Timing of Enteral Nutrition
While enteral feeding is the preferred route of nutrient administration, how soon it should be started after an acute injury or insult is not clear. In critically ill patients, there were eight randomized controlled trials comparing early EN (i.e., that started within 24 to 48 hours of admission to the ICU) to some form of delayed nutrient intake (i.e., delayed EN or oral diet).49 When results from these studies were aggregated, early EN was associated with a trend towards a reduction in mortality (RR 0.52; 95% CIs 0.25, 1.08; p = 0.08) when compared to delayed nutrient intake. Three studies reported infectious complications.50–52 When these were aggregated, early EN was associated with a trend towards a reduction in infectious complications (RR 0.66; 95% CIs 0.36, 1.22; p = 0.19) when compared to delayed nutrient intake. No differences in length of stay were observed between groups. All seven studies that reported nutritional end points showed a significant improvement in the groups receiving early EN (e.g., improvements in calorie intake, protein intake, percentage of goal achieved, and better nitrogen balance achieved). There were no differences in other complications between the groups.
Although the results lack statistical significance, they do suggest a large improvement in clinical outcome and a significant increase in nutrient delivery associated with early enteral feeding. However, before endorsing the concept of early enteral feeding, one must consider the potential risks of such a strategy. Two recent nonrandomized studies suggest that early enteral feeds delivered into the stomach may be associated with increased complications.53,54 In contrast, Taylor and colleagues combined an aggressive early feeding protocol with the use of small bowel feedings and documented that head-injured patients fed aggressively, compared to standard (slower) provision of EN, not only had better nutritional status, but also had fewer complications and a more rapid recovery from their illness.42
Synthesizing these discordant results, it would seem that early EN may be associated with improved clinical outcomes when done in such a way that maximizes the benefits and minimizes the risks (see below). Furthermore, it should be noted that the goal of early EN, while critically ill patients are still early in the acute phase of their illness, is to provide enough critical nutrients to the gut to modulate the disease process and enhance gut barrier structure and function, not to meet their caloric requirements as soon as possible. Thus for some patients with evidence of inadequate oxygen delivery, specific nutrients (e.g., glutamine and antioxidants) may be required in the first few days of critical illness. If patients are still on high-dose inotropes to maintain adequate blood pressure, the risk of providing EN may outweigh the benefits, and EN should be held until patients are weaning off pressor agents. As the patient enters the more stable, chronic phase of their illness (after 7 to 10 days), preserving lean body mass, stimulating protein synthesis, and minimizing nutritional deficits, rather than maintaining GI structure and function, become the primary goals.
Reducing Risk of Aspiration
It is important on initial evaluation to assess the patient's risk for aspiration on EN. Aspiration may occur from the antegrade passage of contaminated oropharyngeal secretions or the retrograde passage of contaminated gastric contents into the larynx. Regurgitation occurs more frequently than aspiration.55 A number of risk factors have been identified that increase risk of aspiration in the ICU.56 While it is difficult to quantify or stratify degree of risk among these factors, a simple categorization differentiates major risk factors for which change in management strategy may be needed, versus additional minor risk factors that may not warrant specific change in therapeutic course. Major risk factors include documented previous episode of aspiration, decreased level of consciousness (including sedation or increased intracranial pressure), neuromuscular disease, structural abnormalities of the aerodigestive tract, need for endotracheal intubation, overt vomiting or regurgitation, need for prolonged supine position, and persistently high gastric residual volumes.56Additional risk factors include presence of a nasoenteric tube, noncontinuous or bolus intermittent feeding, abdominal/thoracic surgery or trauma, delayed gastric emptying, poor oral care, advanced age, inadequate nursing staff, large bore feeding tube, malpositioned enteral tube (back into the esophagus), or transport out of the ICU.1,56 Strategies to prevent aspiration in patients receiving nutrition support who have significant risk factors, as outlined below, should be utilized to minimize the risks associated with EN in this setting.
