Clinical review: optimizing enteral nutrition for critically ill patients - a simple data-driven formula

In modern critical care, the paradigm of 'therapeutic nutrition' is replacing traditional 'supportive nutrition'. Standard enteral formulas meet basic macro- and micronutrient needs; therapeutic enteral formulas meet these basic needs and also contain specific pharmaconutrients that may attenuate hyperinflammatory responses, enhance the immune responses to infection, or improve gastrointestinal tolerance. Choosing the right enteral feeding formula may positively affect a patient's outcome; targeted use of therapeutic formulas can reduce the incidence of infectious complications, shorten lengths of stay in the ICU and in the hospital, and lower risk for mortality. In this paper, we review principles of how to feed (enteral, parenteral, or both) and when to feed (early versus delayed start) patients who are critically ill. We discuss what to feed these patients in the context of specific pharmaconutrients in specialized feeding formulations, that is, arginine, glutamine, antioxidants, certain ω-3 and ω-6 fatty acids, hydrolyzed proteins, and medium-chain triglycerides. We summarize current expert guidelines for nutrition in patients with critical illness, and we present specific clinical evidence on the use of enteral formulas supplemented with anti-inflammatory or immune-modulating nutrients, and gastrointestinal tolerance-promoting nutritional formulas. Finally, we introduce an algorithm to help bedside clinicians make data-driven feeding decisions for patients with critical illness.

ICU patients typically experience catabolic stress and systemic infl ammatory response; in turn, these responses alter both the morphology and function of the gastrointestinal (GI) tract [2]. Up to 60% of ICU patients suff er GI dysfunction due to impaired GI motility, digestion, or absorption [3,4]. Such GI dysfunction, often coupled with inadequate caloric intake, leads many critically ill patients to develop an energy defi cit and lose lean body mass [5][6][7][8][9][10].
ICU patients with poor nutritional status commonly experience immune dysfunction, weakened respiratory muscles and lowered ventilation capacity, and reduced GI tolerance [11,12]. As a result, patients are at risk for a wide range of complications: ventilator dependence; GI dysfunction with gastroesophageal refl ux, esophagitis, or pulmonary aspiration; and infections that can lead to sepsis, multi-organ failure, and even death [1,12,13]. For such patients, supportive nutrition has long been used as adjunctive care; however, such nutrition is often inadequate, including only fundamental macronutrients to sustain patients through periods of metabolic stress.
Feeding an ICU patient now extends beyond choosing the right feeding route, the rate, and the caloric density. In modern critical care, the paradigm of 'therapeutic nutrition' is replacing traditional 'supportive nutrition' [14]. Enteral feeds are now formulated with active nutrients that may help reduce oxidative damage to cells and tissues, modulate infl ammation, enhance benefi cial stress responses, and improve feeding tolerance.
In this review, we summarize the most recent data on feeding critically ill patients. We describe practice guidelines established by US, European and Canadian enteral and parenteral nutrition societies, beginning with recommen dations of how to feed (enteral, parenteral, or both) and when to feed (early versus delayed start). We systema tically discuss what to feed critically ill patients by reviewing mechanisms of action for specifi c pharmaconutrients and by concisely summarizing current guidelines and expert recommendations for feeding various populations in the ICU. We compile clinical evidence on feeding anti-infl ammatory, immune-modulating, and GI tolerance-promoting nutritional formulas in specifi c patient subgroups. Finally, we introduce a straightforward

Abstract
In modern critical care, the paradigm of 'therapeutic nutrition' is replacing traditional 'supportive nutrition' . Standard enteral formulas meet basic macro-and micronutrient needs; therapeutic enteral formulas meet these basic needs and also contain specifi c pharmaconutrients that may attenuate hyperinfl ammatory responses, enhance the immune responses to infection, or improve gastrointestinal tolerance. Choosing the right enteral feeding formula may positively aff ect a patient's outcome; targeted use of therapeutic formulas can reduce the incidence of infectious complications, shorten lengths of stay in the ICU and in the hospital, and lower risk for mortality. In this paper, we review principles of how to feed (enteral, parenteral, or both) and when to feed (early versus delayed start) patients who are critically ill. We discuss what to feed these patients in the context of specifi c pharmaconutrients in specialized feeding formulations, that is, arginine, glutamine, antioxidants, certain ω-3 and ω-6 fatty acids, hydrolyzed proteins, and mediumchain triglycerides. We summarize current expert guidelines for nutrition in patients with critical illness, and we present specifi c clinical evidence on the use of enteral formulas supplemented with anti-infl ammatory or immune-modulating nutrients, and gastrointestinal tolerance-promoting nutritional formulas. Finally, we introduce an algorithm to help bedside clinicians make data-driven feeding decisions for patients with critical illness.
algorithm to help bedside clinicians make feeding decisions for patients with critical illness.

