Adult Chest Surgery

Chapter 6. Critical Care

The thoracic surgery patient population can present significant challenges to care. In general, thoracic surgery patients are older, most are current or former smokers, they are sicker than the general surgical population, and most have a higher rate of comorbidity. It is not uncommon for these patients to present with underlying chronic lung disease, some form of arteriovascular disease, hypertension, diabetes, and baseline renal insufficiency. Thoracic surgery patients also have diminished physiologic reserve and more limited ability to recover from perioperative complications. They are particularly prone to pulmonary complications, which are very poorly tolerated in this patient group. As a result, thoracic surgery patients, more than any other group, require the services of an intensive care unit (ICU) and its highly trained, specialized staff. This chapter reviews critical-care issues specific to thoracic surgery patients, general issues of managing sepsis and acute respiratory distress syndrome (ARDS), and strategies to avoid the common nosocomial complications of critical care.


Secretion Retention, Atelectasis, Pneumonia

Thoracic surgery patients are at risk of secretion retention and atelectasis. General anesthesia, particularly when accompanied by one-lung ventilation, causes a marked decrease in functional residual capacity, promoting atelectasis. Surgical manipulation of the lung can lead to retained blood and secretions with partial or complete airway obstruction. Gas flow is further hindered by bronchospasm and decreased compliance of the operative lung. Splinting from postoperative pain or, conversely, respiratory depression from opiates or benzodiazepines further limits lung expansion. Patients with preexisting chronic obstructive pulmonary disease (COPD), asthma, bronchitis, or pneumonia will be at greatest risk. Similarly, patients with impaired cough reflexes, including those who have had an airway resection with anastomosis (e.g., sleeve resection), would be expected to have greater difficulty clearing secretions. Over time, retained secretions give rise to hypoxemia and hypercarbia once sufficient functional lung volume has been lost. They also predispose patients to pneumonia.

Prevention of secretion retention and atelectasis requires a systematic, multidisciplinary approach. The duration of general anesthesia should be limited to the minimum time required to complete the procedure. Patients should be extubated immediately whenever possible. Fiberoptic bronchoscopy performed immediately before extubation facilitates the removal of blood and secretions from the proximal airways. Excellent analgesia combined with aggressive early ambulation will promote recruitment of lung volumes and clearance of secretions. Chest physiotherapy further aids this process. Any patient with a preoperative pulmonary infection should undergo aggressive, culture-directed antibiotic therapy during the immediate perioperative period.

Treatment of retained secretions and atelectasis includes aggressive chest physiotherapy and mobilization. Humidified oxygen, nebulized saline, and bronchodilators can help to thin secretions and promote gas flow. Patients with copious thick secretions may benefit from nebulized N-acetylcysteine or dornase (DNAse), with bronchodilator pretreatment to mitigate treatment-induced bronchospasm. Any patient having significant trouble clearing secretions should be evaluated for the possibility of vocal cord dysfunction, which is a known complication of certain thoracic surgical procedures and results in a markedly impaired cough. A small subset of patients may require more aggressive interventions, including repeated awake fiberoptic bronchoscopy, intermittent noninvasive ventilation, assisted cough using an in-exsufflator device, or in the most severe cases, intubation and mechanical ventilation.

Postpneumonectomy Pulmonary Edema

In general, 2–4% of pneumonectomy patients will experience early-onset idiopathic acute lung injury. It is characterized by the development of diffuse infiltrates followed by significant hypoxemia in the first 1–3 days postoperatively. In contrast to late-onset acute lung injury, no etiology is readily apparent. Pulmonary capillary wedge pressures are normal to low, and the alveolar fluid has a high protein content. Studies using radiolabeled albumin have been consistent with a pulmonary capillary leak syndrome.

The relative rarity of this event precludes prospective study. Retrospective analyses of acute lung injury have identified multiple factors associated with the development of this syndrome (Table 6-1). The exact etiology of the syndrome remains unknown. It has been postulated that ventilator-induced lung injury, oxygen toxicity, tissue injury with cytokine release, loss of lymphatic drainage, pulmonary hypertension, and shear injury to the capillary endothelium from increased blood flow through the remaining pulmonary artery all contribute to the development of capillary leak and the subsequent accrual of interstitial lung water. Once capillary leak develops, movement of fluid from the pulmonary vasculature to the interstitium is governed by hydrostatic and colloid osmotic pressures, as describe by Starling's law. Few recommendations can be made regarding strategies to prevent this syndrome, but it would seem prudent to use lung-sparing ventilation with lower tidal volumes and plateau pressures, as well as to limit intravascular fluid to the minimum needed to support end-organ perfusion. Likewise, hypercarbia, hypoxia, and pain should be avoided because of their propensity to increase pulmonary arterial pressures. Once established, postpneumonectomy pulmonary edema is treated in the same manner as any other case of acute lung injury or ARDS with lung-sparing ventilation and good supportive care.

