine biochemistry panels may alert the clinician of problems that are developing. The development of persistent hyperglycemia during nutritional support may require adjustment to the nutritional plan (e.g. decreasing the dextrose content in PN) or the administration of regular short-acting insulin. This obviously necessitates more vigilant monitoring.
With continual reassessment, the clinician can determine when to transition the patient from assisted feeding to voluntary consumption of food. The discontinuation of nutritional support should only begin when the patient can consume approximately 75% RER without much coaxing. In patients receiving TPN, transitioning to enteral nutrition should occur over the course of at least 12-24 hours, depending on tolerance of the patient to enteral nutrition.
In recent years, there has been particular focus in the pharmacological role of nutrients in modulating disease in human beings. “Immune-enhancing diets” often include nutrients such as glutamine, arginine, omega-3 fatty acids, antioxidants, and nucleotides. In certain populations of critically ill human patients, these strategies (singly or in combination cocktails) provide significant benefits in reducing complications and even decreasing mortality (14-16). Unfortunately, no such benefits have been documented in veterinary patients. Documentation of depletion of these key nutrients in animals with naturally-occurring disease is also lacking. While such nutrients may indeed confer health benefits to veterinary patients, studies confirming these benefits are unlikely to be forthcoming. As the majority of veterinary patients are only hospitalized for a short term (a relatively low percentage are hospitalized for more than 10 days), pharmacological effects of nutrients will be difficult to discern. While the risk of side-effects from these therapies are likely low, the added cost of such supplements in the face of little supporting evidence in veterinary patients may make use of such products unwarranted at this time. As nutrient requirements vary considerably among species, it is likely that the same holds true in respect to the pharmacological effects of nutrients. Determination of minimal dosages for each nutrient necessary to achieve desired biological effects may be warranted as the next step.
While critically ill patients are often not regarded as in urgent need of nutritional support given their more pressing problems, the severity of their injuries, altered metabolic condition, and necessity of frequent fasting, place these patients at high risk of becoming malnourished during their hospitalization. Proper identification of these patients and careful planning and execution of a nutrition plan can be key factors in the successful recovery of these patients.
1. Wray CJ, Mammen JM, Hasselgren P. Catabolic response to stress and potential benefits of nutritional support. Nutrition 2002; 18: 960-965.
2. Mawby DI, Bartges JW, D’Avignin A, et al. Comparison of various methods for estimating body fat in dogs. J Am Anim Hosp Assoc 2004; 40: 109-114.
3. Buffington T, Holloway C, Abood A. Nutritional Assessment. In: Buffington T, Holloway C, Abood S (eds). Manual of veterinary dietetics. WB Saunders, St. Louis 2004. pp. 1-7.
4. O’Toole E, Miller CW, Wilson BA, et al. Comparison of the standard predictive equation for calculation of resting energy expenditure with indirect calorimetry in hospitalized and healthy dogs. J Am Vet Med Assoc 2004; 255: 58-64.
5. Walton RS, Wingfield WE, Ogilvie GK. Energy expenditure in 104 postoperative and traumatically injured dogs with indirect calorimetry. J Vet Emerg Crit Care 1998; 6: 71-79.
6. Freeman LM, Chan DL. Total Parenteral Nutrition. In: DiBartola SP (ed). Fluid, Electrolyte, and Acid-base Disorders in Small Animal Practice. Saunders Elsevier, St. Louis 2006; 3: 584 – 601.
7. Chan DL. Nutritional requirements of the critically ill patient. Clin Tech Small Anim Pract 2004; 19: 1-5.
8. Abood SK, Buffington CA. Improved nasogastric intubation technique for administration of nutritional support in dogs. J Am Vet Med Assoc 1991; 199: 577-579.
9. Mazzaferro EM. Esophagostomy tubes: don’t underutilize them! J Vet Emerg Crit Care 2001; 11: 153-156.
10. Bright RM. Use of percutaneous gastrostomy tubes and low profile feeding devices. In: Bojrab MJ (ed). Current Techniques in Small Animal Surgery. Williams & Wilkins, Baltimore 1998; 4: 170-176.
11. Devitt CM, Seim HB. Use of jejunostomy and enterostomy tubes. In: Bojrab MJ (ed). Current Techniques in Small Animal Surgery. Williams & Wilkins, Baltimore 1998; 4: 177-182.
12. Johnson PA, Mann FA, Dodam J, et al. Capnographic documentation of nasoesophageal and nasogastric feeding tube placement in dogs. J Vet Emerg Crit Care 2002; 12: 227-233.
13. Zsombor-Murray E, Freeman LM. Peripheral parenteral nutrition. Compend Contin Educ Pract Vet 1999; 21: 512-523.
14. Heyland DK, Novak F, Drover JW, et al. Should immunonutrition become routine in critically ill patients? A systematic review of the evidence. J Am Med Assoc 2001; 286: 944-953.
15. Novak F, Heyland DK, Avenell A, et al. Glutamine supplementation in serious illness: a systematic review of the evidence. Crit Care Med 2002; 30: 2022-2029.
16. Mendez C, Jurkovich GJ, Garena I, et al. Effects of an immuneenhancing diet in critically injured patients. J Trauma 1997; 42: 933-940.