Nutritional Support For Trauma Patient

Many patients who are severely injured from an episode of trauma are unable to eat. There are many reasons for this not least a physical inability. Even those less severely injured patients, who are physically able, may not eat enough to support their metabolic state. Hospital food is notoriously unappealing and the injury associated pain together with the pain relief (usually opioids) may be such to cause marked anorexia and nausea. Metabolic deficits are difficult to make up and their influence on the natural history of the injury depends very much on the underlying nutritional state of the individual. It is important for the medical attendants to address this issue early on in the post-injury period. Key points •

Many patients who are severely injured from an episode of trauma are unable to eat.

The metabolic response to trauma The metabolic response to injury involves a number of well described events triggered and influenced by various mediators. A catabolic state ensues designed to mobilise substrates that can be used by those regions of the body involved in the recovery from injury. This results in fat mobilisation, hyperglycaemia, salt and water retention and net protein breakdown (Meguid et al. 1974, Weissman 1990, Wilmore. 1991, Oppeheim et al. 1980). Fever, tachypnoea, tachycardia and a raised white cell count are the usual clinical correlates. Collectively this is known as the stress response or more recently the term “Systemic Inflammatory Response Syndrome” (SIRS) has been used (AMCP/SCCM 1992). Key points •

A catabolic state ensues designed to mobilize substrates that can be used by those regions of the body involved in the recovery from injury.

Mediators of the metabolic response The mediators of the stress response can be broadly divided into two groups. Firstly, there are those that are released from stimulation of the neuroendocrine axis ie. hormones. Secondly, there are those with a primary immunological function such as cytokines. The response however is complex, involving the release of multiple mediators that are interconnected by amplification loops and negative feedback signals (Winmore 1991, Grimble 1996). Key points •

Mediators are neuroendocrine and immunological

The neuroendocrine axis Hormones are released from afferent neuronal stimulation and also via cytokine release at the site of injury. A list of the major hormones involved in the stress response and their effects is seen in Table 36.1.

Cytokines and other mediators

Cytokines are a group of proteins or peptides released from a variety of cells in response to injury. They influence the immune response as well as producing widespread metabolic changes (Grimble 1996), (see Table 36.3). Cytokines may be considered pro-inflammatory or anti-inflammatory. Some pro- lead to the production ofαinflammatory cytokines such as IL-8, IL-1 and TNF- nitric oxide and other oxidant molecules which may damage the host. Furthermore induce further production of IL-6 and ofαcytokines such as IL-1 and TNF- themselves. Excessive or inappropriate production of cytokines has been associated with increased morbidity and mortality. Other cytokines such IL-4 and production. A list ofαIL-10 are anti-inflammatory and inhibit IL-1 and TNF- cytokines and other immune mediators are shown in Table 36.2.

Alteration of the stress response The value of the stress response is evident by the poor outcome of sympathectomized and adrenalectomized animals when starved. However a prolonged stress response with production of cytokines is believed to contribute to the occurrence of multiple organ failure (Beal and Cerra 1994). The challenge lies in modulation of the stress response to the advantage of the host. This may be achieved by diminishing or preventing the signal at four levels; 1.

The initial stimulus

2.

Gene transcription and translation of the cytokines

3.

Stimulation of effector cells

4.

Transduction of the cell signal

Clinically this has been attempted with variable effect using glucocorticoids (Bone et al. 1987), monoclonal antibodies to cytokines (Fisher et al. 1996), immune enhanced enteral nutrition (Atkinson et al. 1998, Houdijk et al. 1998), early excision of burnt tissue and gut decontamination (Winmore 1991). Nevertheless, it is clear that the form of an individual’s stress response reflects not only the severity and nature of the injury sustained but also the underlying genotype of that individual. Our understanding of the genes controlling the stress response is in its infancy but already various forms of polymorphism for TNF production in septic states has been reported (Stuber 1996, Westendorp 1997). The concept of protective or destructive sets of “trauma genes” has been introduced to explain the individual susceptibility to an injury of a given severity.