Table 6-1. Factors Associated with Postpneumonectomy Pulmonary Edema

·   Right-sided pneumonectomy

·   High intraoperative ventilatory pressures

·   High intraoperative fluid administration

·   Duration of surgery

·   Fresh-frozen plasma administration

·   Advanced age

·   Prior chest irradiation



Atrial Arrhythmias

Atrial fibrillation and atrial flutter are common after thoracic surgery. The incidence can be as high as 35–40% after extrapleural pneumonectomy (see Chap. 103). A number of factors have been associated with the occurrence of atrial arrhythmias (Table 6-2). Etiologies may include myocardial ischemia, pulmonary hypertension, atrial enlargement, hypoxemia, electrolyte imbalances, mechanical displacement of the heart, and vagal nerve irritation. Trials in the cardiac surgery population have shown that prophylaxis with beta blockers, calcium channel blockers, sotalol, or amiodarone may decrease the incidence of atrial fibrillation. There is a general paucity of trials specific to lung resection patients. Hemodynamically stable patients with atrial arrhythmias can be treated with beta blockers or calcium channel blockers to achieve rate control. Amiodarone is a useful agent for patients who maintain a high ventricular response rate despite maximum therapy with other agents or who become hypotensive with first-line agents. Amiodarone can cause primary acute lung toxicity in patients who have undergone thoracic resections and therefore should be reserved as a second-line therapy. Unstable patients may require electrical cardioversion to restore sinus rhythm. Unfortunately, the factors that lead to atrial arrhythmia usually are still present, and recurrence of the arrhythmia is common after either electrical or initial chemical cardioversion. Of note, most patients with new-onset atrial arrhythmia after surgery will return to sinus rhythm within 6 weeks of their surgery. This makes rate control, with anticoagulation, if not otherwise contraindicated, a practical approach in this patient population.

Table 6-2. Factors Associated with Atrial Arrhythmias

·   Underlying cardiac disease

·   Advanced age

·   Electrolyte abnormalities

·   Increased pulmonary vascular resistance

·   Volume of lung resected

·   Intrapericardial dissection

·   Postoperative pulmonary edema



Bronchospasm and COPD Exacerbation

COPD is a common comorbidity in the thoracic surgery patient population. Intubation and airway manipulation can exacerbate the symptoms of COPD. Increased resistance to airflow increases the work of breathing and may result in frank respiratory distress. Associated intrinsic positive end-expiratory pressure (auto-PEEP) or dynamic hyperinflation, resulting from trapped alveolar gas, further increases the work of breathing, adversely affects gas exchange, and causes hemodynamic instability if venous return is impaired. Patients with COPD should be maintained on their home bronchodilator and inhaled steroid regimen throughout the perioperative period. The rare patient will require systemic steroids to treat the acute exacerbation of COPD. Early extubation is critical to avoid the cycle of airway irritability, bronchospasm, and respiratory distress.


Postoperative hypotension can be broken down into problems of pump function or venous return. Myocardial ischemia is a cause of acute ventricular dysfunction. Likewise, acute onset or worsening of pulmonary hypertension can lead to right ventricular failure (see later). Venous return problems that limit effective diastolic filling of the heart are more common. Dehydration or acute hemorrhage leads to absolute hypovolemia. Tension pneumothorax, pericardial tamponade, and pulmonary embolism, on the other hand, limit venous return despite normal intravascular volume. Likewise, mediastinal shift or cardiac herniation, both of which may occur after pneumonectomy, will acutely impair the venous return despite normal intravascular volume. Cardiac outflow tract obstruction also may occur. Finally, some patients will exhibit hypotension with a normal cardiac output and a low systemic vascular resistance. The increase in venous capacitance creates a state of relative hypovolemia. This low-tone state is typical of sepsis but also may be seen with sympathectomies that are either pharmacologically induced from local anesthetics given via a thoracic epidural or even mechanically induced from trauma to the sympathetic chain during surgery. Central venous pressure monitoring can be very helpful in distinguishing between the various causes of postoperative hypotension. Rarely, placement of a pulmonary artery catheter may become necessary.

