Chapter 79. Postoperative Care of the Noncardiac Surgical Intensive Care Unit Patient

1. Demand is growing for postoperative intensive care services due to advances in surgical techniques and the aging of the population. Anesthesiologists and surgical intensivists play a major role in ensuring responsible use of this costly resource.

2. Advances in neuraxial pain management have revolutionized certain types of surgery (eg, thoracic and major vascular), permitting patients to undergo major procedures without the need for prolonged intensive care following surgery.

3. Critical illness polyneuropathy and myopathy are acute illnesses that result in prolonged weakness or paralysis in subsets of intensive care patients.

4. Critically ill patients are at risk for a variety of pulmonary complications including aspiration, ventilator-associated pneumonia and acute lung injury. Intensive care management is directed at minimizing the risk factors predisposing patients to these complications.

5. Pulmonary artery catheter (PAC) guided therapy has not been shown to improve outcomes in critically ill patients with acute lung injury.

6. Conservative fluid management results in improved lung function and shortens the duration of both mechanical ventilation and intensive care stay without altering the rate of extrapulmonary organ failure.

7. The stress response after major surgery or injury is often accompanied by a period of endothelial cell dysfunction and capillary leak with loss of plasma volume into the extracellular "third space." The stress response may be initiated by tissue hypoperfusion due to inadequate fluid resuscitation, ischemia-reperfusion injury, cytokine release, or exposure of the circulating blood volume to an extracorporeal circuit (ie, blood salvage circuits, cardiopulmonary bypass).

8. Renal replacement therapy is a field that has evolved significantly over the past decade, and venovenous diafiltration and hemodialysis have displaced techniques such as arteriovenous hemodialysis.

The surgical intensive care unit (SICU), or a combined medical-surgical ICU, is a specialized patient care floor designed to accommodate and treat critically ill surgical patients in the perioperative period, which may include preoperative, postoperative, and posttrauma injury management. As critical care techniques have evolved, it has become possible to both save the lives of some who might previously have died and prolong the lives of others who will nevertheless still not survive. The percentage of critical care beds has grown in many hospitals, and it is still an increasingly expensive and constrained resource due in part to shortages of qualified physicians, nurses, and ancillary personnel. It is vital, therefore, to find models of efficient and appropriately targeted intensive care.

The types of patients admitted to ICUs have changed considerably over the past several decades. The evolution of trauma systems, with rapid transportation of critically injured patients, has resulted in the concentration of complex multitrauma patients in trauma centers. The evolution of ventricular assist devices, thoracic aortic surgery, and heart and lung transplantation has revolutionized cardiothoracic surgery and changed the nature of perioperative cardiac intensive care. New approaches to the management of head injury and advanced neurosurgical techniques require increased technological sophistication in cerebral monitoring. Advances in the treatment of acute lung injury have resulted from research collaboratives such as the National Institutes of Health ARDS Network and been facilitated by improvements in mechanical ventilation technology. Perhaps the most important component of improvements in critical care has been the increased expertise of physician, nursing, and ancillary personnel in critical care management.

Although many innovations have increased the demands for critical care, endoscopic approaches to thoracic surgery and the use of postoperative epidural analgesia have allowed many patients to go directly to a hospital ward after surgery rather than the ICU. Aortic stents have had similar implications for vascular surgical patients who might once have undergone a laparotomy and several-day ICU stay. Many now go directly to a routine hospital ward.

As our ability to prolong life has improved, so has the complexity of ethical questions arising in the course of ICU care. The fact that we can temporarily or permanently replace the function of the failing heart, kidney, or lung increases the frequency with which we are confronted with questions as to whether it is appropriate to do so.

Critical care bed occupancy can have a major effect on the flow of patients through operating rooms (ORs). OR exit delays often result when ICUs are at a high census level. It is essential to make optimal use of these often scarce and invariably expensive resources. The criteria for admission to an ICU following surgery should be equitable and transparent, and they should prioritize those patients most likely to benefit from unit care.1,2 Ideally, this prevents ICU admission for those who are "too well" or "too sick."

Screening Tools and Resource Allocation

A Task Force of the American College of Critical Medicine and the Society of Critical Care Medicine developed ICU admission screening tools based on 3 different approaches: prioritization, diagnosis, and objective parameters.

• The prioritization approach classifies patients into 4 priorities (Table 79-1).
• The diagnosis approach (Table 79-2) identifies 8 diagnostic criteria: cardiac, pulmonary, neurological, drug ingestion and drug overdose, gastrointestinal disorders, endocrine, surgical, and miscellaneous, and divides them into admissible disease states.
• The objective approach (Table 79-3) is based on 5 objective criteria for admission into the ICU. These 5 categories are vital signs, laboratory values, radiology, electrocardiogram (ECG), and physical findings.

These admission criteria are widely used by health care organizations in the United States and abroad.3,4 Health care organizations also benefit from the development of a triage plan by an internal ICU committee designed to optimally place patients when the demand for ICU beds is greater than the supply.5 An institution-specific ICU committee is typically composed of health care professionals, ethics advisors, and legal advisors. An institutional triage officer can then direct patient triage based on institutional priorities. The principles outlined in the plan should be available to the public and to patients and/or their surrogates.5 The triage plan should not be arbitrary or prejudicial.

The triage officer should also evaluate patients in the ICU to determine which patients are receiving the greatest benefit from ICU care. An example of a patient who might be deemed inappropriate for ICU admission is the patient with an advanced directive directing that cardiopulmonary resuscitation and/or mechanical ventilation not be initiated. In contrast, a brain-dead organ donor patient, although legally dead, might be kept in an ICU until organ harvest to maximize the benefit to potential recipients. The creation and distribution of explicit guidelines for ICU admission, discharge, and triage facilitate bed management when a family or physician wishes to admit a patient to an ICU bed for inappropriate reasons.

The triage plan should outline how to best prioritize admissions to the ICU. Although a hospital has an obligation to treat patients who are already admitted to the institution, a level-1 trauma center has additional obligations to the state and community to admit critically injured patients. These priorities may clash when beds are tight. Floor-based patients who suffer an acute decompensation should be transferred to the ICU as soon as feasible as should emergency department patients who are acutely ill. A hospital may also choose to prioritize admissions of certain patients for specialty care, for example, cardiac transplantation. These competing agendas may lead to the deferral or cancellation of elective surgical cases needing a postoperative ICU bed on rare occasions.

Intraoperative Evaluation for ICU Care

Certain operations such as cardiac surgical procedures and certain neurosurgical procedures routinely require postoperative ICU care for the patients. These patients have predictable postoperative periods of instability during which intensive nursing and physician care are required, and they typically go to the ICU following surgery. However, some patients initially scheduled for ICU admission to the SICU may no longer be considered appropriate for ICU care because the planned procedure was aborted or the surgical course went better than expected. In other circumstances, unanticipated intraoperative events such as airway management problems, intraoperative aspiration of gastric contents, major blood loss, drug reactions, and cardiac events may necessitate unplanned admission to the SICU.