The basics: how and when to initiate feeding in ICU patients
Many ICU patients are incapable or intolerant of a food diet; therefore, a clinician's fi rst nutritional decision is whether to tube-feed enterally, parenterally, or with a combination of both.
Practice guidelines in Europe, Canada, and the US endorse enteral feeding for patients who are critically ill and hemodynamically stable [1,11,15]. Enteral nutrition is preferred over parenteral nutrition (PN) for most ICU patients -an evidence-based practice supported by numer ous clinical trials involving a variety of critically ill patient populations, including those with trauma, burns, head injury, major surgery, and acute pancreatitis [1,16]. For ICU patients who are hemodynamically stable and have a functioning GI tract, early enteral feeding (within 24 to 48 hours of arrival in the ICU) has become a recommended standard of care [1,11,15]. Experts identify these early hours as a window of opportunity to provide nutrition that maintains gut barrier function and supports immune responses [1,11].
Patients with extreme hemodynamic instability -with rising plasma/blood/serum lactate concentrations or escalating requirements for vasopressors -are generally not considered candidates for enteral feeding. However, early fi ndings suggest use of early enteral feeding in other vasopressor-dependent patients. In one study, vasopressordependent patients who were given enteral feeding within the fi rst 48 hours had a signifi cant survival advantage compared to those whose feeding was delayed; in fact, the sickest patients (on multiple vasopressors) experienced the greatest benefi t [17]. It should be noted that this fi nding is based on an observational study. A confi rmatory prospective, controlled study is warranted.
Even though early enteral nutrition is favored for most ICU patients, caloric and protein needs are often not met by enteral feeding [10]. Nutritional intake may be hampered by setting target levels too low, interruption of feeding for procedures, issues of airway management, and poor tolerance of feedings [8,18,19]. To enhance use of enteral feeding in ICU patients, several feeding strategies have been proposed recently: shift from an hourly-rate feeding goal to a 24-hour volume goal, allowing nurses to 'make-up' for interruptions and meet feeding targets [20]; accept gastric residual volumes up to 500 mL to increase the volume of formula delivered in practice [1,[20][21][22]; and use a promotility agent to help reduce gastric residual volume [20,21,23].
PN is necessary in critically ill patients who do not have an intact GI tract, but current guidelines do not agree on when to initiate PN [5]. For patients who are intolerant or have other contraindications to enteral feeding, European guidelines recommend starting PN within 24 to 48 hours if the patient is not expected to be on oral nutrition within 3 days [24]. US guidelines hesitate to recommend PN on admission to the ICU; standard care (intravenous fl uids) is recommended fi rst, with PN reserved and initiated only after 7 days in well-nourished patients [1]. Both the European Society for Parenteral and Enteral Nutrition (ESPEN) and American Society for Parenteral and Enteral Nutrition (ASPEN) guidelines recommend early PN use (within 24 hours of ICU admission) in patients who are malnourished [1,24]. Canadian guidelines state that PN should not be used in any patient with an intact GI tract [25].
When enteral feeding alone is inadequate, some experts are calling for use of PN and enteral nutrition together to meet energy and protein targets [5,7,26,27]. Combination regimens are justifi ed by observations that actual enteral intake typically meets only half of prescribed calories in ICU patients [6,[8][9][10]. For patients who are expected to be mechanically ventilated for more than 72 hours and have body mass index (BMI) scores <25 or ≥35, each additional 1,000 kcal/day or 30 g protein/day was reported to be associated with reduced mortality [27].
However, clinical evidence for combination feeding remains unclear. Two recent randomized trials have helped clarify this subject. Casaer and colleagues [28] conducted a large, single center prospective, random ized trial (EPaNIC trial) comparing outcomes in critically ill patients on enteral nutrition who had early versus late initiation of PN (early, <48 hours after ICU admission, n = 2,312; late, day 8 or later after ICU admis sion, n = 2,328). Results revealed patients on late-initia tion PN had a relative increase of 6% in the likelihood of being discharged alive earlier from the ICU and from the hospital (P = 0.04). Th ose in the late-initiation group also had signifi cantly fewer ICU infections, shorter duration of mechanical ventilation and a shorter course of renal replacement therapy. Several aspects of the study limit generalizability of the fi ndings to all ICU populations. First, patients with chronic malnutrition were not included in the study. Second, the trial PN formulation did not contain either glutamine or immune-modulating ingredients, nor was there any adjustment for presence/ absence of immune-modulating ingredients in enteral feedings. Th ird, patients in the trial received a low protein delivery (median of 0.8 g/kg/day protein (after day 3)) for the study period in the early PN group; this protein target is below what is recommended by most guidelines for critically ill patients (the typical recommendation is 1.3 to 1.5 g/kg/day (ESPEN guidelines)). Finally, the trial examined a low mortality-risk patient group with an average ICU mortality of 6.2% (90-day mortality of 11.2%) and a relatively low severity patient group with an ICU length of stay of 3.5 days, and a mechanical ventilation period of 2 days. Accounting for the afore mentioned limitations, Th e EPaNIC trial is unquestionably a key contribution to the literature on supplemental PN use in critical care. We believe the key conclusion is that early aggressive calorie delivery via PN does not appear to be benefi cial in low mortality risk, non-chronically malnourished patients. In contrast, the recently published, single center TICACOS trial showed hospital and 60-day mortality was reduced (P < 0.02 for both time points) in a higher mortality risk group of ICU patients receiving additional calories via enteral nutrition supplemented with PN [29]. In comparison to the EPaNIC trial, the TICACOS trial was conducted in a higher risk ICU patient group with an ICU mortality of 25.4% (60-day mortality of 47%), ICU length of stay of 12 days, and a mechanical ventilation period of 10.75 days. Th us, it is possible that supplemental PN may improve outcome in high mortality risk ICU patients. Additional trials on the use of enteral nutrition with supplemental PN have recently been completed or are underway. Th ese forthcoming results should continue to clarify the utility of supplemental PN use in the ICU.

