http://surgerybook.net/

Perioperative delivery of enteral nutrition may be challenging in certain circumstances. However,

lessons learned in burn patients reinforce the notion that continuous feeding for nonabdominal

procedures is safe.104 Similarly, presence of an open abdomen is not a contraindication to provision of

enteral nutrition.105

Compared to intravenous dextrose or no nutrition, early enteral nutrition results in lower morbidity

(largely infectious), mortality, and cost in ICU patients.106–108 Bowel rest is not beneficial in ICU

patients. Just as asystole does not rest the heart, starvation does not inhibit bowel function, rather it

decreases splanchnic flow and results in bacterial overgrowth and translocation. Enteral nutrition

promotes blood flow, arguing against the routine practice of discontinuing it in those on pressors.109,110

There is some sense that if full feeding is not possible for whatever reason, that partial or trophic

nutrition be employed. This strategy aims to prevent the overgrowth of bacteria in a static

gastrointestinal tract that could result in translocation into the portal circulation causing sepsis and

fueling multiple organ failure. However, the EDEN study comparing the use of trophic versus full

feeding in ARDS patients did not demonstrate a benefit to the former in terms of infection rate,

ventilator days, ICU length of stay, or mortality. Of course, this study does not address what one should

do if full feeds are not possible – maintain trophic only enteral feeds, add partial or full parenteral

feeds, or employ another strategy.110 Earlier enteral versus parenteral nutrition studies were fraught

with issues of overfeeding, imprecise glucose control, and the potential for substandard intravenous

catheter care, all of which would raise the risk of infection in those parenterally fed.

Strategies When Unable to Feed Enterally/Role of Supplemental Parenteral Nutrition

Several recent studies have addressed the role of supplemental parenteral nutrition to be used in

instances when enteral nutrition might not be possible. In the large EPaNIC study from Europe of over

4,000 patients randomized to “early” or “late” parenteral nutrition if caloric goals were not met by day

2, early patients did worse. Those randomized to receiving late parenteral nutrition (by day 7) had

shorter ICU and hospital length of stay, fewer infections, less ventilator and dialysis days, and lower

cost. The more parenteral nutrition patients received, the worse the outcome.111 In a large Australian

randomized trial, the only parameter that was improved by supplemental parenteral nutrition was a

clinically insignificant, but statistically significant, difference in ventilator days (less than one half

day).112 The Swiss SPN study of supplemental parenteral nutrition demonstrated no difference in

mortality, infections to day 28, length or stay or ventilator days between groups.113 In injured patients,

the use of parenteral nutrition has largely been abandoned, used in fewer than 5% of patients in one

study. Supplemental parenteral nutrition in a cohort of severely blunt injured patients doubled both the

rate of infections and mortality.112 Finally, a recent randomized controlled multicenter trial – CALORIES

– compares the use of early enteral and parenteral nutrition. Interestingly, the authors found no

difference in the infection rate, adverse events, or mortality.114 While some can interpret this study as

one that supports the safety of early parenteral nutrition, an alternate view would be that less expensive

early enteral nutrition is, in fact, able to be administered in most ICU patients.

Delivery of Immunitrition

Glutamine and omega-three fatty acids (fish oils). Previous retrospective studies of immune modulating

nutrients such as glutamine, selenium, and omega-three fatty acids indicated an outcome benefit in

terms of diminished infections and improved survival using high-protein enteral formulas. However,

newer studies are not as conclusive. In the REDOXS trial examining the effect of supplemental

intravenous or oral glutamine in septic patients with multiple organ failure, glutamine administration

resulted in an increased mortality rate of over 5%.115

There is not convincing evidence that utilizing an omega-three-rich lipid composition in parenteral

nutrition affects mortality or ICU length of stay,116,117 but this may be a function of route delivered and

baseline nutritional status. It’s not clear that there is a benefit for enteral administration either, with

some showing great benefit while others do not. Finally, the recent MetaPlus study described a

multicenter, randomized, double-blind trial of 301 mechanically ventilated adult patients using

comprehensive immunonutrition. There was no difference reported in infection rate, ventilator days,

and ICU and hospital length of stay between groups. Further, although not seen in trauma and surgical

patients, in this study, immunonutrition was associated with substantially higher 6-month mortality (54

vs. 35%) in medical ICU patients.118

Nutrition for Specific Indications

Acute Pancreatitis

320

http://surgerybook.net/

Enteral nutrition, if possible, is preferable in acute pancreatitis as it decreases length of stay, improves

resolution of SIRS, decreases infection rate, and does not stimulate exocrine pancreatic secretion to a

great extent. A recent Cochrane review of eight trials revealed a relative risk of death of 0.18 and 0.46

for multiple organ failure for those with severe acute pancreatitis who received enteral nutrition.119

Obesity

Goals in management of the obese patient are to reduce fat mass, improve insulin sensitivity, and

preserve lean body mass. A high-protein, hypocaloric strategy is favored with 22 to 25 kcal/kg/day

based on ideal body weight and 2.0 to 2.5 g/kg/day of protein.