Pulmonary Hypertension and Right Ventricular Failure

Acute onset of pulmonary hypertension with subsequent right ventricular failure is one of the most dreaded perioperative complications of thoracic surgery. Unless it is recognized immediately and managed successfully, it can be rapidly fatal. Preexisting COPD, with its intrinsic loss of pulmonary microvasculature, limits the remaining lung's ability to compensate for an abrupt increase in pulmonary blood flow after pulmonary resection and predisposes the patient to acute perioperative pulmonary hypertension. Preoperative evaluation, including echocardiography and cardiac catheterization, can help to identify patients with significant preoperative pulmonary hypertension. It should be understood, however, that these tests are done at rest and sometimes under conscious sedation and therefore may not accurately predict the pulmonary artery pressures that will occur postoperatively under the conditions of stress and hypermetabolism. Even a reassuring preoperative catheterization with a balloon occlusion trial does not completely rule out the possibility of postoperative right ventricular failure.

Patients with acute right ventricular failure often present with hypotension, evidence of low cardiac output (including oliguria), a high central venous pressure, and peripheral edema. Patients also may complain of dyspnea and lightheadedness, particularly with exertion. Unlike patients with biventricular failure, there is generally an absence of pulmonary edema, and left atrial pressures are low. Right ventricular dysfunction can occur in the setting of right ventricular pressure overload, volume overload, or impaired contractility. An electrocardiogram and echocardiogram will help to exclude right ventricular ischemia or infarction, which, if present, requires urgent therapy for acute coronary syndrome. Likewise, the patient should be assessed for the likelihood of an acute pulmonary embolism because this can cause pulmonary hypertension and right ventricular failure and requires very specific therapy for treatment.

Initial treatment of right ventricular failure is aimed at correcting reversible causes of pulmonary hypertension, including hypoxia, hypercarbia, and acidosis. Pain and fever, because of their associated hypermetabolic state, also can cause pulmonary hypertension and should be managed rapidly. Volume management can be complex. Cardiac output is preload-dependent, but right-sided volume overload can adversely affect left ventricular output through septal shift and intraventricular dependence. The combination of echocardiography and pulmonary artery catheter measurements can be helpful in identifying the optimal volume status for each patient. Right ventricular function and systemic blood pressure can be supported with the use of inotropes (e.g., dopamine, epinephrine, and norepinephrine) and inodilators (e.g., dobutamine and milrinone). A pulmonary artery catheter is helpful when titrating therapy. If an inodilator results in improved cardiac output but worsened systemic hypotension, vasopressin may be used to maintain systemic blood pressure without increasing pulmonary artery pressures. Unfortunately, the inotropes and inodilators are all arrhythmogenic and also increase myocardial oxygen consumption.

Pulmonary vasodilators reduce pulmonary vascular resistance and increase the right ventricular cardiac output without causing arrhythmias or increasing myocardial oxygen consumption. Unfortunately, the common vasodilators, including nitroglycerine, sodium nitroprusside, and hydralazine, usually result in significant systemic hypotension. Inhaled nitric oxide can decrease pulmonary vascular resistance and does not cause systemic hypotension. Likewise, there are case reports of inhaled prostacyclin being used successfully to treat acute pulmonary hypertension. Intravenous prostacyclin, sildenafil, and bosentan may be useful in rare patients.

Massive Hemoptysis

Massive hemoptysis, defined as more than 600 mL of blood loss in 24 hours, is uncommon but can be rapidly fatal. It is associated with pulmonary infection, bronchiectasis, tumor erosion into a pulmonary or bronchial artery, and pulmonary artery rupture during pulmonary artery catheterization. It also can result from traumatic injury to the chest or as a result of a tracheoarterial fistula. The patient with massive hemoptysis is at imminent risk of death from asphyxiation. He or she should be turned bleeding side down to protect the good lung and intubated early rather than late. A bronchial blocker can be used to isolate the site of bleeding. While double-lumen endotracheal tubes do permit lung isolation, difficult positioning plus very small internal lumens make them problematic in this setting. Once the patient is stabilized, focus should shift to determining the source of bleeding. Bronchoscopy, computed tomographic (CT) scanning, and angiography all may be useful. Further management is dictated by the specific cause of the hemoptysis.