Choosing which enteral formulation to feed
For most ICU patients, the next decision is what enteral formula to feed. Critically ill patients are a heterogeneous population, so no one-size-fi ts-all nutritional formula should be expected [30,31]. Feeding formulas to consider are anti-infl ammatory, immune-modulating, GI tolerancepromoting, and standard enteral nutrition.

What are the key functional pharmaconutrients in enteral formulas?
Guidelines from professional nutrition societies around the world identify certain populations of patients who can benefi t from formulations with specifi c pharmaconutrients [1,11,15,25]. For many others, each physician must use his or her own clinical judgment about what formula will best meet the patient's needs. To help guide such choices, the following section reviews functional pharmaconutrients and their roles in critically ill patients.

Anti-infl ammatory enteral nutrients
Critical illness and injury are characterized by oxidative stress and excessive infl ammation, harmful processes that damage cells and impair function of vital organs. Extreme infl ammation -as in patients with systemic infl ammatory response syndrome (SIRS), sepsis, acute lung injury (ALI), or acute respiratory distress syndrome (ARDS) -often progresses to multiple organ dysfunction syndrome and even death. Feeding formulas with specifi c pharmaconutrients can help off set tissue damage and moderate infl ammation. Dietary antioxidants (vitamins A, C, and E and selenium) play important roles in reducing potential for tissue damage by stabilizing free radicals in cells, while dietary fi sh oil and borage oil blunt out-ofcontrol infl ammatory responses by modulating synthesis of pro-and anti-infl ammatory mediators [31][32][33].