Stress Gastrointestinal Bleeding Prophylaxis

Endoscopic evidence of upper gastrointestinal bleeding is extraordinarily common in ICU patients. It is

present shortly after admission in high-acuity ICU patients and thought due to impaired mucosal

protection and less so due to acid hypersecretion and Helicobacter pylori infection. The latter two are

features typical of more distal ulcers that can also result in upper gastrointestinal bleeding. Overt

bleeding, manifested by coffee ground upper intestinal aspirates or hemoccult positive stool, is

appreciated in up to 25% of patients in the ICU. Clinically significant bleeding resulting in the need for

a blood transfusion and/or hemodynamic instability is much less common, noted in less than 5% of ICU

patients, with or without pharmacologic prophylaxis. The strongest risk factor for stress gastrointestinal

bleeding in the ICU is the need for more than 48 hours of mechanical ventilation, although

epidemiologic studies pointing to this observation are dated.120 The next most common risk factor is the

presence of coagulopathy. It seems prudent that pharmacologic prophylaxis of stress gastrointestinal

bleeding in ICU patients be limited to those with these two risk factors. There may, however, be a role

to avoid pharmacologic agents even in the highest-risk patients, particularly if they are being enterally

fed.121 If pharmacologic prophylaxis is prescribed, proton pump inhibitors (PPI’s) appear to be superior

to histamine-2 receptor antagonists in diminishing gastrointestinal bleeding; however, there is no

change in mortality, length of stay, or pneumonia incidence in ICU patients.122 However, PPI’s increase

the risk for infection with Clostridium difficile, although this association has not been appreciated in ICU

patients to date.

ACUTE KIDNEY INSUFFICIENCY (AKI)

Acute renal failure complicates at least 5% and as many as 30% of all surgical ICU admissions with a

mortality of over 50%, as the most prognostic component of multiple organ failure. In analysis of a

cohort of hypotensive, severely blunt injured patients enrolled in the “Inflammation and the Host

Response to Injury” (Glue Grant) database, AKI occurred in one-quarter of patients with a threefold

adjusted risk for hospital death.123 AKI can be graded along a spectrum as described by the RIFLE

classification (from risk to injury to failure to loss to end-stage kidney disease) as defined by the Acute

Dialysis Quality Initiative (ADQI), and may be oliguric or nonoliguric (urine output < or >400 mL/day

in adults). Oliguric renal failure and higher grades of injury portend a higher mortality. An additional

classification system has been developed by the Acute Kidney Injury Network (AKIN) and is also

composed of urine output and serum creatinine criteria as described in Table 10-10. AKIN has the

advantage of considering minor variations in serum creatinine better than does RIFLE and it avoids

interpretation errors in the absence of a baseline creatinine value. Neither classification system,

however, is favored in terms of predicating incidence of and outcome from AKI. However, to simplify

matters, the Kidney Disease Improving Global Outcomes (KDIGO) work group recently proposed a

hybrid definition.124–126

Etiology

Acute kidney injury and ultimately failure is the result of forces extrinsic to the kidney or inherently

parenchymal (about 90%). Prerenal versus postrenal causes can be distinguished with serum and urine

osmolality and basic metabolic panel measurements (Table 10-11). Radiographic diagnostic studies (i.e.,

ultrasound and CT scan) can diagnose hydronephrosis or a distended bladder as causes of postrenal AKI.

Prerenal azotemia, in which the kidney perceives a lack of perfusion, may occur with both hypovolemia

and hypervolemia, the latter with congestive heart failure and impaired cardiac function. Initially, the

kidney is able to maintain perfusion by selectively dilating the afferent arteriole and vasoconstricting

the efferent arteriole of the juxtaglomerular apparatus. As hypotension persists and the renin–

321

http://surgerybook.net/

angiotensin system is activated, systemic vasoconstriction occurs and cortical hypoperfusion occurs. It is

important to note that laboratory markers are able to distinguish between prerenal and renal causes of

AKI, but not between hypovolemic and hypervolemic/cardiogenic states that may be the proximal

cause.

Table 10-10 RIFLE AKIN and KDIGO Classification of Renal Failure

Table 10-11 Prerenal Versus Acute Tubular Necrosis as the Cause for Acute

Kidney Injury

Acute tubular necrosis (ATN) results as a consequence of ischemia (most) or nephrotoxic agents. The

former may occur in the face of sepsis, hypovolemia, or cardiogenic shock. The latter includes common

ICU medications, radiocontrast media, or pigments (myoglobin and hemoglobin). Medications that may

cause ATN include aminoglycosides, cephalosporins, amphotericin, and cisplatin. Contrast nephropathy

occurs in about 5% of those exposed; particularly at high risk are elderly diabetics with pre-existing

kidney disease. Rarely will patients with contrast nephropathy progress to requiring dialysis. Only

administration of intravenous saline precontrast administration has been demonstrated to lessen the

incidence of contrast nephropathy.127 The use of n-acetyl cysteine or bicarbonate has not been

beneficial. In those with pre-existing renal insufficiency who nonetheless require administration of

contrast media, there may be a role for prophylactic continuous dialysis. Hemoglobin pigment rarely

causes ATN in surgical ICU patients. A rare but morbid cause would be the administration of incorrectly

crossmatched blood products. Myoglobinuria as a consequence of rhabdomyolysis is far more common

in the ICU and occurs after severe trauma, compartment syndrome, electrical burns, seizures, or coma

