Nutrition Support For the Patient with Surgery, Trauma, Or Sepsis1

Nutrition Support For the Patient with Surgery, Trauma, Or Sepsis1

Kenneth A. Kudsk

Specialized nutrition support plays an integral role in the preoperative and postoperative management of patients undergoing major surgical procedures or after severe injuries when the patients are unable to take adequate oral intake. Both parenteral and enteral support reduce major wound dehiscence and anastomotic leaks in select populations undergoing major general surgical procedures (see elsewhere in this volume). When provided enterally, nutrition also reduces septic complications, especially in severely injured trauma patients. Nutrition does not play an essential role in the treatment of sepsis itself but rather in the prevention of sepsis; and the role of nutrition in sepsis is not well defined. The key issue is appropriate use of nutrition and choosing the appropriate patients to use this sophisticated therapy because specialized nutrition support can potentially cause injury as well as provide benefit. When the therapy is used in patients not at risk of nutrition-related complications, only the complications of therapy are seen. However, when delivered to patients at risk of wound or septic complications, nutrition can reduce these complications.


Techniques for intragastric feeding have existed for hundreds of years (1), but parenteral nutrition (PN) is a relatively new, highly technical field, which rapidly advanced during the 1970s (2, 3, 4, 5, 6). The goals of nutrition support are to prevent further deterioration of nutritional status, replenish host defenses and lean tissue, improve clinical outcome, and support adjunctive therapies, which otherwise would be impossible in a catabolic, malnourished patient. Patients with total or near total intestinal loss, malnourished patients with chronic inflammatory mucosal disease interfering with normal absorption, or those with fistulas precluding ingestion of adequate oral nutrition could not survive their illness without nutritional support. Indications are less clear when no preexisting malnutrition exists or patients quickly resume oral intake. However, preemptive nutritional therapy reduces the risk of subsequent complications in specific patient populations. Despite limited evidence for use in some patients, nutrition support is commonly prescribed because of the recognized relationships among severe malnutrition, morbidity, and mortality; the high incidence of protein malnutrition in hospitalized patients; the recognition that prolonged starvation impairs healing; and a generalization of data from clinical trials demonstrating benefit in at-risk patients. Fortunately, the risks from nutrition therapy are minimized and benefits increased when experienced
professionals deliver this complex technical therapy to appropriate patient populations.


Identifying the at-risk patient is limited by the tools available. Several scoring systems can quantify the risk of complications—particularly septic complications— following blunt or penetrating trauma. In multiple trials, enteral nutrition improves outcome by reducing sepsis compared with starvation or parenteral feeding (7, 8, 9, 10, 11, 12). Trauma patients are not traditionally considered nutritionally at risk because most are young and well nourished, although alcohol and drug abuse are common. General surgical patients with preexisting nutritional deficits have been harder to stratify because specialized scoring systems do not exist, although several principles apply. Preoperative albumin (ALB) is the single best indicator of postoperative complications and mortality following major surgery (13), but low ALB may reflect liver disease, fluid resuscitation, or inflammation, rather than nutritional status.

The Trauma Patient

In most studies, the Injury Severity Score (ISS) (14), Abdominal Trauma Index (ATI) (15), or both have stratified patients to complication risk. The ISS scores the three most severely injured body regions of the six, which include head and neck, musculoskeletal, soft tissue, abdominal, thoracic, or head. The ISS correlates with mortality and morbidity. In randomized prospective studies, early enteral feeding improves the outcome of patients with an ISS greater than 18 to 20 compared with intravenous (IV) feeding or fasting (15).

The ISS underestimates risk when severe injuries are isolated to a single body area. The ATI identifies risk for infection in patients with intra-abdominal injuries (Table 93.1) (15). Each intra-abdominal organ has a risk factor that when multiplied by the magnitude of that organ’s injury, correlates with risk of sepsis from that injury. Injuries to the pancreas, colon, major vascular structures, duodenum, and liver pose the highest risk. The ATI can be calculated during surgery by summing the scores of each injured organ. An ATI greater than or equal to 20 to 25 poses the greatest risk for sepsis. Sepsis rates are also high with ATI values less than 20 if injuries such as severe pulmonary and chest wall injury, severe closed head injury, spinal cord injury, major soft tissue injuries, or multiple lower extremity fractures exist. These patients have an ISS greater than 20. With ATI greater than 20 to 25 or ISS greater than or equal to 18 to 20, enteral nutrition is usually tolerated and reduces septic complications (9, 12).