Fish oil and borage oil
Dietary intake of certain oils alters the fatty acid composi tion in membranes of cells involved in immune infl ammatory responses, that is, neutrophils and macrophages. Certain cell membrane fatty acids (for example, arachidonic acid (AA)) serve as precursors to infl amma tory eicosinoids and leukotriene mediators, while other fatty acids (eicosapentaenoic acid (EPA), docosohexaenoic acid (DHA), gamma linolenic acid (GLA)) are metabolized to form less pro-infl ammatory mediators [31][32][33][34]. In addition, DHA and EPA are precursors of resolvins and protectins, which help resolve infl ammation and reduce tissue injury [35].
Fish oils are rich in the ω-3 fatty acids EPA and DHA, the active metabolites of alpha-linolenic acid [31][32][33]. While healthy humans have desaturase enzymes that convert alpha-linolenic acid to EPA and DHA, such conversion is negligible in individuals who are critically ill. As a result, enteral nutrition containing EPA and DHA can help quell excessive infl ammatory responses. Th e anti-infl ammatory actions of the ω-3 fatty acids EPA and DHA are thought to occur by (1) blunting production of pro-infl ammatory mediators as a result of replacing AA in macrophage and neutrophil membranes, (2) blocking synthesis of pro-infl ammatory mediators from AA by competing for the metabolic enzymes cyclooxygenase and lipoxygenase, and (3) supporting production of less infl ammatory prostaglandin and leukotriene mediators [34] (Figure 1). Oil from the borage plant contains GLA, an ω-6 fatty acid that similarly competes with pro infl ammatory AA for conversion enzymes; GLA is a precursor to mediators that are less infl ammatory than the prostaglandins and leukotrienes produced from AA [31].
Results vary in studies using fi sh oil-supplemented enteral formulas in critically ill patients, partly because of the diff erent formulations and doses of fi sh oil. However, an enteral formula containing EPA and GLA consistently showed signifi cant benefi ts in three studies of mechanically ventilated patients with ALI/ARDS or severe sepsis/septic shock [36][37][38] (Figure 2). Use of this antiinfl ammatory feeding formulation signifi cantly reduced time on ventilator, ICU and hospital lengths of stay, and incidence of new organ failure. Further, a meta-analysis performed by the authors of the original trials showed that the use of a fi sh-oil-containing formula signifi cantly reduced the risk of 28-day mortality by 49% compared to a high fat control enteral formulation [39]. Results of another large meta-analysis showed that immune-modulating formulas containing fi sh oil signifi cantly reduced by half the risk for mortality and secondary infections and signifi cantly shortened length of hospital stay by more than 6 days in ICU patients with ALI/ARDS or severe sepsis/septic shock; such signifi cant benefi ts were not seen in trauma and burn patients [40]. Moreover, a recent study of critically ill mechanically ventilated, septic patients (not all of whom had ARDS) found that feeding a formula enriched with EPA, GLA, and antioxidants shortened ICU stay but did not show benefi ts in terms of infectious complications, gas exchange, or organ failures [41].

Antioxidant nutrients
Antioxidant enzymes (superoxide dismutase, catalase, and glutathione peroxidase) -in association with selenium, zinc, manganese, and iron -provide intrinsic defense mechanisms to remove reactive oxygen species generated by everyday living [33]. However, critical illness or injury leads to increased oxidative stress, and these enzymes and their cofactors can be depleted.
In critical illness, patients with a history of defi cient dietary intake of antioxidant vitamins were more likely to suff er redox imbalance than were those whose antioxidant vitamin intake was near the recommended daily allowance [42]. Repletion studies showed that survival was improved when such patients were given supple mental antioxidant vitamins and minerals and trace elements (for example, vitamins A, C and E, selenium, copper, and zinc) [25,43]. A multi-center study showed reduced mor tality rates when patients were given high-dose selenium by intravenous supplementation [44]. Th ere were two limitations to this multi-center study: specifi cally, the reduction in mortality with selenium was only apparent when data were analyzed as per-protocol, and the number of exclusions was substantial.