Key points •

It is clear that the form of an individual’s stress response reflects not only the severity and nature of the injury sustained but also the underlying genotype of that individual

Assessment of nutritional status Malnutrition may occur in trauma patients both from the catabolic state induced as part of the stress response and from periods of inadequate intake that occur during treatment eg. surgery. Consequently this may lead to critical loss of body mass and function. Malnourished patients have been shown to have a greater chance of developing complications whilst undergoing treatment (Chandra 1983, Windsor and Hill 1988). This has led to the term nutrition

associated complications (NAC) (Detsky et al. 1984ab). These complications include death, sepsis, abscess formation, pneumonia, wound healing difficulties and respiratory failure (Detsky et al. 1984, Detsky et al. 1994, Windsor and Hill 1988). The aim therefore of nutritional assessment is essentially threefold: 1. Assess the risk of morbidity and mortality from malnutrition. 2. Identify the cause(s) of malnutrition in the patient. 3. Assess if the patient would benefit from nutritional support. Assessment of nutritional status can be divided into two components: a) Analysis of changes in body composition b) Analysis of alteration in physiological function. Whilst the increased risk of NAC is thought to be caused more by functional impairment rather than changes in body composition (Windsor and Hill 1988), there is clearly a correlation between the two components (Detsky et al. 1994). Traditionally many of the tests to assess nutritional status have focused on measurement of body composition (Table 36.4). Whilst these tests have been used successfully to predict morbidity and mortality they often only provide a single snapshot of nutritional status at the moment of measurement. Furthermore they can be expensive and time consuming to perform. A more dynamic and practical test is the subjective global assessment (SGA) (Baker et al. 1982, Detsky et al. 1987). This is a clinical test encompassing historical, symptomatic and physical parameters (Jeejeebhoy 1998). It attempts to identify malnourished patients who are at increased risk of medical complications and would benefit from nutritional therapy. It has been shown to be a better predictor of complications than traditional methods of nutritional assessment (Detsky et al. 1984a, Detsky 1987). The components of this technique are shown in Table 36.5. Once applied the patients are characterized into one of three groups: a) well nourished b) moderate or suspected malnutrition c) severe malnutrition. The characterization is a subjective one. Equivocal information is given a lesser weighting than more definitive data to arrive at the final judgement. Key points •

A more dynamic and practical test is the subjective global assessment (SGA)

Nutritional requirements in the trauma patient Energy Requirements The existing body cell mass is the major determinant of the total caloric requirement (Cerra et al. 1997). This may either be estimated or measured directly. The practice of matching energy input with energy expenditure remains controversial for a number of reasons. Firstly there is often poor utilization of administered nutrients in stressed catabolic patients (Hill and Church 1984, Long et al. 1976). Secondly the excess administration of energy, particularly as carbohydrate can stimulate metabolism leading to an increased production of carbon dioxide. The increased

carbon dioxide may subsequently cause respiratory failure or ventilatory dependency in patients with marginal respiratory reserve. Finally, there is evidence that a hypocaloric nutritional regimen in the early post-traumatic period may be associated with an improved outcome (Battistella et al. 1997, Zaloga and Roberts 1994). Administering 2535 kcal (105-145 kJ) per kg of body weight appears adequate for most patients. Key points •

The existing body cell mass is the major determinant of the total caloric requirement

Sources of energy The three sources of energy are carbohydrate, lipid and protein. Traditionally however protein or nitrogen requirements have been considered separately and are often not included in the overall energy intake. Carbohydrate Glucose is the predominant source of carbohydrate and is metabolized by all body tissues including the brain. It is also required for protein anabolism. Glucose usually accounts for 30-70% of total caloric intake ie. 2-5 g/kg/day. However excess glucose administration results in lipogenesis and increased carbon dioxide production (Wolffe et al. 1980). In the early post-traumatic period insulin may be required, particularly in diabetics, to keep blood sugar levels within normal physiological limits. This is due to the relative insulin deficiency for the level of hyperglycaemia seen in the stress response (Weissman 1990). Furthermore, since hyperglycaemia is a potent inhibitor of neutrophil function, maintaining normoglycaemia is an important priority.