Transport between the OR and the ICU

Intrahospital transport of critically ill patients is well-recognized as a possibly hazardous period potentially associated with complications such as hypoxia, hypotension, cardiac arrest, or ventilator-associated pneumonia.6-8 See Chapter 78 on transport. Studies have shown that most transport-related complications occur during transit between the ICU and the OR or the radiology suite.8 Before transporting a patient, it is imperative that the risk transport be weighed against its goal to see if it is possible to bring the study or procedure to the patient rather than the other way around. A review by Braxton et al summarizes 3 studies assessing the risks and benefits of transporting a patient from the ICU for the purposes of diagnostic testing. In 2 of the studies, the care plans changed within 48 hours of the transport in 24% and 39%, respectively, of the patients transported, and two-thirds of the patients experienced physiologic deterioration during transport. In the third study, only 30% of the diagnostic studies contributed to patient management.9

After determining the requirement for patient transport to or from the ICU, a comprehensive transfer plan should be designed including patient preparation, required personnel, and necessary monitoring. Pretransport preparation is a crucial step for safe patient transport. A checklist is recommended and should include oxygen supply, emergency resuscitation drugs, intravenous (IV) fluids and infusions (if needed). The monitor should be checked for function and level of battery charge. Arrangements should be made to ensure that any special mechanical ventilatory requirements (eg, bilevel ventilation) can be met with the portable ventilator; and all of the foregoing should be documented on a pretransport checklist.

It is essential that the appropriate personnel are available for transport. The team may include a critical care nurse and respiratory therapist or specially trained transport technicians. Some hospitals have developed specialized transport teams trained to care for critically ill patients. Studies have shown that a major cause of transport complications is failure to follow protocol and failure to recognize that a problem has occurred during transport.9 If the patient is physiologically unstable at the time of transport, as during movement to and from the OR, a physician should also be present.

Efficient communication is another important aspect to providing safe transport. The transport team must notify the receiving unit or location at the time of transport to ensure adequate preparation at the receiving end and communicate details about the patient's history and present status. The receiving site should be ready with appropriate monitors, equipment, and drugs.

The last essential to safe patient transport is appropriate monitoring, which should include continuous ECG, blood pressure (BP) (invasive or noninvasive, as appropriate), oxygen saturation, and end-tidal carbon dioxide (CO2) if the patient is ventilated. All portable monitoring devices should have fully charged batteries and working charge indicators.

Each patient care unit should have written transport policies as recommended by the American College of Critical Care Medicine, Society of Critical Care Medicine, and the American Association of Critical Nurses.10 A review of the literature and recommendations for intrahospital transportation has been published by Fanara et al.11 These policies should be modified for each health care institution.

Postoperative Laboratory Tests and Vital Signs

Standard monitors for the intensive care patient include ECG, BP, pulse oximetry, and respiratory rate (often obtained by impedance analysis through the electrocardiographic leads). Laboratory assays that are obtained routinely include complete blood count (CBC), electrolytes, blood urea nitrogen and creatinine, coagulation panel, and arterial blood gas. Capnography is used in some ICUs for mechanically ventilated patients to assess respiratory dead space in acute and chronic diseases of the critically ill such as adult respiratory distress syndrome (ARDS), chronic obstructive pulmonary disease (COPD), and pulmonary embolism. Serial plasma lactate levels may be followed to assess the adequacy of systemic perfusion and resuscitation or as a marker of ischemia in large tissue beds such as the gastrointestinal tract. Serum calcium (ionized or total) and magnesium are important physiologic cations that may be included in intensive care laboratory panels. The absolute levels of these values as well as their trends can provide useful information during the course of resuscitation and treatment.

Pain and sedation management in the ICU is important both for patient comfort and satisfaction, as well as an increasingly recognized contributor to patient outcome. Patients in the ICU routinely report pain and discomfort during routine nursing care (eg, airway care, dressing changes, and physical therapy), and resulting from the presence of catheters, drains, and endotracheal tubes (ETTs).12,13 The potential complications of poorly managed analgesia may include prolonged intubation, ICU delirium, myocardial injury due to increased stress, and increased ICU and hospital length of stay with resulting increased cost. Effective pain management varies from patient to patient and involves an appreciation of the considerable variation in the population in pain tolerance and willingness to acknowledge suffering.

Pain assessment tools have been developed to make measurement as objective as possible and include the Visual Analog Scale (Fig. 79-1) and Numerical Rating Scale (NRS) (Fig. 79-2). The Society of Critical Care Medicine published a practice guideline in 2002 for sedation and analgesia in critically ill patients that recommends assessing pain at regular intervals, using the NRS or a behavioral pain assessment scale in conjunction with physiologic parameters (BP and heart rate) in patients who cannot communicate.

There are many pharmacologic alternatives for the management of postoperative pain including patient-controlled epidural analgesia (PCEA) and IV patient-controlled analgesia (PCA). The patient must be awake, alert, and oriented to use these pumps effectively. PCA can result in somnolence and narcotization, but this is rare if the basal and bolus limits are set appropriately. PCEA is extremely effective after certain surgical procedures, particularly those involving the chest and upper abdomen; but hypotension and bradycardia may result from sympathectomy even at low infusion levels. These physiologic responses to pain control must be differentiated from pathophysiologic problems (ie, evolving myocardial infarction, bleeding, or sepsis). Hypotension and bradycardia do not mandate discontinuation of an epidural infusion; rather the level of the block should be evaluated and the infusion rate adjusted accordingly. The concentration of local anesthetic can be decreased or the BP can be supported with fluid boluses or pressors if the benefits of pain control outweigh manageable complications of the analgesic regimen. In a recent meta-analysis by Wu et al, epidural analgesia provided better analgesia than IV PCA.14

If a patient is unable to use a PCA or PCEA, continuous infusion or intermittent boluses of opioids can be administered as needed. Continuous infusions are usually reserved for patients with a controlled and stable airway (ie, ETT or tracheostomy tube). However, in patients with chronic pain issues or high opioid tolerance, continuous infusions may be used as long as appropriate respiratory monitoring tools are in place.