Role of Small Bowel Feeding
A number of strategies may be employed to maximize the delivery of EN while minimizing the risks of gastric colonization, gastroesophageal regurgitation, and pulmonary aspiration (Table 11-1). By delivering enteral feeds into the small bowel, beyond the pylorus, the frequency of regurgitation and aspiration, as well as the risk of pneumonia, is decreased while at the same time nutrient delivery is maximized.57 There are seven randomized trials that evaluated the effect of route of feeding on rates of ventilator-associated pneumonia (VAP).58 When these results were aggregated, there was a significant reduction in VAP associated with small bowel feedings (RR 0.76; 95% CIs 0.59, 0.99) compared to gastric feeding. Therefore, the converse is also true. In some patients intragastric feeding may be associated with inadequate delivery of nutrition, increased regurgitation, pulmonary aspiration, and pneumonia, particularly if patients are cared for in the supine position.
Table 11–1. Summary of Strategies to Optimize the Benefits and Minimize the Risks of Enteral Nutrition and Total Parenteral Nutrition |Favorite Table|Download (.pdf)
Table 11–1. Summary of Strategies to Optimize the Benefits and Minimize the Risks of Enteral Nutrition and Total Parenteral Nutrition
|Enteral Nutrition||Total Parenteral Nutrition|
|Initiate early, within 24–48 hours of admission||Hypocaloric dose|
|Use small bowel feedings||Do not use lipids for short term use (<10 days)|
|Elevate head of the bed||Tight control of blood sugars|
|Use motility agents||Supplement with glutamine|
|Use feeding protocol that enables consistent evaluation of gastric residual volumes and specifies when feeds should be interrupted||Continue to trickle concentrated amounts of enteral nutrition if able|
|Use concentrated feeding formulae in cases of intolerance|
|Consider formulae with immune additives|
The clinical implications of these findings are influenced by the inherent difficulties in obtaining small bowel access. Given that some patients will tolerate intragastric feeds, it seems more prudent to reserve small bowel feeds for patients at high risk for intolerance to EN (due to use of inotropes, continuous infusion of sedatives, paralytic agents, high gastric residual volumes, or patients with high nasogastric drainage) or at high risk for regurgitation and aspiration (nursed in prolonged supine position).
While several studies document that elevation of the head of the bed is associated with less regurgitation and pulmonary aspiration, only one randomized controlled trial compared the frequency of pneumonia in critically ill patients assigned to semirecumbent or supine position.59 Drakulovic and colleagues demonstrated that providing EN into the stomach in patients kept in the supine position was associated with a much higher risk of pneumonia compared to feeding patients with the head of the bed elevated to 45° (23% vs. 5%, p <0.05). Thus, a simple maneuver (i.e., elevating the head of the bed to 30° to 45°) can reduce the risks associated with enteral feedings.
Gastrointestinal prokinetic agents improve gastric emptying, improve tolerance to enteral nutrition, reduce gastroesophageal reflux and pulmonary aspiration, and therefore may have the potential to improve outcomes in critically ill patients.60 While no study has demonstrated an impact from use of these agents on clinical outcomes, their low probability of harm and favorable feasibility and cost considerations warrant their use as a strategy to optimize nutritional intake and minimize regurgitation. Since cisapride is no longer available and due to the concerns of bacterial resistance with the use of erythromycin, metoclopramide is probably the drug of choice. It can be prescribed with the initiation of enteral feeds or reserved for patients who experience persistently high gastric residuals. It can be discontinued after four doses if there is no benefit observed, or after tolerance to EN is no longer a problem clinically. Reducing narcotic dosages and potentially reversing their effect at the level of the gut by infusing naloxone through the feeding tube, and switching from bolus intermittent feeds to continuous infusion may also be effective in improving gastric function and tolerance to EN, while reducing risk of aspiration.56 Methods not recommended solely to reduce risk of aspiration include switching to PN, adding acid to the enteral formula, switching from a large bore to a small bore nasoenteric tube, or converting a nasogastric tube to a percutaneous endoscopic gastrostomy tube.56
Several observational studies document that EN is frequently interrupted for high gastric residual volumes, procedures, nausea and vomiting, and other miscellaneous reasons.61 Over the duration of ICU stay, this may result in inadequate delivery of EN to a critically ill patient and the associated complications of inadequate nutrition. Nurse-directed feeding protocols or algorithms have been shown to increase the amount of EN delivered on a daily basis.62 Instituting a feeding protocol in ICUs that provides specific instructions on the patient's management related to EN to the bedside nurse has the potential to improve nutrient delivery and decrease complications.