Immune-modulating enteral nutrients
Arginine, glutamine and nucleotides are conditionally essential nutrients that can be depleted by stresses of critical illness and injury. Th ese nutrients usually support immune homeostasis by maintaining the functional integrity of immune cells and helping in wound healing and tissue repair processes. With nutritional defi cits, the function of the immune system is impaired. For instance, arginine defi ciency develops in patients who have ex perienced surgery or trauma [45]. Glutamine defi ciency occurs in many ICU patients, including those with sepsis, trauma, surgery, or burns [31,33,46,47]. Similarly, during times of stress and rapid cell proliferation, nucleotide availability may be rate-limiting for synthesis of nucleotide-derived compounds. In the next sections, we review pivotal roles of these three nutrients in modulating immune responses. Arginine Arginine is a conditionally essential amino acid that is supplied by the diet, synthesized endogenously, and salvaged from body proteins [45]. It supports polyamine synthesis (for cell growth and proliferation), proline synthesis (for wound healing), and is a biosynthetic substrate for nitric oxide (a signaling molecule for immune and other cells) [31,33,45]. Results of recent investigations revealed how immune function is tied closely to arginine metabolism [49]. Soon after injury or surgery, immature cells of myeloid origin appear in the circulation and in lymph tissues. Th ese cells produce arginase-1, an enzyme that breaks down arginine. Increased arginine breakdown, coupled with low intake and limited endogenous production, results in overall arginine defi ciency. Th is defi ciency is associated with suppression of T-lymphocyte function. T-lymphocyte dysfunction after surgery or trauma is characterized by a decrease in the number of circulating CD4 cells, blunted T-cell proliferation, decreased production of IL-2 and inter feron gamma, and loss of the zeta (ζ) chain, a peptide essential in the T-cell receptor complex [48].
Results of several clinical studies showed that repletion of arginine, along with ω-3 fatty acids, helps restore Tlymphocyte numbers and function -that is, CD4 cell counts and IL-2 production [48,49]. Dietary ω-3 fatty acids also blunt expression of arginase 1 [50]. Th us, substantial evidence supports the concept that immunemodulating diets exert their benefi cial eff ects by restoring T-cell function that was impaired by arginine depletion.
Clinical outcome data (30 trials; 3,000 patients) showed a signifi cant treatment eff ect of arginine supplementation following major surgery; arginine treatment reduced the risk of infection (relative risk = 0.58; 95% confi dence interval (CI) of 0.48 to 0.69; P < 0.00001) and overall length of stay (weighted mean diff erence = -2.09; 95% CI of -3.20 to -0.97; P = 0.0002) versus standard enteral nutrition [51]. In addition, arginine (at doses given in immune-modulating nutrition, 12 to 15 g/day) is potentially helpful for patients for 5 to 7 days prior to major surgery (as studied in abdominal, ear/nose/throat, and cardiac surgery) [45,51]. However, very little benefi t, and perhaps harm, is observed in patients with severe sepsis [52,53]. Th is potential harm may be caused by promotion

Conditions Features Treatments Outcome if untreated
of excessive nitric oxide production in patients with sepsis, in turn worsening SIRS and increasing risk for mortality [54]. Nevertheless, these mechanisms are considered expert opinions rather than proven eff ects. For instance, lesser doses of arginine (6 g intravenously or orally) appear to have undetectable eff ects on blood pressure in healthy subjects [55]. Taken together, clinical evidence supports the concept that arginine-containing formulas should be considered as a standard of care in the peri-operative period, that is, both before and after major surgery (Figure 2). Trauma and post-myocardial infarction patients, and those with pulmonary hypertension may also benefi t from supplemental arginine following injury [45]. For patients with severe sepsis, arginine-supplemented enteral formulas should be used with extreme caution [1] or avoided due to potential harm [15,25,40,54].

Glutamine
Glutamine is the most abundant free amino acid in the body, but its stores are rapidly depleted during critical illness or injury, including burns [31,46,56]. Glutamine serves as a metabolic substrate for enterocytes and immune cells, thus supporting intestinal barrier function and immune responses [56]. Glutamine was recently proposed to serve as a signaling molecule in states of illness and injury, that is, a messenger to turn on genes involved in cell protection and immune regulation [57].
Glutamine is expected to have the greatest benefi t in severely ill patients with sepsis and organ failure, that is, those with the most severe glutamine defi ciency. Further, glutamine defi ciency at ICU admission has been correlated with increased mortality [58]. Clinical evi dence for glutamine-supplemented parental nutrition comes from randomized controlled trials (900 patients; 4 level 1 and 13 level 2 randomized controlled trials); together these studies show signifi cant reductions in mortality, infection, and hospital length of stay [25]. Glutamine use is supported by clinical nutrition guide lines worldwide [1,11,15,25].
Enteral formulations with glutamine appear to benefi t patients with burns or trauma [25]. Benefi ts are dosedependent, with >0.3 g/kg/day required for benefi t; 0.5 g/kg/day (in divided portions) may be more eff ective [25]. A meta-analysis of all glutamine randomized controlled trials (both enteral and parenteral) showed a statistically signifi cant reduction in mortality in ICU patients of all types (21 studies, more than 1,500 patients) [25]. More data are needed to support the routine use of enteral glutamine in other critically ill patients.