Agitation is common in the ICU patient and may result from the unfamiliar setting or the inability to communicate (eg, due to the presence of an ETT). Agitation resulting from delirium is a significant and common problem in the intensive care setting. Delirium is defined in the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition as a disturbance of consciousness or cognition that has developed over a short period of time and is caused by direct physiologic consequences of a general medical condition. The incidence of delirium in the ICU is reported to be from 16% to 80%, depending on the assessor and assessment tool used.15,16

To evaluate a patient's degree of anxiety and discomfort objectively and reproducibly, intensive care providers typically use a sedation/analgesia scale. Plasma concentration levels of sedatives, frontalis electromyogram, and Bispectral Index Sensor (BIS) electroencephalographic analysis are more objective but less reliable alternatives. Plasma drug levels are not sufficient because the concentration of drug needed for analgesia varies from patient to patient. Frontalis electromyography (EMG) is not sensitive for monitoring sedation. The BIS analysis method has been studied as a monitor for level of sedation, but its results have not been confirmed in critically ill patients.

The measures in common ICU use include the Ramsay, the Richmond Agitation Sedation Scale (RASS), modified Glasgow Coma, visual analog pain, and sedation-agitation scales. The Ramsay scale is the most widely used scale for sedation monitoring (Table 79-4).17

The Ramsay scale has been found to have strong interobserver agreement and reliability as has the Richmond Analgesia and Sedation Scale (Table 79-5). The Glasgow Coma Scale modified by Cook and Palma (Table 79-6) provides a numerical score to evaluate neurologically impaired patients who are mechanically ventilated. Sedation tools are not appropriate for the evaluation of delirium in the ICU. Instruments currently available and validated to detect delirium in the ICU are the Confusion Assessment Method-ICU (CAM-ICU) and the Intensive Care Delirium Screening Checklist (ICDSC).

The γ-aminobutyric acid (GABA) receptor agonists such as propofol and benzodiazepines are frequently used to sedate agitated patients, although the latter may also be a cause of delirium. Alternative sedatives for delirious patients include haloperidol and dexmedetomidine. Dexmedetomidine is a centrally acting α-2 agonist that is more potent than clonidine with a faster onset and shorter length of action. It is also a potent analgesic and anxiolytic.

The potential complications of poorly managed analgesia and sedation include prolonged intubation, ICU delirium, and myocardial ischemia. Overly sedated patients may require longer periods of ventilator support, with a consequently longer ICU and hospital stay. Prolonged ventilation also increases the patient's risk of ventilator-associated pneumonia, and studies have also shown that inadequate pain control can cause increased morbidity and hospital stay.

Neurologic Issues

Management of the postoperative neurosurgical patient is beyond the scope of this chapter. This section addresses the management of common neurologic disorders seen in other postoperative ICU patients.

New-onset seizures are a major postoperative complication in ICU patients with a reported incidence of 12% in ICU patients without a primary neurologic diagnosis.18 The causes of postoperative seizures include medications, metabolic disorders, primary neurologic disorders, infection, hypoxia, and drug withdrawal (Table 79-7).

Accurate identification of new-onset seizures is important. Seizures must be differentiated from myoclonus and other rhythmic movement disorders (ie, tremor). Electroencephalographic testing is the standard test used to identify seizure activity. Once a new-onset seizure has been diagnosed, laboratory tests, computed tomography or magnetic resonance imaging of the brain and, in patients suspected of an infectious etiology, lumbar puncture are often performed to elucidate the etiology. The most common causes of new-onset seizures in ICU patients are metabolic disturbances with an incidence of 28.6%.14

The initial treatment for new-onset seizures is typically an IV sedative-anticonvulsant such as a benzodiazepine, barbiturates, or propofol.19 The patient is then started on a primary anticonvulsant medication. First-line anticonvulsants include phenytoin, levetiracetam, and carbamazepine.

Critical illness neuromuscular diseases are a relatively rare but important complication in ICU patients. These abnormalities are divided into polyneuropathies and myopathies of critical illness. Critical illness polyneuropathy is an acute degenerative disorder of motor and sensory nerve axons. Symptoms are flaccid tetraparesis and failure to wean from the ventilator. Critical illness myopathy is an acute degenerative illness of the myocyte that results in weakness and paralysis. There are 3 histologic types of myopathy: diffuse necrotizing, necrotizing, and myosin thick filament loss. The risk factors for development of critical illness neuromuscular abnormalities include sepsis, multiorgan dysfunction, hyperglycemia, and treatment with corticosteroids, nondepolarizing neuromuscular blocking agents, and aminoglycosides. The mechanisms by which these risk factors contribute to neuromuscular abnormalities are unknown. Bolton et al suggests that cytokines, free radicals, and activated complement fragments released during sepsis cause axonal injury.20

The incidence of critical illness polyneuropathy varies substantially depending on the diagnostic method used, the patient population, and the timing of the diagnosis. In septic patients, the incidence has been stated to be as high as 70% to 80%.21 Studies have shown an association between corticosteroids and critical illness myopathy in patients with severe asthma;22 however, no specific mechanism has been identified. Neuromuscular blocking agents (NMBAs) are also associated with acute myopathies. Concurrent administration of NMBA with steroids or aminoglycosides increases the incidence of myopathies. High blood glucose levels were associated with critical illness polyneuropathy in one study of septic patients with multiorgan dysfunction, and van den Berghe et al reported that axonal injuries were decreased by 50% in patients who received intensive insulin therapy.23 It is not clear if catecholamines and hyperglycemia lead to polyneuropathy or if insulin or euglycemia protect neuronal axons.

Encephalopathy or sedation can mask critical illness neuromuscular diseases, complicating diagnosis. Objective studies, such as nerve and/or muscle biopsies, in conjunction with EMG, can facilitate diagnosis. Nerve and muscle biopsies can provide pathologic evidence of disease. Nerve biopsies, however, can lead to permanent neurologic deficits. Electrophysiologic studies including nerve conduction studies and EMG are the current gold standard for diagnosis of critical illness neuromuscular abnormalities. Nerve conduction studies measure the latency and amplitude of the conduction and the muscle's electrical response measured with EMG: the compound muscle action potential (CMAP). In critical illness polyneuropathy CMAPs are reduced and abnormal spontaneous activity is present, indicative of axonal neuropathy. EMG records the electrical activity of active and resting muscle.

In patients affected by critical illness neuromuscular abnormalities, abnormal muscle fibrillation occurs. To differentiate between a neuropathy and a myopathy, motor unit action potentials are evaluated using EMG. If a motor unit shows increased amplitude and latency, critical illness polyneuropathy is diagnosed. If the motor unit shows decreased amplitude and latency, critical illness myopathy is diagnosed.

The treatment alternatives for critical illness neuromuscular abnormalities are limited. Intensive physical therapy with discontinuation of steroids and NMBAs and tight glucose control are the mainstays. The long-term outcome is not well studied, and recovery depends on the severity of the disease. Many patients with critical illness neuromuscular abnormalities may regain normal strength after weeks to months; however, some patients never recover.