Bronchopleural Fistulas

Breakdown of an airway anastomosis or stump can result in formation of a large proximal bronchopleural fistula (BPF). Alveolar rupture, persistent leak at a resection margin, or traumatic injury to the pulmonary parenchyma will create a more distal air leak. Empyema, tumor recurrence, irradiation, and poor wound healing all contribute to BPF formation. Early aggressive treatment of infection is critical. Maximizing the nutritional status is a priority. Every effort should be made to keep these patients ambulatory and breathing without mechanical assistance. If a patient requires mechanical ventilation, volume loss through the BPF can be quantified by comparing the inspiratory and expiratory tidal volumes recorded on the time-versus-volume loop of the ventilator graphics. Patients with significant volume loss are easier to ventilate using pressure modes of ventilation. If the patient is awake, pressure support can be used, provided that the ventilator mode permits adjustment of the expiratory flow cutoff. Failure to increase the expiratory flow cutoff above the traditional 25% may result in a sustained inspiration and subsequent ventilator dyssynchrony. No matter what ventilator mode is used, inspiratory pressure and volume should be minimized to limit the airflow across the BFP. In some cases, lung isolation may be required to permit adequate ventilation. This is particularly true of a BPF occurring after pneumonectomy, particularly if the remaining lung has been compromised. In this setting, double-lumen endotracheal tubes, endobronchial tubes, or specially made tracheotomy tubes that are custom measured to sit below the carina will permit exclusion of the BPF. Unfortunately, it is difficult to maintain any of these tubes in constant position, and the emergent need to reposition these tubes occurs with some frequency.


Meticulous supportive care is the mainstay of treatment for patients with ARDS. It is vitally important that mechanical ventilation be managed with a view toward preventing ventilator-induced lung injury, which occurs with both alveolar overdistention and repetitive opening and closing of alveoli. In 2000, the Acute Respiratory Distress Syndrome Network (ARDSnet) published a trial showing that low-tidal-volume ventilation (6 mL/kg of ideal body weight), with a plateau pressure of 30 cm H2O or less, resulted in an 8% decrease in absolute mortality.1 This practice has become the standard of care for patients with or at risk of developing ARDS. Ideally, the level of positive end-expiratory pressure (PEEP) should be set to prevent cyclic alveolar collapse at end expiration. Unfortunately, there is no clear agreement on the method one should use to determine optimal PEEP in routine clinical practice. A pragmatic approach is to titrate the PEEP upward until maximum pulmonary compliance and best oxygenation are achieved. One must be cognizant of the fact that increased intrathoracic pressures will decrease venous return and may adversely affect cardiac output and arterial blood pressure.

During the late 1970s and early 1980s, a number of trials were undertaken to examine the benefit of early high-dose corticosteroids in early ARDS. These were largely negative studies, with some actually reporting increased mortality. For years thereafter, steroids were considered absolutely contraindicated. The steroid debate reemerged in 1998 with the publication of a very small single-center trial of prolonged methylprednisolone for late-stage ARDS.2 This trial reported a significant decrease in mortality in the treatment group but was plagued by serious methodologic problems in design and execution. Finally, in 2006, the ARDSnet Trials Group published the results of a large, multicenter randomized trial of steroids for late ARDS.3 In the initial 4 weeks of treatment, the treatment group did have more ventilator-free and shock-free days, as well as improved oxygenation and pulmonary compliance. There was no difference in mortality at 60 or 180 days. However, if methylprednisolone was started more than 14 days after the onset of ARDS, there was a significant increase in mortality at both 60 and 180 days. The only subgroup that appeared to benefit from treatment was the group that had elevated procollagen III levels in the alveolar fluid at the time of enrollment. Hence, at this time, corticosteroids cannot be recommended for the routine treatment of ARDS.

The ARDSnet Trials Group recently completed a complex but well-designed trial comparing a liberal versus restrictive fluid policy for patients with ARDS.4 The restrictive protocol targeted central venous pressures of less than or equal to 4 mm Hg, provided that the patient was not in shock and did not display signs or symptoms of end-organ hypoperfusion. The liberal protocol targeted pressures of 10–14 mm Hg. There was no difference in mortality, but the restrictive protocol resulted in earlier liberation from the ventilator and earlier discharge from the ICU without any increase in organ failure. Another arm of the study executed the same fluid interventions but used a pulmonary artery catheter rather than a central venous catheter. There was no advantage to pulmonary artery catheter use in ARDS. This study provides reassurance and support for the common clinical practice of moderate fluid restriction.


Pneumonia, line sepsis, empyema, and esophageal leak are among the most common causes of sepsis in the thoracic surgery population. Traditionally, mortality for septic shock has been upward of 40%. In 2001, Rivers and colleagues published a randomized, prospective clinical trial of early goal-directed therapy. They focused on protocol-driven resuscitation within the first 6 hours of presentation.5 Mortality was decreased from 46.5 to 30.5%. Also in 2001, Bernard and colleagues published the results of a randomized, prospective trial of activated protein C (APC). APC led to a 6% absolute reduction in mortality,6 making APC the first pharmacologic substance proved to have efficacy in sepsis.