General Surgical Patients

Severely malnourished patients are susceptible to wound dehiscence, infections, anastomotic leaks, and so on. With no gold standard to determine nutritional status, a complete history and physical examination, determination of body weight changes, and the use of select serum tests help identify risk for nutrition-related complications.

The simplest screen is a history and physical with identification of unintentional weight loss because a strong correlation exists between impaired protein status and postoperative complications (16). Unintentional weight loss greater than 10% occurring over 6 months with increased metabolic requirements indicates nutritional risk. Two calculations are commonly used (17):


Symptoms of abdominal pain, chronic diarrhea, anorexia, or lethargy often accompany weight change. Anthropometric measurements, creatinine-height index, and delayed cutaneous hypersensitivity to a battery of antigens are rarely used in practice currently (18, 19, 20). Assessment of lymphocyte count or lymphocyte transformation also is not specific. Protein-calorie malnutrition decreases ALB synthesis, but a decrease in protein degradation can maintain serum levels. This occurs with marasmus when protein and calorie intake are severely restricted. Lower levels of constitutive transport proteins such as ALB (t½ = 21 days), transferrin (TFN; t½ = 8 days), or thyroxin-binding prealbumin (t½ = 2-3 days) may reflect the degree of malnutrition (21). However, inflammatory conditions (e.g., trauma, sepsis, peritonitis) increase serum interleukin-6 (IL-6), which stimulates the acute phase protein response (22) to increase C-reactive protein (CRP) and α-1-acid glycoprotein (AAG) and reduce constitutive protein production. Therefore, initial protein assessment should include CRP with ALB or prealbumin. Low constitutive proteins with a low CRP more likely indicate preexisting malnutrition. Elevated CRP with depressed constitutive proteins may reflect inflammation, protein-calorie malnutrition, or both.

Combinations of these parameters have been used in predictive models to quantify risk. The Prognostic Nutritional Index (PNI) (23) is calculated as follows:

PNI (%) = 158 − 16.6 (ALB) − 0.78 (TSF) − 0.20 (TFN) − 5.8 (DH) where PNI is the percentage of risk of complication, ALB is serum ALB in g/dL, TSF is the triceps skinfold thickness in millimeters, TFN is the serum TFN in mg/dL, and DH is delayed hypersensitivity reactive to one of three recall antigens. With DH, 0 = nonreactive; 1 = less than 5 mm induration; and 2 = greater than 5 mm induration. Because DH is rarely used, an alternative applies a lymphocyte score of 0 to 2, where
0 = less than 1000 total lymphocytes/mm3; 1 = 1000 to 2000/mm3 and 2 = greater than 2000/mm3. ALB drives the results, rendering it susceptible to nonnutritional factors such as inflammation, preexisting liver disease, and edema. The PNI predicts complications better than ALB alone (24).