Pulmonary Issues in the Postoperative ICU Patient

Postoperative critically ill patients often undergo a period of relative hypoxemia due to the effects of anesthetics and/or surgery superimposed on any primary disease process. Respiratory management in the ICU is often focused on the initiation, termination, or prevention of complications of mechanical ventilation.

Postoperative patients may be admitted to the ICU following surgery with an ETT in place. The newly arrived patient in the ICU should undergo a chest radiograph (CXR) to determine the location of any central lines placed in the OR, the position of the ETT, and identify pulmonary complications that may have occurred in the OR such as pneumothorax or aspiration. The ETT may need to be advanced (if the tip is above the level of the clavicle on CXR) or withdrawn (in the event of mainstem intubation).

So-called fast-track weaning, in which ventilatory support is weaned automatically by respiratory therapists as the patient meets certain milestones has become prevalent, and it has been shown to reduce the duration of mechanical ventilation in patients.24 New approaches to noninvasive ventilation and their application in a variety of patient populations have reduced the threshold for early extubation. Continuous positive airway pressure and bilevel positive airway pressure (BiPAP) are effective in reducing the need for reintubation of patients who develop respiratory insufficiency after extubation. BiPAP can be administered by face or nasal mask and is used to treat patients with marked obesity, COPD, sleep apnea, or respiratory muscle weakness by allowing them to be extubated and then rested with periods of noninvasive positive pressure ventilatory support.25

Nosocomial pneumonias include ventilator-associated and other hospital-acquired pneumonias, such as in-hospital aspiration pneumonia. The definition of the former is an infection occurring greater than 48 hours after hospital admission that is radiographically consistent with pneumonia. Aspiration pneumonia can develop as a complication of endotracheal intubation in patients who regurgitate gastric contents, and it is more likely to occur when the pH of the aspirated material is low and the volume greater than 0.3 mL/kg of body weight. Critically ill patients are at risk for the development of both of these conditions, and predisposing factors include inadequate endotracheal cuff seal, the supine position, increased gastric contents, and bacterial colonization of the oropharynx.

Acute lung injury (ALI) and ARDS can occur in the postoperative patient. These diseases represent points on a continuum of acute lung disease and result from a multitude of etiologies including trauma, transfusion reactions, and sepsis. Although a variety of approaches have been investigated for the prevention and treatment of these diseases, the standard therapy is "protective ventilation," in which low tidal volumes are given to prevent overdistension of injured as well as healthy alveoli. The ARDS network funded by the National Institute of Health published the results of a landmark study of ALI/ARDS in 1999,26 which showed that the use of low (6 mL/kg) tidal volume breaths conferred a mortality benefit when compared with the then more traditional (12 mL/kg) tidal volume ventilation. The same group recently published data from a factorially designed study comparing ARDS patients managed with a pulmonary arterial catheter (PAC) versus a central venous line and patients with a conservative versus liberal approach to fluid management.27 The results of this study showed that PAC-guided therapy did not improve outcomes and was associated with more complications, specifically those associated with placement of the PAC. The study also showed that a conservative fluid management approach resulted in improved lung function and shortened the duration of both mechanical ventilation and the ICU stay (when compared to liberal fluid management) without altering the rate of nonpulmonary organ failures.

Cardiovascular Issues in the Postoperative ICU Patient

Postoperative cardiac management is directly comparable with intraoperative management in the indications for and methods of treating abnormalities such as intravascular volume depletion, hemorrhage, bradycardia, and tachydysrhythmias. As in the OR, intensive care monitoring invariably includes noninvasive or invasive BP measurement, continuous ECG, and pulse oximetry. The stress response after major surgery or injury is often manifested by a period of impaired endothelial cell function and the resulting loss of plasma volume into the "third space," or extravascular space. Precipitating factors include tissue hypoperfusion due to inadequate fluid therapy, ischemia-reperfusion injury, and cytokine and complement activation.

Postoperative patients require ongoing fluid resuscitation commensurate with the magnitude of the surgical or traumatic injury, and in some cases that requirement may exceed 10 mL/kg/h. In healthy patients, the need for fluid resuscitation typically ends after a period of 24 to 48 hours, and the accumulated excess volume and solute load is then eliminated through the kidneys spontaneously over the succeeding several days. The time courses of these phases may be altered in patients with underlying illnesses such as hepatic or renal failure. As distinct from a pure volume requirement, a progressive decrease in the blood hemoglobin concentration or hematocrit level necessitates a search for a bleeding source.

Patients with labile BP may require an intra-arterial, central venous, or pulmonary arterial catheter to better assess and manage their intravascular volume status and cardiac performance. The utility and safety of the pulmonary artery catheter has been the subject of much recent debate,28 but current-generation catheters provide advanced functions such as continuous measurement of cardiac output, right ventricular performance, and mixed venous oxygen saturation, all of which can be used to guide short-term fluid resuscitation. Monitors such as the central line (venous pressure), pulmonary artery catheter (pulmonary pressure, occluded wedge pressure, mixed venous oximetry, cardiac output) or echocardiography (to assess ventricular filling and performance, valve abnormalities) can be used to guide resuscitation and the initiation and titration of inotropes or vasopressors.

Cardiac dysrhythmias are relatively frequent in the OR and ICU, where metabolic, ischemic, and neurohormonal stresses may cause premature atrial and ventricular contractions, conduction block, or atrial fibrillation. Cardiac rhythm abnormalities are more prevalent in patients with structural heart disease29 and can be caused by surgical stress, electrolyte abnormalities, sympathetic stimulation, direct mechanical irritation of the heart (as by intracardiac catheters), or device malfunction.30 Dysrhythmias can be separated into narrow and wide-complex QRS rhythms. Narrow QRS rhythm abnormalities typically originate in the atria and include sinus tachycardia, premature atrial complexes, atrial flutter/fibrillation, accessory pathway tachycardias, and sinus bradyarrhythmias.

In patients with wide complex QRS rhythms, it is may be difficult to differentiate between supraventricular dysrhythmias and ventricular dysrhythmias. Premature ventricular complexes (PVCs) are common and caused by structural heart disease, electrolyte imbalances, acidosis, and hypoxia. PVCs are usually benign and do not require antiarrhythmic treatment in patients without structural heart disease.31 However, studies have shown that when PVCs have a frequency of 10 or more per hour or multifocal origins, there is an increased risk of developing a life-threatening dysrhythmia, especially in patients with structural heart disease.32,33

Atrial fibrillation is the most common perioperative dysrhythmia.34 The differential diagnosis for postoperative atrial fibrillation is extensive and includes increased catecholamine levels due to stress response, electrolyte disorders, hypo- and hypervolemia, and hypoxia. The incidence in postoperative cardiac surgical patients is as high as 30% to 40%, and 3% to 4% of routine perioperative patients may develop the problem. Electrolyte and ventilatory abnormalities should be identified and corrected. Pharmacologic treatments include β-blockers, calcium channel blockers, and amiodarone. In most instances the atrial fibrillation is benign and self-limited, but for some patients in whom the arrhythmia persists, anticoagulation may be appropriate; and when accompanied by hemodynamic instability, electrical cardioversion is indicated.