In 2004, the Society of Critical Care Medicine, in conjunction with multiple other organizations, launched the Surviving Sepsis Campaign.7 This campaign integrates the findings of all recent high-quality sepsis trials and establishes guidelines for the initial treatment of sepsis. Care goals are bundled into the first 6 hours and the first 24 hours, reflecting increasing evidence that outcomes from sepsis depend on the timeliness of treatment. Within the first 6 hours, the focus should be on obtaining cultures, starting an appropriate broad-spectrum antibiotic regimen that takes into account any prior knowledge of likely source of pathogens, and executing rapid and effective fluid resuscitation. Specific recommendations for fluid resuscitation include the early placement of an arterial line and central venous line. Either crystalloids or colloids may be used to reach a central venous pressure target of 8–12 mm Hg. If the patient is mechanically ventilated, a higher goal of 12–15 mm Hg is appropriate. Mean arterial pressure should be maintained at levels greater than 65 mm Hg. If the target central venous pressure has been reached and the mean arterial pressure is still low, the patient should be started on vasopressors. Urine output (at least 0.5 mL/kg/h), trends in lactic acid levels, and the central venous oxygen saturation (SvcO2) are followed for evidence of tissue hypoperfusion. If the SvcO2 is less than 70% and the urine output remains low or lactic acid levels are climbing, the patient may be a candidate for inotropic therapy and/or transfusion to a hematocrit of 30%, with the goal of increasing cardiac output and oxygen delivery.5 At this time, norepinephrine and dopamine are considered the vasopressors of choice, and dobutamine is the inotrope of choice, although these recommendations may change with the completion of ongoing clinical trials focused on vasopressor choice in sepsis. Source control is vital to the successful treatment of sepsis, and any focal area of infection amenable to drainage or debridement should be attended to at the earliest possible moment.

A number of other important interventions should be initiated as soon as possible within the first 24 hours of presentation. All vasopressor-dependent patients should be evaluated for adrenal insufficiency. One may either treat all of these patients empirically with hydrocortisone (200–300 mg/day in divided doses) for 7 days or perform an adrenocorticotropic hormone challenge test and treat only those deemed to have adrenal insufficiency by established criteria.8 Patients at high risk of death, as defined by the presence of acute organ dysfunction in one or more organs and an APACHE II score of 25 or higher, benefit from APC. Patients who are less severely ill have not been shown to benefit and actually may have increased mortality with the administration of APC. The benefit depends on early administration. Unfortunately, the major complication of APC is bleeding, with a 3.5% rate of serious bleeding, including an increased incidence of intracranial hemorrhage. Patients who are fully anticoagulated, have had a stroke or intracranial procedure within 3 months, have a platelet count of less than 30,000/mm3, or have active bleeding should not receive APC. A full list of exclusion criteria can be found in the original paper reporting the use of APC.6 At my institution, APC is administered to qualifying patients who are at least 12 hours out from surgery, provided that there is no evidence of ongoing bleeding. The decision to administer APC in a surgical patient, however, must be carefully weighed against the risk of major bleeding. If the risk of bleeding is thought to be high enough that full systemic anticoagulation would be contraindicated, then APC probably should not be administered. Other important interventions in the first 24 hours include the institution of lung-sparing mechanical ventilation to minimize ventilator-induced lung injury and tight glucose control using an insulin infusion (see below). Finally, care should be taken to prevent secondary complications of critical care. Specific interventions are discussed below.


In many situations, a critically ill patient may survive his or her initial illness only to succumb to complications of his or her care. There are a number of well-recognized nosocomial complications common to ICU patients throughout the world. Fortunately, advancing medical knowledge has led to an increased understanding of many of the factors that contribute to these complications and has facilitated efforts aimed at prevention. Indeed, attention to detail and forward-thinking care that anticipates and moves to prevent complications is truly the mainstay of critical care. The most common and significant nosocomial complications are discussed below.


Deep venous thrombosis is initially a silent disease. When accompanied by pulmonary embolism, it can be quite morbid. This is particularly true in the thoracic surgery population secondary to both limited pulmonary reserve and a significant tendency toward hypercoagulability. Virtually all thoracic surgery patients should receive aggressive prophylaxis against deep venous thrombosis while in the hospital, including mechanical venous compression devices and pharmacologic therapy with either low-dose unfractionated heparin or low-molecular-weight heparin. Even with appropriate prophylaxis, deep venous thrombosis can occur, and it is necessary to maintain a high index of suspicion for thrombotic and thromboembolic events.