Lipid Lipid provides more energy per unit mass than carbohydrate (9.3 kcal/g compared with 4.1 kcal/g). It is important for the function of lipid soluble vitamins, cell wall integrity and prostaglandin synthesis. The amount that lipid should contribute to the total caloric intake is not known. Indeed one study (Battistella 1997) showed an improved outcome, lower infection rate and a reduced period of respiratory failure if lipid was withheld in trauma patients requiring total parenteral nutrition. Furthermore, there is considerable in vitro evidence of immunosuppression associated with the available intravenous lipid preparations (Seidner et al. 1989, Robin et al. 1989, Gogos et al. 1990), Atkinson and Bihari 1994) and some in vivo evidence (Freeman et al. 1990). Throughout the late 1980s and 1990s, as a consequence of pharmaceutical advances improving the stability of lipid in the “three in one” TPN solutions, lipid has accounted in general for 15-30% of total daily caloric intake delivered by the intravenous route. Intravenous nutrition delivered via a peripheral vein (“peripheral TPN”) has used even higher proportions (up to 70-80%) so as to reduce the osmolality and hence venous toxicity of the TPN solution (Kolhardt et al. 1994). What type of lipid to use for nutrition is also not clear. However omega 6 polyunsaturated fatty acid triglycerides should contribute at least 7% of total caloric intake to prevent essential fatty acid deficiency (Cerra et al. 1997). Fish oils such as the omega-3 fatty acids have generated much interest due to their ability to moderate the stress response in experimental models (Teo et al. 1991). A decreased number of infections and reduced length of stay has

been shown in trauma patients (Bower et al. 1995) and in general ICU patients (Atkinson et al. 1998) when given feeds containing omega-3 fatty acids. However no difference in overall mortality was demonstrated and the fact that the feeds contained other immune enhancing components (arginine and purine nucleotides) makes it difficult to interpret the exact role of omega-3 fatty acids by themselves. A theoretical advantage for the use of medium-chain triglycerides is they may be absorbed from the small intestine independent of pancreatic lipase.

Protein Protein or nitrogen intake should be sufficient to allow protein synthesis and promote nitrogen retention rather than be used as a source of energy due to inadequate intake of other caloric sources. This is usually achieved with 1.2 to 1.5 g/kg/day or 15-20% of total caloric intake. In the past, a reduction in dose was considered necessary if the blood urea nitrogen and/or blood ammonia was rising together with a clinical encephalopathy. Nowadays, the widespread availability of haemodiafiltration for the treatment of renal failure and the recognition that withholding nitrogen only exacerbates endogenous breakdown have changed this view. The ideal composition of the nitrogen source is not clear. Obviously it must include some quantity of essential amino acids and preferably an amount of non essential amino acids, but the precise quantities of each are undetermined (Table 36.6). Blood, plasma and albumin are poor sources of nitrogen as firstly they must be catabolized into their constituent amino acids and secondly they do not contain all the essential amino acids. A lot of attention has focused on the use of particular amino acids such as glutamine, arginine or the branched chain amino acids (BCAA). The BCAA ie. isoleucine, leucine, and valine are essential amino acids, which if given in relatively greater amounts than other amino acids (ie. > 0.5 g/kg/day) improve nitrogen balance and protein synthesis in trauma patients (Vente et al. 1991). A difference in outcome however has yet to be demonstrated. Glutamine is a non-essential amino acid, which has stimulated interest for its role in enhancing immunity and preventing muscle loss in critically ill patients (O’Leary and Coakley 1996). Indeed an early report has shown a lower frequency of pneumonia, sepsis and bacteraemia in trauma patients who received glutamine-supplemented enteral nutrition (Houdijk et al. 1998). However, this reduction in infection rate did not translate into a reduction in duration of ventilation or length of ICU stay begging the question of its clinical relevance. Arginine is another non-essential amino acid, the role of which in nitric oxide synthesis, urea synthesis and immune function has also created much interest (Cerra et al. 1997). As previously referenced, it has been used together with omega-3 fatty acids and purine nucleotides (derived from yeast RNA) as part of an enteral nutrition immune enhancing regimen. This regimen has been shown to decrease infections and length of stay in trauma and other critically ill patients (Atkinson et al. 1998, Bower et al. 1995) although once again no change in mortality was seen. A meta-analysis of immune enhancing enteral nutrition using this regimen is now available and supports the view that the combination of immune enhancing nutrients has a positive effect on clinical outcome (Beale et al. 1999). Elemental feeds provide protein in the form of free amino acids or peptides. They are expensive and offer no advantage over standard feeds in trauma patients (Mowatt-Larssen et al. 1992).