Myocardial ischemia and infarction can occur in the postoperative period due to anemia, underresuscitation, tachycardia, volume overload, or the relatively hypercoagulable state that occurs 2 to 3 days after surgery. Proactive management of cardiac stressors including hyper- or hypovolemia and tachydysrhythmias improves the myocardial oxygen demand-to-supply ratio.

Gastrointestinal Issues in the Postoperative ICU Patient

Postoperative and traumatically injured patients have increased metabolic demand related to energy consumption by tissues during the repair process. Patients enter an inflammatory phase following surgery or trauma proportionate to the magnitude of the injury and become hypermetabolic as the reparative process proceeds. In the absence of adequate glucose, glycogen-based energy stores are rapidly depleted, forcing the body to break down protein from muscle to meet energy demands. Supplemental nutritional therapy is designed to meet daily energy requirements in critically ill postsurgical patients and thereby prevent protein loss. Several questions typically arise when discussing nutrition, including when to begin, how to assess nutritional requirements, and by what route to administer the feeds.

The timing of initiation of supplemental nutrition depends on both the baseline nutritional status of the patient and the disease process. Patients with disease states such as burns, sepsis, and cancer, and malnourished patients have higher caloric requirements than most other ICU patients. They require higher than standard nutritional support to heal and prevent infection. Laboratory parameters are often used to guide nutritional supplementation. Albumin, prealbumin, and transferrin levels are typically measured to guide the patient's nutritional status; however, the inflammatory response may alter the reliability of these laboratory values. Serum albumin levels often drop precipitously following surgery.

Depending on the route of delivery, nutritional support has differing complications. Enteral feeding may be complicated by pulmonary aspiration in patients who are unable to cough or gag. Additional complications include enteral anastomotic leak after gut anastomosis or intestinal ischemia in patients on vasopressors. Enteral nutrition should not be given to patients with intestinal obstruction. Supplemental feeds are typically started 3 to 5 days following surgery in patients who are unable to take adequate nutrition by mouth.

Enteral feeding is preferred over IV feeds when possible because it is physiologic and preserves gut mucosal barrier function and thereby reduces translocation of bacteria from the intestinal lumen into the vasculature. Gut feeding also preserves the immunologic function of gut-associated lymphoid tissue.

Parenteral nutrition can be administered through a peripheral or central vein. Central delivery of nutrition permits delivery of a high concentration of carbohydrates and protein, one that would irritate and sclerose peripheral veins. The disadvantages of parenteral nutrition include higher risks of infection and sepsis, intestinal atrophy, and complications associated with central venous access.

Abdominal compartment syndrome (ACS) has become an increasingly recognized gastrointestinal problem in critically ill patients. The abdominal compartment is a potential space that can fill with fluid or blood, compressing the organs and vessels within. In mechanically ventilated patients, elevated abdominal pressure can interfere with breathing by impeding downward excursion of the diaphragm. Unintubated patients may become dyspneic due to the same mechanism. Intra-abdominal pressure is assessed by measuring bladder pressure. The pressure of a column of fluid (typically urine) in continuity with the bladder can be measured using a urinary catheter, and the diagnosis of ACS should be considered when pressures exceed 20 mm Hg.

The increased intra-abdominal pressure resulting from ACS can also impair perfusion to any of the visceral organs as well as venous return to the heart. Abdominal compartment syndrome causes decreased venous return due to compression of the inferior vena cava, decreased ventricular compliance, and increased peripheral vascular resistance, all of which result in decreased cardiac output.

Renal function can be compromised as a secondary effect of impaired cardiac performance or due to direct compression of the kidneys, and it is indicated by decreased urine output. This is due to decreased renal blood flow glomerular filtration rate. Compression of the renal parenchymal and vessels can result in renin, aldosterone, and antidiuretic hormone release, leading to free-water retention that may exacerbate the primary problem.

Successful management of ACS requires that clinicians be aware of which patients are at greater risk for this complication of surgery and trauma, attentive monitoring, and early intervention. Trauma patients who have undergoing large-volume resuscitations, even in the absence of intra-abdominal injury, and surgical patients with edematous bowel following gastrointestinal surgery are at increased risk for the development of ACS. Pancreatitis, septic shock, acidosis, hypothermia, ileus, hemo- and pneumoperitoneum can predispose a critically ill patient to the development of intra-abdominal hypertension. The differential for ACS also includes a variety of medical, surgical, and traumatic conditions.

The detrimental effects of ACS can be treated in the ICU with fluids, vasopressors, and pharmacologic muscle relaxation to relax the abdominal wall. Definitive therapy is a decompressive laparotomy, although there is some controversy as to the appropriate threshold for this invasive procedure. Opening the abdomen incurs many risks including infection, sepsis, dehydration from large insensible fluid losses, and fistula formation. A grading system has been proposed to guide the evaluation and treatment of ACS (Table 79-8).

Following laparotomy in patients with ACS, there is often a dramatic improvement in cardiac output and organ perfusion. The fascia may then be secured and the viscera covered with mesh, which can be drained with a VAC dressing, or a Bogota bag. The VAC dressing helps to quantify abdominal fluid losses, protects the abdominal contents from infection, and helps prevent the abdominal wall musculature from contracting laterally, thus preventing future closure. After the ACS has resolved, the abdomen can be closed primarily or with mesh.

Hematologic Issues in the Postoperative ICU Patient

Anemia is a common problem in the SICU, but transfusion can be associated with complications. Blood product administration can result in transmission of viruses, transfusion reactions, graft-versus-host disease, and depression of the recipient's natural killer cell function. Poor wound healing, risk of anastomotic leak, and postoperative infections have also been associated with perioperative blood transfusion. However, although treatment thresholds have risen as new research has shown that critically ill patients tolerate hemoglobin levels much lower than the traditionally treated trigger of 10 mg/dL with few adverse consequences, transfusion may be necessary to prevent end-organ ischemia.

In 1999, Hébert et al35 conducted a randomized multicenter clinically controlled trial of transfusion requirements in critical care patients (the TRICC trial). The purpose of this study was to determine if a restrictive transfusion protocol that maintained circulating levels at 7.0 to 9.0 g/dL had an equivalent risk of morbidity and 30-day mortality as a liberal alternative that maintained hemoglobin (Hgb) levels of 10 to 12 g/dL. The results showed that there was no mortality difference between the 2 groups; however, in a subgroup of patients who were younger than 55 years with a low predicted mortality risk, the 30-day mortality was decreased and survival rate increased in the restrictive transfusion group.