High risk



  1. Tangential

  2. Through-and-through (duct intact)

  3. Major debridement or distal duct injury

  4. Proximal duct injury

  5. Pancreaticoduodenectomy

Large intestine


  1. Serosal injury

  2. Single wall injury

  3. ≤25% wall injury

  4. >25% wall injury

  5. Colon wall and blood supply

Major vascular


  1. ≤25% wall

  2. >25% wall

  3. Complete transection

  4. Interposition grafting or bypass

  5. Ligation

Moderately high risk



  1. Single wall

  2. ≤ 25% wall

  3. >25% wall

  4. Duodenal wall and blood supply

  5. Pancreaticoduodenectomy



  1. Nonbleeding, peripheral

  2. Bleeding, central, of minor debridement

  3. Major debridement

  4. Lobectomy

  5. Lobectomy with caval repair or extensive bipolar débridement

Moderate risk



  1. Single wall

  2. Through-and-through

  3. Minor debridement

  4. Wedge resection

  5. >35% resection



  1. Nonbleeding

  2. Cautery or hemostatic agent

  3. Minor debridement or suturing

  4. Partial resection

  5. Splenectomy

Low risk



  1. Nonbleeding

  2. Minor debridement or suturing

  3. Major debridement

  4. Pedicle or major calyceal injury

  5. Nephrectomy



  1. Contusion

  2. Laceration

  3. Minor debridement

  4. Segmental resection

  5. Reconstruction



  1. Single wall

  2. Through-and-through

  3. Debridement

  4. Wedge resection

  5. Reconstruction

Extrahepatic biliary


  1. Contusion

  2. Cholecystectomy

  3. ≥25% wall

  4. >25% wall

  5. Biliary enteric reconstruction



  1. Periosteum

  2. Cortex

  3. Through-and-through

  4. Intra-articular

  5. Major bone loss

Small bowel


  1. Single wall

  2. Through-and-through

  3. ≤25% wall

  4. >25% wall

  5. Wall and blood supply or >5 injuries

Minor vascular


  1. Nonbleeding small hematoma

  2. Nonbleeding large hematoma

  3. Suturing

  4. Ligation of isolated vessels

  5. Ligation of named vessels

a The Abdominal Trauma Index is calculated by multiplying the risk of sepsis (column 2) by the severity of injury (column 3) for each individual organ injured and summing the individual scores for all injuries.

Data from references 8, 15, and 57, with permission.

The Prognostic Inflammatory and Nutrition Index (PINI) (23, 24, 25) correlates recovery with acute phase and constitutive proteins as follows:

CRP, AAG, and prealbumin are measured in mg/dL, whereas ALB is measured in g/dL. Because the AAG elevation and ALB depression are prolonged and slow to recover, CRP and PA drive the equation, although sensitivity and specificity are lost when AAG and ALB are not included.

The Subjective Global Assessment (SGA) (26, 27) examines changes in organ function and body composition, the disease process, and the restriction of nutrient intake to predict nutrition status. The SGA is more valuable than anthropometry, which suffers from interob-server variability, hydration state, and age.

Fig. 93.1. Complications increase as albumin levels drop in patient surgical populations (solid line). Complication rates, however, vary by surgical procedure. Patients undergoing esophageal (squares connected by dots) and pancreatic (circles connected by dots) procedures have developed complications at a higher rate at the same albumin level compared with patients undergoing gastric (triangles connected by dashed lines), or colon (squares connected by dashed lines). (Reprinted from Kudsk KA, Tolley EA, DeWitt C et al. Preoperative albumin and surgical site identifies surgical risk of major post-operative complications. JPEN J Parenter Enteral Nutr 203;27:1-9, with permission from the American Society for Parenteral and Enteral Nutrition [ASPEN]. ASPEN does not endorse the use of this material of any form other than its entirety.)

The most stressful gastrointestinal (GI) operations are esophagectomy and pancreatic surgery. Complications increase as preoperative ALB levels drop in elective surgery on these organs (Fig. 93.1) (28). Patients undergoing esophagectomy with ALB less than 3.5 g/dL or pancreatic or gastric operations with ALB less than 3.25 g/dL have increased risk for major postoperative complications; risk increases as ALB drops.


The metabolic, physiologic, inflammatory, and cytokine responses to surgery and injury have been well described (29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58). These are also covered in detail elsewhere in this volume.


Estimating Total Caloric Requirements

The nutrient prescription should meet the metabolic demands of the patient. Overfeeding increases oxygen consumption, generates hepatic lipogenesis, produces immunosuppression (secondary to hyperglycemia or lipid deposition), and increases CO2 production, and should be avoided.

The most common way to determine basal energy expenditure (BEE) is the Harris-Benedict formula:

In earlier decades, these values were multiplied by stress and activity factors, but indirect calorimetry shows that stress and activity factors often result in overfeeding when they are used (59, 60).