Key points •

The BCAA if given in relatively greater amounts than other amino acids improve nitrogen balance and protein synthesis in trauma patients

Water and electrolytes The normal adult daily water requirement is 30-35 mL/kg, although additional losses from diarrhoea, upper gastrointestinal losses, sweating and fever must be considered. Ideally this water is given as part of the normal feeding regime. However the electrolyte content of the nutritional regimes can differ as well as the electrolyte requirements of individual patients. Thus, frequent measurement of electrolytes is usually required until the nutritional regime is stabilised to prevent physiological dysfunction. The normal adult electrolyte requirements are shown in Table 36.7.

Vitamins, and trace elements The exact requirements of micronutrients are not yet determined. They do however have many important roles and deficiency can result in serious illness or physiological dysfunction. Vitamins are important for optimal utilisation of nutritional components. A list of vitamin requirements is shown in Table 36.8. Trace elements also must be considered, zinc for example is lost rapidly in critical illness (Shenkin 1986). It is an important component in many enzymes and deficiency results in hair loss, skin rashes, poor wound healing and infections. A list of trace element requirements as recommended by the Australian Society of Parenteral and Enteral Nutrition (AuSPEN) is shown in Table 36.9.

Key points •

The exact requirements of micronutrients are not yet determined

Growth hormone and other anabolic hormones There has been much interest in the use of anabolic hormones to manipulate the stress response of trauma and thereby reducing morbidity. Growth hormone has been shown to increase net protein synthesis and improve nitrogen balance when infused into burned patients (Gore et al. 1991). However, early reports of randomized control trials in critically ill patients have shown a higher mortality in the treatment group. Similarly insulin, testosterone, and IGF-1 have all been shown to restore anabolism in stressed patients but their effect on clinically is yet to be evaluated (Ferrando 1999).

Enteral versus parenteral nutrition The enteral route is the preferred route for nutrition for a number of reasons. Firstly there is the economical consideration. It has become clear that feeding by the enteral route is more cost effective than TPN (Frost and Bihari 1997). Secondly, there may be a reduced incidence of gastrointestinal bleeding (Pingleton et al. 1983) associated with its use but this has been difficult to document. However, more importantly a number of recent studies have shown a decreased septic morbidity in trauma and other critically ill patients fed enterally as opposed to via the parenteral route (Kudsk et al. 1992, Moore et al. 1992, Moore et al. 1989). One physiological mechanism has been postulated to explain these observations ie. the gut relies on adequate supply of oxygen and nutrients to maintain

normal structure and function. Animal studies (Heyland et al. 1993, Goodlad et al. 1988) suggest that TPN may be associated with loss of mucosal integrity and immunological function. It has been suggested that these changes might lead to bacterial translocation and the passage of endotoxin into the systemic circulation (Blue 1993). This may result in SIRS and ultimately single or multiple organ failure (MOF). A common criticism of enteral feeding is the inability to meet the estimated nutritional requirements of the patient. This is due to a number of reasons, such as impaired gastric emptying or feed absorption, diarrhoea or fasting for procedures. However, the use of specific feeding protocols and aggressive instigation of prokinetic agents results in a greater volume of food delivered (Adam et al. 1997, Frost et al. 1997). TPN however can be increased easily to meet nutritional goals. Interestingly, a study by Moore et al (1986) demonstrated a similar nitrogen balance and caloric intake in patients fed by the enteral route compared to those receiving TPN, whilst another study (Adams et al. 1986) actually demonstrated an improved nitrogen balance in enterally fed patients. Additionally a study in children with burns showed better nitrogen balance, less bacteraemia and improved survival in those fed enterally compared to those fed intravenously (Alexander et al. 1980). Key points •

The enteral route is the preferred route for nutrition for a number of reasons.

Enteral nutrition Access to the gastrointestinal tract A list of the methods of access to the gastrointestinal tract for enteral nutrition is given in Table 36.10. The nasogastric route is relatively contraindicated in trauma patients with suspected or known base of skull fracture due to the risk of intracranial placement. In all other patients it is the preferred route as it is better tolerated than the oral route. The complication rate from insertion of nasogastric feeding tubes is relatively low but risks include nasopharyngeal bleeding and trauma, sinusitis, oesophageal perforation and tracheobronchial misplacement. If nasogastric aspirates remain large despite the use of prokinetic agents (cisapride, erythromycin) then a fine-bore nasojejunal tube should be considered. These can be inserted blindly with the help of prokinetic agents or with the use of gastroscopic or X-ray guidance. Many trauma patients undergo a laparotomy as part of their management. This provides the opportunity for placement of a percutaneous jejunostomy. This is particularly useful in the trauma patient as it bypasses the problem of gastroparesis often seen early on in the ICU course of these patients. Many studies have shown that needle catheter jejunostomies inserted in patients with abdominal trauma allow successful early feeding with minimal complications (Jones et al. 1989, Moore et al. 1981). Percutaneous gastrostomy performed via endoscopy is usually reserved for patients requiring long-term enteral nutrition. Timing of enteral nutrition