Nucleotides
Nucleotides play an active role in cell proliferation and immune modulation. Th ey are low molecular weight, intra cellular compounds that serve as building blocks for DNA, RNA, and ATP, and are components of several key metabolic coenzymes [33]. Nucleotides can be synthesized de novo, or they can be retrieved through salvage pathways. During times of stress and rapid cell proliferation, nucleotide availability may be rate-limiting for synthesis of nucleotide-derived compounds. Dietary supplies help compensate for these shortfalls, in turn supporting cell proliferation, diff erentiation, and function, especially for cells of the immune system [33]. Evidence for benefi ts of dietary nucleotides comes from preclinical studies, so clinical outcome trials are still needed. While nucleotides are unlikely to cause harm, their inclusion in therapeutic nutrition cannot yet be recommended.

Tolerance-promoting prebiotic formulas
Enteral-feeding intolerance is common in critically ill patients [59]. Such intolerance is attributed to illnessassociated intestinal dysfunction, such as impaired motility, inadequate digestion, and reduced absorption, as well as to side eff ects of treatment drugs. Half of all ICU patients on enteral nutrition have some symptoms of intolerance, that is, increased residual volume, bowel dilation, vomiting, or diarrhea [3,4]. About 80% of patients with head injury and 50% of mechanically ventilated and post-surgical patients have delayed gastric emptying [60][61][62].
Certain diseases and conditions predispose patients to GI problems. Not surprisingly, abdominal surgery contributes to GI tract dysfunction, including laparoscopic procedures where the gut is not handled at all [63]. Patients with burns and traumatic injuries may experience GI motility problems [63,64]. In sepsis, exocrine pancreatic function is reduced, which can impair protein digestion [65]. Treatments for cancer and other conditions -for example, opiates, radiotherapy, or chemotherapy -have side eff ects of anorexia and impaired GI function [66].
In such situations, tolerance-promoting enteral formu las are a rational feeding choice. Enteral formulas with extensively hydrolyzed proteins, that is, peptide-based feeding formulas, can be fed to patients intolerant of whole proteins. Enterocytes normally absorb small peptides (primarily di-and tripeptides), so proteins do not need to be fully digested to single amino acids for uptake [67]. It will be important to confi rm that this mechanism for peptide uptake remains active in critically ill patients.
Further, tolerance-promoting formulas often supply fats as medium-chain triglycerides, that is, lipids that are small enough to be water-soluble with little or no bile salt. Compared to long-chain fats, the medium-chain triglycerides appear to be tolerated because they are quickly absorbed by the intestine and metabolized by the liver. Th e length of a fatty acid chain may infl uence gastric emptying, at least in part, via a reduction in GI hormones such as cholecystokinin and glucagon-like peptide-1 [68].
Prebiotic ingredients also help improve GI tolerance of enteral formulations; prebiotics are non-digestible food ingredients, mainly carbohydrates, that stimulate growth of benefi cial bacteria in the digestive system, in turn benefi ting the health of the host [69,70]. In one study of patients with acute pancreatitis, feeding an enteral formula with prebiotics (a mix of fructooligosaccharides and other soluble and insoluble fi bers) signifi cantly improved clinical outcomes and lessened disease severity compared to feeding a fi ber-free formula [71]. More studies are warranted and needed to determine the clinical eff ects of prebiotic-supplemented enteral formulations in critically ill patients [72].

Expert guidelines for therapeutic enteral nutrition in the ICU
Evidence-based nutrition guidelines for critically ill patients have been developed in North America and Europe [1,15,25]. While the guidelines agree in principle regarding pharmaconutrients, there are some diff erences in specifi c recommendations (Table 1).