Role of Immune Stimulants and Antioxidants
An additional strategy to maximize the benefits of enteral nutrition is to consider using products supplemented with specific nutrients that modulate the immune system, facilitate wound healing, and reduce oxidative stress. Enteral formulas developed to such an extent contain selected substrates such as glutamine, arginine, and omega-3 fatty acids, as well as selenium, vitamins E, C, and A, and beta-carotene in supraphysiologic concentrations. Unfortunately, with the possible exception of glutamine, the nutrients by themselves have not been adequately studied in critically ill patients, so their individual efficacy remains unknown. Nevertheless, these nutrients have been combined together and marketed as an immune-enhancing diet. We use the term immunonutrition as a general term to describe all these enteral products, but attempt to make summary recommendations based on the specific nutrients by themselves.
Supplementing arginine in the diet has a variety of biologic effects on the host63,64 (Fig. 11-3). L-arginine is an active secretagogue that stimulates the release of growth hormone, insulin growth factor, and insulin, all of which may stimulate protein synthesis and promote wound healing. Conversion of arginine to ornithine by arginase provides two further functions. This pathway enables shuttling of nitrogen to urea, and ornithine is utilized in polyamine synthesis (which is involved in deposition of hydroxyproline, collagen, and the laying down of connective tissue to heal wounds). Arginine has also been shown to have significant immunostimulatory effects. Arginine has a trophic effect on the thymus gland that promotes the production and maturation of T lymphocytes. In the nitric oxide synthase pathway, the precursor arginine may contribute to improved bacterial killing.63
Arginine metabolic pathways.
Of interest is the fact that the arginase pathway is driven by a Th2 cytokine profile, mediated by further release of IL-4, IL-10 and TGF-β.64 The Th2 cytokine profile has the effect of reducing the overall inflammatory immune response. In contrast, the nitric oxide synthase pathway is mediated by a Th1 cytokine profile, and is perpetuated by further release of IL-1, TNF, and IFN-γ.64 This pathway has the capability of promoting the inflammatory response and inducing the formation of nitric oxide. Increased levels of nitric oxide may exert a negative inotropic and chronotropic effect on the cardiovascular system, and promote vasodilation (which may contribute to the hypotension and shock associated with sepsis syndrome). Nitric oxide in larger amounts may act as a mitochondrial toxin and inhibit several steps in the oxidative phosphorylation chain. Nitric oxide may also damage gut epithelium, increasing bacterial translocation and reducing overall gut integrity.63 Nitric oxide can also have nonspecific cytotoxic effects of inhibiting growth or killing cells indiscriminately.63
There are no randomized studies of pure arginine supplementation in critically ill patients that evaluate clinically important outcomes. All studies in critically ill patients have combined arginine with other immune-modulating nutrients. When the results of these 15 trials were aggregated, there was no effect on mortality (RR 1.05; 95% CIs 0.82, 1.35; p = 0.7), no overall effect on infectious complications (RR 0.94; 95% CIs 0.76 to 1.16; p = 0.6), and a trend towards reduction in hospital length of stay (weighted mean difference –3.5; 95% CIs –8.8, 1.9; p = 0.20). The presence of significant statistical heterogeneity across studies weakens the estimate of effect on length of stay.