Stress Ulcers

Shallow gastric mucosal ulcerations, usually located in the proximal stomach, are common in ICU patients and develop rapidly after the onset of acute illness. In a small percentage of patients, the ulceration will progress into the submucosal layers and can result in bleeding or perforation. Histamine-2-receptor antagonists are the mainstay of prophylaxis in the ICU setting. They have been shown to effectively decrease the rate of clinical bleeding from stress ulcers. Proton pump inhibitors also have been used for this purpose. Unfortunately, both these classes of drugs result in alkalization of the stomach and overgrowth with gram-negative bacilli. There is a growing body of evidence that stress ulcer prophylaxis with pH-modifying drugs increases the rate of ventilator-associated pneumonia (VAP). Newer studies suggest that sucralfate, which does not change gastric pH, is nearly as effective for preventing bleeding episodes while resulting in fewer pneumonias.

Pressure Sores

Pressure sores remain a troublesome issue for ICU patients. Many patients have multiple risk factors, including immobility, heavy sedation, poor nutrition, impaired tissue perfusion, and frequent incontinence. Pressure sores are both morbid and costly to treat. Early identification of at-risk patients, combined with aggressive efforts at prevention, is needed. Preventive measures include urgently correcting nutritional deficits, restoring tissue perfusion, minimizing sedation, emphasizing early return to mobility even for ventilated patients, and for those who are immobile, redistributing body weight over bony prominences at least every 2 hours. Additionally, skin must be kept clean and dry, and great care must be taken to avoid shear injury to the skin during turns. High-risk patients may benefit from special low-air-loss therapy beds. Fortunately, it is now possible to get low-air-loss therapy beds with other key functions built into the design. Each center should review the beds or mattress overlays available to it and select the default bed most appropriate for use with high-risk patients.


Inadequate nutrition is an insidious problem that ultimately can lead to severe consequences. It contributes to wound breakdown and infection, the development of pressure sores, immune suppression, and generalized weakness, which, in turn, limits mobility and impairs weaning from mechanical ventilation. Additionally, malnourished patients tend to have lower colloid osmotic pressures and more trouble mobilizing edema fluid. Corticosteroid use, when necessary, will further compound muscle catabolism and amplify the systemic effects of malnutrition. It is vitally important to begin nutritional repletion as soon as possible in all ICU patients. Enteral feeding is strongly preferred. The decision to use an enteral route should be made on an individual patient basis, determining the best option for accessing the gastrointestinal system. At my institution, there is a strong bias in favor of a feeding jejunostomy for critically ill patients who are thought to be at high risk of aspiration or have very limited reserve to tolerate an aspiration event. At this time, there is a lack of good literature supporting this practice, but clinical experience to date at my institution would support it. Patients who are not candidates for enteral feeding in the short term should be supported with total parenteral nutrition. Since total parenteral nutrition is associated with significant complications, initiation of total parenteral nutrition should be done concomitant with reassessment of options to promote enteric function and allow for a rapid transition to enteric feeding. Critically ill patients should be reevaluated regularly to determine if their nutritional needs are being met.


Surgery and critical illness both result in a systemic stress response. Even nondiabetic patients will develop insulin resistance and hyperglycemia. Traditionally, blood glucose has been loosely controlled in the ICU setting, with a typical goal of maintaining the blood glucose concentration near 200 mg/dL. It was thought that tighter control was not necessary and might even be dangerous if it were to lead to unrecognized hypoglycemia. Research done in the 1990s in cardiac surgery patients, however, began to challenge the notion of loose glucose control, hinting at improved outcomes with lower glucose targets. van Den Berghe and colleagues published the results of a prospective randomized controlled study of glucose control in surgical patients in 2001.9 The control group had a target glucose level of 180–200 mg/dL, whereas the intervention group targeted a range of 80–110 mg/dL. Intensive insulin therapy was found to reduce mortality in patients who stayed in the ICU for more than 5 days. The intervention group also had fewer episodes of sepsis, renal failure, and critical illness polyneuropathy; a shorter duration of mechanical ventilation; and shorter ICU stays. These benefits accrued whether or not the patient had any form of diabetes at baseline.