Enteral formulas with anti-infl ammatory nutrients
Enteral formulas supplemented with infl ammationmoderat ing fats (ω-3 fi sh oils DHA and EPA, and ω-6 borage oil GLA) and antioxidants are recognized in evidence-based guidelines to blunt infl ammatory responses and protect against free radical damage [1,15,25]. All guidelines specifi cally recommend the use of ω-3 fatty acids in ALI/ARDS patients [1,15,25], and all guidelines support enrichment of antioxidant vitamins and trace elements in enteral formulas [1,15,25].

Enteral formulas with immune-modulating nutrients
Formulas with immune-modulating nutrients, such as arginine, glutamine, and possibly nucleotides, support immune responses and lower risks of infectious complications in many patients. While some recom men dations group these nutrients together, recent fi ndings identify specifi c indications for each nutrient. Enteral formulations with arginine are recommended for surgical patients [1,15]. By contrast, guidelines advise using arginine with caution in patients who are experiencing infl ammatory responses following infection, such as those with severe sepsis or septic shock, due to arginine's potential for increasing endogenous production of nitric oxide [1,15,25].
Glutamine-supplemented enteral nutrition is specifi cally recommended for burn and trauma patients; glutamine is recommended for all patents on PN [1,15,25].
Nucleotides may be conditionally essential for cell proliferation during times of metabolic stress, but only European and US guidelines mention theoretical benefi ts from nucleotide supplementation in immune-modulating formulas [1,15]. Canadian guidelines do not support this practice [25].

Enteral formulas designed to promote gastrointestinal tolerance
Peptide-and medium-chain triglyceride-based formulas with prebiotics lessen GI tolerance problems related to enteral nutrition. Guidelines specifi cally recommend hydrolyzed protein formulas for patients with GI dysfunction, such as short bowel syndrome, pancreatitis, or persistent diarrhea [1,15,25]. Guidelines have not yet made recommendations regarding fi ber content for enteral feeding in the ICU.

Choosing the right feeding formula for an ICU patient
When considering nutritional therapy for ICU patients, three main decisions must be made on the basis of the patient's medical condition, as indicated in the algorithm provided ( Figure 3): (1) route of feeding -enteral versus parenteral; (2) when to feed -begin within 24 to 48 hours of ICU admission preferred (early enteral feeding); and (3) what to feed -a standared enteral formula or one with targeted functional ingredients (anti-infl ammatory, immune-modulating, or tolerance-promoting nutrients).
Th is nutrition algorithm is intended as a guide for choosing the right therapeutic nutrition formula; as It highlights targeted use of anti-infl ammatory and immune-modulating formulas, as well as tolerance-promoting formulas. ALI, acute lung injury; AOX, anti-oxidants; ARDS, acute respiratory distress syndrome; ARG, arginine; BMI, body mass index; GI, gastrointestinal; GLN, glutamine; IV, intravenous; ω-3, omega-3 fatty acids.  always, clinical judgment also plays a very important role in decision-making. In choosing, clinicians may need to consider other factors, such as whether formula cost is a limiting factor and what to choose when conditions coincide -for example, sepsis and GI intolerance. In addition, the health and nutritional status of a patient can change rapidly -for example, a trauma patient develops sepsis. As a result, the nutritional needs of a critically ill patient should be reassessed routinely and often.

Conclusions
Nutrition -given early and by the enteral route -is now well established in the ICU. We believe future ICU nutrition algorithms need to focus on evidence-based therapeutic nutrition as a new standard of care. In our opinion, the modern ICU clinician now has the opportunity to make full use of specialized nutrition in the ICU, that is, formulas supplemented with anti-infl ammatory, immune-modulating, and tolerance-promoting nutrients. Each formula type contains a combination of pharmaconutrients that have the potential to enhance natural recovery processes, prevent or blunt certain disease-related complications, and provide vital energy as well-tolerated macronutrients. Antioxidant vitamins and minerals, arginine, glutamine, certain ω-3 long-chain fatty acids, hydrolyzed proteins, and medium-chain triglycerides are functional nutrients. By providing full macro-and micronutrient support along with these pharmaco nutrients, therapeutic nutrition off ers fresh hope for improved outcomes in ICU patients.