Concerns have been expressed that based on the scientific rationale presented above, arginine-containing products may worsen outcomes in critically ill septic patients.65 There are now three reports in the literature of excess mortality associated with critically ill septic patients who received arginine-supplemented diets.66–68 In contrast, Galban and colleagues69 demonstrated an increase in survival associated with arginine-supplemented diets in critically ill patients with infection. However, in this study, it was apparent that all the treatment benefit was in the least sick patients (baseline APACHE II score <15). The effect of arginine-containing products on critically ill patients with a high severity of illness remains unanswered. To the extent that in sepsis endotoxin exposure and cytokine activation have led to elevated levels of inducible nitric oxide synthesis, supplemental arginine may lead to the production of excessive amounts of nitric oxide, shock, and early death. Thus arginine-supplemented specialized diets should not be used in critically ill patients who are clearly septic. If a critically ill patient receiving an arginine-supplemented diet develops sepsis, the arginine-containing diet should be discontinued. Which subpopulations of critically ill patients benefit from these diets remains to be elucidated.
Omega-3 fatty acids may be provided in the form of fish oil or canola oil. These agents do not have direct stimulatory effects, but instead have an indirect effect by modifying phospholipids in cell membranes throughout the body.70 Omega-6 fatty acids are involved in the cyclooxygenase pathway, generating prostaglandin E2 (PGE2) and leukotriene B4 (LTB4) from arachidonic acid. These are proinflammatory cytokines that lead to immune suppression and nosocomial infection, SIRS, and organ dysfunction. Through diet supplementation, omega-3 fatty acids compete with the omega-6 fatty acids for incorporation into cell membranes. Upon activation of the cyclooxygenase pathway, omega-3 fatty acids instead lead to the formation of PGE3 and LTB5. These compounds have $110 the biologic activity of the PGE2 and LTB4 series, and as a result have a much less immunosuppressive effect.70 Borage oil is unique as an omega-6 fatty acid, because it is metabolized to the PGE1 series. PGE1 possesses both anti-inflammatory and antiproliferative (reduced thrombosis) properties, and will attenuate the biosynthesis of arachidonic acid metabolites.71
The only study of omega-3 fatty acids was conducted in patients with acute lung injury (ALI)/acute respiratory distress syndrome (ARDS).72 The pathophysiology of this syndrome is thought be related to the release of arachidonic acid–related metabolites from inflammatory cells. Gadek and colleagues performed a multicenter, randomized, double-blind, clinical trial to evaluate whether a diet supplemented with eicosapentaenoic acid, docosahexanoic acid (DHA), borage oil, and antioxidants (Oxepa, Ross Products, Columbus, OH) would have a favorable effect on markers of inflammation in the lung and an improvement in clinical outcomes.72 In this study, 146 patients meeting the standard definition of ARDS with evidence of active pulmonary inflammation as indicated by fluid from a bronchoalveolar lavage (BAL) that contained a neutrophil count >10%, were randomized to the experimental diet or a high-fat, low carbohydrate control feed.
In the subset of “evaluable” patients, those who received the experimental diet had higher plasma levels of dihomo-gamma linolenic acid, eicosapentaenoic acid, and an increased eicosapentaenoic/arachidonic acid ratio. With respect to the clinically important outcomes, patients fed the experimental diet experienced a reduction in days receiving supplemental oxygen (13.6 vs. 17.1; p = 0.078), required significantly fewer days of ventilatory support (9.6 vs. 13.2; p = 0.027), and less time in the ICU (11.0 vs. 14.8; p = 0.016), and had fewer new organ failures (10% vs. 25%; p = 0.018). There was also a trend towards a reduction in mortality associated with the experimental diet (16% vs. 25%; p = 0.17).
This study confirms that short-term administration of dietary lipids in critically ill patients can modify fatty acid levels with a resultant favorable effect on neutrophil recruitment in the lung and subsequent clinical outcomes. However, given that a high-fat diet in itself may be harmful,73 the use of a high-fat control formula and the requirement for a BAL to identify patients with a high degree of neutrophils limits the applicability of these findings to clinical practice. Furthermore, it is difficult to attribute the beneficial effects of the experimental diet to fatty acid composition when it was combined with antioxidants as well.