At this time, all long-stay surgical ICU patients should have their blood glucose tightly controlled. This is best done with a continuous insulin infusion. Therapy should be initiated at the time of admission. Once the patient's condition improves, treatment may be transitioned to a combination of short- and long-acting insulin delivered by the subcutaneous route. The target blood glucose concentration, however, should remain low. At an institutional level, development of intravenous insulin protocols for institution-wide use facilitates bedside treatment. It is important to recognize that it is nearly impossible to predict ICU stays accurately for most patients. Therefore, to make sure that those who most benefit from tight glucose control receive the appropriate therapy from the time of admission, it is necessary to treat a number of patients who actually end up being discharged from the ICU in fewer than 5 days.

Nosocomial Bloodstream Infection

Catheter-related bloodstream infections are associated with an increased length of ICU stay, significant increases in cost of care, and significant attributable mortality. Recent research indicates that catheter-related bloodstream infections are nearly completely preventable. The Centers for Disease Control and Prevention reviewed the literature and developed practice-based guidelines with which all clinicians caring for patients with central lines should become familiar.10 Johns Hopkins conducted a clinical trial of system-based changes in the placement of central lines in the ICU.11 Their interventions included provider education, creation of a line-insertion cart, daily evaluation of each patient to see if lines could be removed, and a nursing-administered real-time check list to ensure proper aseptic technique during line insertion. Emphasis was simply on the universal use of appropriate aseptic technique, including hand hygiene, appropriate skin preparation, use of sterile gloves and gown for all operators, and full-body draping. These simple interventions netted a sustained 20-fold reduction in line-infection rates.

Ventilator-Associated Pneumonia

Many thoracic surgical patients who have complicated postoperative courses will require a period of mechanical ventilation. Up to 25% of ventilated patients develop VAP. VAP is associated with significant morbidity and mortality, and therefore, prevention of VAP should be a priority for all ICUs. A number of simple interventions can reduce the incidence of VAP. Patients should be nursed in the semirecumbent position with the head of the bed elevated at least 30 degrees at all times. Unless contraindicated, consideration should be given to the use of sucralfate rather than pH-modifying agents for stress ulcer prophylaxis. Subglottic suctioning via specially designed endotracheal tubes can be used for patients likely to be intubated for more than 72 hours. Oral care has been shown to have a significant impact on VAP rates, and each institution should develop a standardized policy for mouth care of intubated patients. Oral chlorhexidine applied twice daily has been shown to decrease VAP rates in postoperative cardiac surgery patients. Tight glucose control, as discussed earlier, is associated with lower VAP rates, and daily interruption of sedation minimizes overall time on the ventilator. Finally, the Centers for Disease Control and Prevention has published guidelines on the care of respiratory equipment, including humidifiers, nebulizers, and tonsil-tip suction devices. Institutional practices should be reviewed periodically to verify compliance with these guidelines to prevent such equipment from serving as a nidus of infection.

Delirium and Oversedation

For years, physicians and nurses have sedated mechanically ventilated patients to minimize the physiologic stress response and maximize patient comfort. While these are laudable goals, our traditional practices have resulted in several unexpected complications. Heavy sedation increases the incidence of delirium, slows weaning from the ventilator, increases the risk of pneumonia, and slows the time to return of ambulatory status. Delirium, in turn, is associated with increased length of stay and higher mortality. Implementation of a formalized delirium evaluation and a standardized sedation scale leads to early identification of patients needing treatment for their delirium and helps to prevent oversedation. The Confusion Assessment Method for the ICU is a delirium-assessment tool that is simple and rapidly applied. The bedside nurse should use it once per shift. Patients who are found to be delirious via the Confusion Assessment Method for the ICU tool need to have a comprehensive review of their treatment plan, with an emphasis placed on minimizing delirium-producing medications and environmental factors and treating their agitation with medications that are likely to improve delirium rather than worsen it. A number of sedation scores have been published. I find the Richmond Agitation-Sedation Scale to be easy to implement and very useful. Writing sedation orders to titrate to a target sedation score helps to minimize oversedation and clearly communicates the targeted sedation level. Previous research has indicated that daily interruption of sedative infusions speeds the time to liberation from mechanical ventilation. It is likely that sedation-scale-targeted sedation, with emphasis on the calm and only lightly sedated end of the scale, will make the practice of sedative interruption altogether unnecessary.