Indirect calorimetry measures expired CO2 and O2 consumption via expired gas to determine the overall resting energy expenditure via the Weir equation. Measurements under controlled conditions approximate the Harris-Benedict equation within 5% to 10% demonstrating that large stress factors are unnecessary. The respiratory quotient (RQ) ratio analyzes the substrate used by the patient because each fuel has a characteristic RQ during metabolism (carbohydrate RQ = 1.0; protein RQ = 0.8; fat RQ = 0.7). Lipogenesis has an RQ value of approximately 8, and a calculated RQ greater than 1 is diagnostic of overfeeding. Unfortunately, small errors in measurement of inspired or expired [V with dot above]O2 because of chest tube losses or leaks at a tracheostomy in patients administered a high O2 concentration can produce a 100% error (61). Thus, the patients who could benefit most from these measurements—the most critically ill, ventilated patients—are those most likely to have tainted values. The technique is labor intensive and requires defined protocols. Because critically ill surgical patients rarely have needs greater than or equal to 15% above the calculated BEE, providing 20 to 30 kcal/kg/day results in 90% of patients receiving adequate nutrition with overfeeding in only 10% to 20% (59, 60). Total requirements are met by administering fat (10 kcal/g IV or 9.1 kcal/g enteral), carbohydrate (4.0 kcal/g enteral and 3.4 kcal/g hydrated glucose), and protein (4 kcal/g) (59).

A modest caloric intake may yield better clinical outcomes in some critically ill patients. Medical intensive care unit (ICU) patients receiving between approximately 9 and 18 kcal/kg/day through enteral and/or parenteral feedings achieved spontaneous ventilation before ICU discharge and had greater survival to hospital discharge than patients receiving 18 to 28 kcal/kg/day (62). The modest caloric intake is also improved outcome compared with patients given a caloric intake of 0 to 9 kcal/kg/day. A specific range for caloric intake may exist for which calories exceeding or failing to meet this intake can have a negative impact on patient outcome. Trauma studies have demonstrated that patients receiving only 40% of goal enteral nutrition intake exhibited significantly fewer infectious complications compared with PN patients receiving greater than 50% of the goal calories (9). Whether these results are because of avoiding
overfeeding or some other mechanism is unknown, but providing less than 100% of calculated needs may be beneficial.

Evidence suggests that obesity is an independent risk factor for ICU death (63, 64), and “permissive underfeeding” has been used in obese postoperative patients requiring PN. A hypocaloric, high-protein regimen promotes use of endogenous fat in stressed obese patients while maintaining lean tissue mass. One study showed that hypocaloric, high-protein PN administered to obese patients produced an average weight loss of 2.3 kg/week over 48 days while maintaining positive nitrogen balance with complete healing of wound dehiscence, abscesses, and fistulae (65). Typical hypocaloric PN formulations provide total 22 kcal/kg/day and 2 g protein/kg/day based on ideal body weight for patients with BMI between 30 and 40 and 25 kcal/kg/day and 2.5 g protein/kg/day for patients with a BMI greater than 40. More research on caloric dosing in catabolic patients is clearly needed because of the relative lack of rigorous controlled clinical trials in this area (66).

Glucose Requirements

Hepatic gluconeogenesis produces hyperglycemia as glucose production increases from 2 to 2.5 mg/kg/minute normally to 4 to 5 mg/kg/minute during stress (56, 67). Maximal rate of glucose oxidation is 5 mg/kg/minute (7.2 g/kg/day), which is easily exceeded (68). In a 70-kg person, 2 L of 25% dextrose contains 500 g of glucose, which reaches this maximal level. Traditional recommendations have been to maintain blood glucose values lower than 200 mg/dL because of effects on neutrophils, but data suggest that even tighter control (80 to 120 mg/dL) with insulin improves clinical outcome (53). Mortality dropped in a large group of ICU patients (primarily cardiac surgery patients). It was unclear whether insulin had the primary effect on the cardiac response or through other effects because general surgical patients in that study (including trauma, vascular, and other intra-abdominal procedures) showed no significant improvement with the aggressive insulin treatment. Further work is necessary, but current recommendations are that glucose should be maintained much lower than 180 mg/dL and ideally 140 to 180 mg/ dL (and possibly <150 mg/dL in surgical ICU patients) via insulin infusion and very close clinical monitoring to prevent hypoglycemia (56, 69).