The timing of the initiation of enteral nutrition seems to be important. It has been hypothesized that feeding very early after trauma attenuates the stress response and leads to improved patient outcome. Indeed animal studies show that early enteral nutrition, compared to delayed enteral nutrition, is associated with greater wound strength after abdominal surgery (Zaloga et al. 1992) and a reduction in the metabolic response to injury (Mochizuki et al. 1984). Additionally very early (< 2 hours) enteral nutrition in burns patients is associated with a reduced metabolic response (Chiarelli et al. 1990). However a randomized trial in blunt trauma patients comparing early and delayed enteral feeding failed to confirm these findings (Eyer et al. 1993). It showed no difference in the metabolic response, complications or mortality between the two groups. It has also been hypothesized that bowel rest associated with delayed enteral nutrition is associated with gastrointestinal mucosal atrophy and loss of the mucosal barrier function (Minard and Kudsk 1994). This leads to bacterial translocation and possible exposure to endotoxin. As previously noted, a number of studies have showed the feasibility of early enteral nutrition in trauma patients. Several of these studies have shown a reduction in septic complications when early enteral nutrition was compared to delayed (> 5 days) enteral feeding (Alexander 1999). Similarly, a study of head injured patients randomized to early (< 36 hours) or delayed (3-5 days) feeding showed a decreased number of infections and reduced length of stay in the early fed group (Grahm et al. 1989). Thus, it is generally accepted that enteral nutrition should be started as soon as the trauma patient is fully resuscitated with a stable cardiovascular system. Total parenteral nutrition (TPN) TPN is associated with an increased risk of infectious complications related to the direct immunosuppressive effects of the TPN solutions plus additional catheter related complications. Furthermore it does not provide the benefit of maintaining gastrointestinal structure and function that enteral nutrition does. Therefore it should only be used whenever enteral nutrition is not feasible (eg. total small bowel resection) or requires supplementation. The numbers of patients receiving TPN are dwindling as intensive care physicians and surgeons have become more aware of the complications associated with this form of nutritional support. Nowadays, it is used in fewer than 15% of patients who require long term support in the ICU. The benefits of TPN have been difficult to demonstrate. One large study (Buzby et al. 1991) suggested it should be limited to patients who are severely malnourished, whilst a recent meta-analysis (Heyland et al. 1998) failed to demonstrate any mortality advantage in surgical or critically patients. The meta-analysis did suggest however, that there may be a reduction in nutrition associated complications in patients who were initially malnourished but the quality of those studies demonstrating these improved outcomes were somewhat doubtful. The obvious advantage for the parenteral route for nutrition is the ease of achieving nutritional goals. It has to be remembered of course that TPN has been used successfully for long periods in that small minority of patients unable to receive enteral nutrition but careful monitoring is required in order to prevent complications (Table 36.11). Key points •

TPN should only be used whenever enteral nutrition is not or requires supplementation.

Conclusion Whilst starvation is not an option in the care of the critically ill, trauma patient, controversy still surrounds the route and timing of nutritional support together with the exact nature of the formulation of nutrition delivered. Nowadays, parenteral nutrition is out of fashion primarily because of its cost and the lack of efficacy studies in the critically ill, trauma population. On the other hand, there does appear to be evidence emerging that early enteral nutrition, that is feeding established within 48 hours of admission to the ICU, either using a surgical jejunostomy or the more simple nasogastric tube (in combination with prokinetics) is associated with improvements in outcome. Similarly, there are now a number of studies emphasising the benefits of using specific immune enhancing nutrients – glutamine, arginine, omega 3 fatty acids and purine nucleotides. It appears likely that nutritional support with one of these enhanced enteral formulations will become the standard of care for the critically injured trauma patient in the ICU