In 2001,36 the same group published a post hoc subgroup analysis of the TRICC trial to compare the morbidity and mortality in critically ill patients with cardiovascular disease comparing the restrictive transfusion strategy group with the more liberal transfusion strategy group. Cardiovascular disease was defined as patients with a primary or secondary diagnosis of myocardial infarction, angina, congestive heart failure, dysrhythmias, cardiogenic and other forms of shock, vascular procedures, and cardiac procedures (except open heart surgery). The study results showed that there was no statistical difference with respect to survival, ICU mortality, and 30- and 60-day mortality, suggesting that lower Hgb levels did not produce additional harm in the subpopulation with cardiovascular disease.

Several other groups have subsequently examined transfusion strategies in patients with acute coronary syndrome (ACS). Patients with ACS suffer myocardial ischemia, angina, or myocardial infarction and it was believed this was due to the imbalance of oxygen supply and demand, perhaps attributable to anemia. In 2001, Wu et al37 conducted a retrospective study of 78974 Medicare patients older than 65 years with a confirmed acute myocardial infarction to determine the risk associated with anemia in these patients and the effect of blood transfusion on mortality. The results showed that 3680 patients (4.7%) received blood transfusions. It also showed that transfusion was associated with a lower 30-day mortality in patients who had an admission hematocrit 33% or lower and an increased mortality in patients with hematocrits 36.1% or greater.

In a 2005 study, Rao et al38 examined the association between blood transfusions and mortality among patient with ACS who developed bleeding, anemia, or both during their hospital course. This retrospective analysis of 24112 patients used data from 3 separate glycoprotein IIb/IIIa inhibitor trials: GUSTO IIb, PURSUIT, and PARAGON B. The study showed that 2401 of the patients (10%) had received transfusions and that these patients had a significantly higher unadjusted risk of 30-day mortality and 30-day myocardial infarction. Patients transfused to a hematocrit of 30% or greater were found to have a significantly higher risk of 30-day mortality.

In 2009, a task force consisting of representatives from the Society of Critical Care Medicine, the American College of Critical Care Medicine, and the Eastern Association for Surgery for Trauma developed clinical practice guidelines for the transfusion of red blood cells in adult trauma and critically ill patients.39 Their recommendations included the following:

1. Transfusion of red blood cells in patients with hemorrhagic shock (level 1)

2. Red blood cells may be administered to patients with acute hemorrhage, hemodynamic instability, or a low SVO2 (level 1)

3. Avoidance of the use of a numerical hemoglobin threshold for transfusion: Transfusion should be based on the patient's intravascular volume, hemodynamic parameters and cardiopulmonary status (level 2)

4. Red blood cells treatment with single units unless the patient is actively bleeding (level 2)

5. Avoidance of transfusion in patients with acute lung injury (ALI) or acute respiratory distress syndrome (ARDS) because transfusions can exacerbate ALI and ARDS (level 2)

6. The analysis concluded that there is no benefit in transfusing stable patients with moderate to severe traumatic brain injury who have a hemoglobin less than 10g/dL (level 2)

More clinical studies are still needed to clarify the roles for transfusion in patients with ACS and subarachnoid bleeding.

Coagulopathies are another common problem in postoperative ICU patients and include dilutional coagulopathy, heparin-induced thrombocytopenia (HIT), and disseminated intravascular coagulation (DIC). Dilutional coagulopathy may complicate massive transfusion in patients with large-volume blood loss. The transfusion of packed red blood cells may result in dilution of clotting factors, platelets, and consequent prolongation of the prothrombin and partial thromboplastin times. Clinical findings include diffuse oozing of blood from mucosal and serosal surfaces, as well as from wounds and vascular access sites.40 A study by Cosgriff41 described several risk factors for severe coagulopathy. In a 2-year prospective study that evaluated patients who received more than 10 units of transfusion of packed red blood cells, a multiple logistic regression analysis revealed 4 significant risk factors: (1) blood pH higher than 7.10; (2) rectal temperature lower than 34°C; (3) injury severity score higher than 25; and (4) systolic BP lower than 70 mm Hg. Patients with no risk factors had a 1% chance of a life-threatening coagulopathy, whereas patients with 1 risk factor had a 10% to 40% chance, and patients with all 4 risk factors had a 100% incidence of a life-threatening coagulopathy.

DIC is a consumptive coagulopathy that can occur in the postoperative ICU setting. Triggers for DIC include sepsis, bone marrow and fat emboli, amniotic fluid emboli, and brain tissue embolization after traumatic injury. These tissues contain hematologically active factors and thromboplastins that trigger the clotting cascade and subsequent consumption of clotting and anticlotting factors. There is a resulting inhibition of localized clot formation and simultaneous intravascular thromboembolism. DIC is treated by addressing the precipitating problem (eg, treating sepsis when possible). Replacement blood components may be administered if the patient is actively bleeding.

Hemodilutional thrombocytopenia resulting from crystalloid administration paradoxically causes increased coagulation, as measured by thromboelastogram.42-44 The mechanism responsible for this hypercoagulability is not known.

Thrombocytopenia is common in critically ill patients. Although sepsis and hemodilution are the most common etiologies, HIT is a relatively unusual but significant platelet-based problem. There are 2 types of HIT. Type 1 has an incidence of 10% to 20% in heparin-treated patients, a nonimmune mechanism, and is not associated with thrombosis. Type 2 HIT has a 30% to 80% incidence and is an autoimmune-mediated thrombocytopenia that is associated with thrombosis. Treatment for HIT mandates the discontinuation of all forms of heparin and treatment with alternative anticoagulants such as argatroban or lepirudin.

Therapeutic anticoagulation with oral agents (as for the patient with atrial fibrillation) is typically interrupted during the perioperative period, when the risks of bleeding outweigh those of clotting. Patients at high risk for thrombosis or embolism may be started on IV heparin concurrently with discontinuation of oral anticoagulation. Heparin is then discontinued during the immediate operative and perioperative period and resumed as soon as the risk of bleeding from operative sites has abated, typically 12 hours following surgery. Oral anticoagulation can be resumed at the same time and heparin is discontinued when the prothrombin time international normalized ratio has reached the desired target level.