Protein Requirements

Adult patients without renal dysfunction should generally receive 1.2 to 1.5 g/kg/day of protein (or amino acids), although this recommendation is not based on rigorous clinical trial data (59). Higher amounts of protein may be indicated in certain conditions (e.g., renal failure patients receiving renal replacement therapy, burn injury). Children also require higher levels of protein per kilogram to account for growth needs (59). If blood urea nitrogen exceeds 100 mg/dL, protein should be decreased (e.g., to 1.0 to 1.3 g/kg/day). With hemodialysis or renal supportive techniques such as continuous arterial venous hemodialysis or continuous venovenous hemodialysis, protein requirements actually increase to 1.5 to 2.0 g/kg/day because of protein losses across the dialysis membranes. Burn patients typically require 2.0 to 2.5 g/kg/day owing to urinary and wound losses (59). Table 93.2 summarizes European and US clinical practice guidelines for protein and amino acid needs in catabolic patients (70, 71).





1.0 g/kg/d actual BW

Stress or repletion

1.3-2.0 g/kg/d actual BW

Renal failure/before dialysis

0.8-1.0 g/kg/d dry BW

Renal failure/hemodialysis

1.2-1.5 g/kg/d dry BW

Renal failure/peritoneal/CVVHD

1.5-2.0 g/kg/d dry BW

Burn injury

2.0-2.5 g g/kg/d dry BW

Hepatic failure

0.6-1.2 g/kg/d dry BW

Liver transplant

1-1.5 g/kg/d dry BW

Bone marrow transplant

1.5-2.0 g/kg/d dry BW

BW, body weight; CVVHD, continuous venovenous hemodialysis.

Fat Requirements

Glucose should provide approximately 50% to 60% of total calories (approximately 70% to 80% of nonprotein calories) (59, 70, 71). The balance of nonprotein calories should be given as 1 to 1.5 g/kg/day of fat with triglyceride levels less than 300 mg/dL (70, 71). Hyperlipidemia with triglycerides greater than 500 mg/dL mandates withholding IV lipid emulsion (IVLE) in PN until triglyceride levels decrease to a safer range. The maximum recommended dose of IV lipid is 2.5 g/kg/day in the adult, but this should be used rarely (70, 71). Fat calories can be increased to 50% of requirements in select patients with severe hyperglycemia or high CO2 production, but with risks of hyperlipidemia, cholestasis, immunosuppression, and increased infection (59, 70, 71). Suspected overfeeding with increased CO2 should be treated by reduction in total calories (59, 70, 71).

Vitamin Requirements

In April 2000, the Food and Drug Administration (FDA) modified requirements for adult parenteral multivitamins recommending changes to the earlier 12-vitamin formulation (73). Changes included higher dosages of vitamins B1 (thiamin), vitamin B6 (pyridoxine), vitamin C (ascorbic acid), and folic acid, and the addition of vitamin K (phylloquinone) (Table 93.3). The vitamin content of the original formula was based on the known nutritional needs of healthy individuals to prevent nutrient deficiency. Clinicians suggested that
requirements may be greater in seriously ill patients requiring specialized nutrition (74). Higher dosages of vitamin C or vitamin E may play a critical role in antioxidant defense. Patients undergoing abdominal aortic aneurysm surgery (75) received 600 IU of oral vitamin E daily for 8 days before surgery. The vitamin E reduced ischemic-reperfusion tissue injury on muscle biopsies compared with patients without supplementation. Patients following trauma or emergency surgery were randomized to vitamins C and E (α-tocopherol 1000 IU every 8 hours orally and ascorbic acid 1000 mg every 8 hours IV during an ICU stay up to 28 days) or nothing (76). No difference in pneumonia or acute respiratory distress syndrome (ARDS) was detected but multiple organ dysfunction syndrome (MODS) was significantly lower with vitamin treatment, although the incidence of MODS occurring in both groups was low. Although micronutrient requirements for critically ill patients are unknown, additional studies suggest that higher dose for a variety of vitamins, such as vitamins A, C and D, are likely higher than as recommended currently in Table 93.3 (59, 71, 74, 77).

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Jul 27, 2016 | Posted by in PUBLIC HEALTH AND EPIDEMIOLOGY | Comments Off on Nutrition Support For the Patient with Surgery, Trauma, Or Sepsis1

Full access? Get Clinical Tree

Get Clinical Tree app for offline access