The amino acid glutamine plays a central role in nitrogen transport within the body, is a fuel for rapidly dividing cells (particularly lymphocytes and gut epithelial cells), is a precursor to glutathione, and has many other essential metabolic functions. As noted previously, plasma glutamine levels drop during critical illness, and lower levels of glutamine have been associated with immune dysfunction74 and increased mortality.75 Human studies suggest that glutamine supplementation maintains gastrointestinal structure76 and is associated with decreased intestinal permeability compared to standard PN.77,78 In humans, glutamine-supplemented formulas have resulted in improved nitrogen balance,79 and higher intramuscular glutamine levels.80 Glutamine plays a crucial role in enhancing immune cell function81 with no elevation in proinflammatory cytokine production.82,83
There have been several randomized trials of perioperative or critically ill adults reporting on clinically important outcomes.84 When the results of these trials were aggregated, a significant reduction in mortality (RR 0.78; 95% CIs 0.61 to 0.99; p = 0.04), a trend towards a reduction in infectious complications (RR 0.89; 95% CIs 0.73 to 1.08; p = 0.2), and no overall effect on length of stay (LOS in days –1.30; 95% CIs – 4.77, 2.17; p = 0.5) were observed.85 Subgroup analysis suggested that with respect to mortality and infectious complications, the majority of the treatment effect observed was associated with parenteral glutamine in patients receiving PN compared to enteral glutamine supplementation. The majority of glutamine provided enterally will be metabolized in the gut and liver, and therefore may not have a systemic effect. The only study that demonstrated a mortality effect with enteral glutamine was a small study in burn patients.86 In a study of trauma patients, enteral feeds supplemented with glutamine were associated with a trend towards a reduced rate of infection compared to control feeds (20/35 [57%] vs. 26/37 [70%], p = 0.24).87
Therefore, for critically ill patients requiring PN, we recommend parenteral glutamine supplementation as long as the patient remains on PN. For patients with major burns or trauma, enteral diets supplemented with glutamine could be considered. Recommendations about glutamine supplementation (enteral or parenteral) in other critically ill patient populations fed enterally are premature and warrant further study.
Antioxidant Vitamins and Trace Minerals
While there is a putative beneficial role of reactive oxygen species in modulating cell signaling (redox signaling), and thus regulating proliferation, apoptosis, and cell protection, oxygen-derived radicals may cause cellular injury by numerous mechanisms, including destruction of cell membranes through the peroxidation of fatty acids, disruption of organelle membranes such as those covering lysosomes and mitochondria, degradation of hyaluronic acid and collagen, and disruption of enzymes like Na+,K+-ATPase or alpha1-proteinase inhibitor.
To protect tissues from oxygen free radical–induced injury, the body maintains a complex endogenous defense system that consists of a variety of extra- and intracellular antioxidant defense mechanisms. The first line of intracellular defense is comprised of a group of antioxidant enzymes such as superoxide dismutase, catalase, glutathione peroxidase, and glutathione reductase, including their metal cofactors selenium, copper, and zinc. When these enzymatic antioxidants are overwhelmed, oxygen free radicals (OFRs) are free to react with susceptible target molecules within the cell (i.e., unsaturated fatty acids of the cell membrane). Thus there is a need for a second line of defense scavenging OFRs by means of nonenzymatic antioxidants that are either water soluble, such as glutathione and vitamin C, or lipid soluble, such as vitamin E and beta-carotene.88
In critical illness, oxidative stress arises when the balance between protective antioxidant mechanisms and the generation of reactive oxygen species (ROS) is disturbed. This imbalance may be caused by excess generation of ROS by means of ischemia/reperfusion injury, inflammation, infection, and toxic agents (chemotherapy or drugs), or by low antioxidant capacity (secondary to comorbid illnesses, malnutrition, and excessive losses such as in the case of burns). Many studies have demonstrated low plasma and intracellular concentrations of the various antioxidants in critically ill patients, and the clinical consequence of these low endogenous stores of antioxidant levels is increased morbidity and mortality.89–91
Most of the immune formulas are fortified with vitamins and minerals that have increased antioxidant capabilities. Vitamins A, E, and C, and the trace mineral selenium have antioxidant capabilities and are added in different amounts to the various formulas. The exact doses of these components have not been standardized.