Renal Issues in the Postoperative ICU Patient

Acute renal failure (ARF) can increase the morbidity, mortality, and length of hospital stay in affected patients. In critically ill patients, the mortality associated with ARF ranges from 23% to 64% depending on the criteria used to define ARF. The incidence can range from 17.2%45 to 24.7%46 in critically ill patients. There are many precipitants for ARF in ICU patients, but in postoperative patients the most common are sepsis, ischemic acute tubular necrosis, drug-induced acute tubular necrosis, and pigment nephropathy, all of which are subgroups of acute tubular necrosis (ATN). ATN results from the injury and death of tubular epithelial cells that then slough off and obstruct the tubule, prompting the renal vasculature to constrict.

Common causes of ischemic ATN are cardiac arrest, hypotension from shock, and hypovolemia. The latter can occur postoperatively after large abdominal procedures with third space losses or vascular procedures with large fluid shifts. The treatment of ischemic ATN is to improve renal perfusion by increasing mean arterial BP with fluid administration for hypovolemic patients, the addition of vasoactive pressors to increase vascular tone in shock patients, or inotropes to improve cardiac function in patients with cardiac insufficiency.

• Drug-induced ATN is commonly caused by radiocontrast agents. Although the exact mechanism of injury is unknown, it is believed that contrast induces the production of free radicals that injure the nephron. Several studies have looked at ways to prevent ATN from radiocontrast agents. Tepel et al in 200047 showed that oral administration (pretreatment) of the antioxidant n-acetylcysteine actually decreased the serum creatinine in patients with chronic renal insufficiency. A 2004 study by Merten et al48 showed that treatment with sodium bicarbonate before and after exposure to IV radiocontrast decreased the incidence of contrast-induced nephropathy when compared with a placebo infusion. Merten's hypothesis was that free radical production is increased in an acidic environment and that by raising the pH, the production of free radicals would be decreased. Other treatments such as mannitol and furosemide have been shown to be ineffective in preventing ATN.
• Dopamine has also been used to treat various causes of acute renal failure. Extensive research has shown that dopamine increases renal blood flow, glomerular filtration, and urine output. The rationale for the use of "renal dose" dopamine (1-3 μg/kg/min) stems from the hypothesis that increasing blood flow results in improved oxygen delivery to hypoxic areas of the kidney to help treat or prevent ATN. In 2001, Kellum and Decker48 published a meta-analysis to determine whether low-dose dopamine decreased the incidence of acute renal failure, the need for dialysis, or the mortality in critically ill patients. They showed that dopamine did not significantly decrease the risk of mortality (relative risk [RR]: 0.83 [0.39-1.77]), development of acute renal failure (RR: 0.79 [0.54-1.13]), or the requirement for dialysis (RR: 0.89 [0.66-1.21]).
• Patients dependent on renal replacement therapy are defined as having acute renal failure or end-stage renal disease (ESRD). The ICU mortality of ESRD ranges from 11% to 40% in various studies.40,50 It is interesting to note the ICU mortality associated with ESRD is less than the mortality associated with ARF. In a study by Clermont et al, ICU mortality was 5% for patients with no renal failure, 20.4% for patients with ARF without renal replacement therapy (RRT), 57% for patients with ARF requiring RRT, and 11% for patients with ESRD.51
• There is no consensus regarding the appropriate timing for initiation of RRT in the ICU. There are, however, some common indications for starting RRT: (1) excessive intravascular volume in a patient with ventilatory or hemodynamic compromise, (2) electrolyte abnormalities (ie, hyperkalemia), (3) metabolic acidosis, (4) hyperuremia, and (5) treatment for an overdose of a dialyzable toxin or drug. The dialysis techniques used most frequently in the ICU are intermittent hemodialysis (IHD) and continuous venovenous hemodialysis (CVVHD). Continuous arteriovenous dialysis and peritoneal dialysis were once prevalent but are no longer standard therapies. There are conflicting opinions concerning the comparative survival advantage of CVVHD versus IHD in acutely ill patients, and studies have shown no definitive advantage for either technique in critically ill patients. A retrospective study by Gangji, in 2005,52 demonstrated that the use of CVVHD was associated with decreased mortality in the subgroup of patients with multiple organ dysfunction syndrome.

Endocrine Issues in the Postoperative ICU Patient

Postoperative patients have a complex endocrine response to surgical stress. The normal sympathetic response to surgery results in the release of epinephrine, glucagon, and cortisol to help repair injured tissue and fight off infection. However, critically ill postoperative patients often have an abnormal response to stress that leads to increased morbidity and mortality in the ICU. Tight glucose control and steroid replacement therapy have been used in the ICU to help decrease morbidity and mortality.

In 2001, van Den Berghe et al53 studied the effects of intensive insulin therapy on critically ill postsurgical patients. The authors hypothesized that hyperglycemia during critical illness would increase the risk of severe infections, multiple organ failure, and death. They found that by maintaining glucose levels between 80 and 100 mg/dL, the risk of mortality in the ICU decreased by 32%. This observed risk reduction occurred in patients who stayed longer than 5 days in the ICU. A large subsequent meta-analysis, however, showed that "tight glucose control is not associated with significantly reduced hospital mortality but is associated with an increased risk of hypoglycemia."

Corticosteroid replacement therapy was first studied in 1952 when Fraser et al54 described a patient with perioperative shock secondary to adrenal insufficiency. Cortisol is an endogenous corticosteroid that is produced by the adrenal glands. It is integral to the maintenance of vascular tone, vascular integrity, distribution of total body water, glucose metabolism, electrolyte homeostasis, catecholamine production and immunity, and many others In healthy patients during nonstress periods, plasma cortisol levels follow a circadian rhythm, increasing in the early morning and decreasing in the evening. After an operation, cortisol levels increase and the circadian rhythm disappears. Plasma cortisol concentrations normally reach their highest level during periods of severe stress (following burns, trauma, and sepsis).

When cortisol production is insufficient (eg, in patients suffering from Addison disease), the body appears to develop signs of shock (decreased SVR, decreased myocardial contractility, and decreased cardiac output) under the stresses of illness, surgery, or injury. In the postsurgical patient, as well as the critically ill patient, the incidence of total adrenal insufficiency is rare (2%-3%).55 However, functional or relative adrenal insufficiency is common with an incidence of 30%.56 The serum cortisol level fails to increase appropriately during stress states in patients with relative adrenal insufficiency. Steroid supplementation may be necessary for patients with functional adrenal insufficiency who develop shock that does not respond to volume resuscitation or vasopressors. A corticotrophin stimulation test can be used to determine the adrenal reserve, in which 250 μg of cosyntropin, a synthetic derivative of adrenocorticotrophic hormone, is given IV. Cortisol levels are then drawn at t0, t30, and t60 minutes. If the cortisol level 1 hour following "stimulation" is 9 μg/dL or less, the patient is considered a nonresponder.