Renal Dysfunction

Acute renal failure is a common complication in critically ill patients. Unfortunately, it is also a highly morbid one, with reported mortality rates exceeding 50% in some trials. Prerenal azotemia refers to compromised renal function secondary to impaired renal perfusion without actual damage to the nephron. This sequela occurs in patients who are hypovolemic or have a severely decreased cardiac output. It is rapidly reversible provided that the underlying condition is recognized and corrected. Persistent renal hypoperfusion, such as occurs in shock states, can result in cellular damage to the nephron or glomerulus, thus converting prerenal azotemia into intrinsic renal failure. Administration of nephrotoxic agents can amplify the injury or even result in de novo injury to the nephron in the absence of renal ischemia. Finally, the clinician must remember to evaluate every patient with acute renal failure for the potential for postrenal obstructive processes.

Close attention should be paid to avoiding renal injury in critically ill patients. Nephrotoxic drugs should be avoided whenever possible. Consideration should be given to pretreatment with N-acetylcysteine or sodium bicarbonate before intravenous contrast material administration. Hypotension, hypovolemia, and hypoxia must be corrected rapidly. Patients with new onset of acute renal failure should have urine electrolytes and urine sediment evaluated. Central venous pressure monitoring may be helpful in detecting hypovolemia. An echocardiogram can provide a rapid assessment of overall cardiac function. If there is any concern for obstructive pathology, an immediate renal ultrasound should be performed. While positive findings are rare, they are often amenable to intervention. It is important to carefully review the patient's medication list, eliminating all potentially nephrotoxic agents and renally dosing the rest. Should renal replacement therapy be needed, most critically ill patients best tolerate a continuously administered replacement. It is important to appreciate that it may take a matter of weeks for renal function to recover, and it is impossible to predict renal recovery accurately.

Critical Illness Polyneuropathy and Acute Myopathy

Critical illness polyneuropathy occurs in up to 50% of patients who are septic for more than 2 weeks. The etiology remains unknown. Clinically, it presents as severe (distal worse than proximal) muscle weakness associated with difficulty weaning from mechanical ventilation and decreased deep tendon reflexes. Pathologically, it is a disease of axonal degeneration and denervation. Nerve conduction studies show normal conduction velocities with prominent denervation. Patients who survive their underlying illness will regain muscle strength over a period of weeks to months. Treatment and prevention focuses on avoiding compounding factors such as unnecessary use of corticosteroids or non-depolarizing neuromuscular blockers, both of which are thought to cause acute myopathy. If use of paralytic agents is unavoidable, every effort should be made to limit the duration of use to less than 48 hours. Beyond 48 hours, the incidence of myopathy becomes quite high. Blood glucose concentration should be tightly controlled with a target of 110 mg/dL or less. Malnutrition, heavy sedation, and prolonged immobility need to be avoided. At-risk patients should receive daily physical therapy with a goal of early resumption of mobility even in the setting of prolonged mechanical ventilation. Ventilated patients may be ambulated with the help of a walker and either an Ambu bag or portable mechanical ventilator.


1. ARDSnet: Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med 342:1301, 2000. 

2. Meduri GU, Headley AS, Golden E, et al: Effect of prolonged methylprednisolone therapy in unresolving acute respiratory distress syndrome: A randomized controlled trial. JAMA 280:159, 1998.[PubMed: 9669790]

3. Steinberg KP, Hudson LD, Goodman RB, et al: Efficacy and safety of corticosteroids for persistent acute respiratory distress syndrome. N Engl J Med 354:1671, 2006.[PubMed: 16625008]

4. Wiedemann HP, Wheeler AP, Bernard GR, et al: Comparison of two fluid-management strategies in acute lung injury. N Engl J Med 354:2564, 2006.[PubMed: 16714767]

5. Rivers E, Nguyen B, Havstad S, et al: Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 345:1368, 2001.[PubMed: 11794169]

6. Bernard GR, Vincent JL, Laterre PF, et al: Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 344:699, 2001.[PubMed: 11236773]

7. Dellinger RP, Carlet JM, Masur H, et al: Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock. Crit Care Med 32:858, 2004.[PubMed: 15090974]

8. 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. JAMA 288:862, 2002.[PubMed: 12186604]

9. van den Berghe G, Wouters P, Weekers F, et al: Intensive insulin therapy in the critically ill patients. N Engl J Med 345:1359, 2001. 

10. O'Grady NP, Alexander M, Dellinger EP, et al: Guidelines for the prevention of intravascular catheter-related infections. Centers for Disease Control and Prevention. MMWR Recomm Rep 51:1, 2002. 

11. Berenholtz SM, Pronovost PJ, Lipsett PA, et al: Eliminating catheter-related bloodstream infections in the intensive care unit. Crit Care Med 32:2014, 2004.[PubMed: 15483409]

If you find an error or have any questions, please email us at Thank you!