Single Antioxidant Nutrients
Selenium is an important co-factor in glutathione enzymatic function and has favorable effects on cellular immune function. In critically ill patients, there have been only a few randomized controlled trials looking at the effects of selenium supplementation alone.92–95 In a poorly designed trial of 17 patients with acute necrotizing pancreatitis, parenteral supplementation of 500 μg of selenium was associated with a significant reduction in ICU mortality (0% vs. 89%).93 In a prospective randomized trial, Zimmermann and colleagues reported a reduction in mortality (15% vs. 40%) after IV administration of 1000 μg of sodium selenite for 28 days in patients with SIRS compared to placebo.94 However, no difference in mortality or pneumonia was seen in critically ill trauma/surgical patients given IV selenium supplementation (2.9 μmol/d) compared to placebo.92 In a trial of 42 patients with SIRS, subjects that received a higher dose of parenteral selenium (535 μg/d × 3 days, 285 μg/d × 3 days, 155 μg/d × 3 days, and 35 μg/d thereafter) versus a lower dose (35 μg/d) had a trend towards reduced hospital mortality (33% vs. 52%; p = 0.13).95 When the results from the four trials that compared supplementation of selenium alone to standard were aggregated,92–95 selenium was associated with a trend towards a reduction in mortality (RR 0.52; 95% CIs 0.21, 1.30; p = 0.16).85
Zinc is an essential trace element necessary for normal protein metabolism, membrane integrity, and the function of more than 200 metalloenzymes including enzymes involved in oxidative capacity. In a randomized, prospective, double-blinded controlled trial in severely head injured, ventilated patients, those receiving a higher zinc supplement (12 mg elemental zinc via PN for 15 days, then progressing to 3 months of oral zinc) had a trend towards a reduction in mortality (p = 0.09) when compared to those receiving a placebo (2.5 mg elemental zinc).96
Combined Antioxidant Nutrients
Many randomized controlled trials have chosen to administer a combination of antioxidants via various routes of administration, thereby making it impossible to attribute the outcomes to a specific nutrient. When 11 trials of single and combined antioxidants were aggregated, overall antioxidants were associated with a trend towards a reduction in mortality (RR = 0.73; 95% CIs 0.47, 1.12; p = 0.15) and no effect on infectious complications (RR = 0.94; 95% CIs 0.63, 1.40; p = 0.8).97 Thus, for critically ill patients, selenium supplementation in combination with other antioxidants (vitamin E/alpha tocopherol, vitamin C, N-acetylcysteine, and zinc) may be beneficial, but insufficient data currently exist to support clinical recommendations.
Role of Parenteral Nutrition
Several trials and meta-analyses have evaluated the treatment effect of parenteral nutrition in the last few years, and none has shown a positive result, while some have suggested increased harm associated with PN in the critically ill patient.11,98,99 We have already stated that EN is used preferentially to PN. However, to optimize the delivery of nutrients, some prescribe PN at the same time EN is initiated, to provide nearly all required calories and protein immediately. Then, as EN becomes successfully established, PN is reduced and eliminated. There are five randomized trials that address the clinical benefits of such a strategy in critically ill patients.100 All five studies reported on mortality and the aggregated results demonstrated a trend towards an increased mortality associated with the use of combination EN and PN (RR 1.27; 95% CIs 0.82 to 1.94; p = 0.3). In one study, there was a significant increase in mortality associated with supplemental PN.101 Supplemental PN was not associated with a difference in the incidence of infections (RR 1.14; 95% CIs 0.66 to 1.96; p = 0.6), had no effect on hospital stay (standardized mean difference 0.12; 95% CIs 0.45, 0.2; p = 0.5), and had no effect on ventilator days. Thus there appears to be no clinical evidence to support the practice of supplementing EN with PN when EN is initiated. If anything, there appears to be a signal of excess harm.