Annane et al57 published data showing that nonresponders have a decreased risk of death when given 50 mg of hydrocortisone every 6 hours and 50 μg of fludrocortisone daily for 7 days. However, in 2008, Sprung et al published a multicenter randomized, double-blinded, placebo-controlled trial to evaluate the safety and efficacy of hydrocortisone in patients who had a response to cosyntropin (The CORTICUS58 [Corticosteroid Therapy of Septic Shock] trial). A total of 499 septic shock patients were randomized to receive 50 mg of hydrocortisone or a placebo. The study showed that there was no difference in mortality at 28 days between the 2 study groups or between cosyntropin responders and nonresponders. There was a subpopulation of patients with persistent hypotension unresponsive to vasopressors and fluids for whom hydrocortisone treatment was associated with decreased mortality.

Vasopressin, an endogenous endocrine hormone, is frequently used as an adjunct to catecholamines in the treatment of shock due to the evidence of its relative deficiency in septic patients.59 The Vasopressin and Septic Shock trial (VASST)61 was a multicenter randomized, double-blinded trial designed to compare low-dose vasopressin with norepinephrine to see if vasopressin conferred a mortality benefit in septic shock patients. A total of 778 septic shock patients were randomized to receive either low-dose vasopressin or norepinephrine titrated to a mean arterial pressure of 65 to 75 mm Hg. The study showed that there was no significant difference in 28- or 90-day mortality in either group. However, in patients with less severe septic shock (requiring 5-14 μg/kg/min of norepinephrine or its equivalent before randomization), the 28-day mortality was lower in the group treated with vasopressin.

Nosocomial infections are major sources of morbidity and mortality in the ICU. Hospital-acquired infections result in increased ICU length of stay and health care costs. In a report from the Centers for Disease Control and Prevention (CDC) on nosocomial infection control, a third of all such complications may be preventable through rigorous infection control practices.60 As this data has emerged, a series of best practice infection control measures have been developed including handwashing before and after examining patients, periodic surveillance of patients for antibiotic resistant organisms, co-location of patients infected with or colonized with antibiotic resistant organisms, and gowning and gloving when caring for patients with antibiotic-resistant organisms. In addition, the CDC recommends wearing a sterile gown, gloves and fully draping a patient when placing any invasive monitoring devices. Finally, the development of institutional antibiotic guidelines is important, and the guidelines will vary depending on the bacterial flora of a given institution.

The handwashing and barrier guidelines are designed to decrease the horizontal transmission of infections from health care worker to patient. Antibiotic guidelines are designed to help physicians choose the appropriate antibiotic therapy and length of therapy for a particular infectious disease and thereby decrease the inappropriate use of antibiotics and decrease the risk of developing antibiotic-resistant organisms. Other prophylactic measures that help decrease nosocomial infection rates relate to central line placement and maintenance. The following measures have been shown to decrease catheter-related bloodstream infections: full barrier precautions during central line insertions, preparation of the insertion site with chlorhexidine,62 and the use of single-lumen catheters rather than multilumen catheters when possible.63

Stress-related gastrointestinal mucosal damage occurs in many critically ill patients and may develop within 24 hours of ICU admission The incidence of clinically significant gastrointestinal bleeding with hemodynamic changes and a decreasing Hgb level due to stress ulceration has been estimated to be 1.5% in the ICU population. ICU stays are prolonged by an average of 8 days due to clinically significant bleeding, and mortality is increased 4-fold. Gastric hypoperfusion plays a major role in the pathogenesis of stress ulcers. Reduced gastric blood flow impairs the ability of gastric mucosal cells to regenerate and the consequent development of ulcers in the acid milieu of the stomach, but whereas gastric acidity may exacerbate stress ulcers, it does not cause them. Risk factors for stress ulcers in critically ill patients include major trauma, burns, respiratory failure, coagulopathy, hypotension, sepsis, hepatic failure, renal failure, and surgery. Prophylaxis is the preferred option for ICU patients at risk for stress ulcers. Therapies that have been shown to decrease the risk of stress ulceration include antacids, sucralfate, H2 blockers, and proton pump inhibitors. The goal of the therapies that alter gastric pH is maintenance of a gastric pH of 4.0 or greater,64 whereas sucralfate binds to and protects ulcerated gastric tissue in the presence of acid and should not be used in combination with antacids.

Deep venous thrombosis (DVT) is a common complication in hospitalized patients, especially the critically ill. The risk of developing DVT in general surgery patients is 15% to 40%. Patients who have sustained major trauma, spinal cord injury, or who are critically ill have a risk as high as 80%. DVTs occur more frequently with increasing age, with an incidence of 200 per 100000 in persons older than 70 years.65 Clots formed in the deep venous system of the lower extremities may break away and migrate into the venous circuit, so prevention of DVTs decreases the risk of fatal pulmonary embolism. Prophylactic treatment also prevents the long-term sequelae of DVT including leg swelling, dermatitis, leg ulcers, and a decreased quality of life. Therapy for the prevention of DVT varies and has been stratified based on risk of acquiring a DVT. Table 79-9 describes the thrombotic risk stratification for surgical patients and evidence-based guidelines for the use of antithrombotic prophylaxis.

Intensive care is a rapidly evolving specialty, particularly in the perioperative population. As new surgical procedures are developed, and as anesthesiologists are increasingly able to safely anesthetize patients at high risk such as the elderly, morbidly obese, and those with respiratory disease or severe heart failure, the demand for critical care services has escalated. Critical care specialists include anesthesia and surgery-based intensivists focusing on the care of patients in SICUs. Areas of subspecialization have emerged such as cardiothoracic critical care, neurocritical care transplant, burns, and trauma. As the promise of, and the demand for critical care services have increased, so has its expense. Modern intensivists must act as responsible stewards of these life-supporting, and sometimes merely life-prolonging, but highly expensive and in-demand ICU resources.

• Annane D, Sebille V, Charpentier C, et al. Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock.[see comment]. JAMA. 2002;288(7):862-871.
• Clearmont G. Renal failure in the ICU: comparison of the impact of acute renal failure and end-stage renal disease on ICU outcomes. Kidney Int. 2002;62:986-996.
• Hébert PC, Wells G, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. N Engl J Med. 1999;340(6):409-417.
• Hébert PC, Yetisir E, et al. Is low transfusion threshold safe in critically ill patients with cardiovascular diseases? Crit Care Med. 2001;29(2): 227-234.
• Mirski MA, Varelas PA. Diagnosis and treatment of seizures in the adult intensive care unit. Cont Crit Care. 2003;1(4):1-12.
• Society of Critical Care Medicine Ethics Committee. Consensus statement on the triage of critically ill patients. JAMA. 1994;271(15):1200-1203.
• van den Berghe G, Wouters P, et al. Intensive insulin therapy in the critically ill patients. N Engl J Med. 2001;345:1359-1367.