322

http://surgerybook.net/

with prolonged immobilization. Myoglobinuria should be suspected in patients with dark urine and a

positive urine dipstick without red cells on microscopy. Patients with a serum myoglobin level

exceeding 10,000 ng/mL or a creatinine phosophokinase (CPK) greater than 20,000 IU/L are a risk for

pigment nephropathy. Myoglobin is both a potent vasoconstrictor (FeNa is <1%) and a direct tubular

toxin when converted to ferrihemate in an acid environment. However, alkalinization of the urine (with

administration of sodium bicarbonate) has not been proven to be more effective than volume

administration and forced dieresis to achieve a urine output of 2 mL/kg/hr.128

Postobstructive uropathy is rare in the ICU and leads to AKI as the result of a mechanical obstruction

to urinary flow. An obstructed urinary catheter is the most common cause and should prompt ICU

personnel caring for a patient with an abrupt cessation of urinary output to interrogate the patency of

the catheter.

10 Prevention of acute renal failure is most effectively achieved by controlled fluid resuscitation of

volume repletion. There is little evidence to benefit of colloid over crystalloids. Higher–molecularweight hydroethyl starch preparations should be avoided, however, as they are nephrotoxic through

tubular obstruction. If a mean arterial pressure (MAP) of greater than 65 mm Hg cannot be maintained

solely with fluid administration, it is reasonable to initiate norepinephrine therapy. There is no role for

the use of low-dose dopamine for protection against AKI or improvement of diuresis, as noted by

several meta-analyses.129,130

Treatment of Acute Renal Failure

General Care

Attention must be directed toward management of electrolyte disturbances and hypervolemia in those

with renal failure or impending failure. Hyperkalemia, hyperphosphatemia, metabolic acidosis, uremia,

and congestive heart failure may all be of consequence. It is important to remember that treatment for

hyperkalemia can either be temporizing or definitive. Temporizing treatment such as the administration

of calcium or insulin will stabilize cell membranes and lessen cardiac irritability by forcing potassium

intracellularly. However, only removal of potassium through the gastrointestinal tract, the urine or

dialysis offers definitive therapy. Thus, if a surgical ICU patient is oliguric and not a candidate for

enteral or rectal administration of a potassium binder, then dialysis must be promptly administered to

those with markedly elevated potassium levels or ECG changes. Similarly, uremia causing a significant

pericardial effusion or encephalopathy may be temporized or treated definitively with dialysis. Platelet

dysfunction is also associated with AKI, particularly in face of a BUN in excess of 100 mg/dL. The

administration of desmopressin (DDAVP) or analogues may be of benefit. Finally, as the strategy for

renal replacement therapy is developed, attention must be directed toward optimizing nutritional status

without worsening electrolyte disturbances or hypervolemia. Prior to initiation of dialysis, it may be

necessary to limit protein administration so as to avoid exacerbating uremia. However, as renal failure

typically occurs in the catabolic milieu of multiple organ failure, it is vital to administer an adequate

protein substrate as well as calories to prevent excessive gluconeogenesis and protein breakdown. This,

in turn, leads to worsening uremia.

Diuretics

There may be a theoretical attraction to using diuretics to ameliorate AKI. However, in practice

diuretics have not been successful in either preventing or ameliorating AKI. Rather, prophylactic

diuretics often raise serum creatinine levels. Systematic reviews and meta-analyses

131,132 have

demonstrated that diuretic use does not alter outcome from AKI, but has a significant risk of morbidity

(such as hearing loss) and even mortality in some. Finally, if diuretics are used, there is no evidence in

systematic review and meta-analysis that continuous dosing is superior to bolus administration.133,134

Renal replacement therapy (RRT) is indicated for intractable fluid overload, hyperkalemia

(potassium concentration greater than 6.5 mmol/L or 5.5 with ECG changes), severe metabolic acidosis

(pH <7.1), uremic encephalopathy or pericarditis and overdose with a dialyzable agent. Hemodialysis

is typically performed in ICU patients and may be continuous or intermittent. Both the Randomized

Evaluation of Normal versus Augmented Level Renal Replacement (RENAL) study and the Veterans

Administration/National Institutes of Health Acute Renal Failure Trial Network (ATN) failed to

demonstrate a benefit from increasing the intensity of RRT beyond 20 to 25 mL/kg/hr of effluent flow,

the former trial utilizing continuous modes, the latter intermittent.135,136

Continuous renal replacement therapy (CRRT) includes slow continuous ultrafiltration (SCUF)

and continuous venovenous and arteriovenous modes. Patients who are placed on CRRT must have a

323