What about the patient who has been started on EN, and after several days is only tolerating inadequate amounts of EN? Does PN have a role in this patient population? The preferred approach is to continue with EN and standard IV therapy. However, at some point (probably between 7 and 14 days postinjury) the risk from further deterioration of nutritional status outweighs the risk of providing PN, due to the cumulative effect on immune function, continued losses to the lean body mass, and development of specific key nutrient deficiencies in the critically ill patient receiving inadequate nutritional support by EN. This time frame may be considerably shortened in patients at tremendously increased risk for deterioration of nutritional status due to the presence of large open wounds, enteric fistula, or short bowel syndrome. Unfortunately, there are no randomized trials to guide practitioners as to when PN should be initiated in patients tolerating inadequate amounts of EN. While the results of our previous reviews suggest that PN is associated with no clinical benefit or increased harm, prolonged starvation (more than 14 days) is equally associated with poor outcomes.102
In summary, PN has a very limited role in the critical care setting. PN should not be started in critically ill patients until all strategies to maximize EN delivery (such as the use of small bowel feeding tubes and motility agents) have been attempted. Waiting 2 weeks in someone tolerating inadequate amounts of EN is probably too long, but practitioners will have to weigh the safety and benefits of initiating PN in patients not tolerating EN on an individual case-by-case basis.
Maximizing the Benefits and Minimizing the Risks of Parenteral Nutrition
If PN is associated with harm in critically ill patients, it may be due to a variety of potentially avoidable pathophysiologic mechanisms, including overfeeding, the immunosuppressant effects of lipids, hyperglycemia, absence of key nutrients like glutamine, and the association of gut disuse and systemic inflammation. Understanding these potential mechanisms can guide practitioners when they use PN to use it in such a way that its benefits are maximized and its risks are minimized.
Role of Hypocaloric Parenteral Nutrition
Because of the degree of insulin resistance so commonly observed in stressed critically ill patients, providing large amounts of dextrose intravenously results in hyperglycemia and predisposes critically ill patients to risk of infection. Other attendant complications associated with overfeeding carbohydrates include hepatic steatosis, hypertriglyceridemia, and hypercapnia. This has given rise to the notion of hypocaloric or hypoenergetic PN as a strategy to minimize complications associated with PN. There are only two small studies that have evaluated the effect of hypocaloric feeding in critically ill patients. To achieve a hypocaloric dose of PN, Choban and associates103 reduced both carbohydrates and lipids in morbidly obese critically ill patients, while McCowen and colleagues104 withheld lipids in a heterogeneous group of patients, including critically ill patients. Only one study reported infectious complications, and in that study104 hypocaloric feeding was associated with a trend toward a reduction in infectious complications (p = 0.2). There were no significant differences in mortality or length of stay between groups in either study. Given the lack of positive treatment effect from standard PN, minimizing the dose of PN seems reasonable until further data emerge to prove the contrary.
There are several reports that demonstrate that intravenous lipids may adversely affect immune status and clinical outcomes.105,106 The results of previously described meta-analysis of PN11 suggest that the adverse effects of lipids may negate any beneficial effect of nonlipid parenteral nutritional supplementation. There are two studies reviewed that compared the use of lipids to no lipids in parenteral nutrition.104,107 A significant reduction in pneumonia (48% vs. 73%; p = 0.05), catheter-related sepsis (19% vs. 43%; p = 0.04), and a significantly shorter stay in both ICU (18 vs. 29 days; p = 0.02) and hospital (27 vs. 39 days; p = 0.03) was observed in trauma patients not receiving lipids compared to those receiving lipids.107 In the McCowen study mentioned previously, the group that received no lipids (hypocaloric group) showed a trend towards a reduction in infections (29% vs. 53%; p = 0.2). No difference in length of stay was seen in this study, and it did not report on ventilator days. Combining these two studies, the meta-analysis done showed a significant reduction in infections in the group that received no lipids (RR 0.63; CIs 0.42 to 0.93; p = 0.02) and no difference in mortality (RR 1.29; CIs 0.16 to 10.7; p = 0.8).
It is unknown what the effects of long-term fat-free parenteral nutrition would be, and there is a paucity of data in malnourished patients. Given these caveats, lipid-free PN is probably best indicated for those patients requiring PN for a short time (<10 days), where the risk of fatty acid deficiency would be minimal. This recommendation cannot be extrapolated to those who have an absolute contraindication to EN and need PN for a longer duration.
Hyperglycemia, which occurs more often with PN than EN, is associated with increased infectious complications. Van den Berghe and associates108 compared intensive insulin therapy (target range 4.4 to 6.1 mmol/L) vs. conventional treatment (10.0 to 11.1 mmol/L) in critically ill patients receiving nutrition support. This was a large study (n = 1548) of surgical ICU patients (predominantly elective cardiovascular surgery) with a relatively low APACHE II score (median 9). Study patients were started on a glucose load (200 to 300 g/d) and then were advanced to PN, combined PN/EN, or EN after 24 hours of admission. Intensive insulin therapy was associated with a lower incidence of sepsis (p = 0.003), a trend towards a reduction in ventilator days, and a reduced ICU (p <0.04) and hospital mortality (p = 0.01), compared to conventional insulin therapy. From this study, one can infer that intensive insulin therapy to achieve tight glycemic control may be associated with improved clinical outcomes in critically ill patients. The corollary to this is that high glucose loads in patients who are insulin resistant is associated with excess complications and increased mortality, which can be reduced by insulin. Whether insulin has any therapeutic effects in patients who do not receive such high glucose loads, or whether these results apply to other medical, sicker ICU patients is unknown. Despite these limitations, in the absence of further studies, patients prescribed PN should receive intensive insulin therapy to obtain tight glycemic control. This can best be accomplished by using an insulin protocol or nomogram.109
Supplementation with Glutamine
Perhaps the lack of treatment effect of PN relates to the lack of key nutrients necessary for repair and recovery following critical illness. As noted previously, there are data that suggest that PN supplemented with glutamine is associated with increased survival in seriously ill hospitalized patients.12 It is difficult to provide high-dose free glutamine intravenously to critically ill patients due to problems with limited solubility and stability, especially in critically ill patients with volume-restricted conditions. However, recent advances in parenteral glutamine delivery have overcome some of these challenges, making the provision of bioavailable glutamine practical, even at higher doses.110 The treatment effect is likely greatest when high-dose (>0.2 g/kg per day) glutamine is given parenterally to patients with gastrointestinal failure (and thus prescribed PN). Whether parenteral glutamine has a beneficial effect on patients receiving enteral nutrition is unknown.
Use of Enteral Nutrition in Patients on Parenteral Nutrition
The adverse effect of PN may be related to the absence of nutrients in the bowel. The gastrointestinal mucosa is metabolically very active and the lack of enteral nutrients (as in the case of PN) would result in mucosal atrophy, increased permeability, bacterial overgrowth, translocation of bacteria and/or gut-derived factors that activate the immune system, atrophy of the gut-associated lymphoid tissue, and increased production of proinflammatory cytokines.
An observational study suggested that low-volume EN is associated with less toxicity compared to PN alone.111 Clearly our recommendation is that EN is used preferentially to PN, but in the patient who is not tolerating adequate amounts of EN over a prolonged period of time, if PN is going to be used, we suggest that attempts to provide EN be continued until EN is successful and the PN can be discontinued.