PNEUMONIA IN THE INTENSIVE CARE UNIT
Pneumonia is the leading cause of death from nosocomial infection in the United States. The incidence of nosocomial pneumonia in community hospitals and general medicine wards is about 10 cases per 1,000 admissions but is 10- to 20-fold higher in the ICU. The overall mortality rate of pneumonia in the ICU in one study was 50%. Factors such as ventilator-associated pneumonia and development of adult respiratory distress syndrome significantly worsen the prognosis.
In the normal host, various defense mechanisms, such as filtration of inspired air by the upper respiratory tract, intact cough reflex, secretion of mucus in the tracheobronchial tree, intact humoral immunity, and pulmonary macrophage clearing of bacteria, keep the lower respiratory tract sterile. When these defenses break down or are overwhelmed by a large inoculum of a virulent organism, pneumonia can occur. Compromise of host defenses is common in the typical ICU patient, who is often debilitated or traumatized. Indeed, stroke, seizure, and drug intoxication (common ICU admitting diagnoses) compromise epiglottic closure and cough, resulting in aspiration of oropharyngeal bacteria. Endotracheal intubation or tracheostomy bypasses upper respiratory filtration defenses. The duration of endotracheal intubation is also important as a risk factor in nosocomial pneumonia, having an estimated additive risk of 1% for each day of mechanical ventilation. Impaired pulmonary macrophage function occurs in hypoxemia, uremia, malnutrition, and heart failure. One study also implicates nasogastric tubes, upper abdominal and thoracic surgery, and bronchoscopy as independent risk factors.
In the ICU, bacteria may gain access to the lung by one of these mechanisms: (a) hematogenous spread (e.g., Staphylococcus aureus may cause pneumonia when an infected intravenous line results in bacteremia and pulmonary seeding); (b) aerosolization, as may occur during respiratory therapy, particularly with reservoir nebulizers (e.g., Pseudomonas aeruginosa pneumonia); and (c) aspiration of endogenous bacteria colonizing the oropharynx.
Whereas pneumonia acquired in the community is typically caused by pathogens such as Streptococcus pneumoniae, Haemophilus influenzae, Moraxella (Branhamella) catarrhalis, Legionella pneumophila, and Mycoplasma pneumoniae, pathogens causing pneumonia in the ICU are somewhat different. Most nosocomial, and particularly ICU-acquired, pneumonia is caused by gram-negative bacilli or S. aureus. These bacteria can be found colonizing the oropharynx of most ICU patients, and the prevalence of gram-negative colonization increases with severity of illness. Factors that have been shown to increase colonization include coma, hypotension, acidosis, azotemia, and endotracheal intubation.
The spectrum of microorganisms known to cause nosocomial pneumonia has increased rapidly to include several low-virulence organisms, such as Staphylococcus epidermidis, Corynebacterium, nontypeable H. influenzae, and M. catarrhalis. However, the etiologic agents responsible for pneumonia in the ICU have remained predominantly P. aeruginosa and other gram-negative bacilli, such as Klebsiella, Enterobacter, Escherichia coli, Proteus, and Serratia, with S. aureus a close second. Table 7-1 summarizes these and other agents that cause pneumonia in an intensive care setting.
Table 7-1. Frequency of etiologic agents in ICU pneumonia
There are some pathogens and circumstances worth special mention. S. pneumoniae, the common community-acquired pathogen, is known to cause pneumonia in the ICU occasionally, and superinfection with gram-negative bacilli is not uncommon. Severe community-acquired pneumococcal pneumonia may frequently result in ICU admission, particularly if the organism is penicillin-resistant and inappropriate therapy has been started. S. pneumoniae is by far the most common organism to cause severe, life-threatening pneumonia in elderly patients. Other streptococci, particularly group A and group B b-hemolytic streptococci, as well as enterococci (group D streptococci), can cause pneumonia in the ICU. Predisposing factors for these pathogens include preceding viral infections, coinfection with S. aureus, and previous treatment with broad-spectrum antimicrobials. H. influenzae, usually nontypeable, is an important respiratory pathogen in patients with chronic obstructive pulmonary disease and can lead to bronchopneumonia causing respiratory failure. M. catarrhalis, now clearly recognized as an important respiratory pathogen, also tends to occur in those with underlying lung disease, many of whom are receiving corticosteroids. Legionella species have also been implicated in hospital- and ICU-acquired pneumonia; spread of L. pneumophila by aerosol from hospital cooling or hot water systems can potentially make this a serious consideration, especially when sputum Gram’s stain and culture are unrevealing.
Hantavirus is an additional cause of life-threatening pneumonia, first described in 1993 when an outbreak was reported in New Mexico. Hantavirus pulmonary syndrome is characterized by a flulike illness followed by noncardiogenic pulmonary edema. Symptoms of fever, myalgia, nausea, and diarrhea are accompanied by tachycardia, tachypnea, and hypotension.
Finally, the nonbacterial opportunistic pathogens such as Candida albicans, Pneumocystis carinii, Aspergillus, Phycomycetes, herpes simplex virus, cytomegalovirus, and varicella-zoster virus occur in a selected population of patients (i.e., the immunosuppressed) and have been associated with poor outcome.
Pneumonia, especially in the debilitated, critically ill patient, can be complicated by several life-threatening conditions. For instance, when a response to appropriate antimicrobial therapy cannot be obtained, empyema should be considered. When pleural fluid is noted on physical examination or chest x-ray films in the patient with pneumonia, thoracentesis is necessary to rule out infection of the pleural space. Patients with a pleural fluid pH of 7 or less, a glucose level below 40 mg/dL, or gross pus in the pleural space usually require chest tube placement. The organism most prone to cause pleural space infection is S. aureus. However, gram-negative bacilli, anaerobes, H. influenzae, S. pneumoniae, and b-hemolytic streptococci can all produce empyema. Purulent pericarditis, a common complication of pneumococcal pneumonia in the era before antimicrobial therapy, is now rarely seen as a complication of pneumonia. When encountered, purulent pericarditis is most likely to be associated with S. aureus or gram-negative bacilli and to occur in the critically ill patient, particularly after thoracic surgery. The diagnosis of purulent pericarditis should be considered in the pneumonia patient who has not responded to antimicrobial therapy and has signs of an expanding cardiac silhouette or cardiac tamponade, atrial arrhythmias, and chest pain. Echocardiography will demonstrate pericardial fluid, and prompt drainage is almost always required.
An equally important complication of pneumonia in the ICU is meningitis. In patients with pneumonia, abnormal mental status or coma is sometimes attributed to hypoxia or sepsis, and central nervous system infection is not considered. Lumbar puncture must be performed in the patient with pneumonia who is comatose or shows a rapid change in mental status, even if meningeal signs are absent. Initial treatment with vancomycin and ceftriaxone is required if resistant pneumococci are suspected.
Superinfection is likely to occur in ICU patients being treated for pneumonia, especially those on broad-coverage antimicrobials. When an ICU patient deteriorates after initial improvement or becomes febrile, or if a new pulmonary infiltrate develops, one must consider the possibility of pneumonia with a secondary organism.
The diagnosis of pneumonia in the ICU should be entertained in the patient who has (a) a new or progressive asymptomatic infiltrate, (b) fever, (c) leukocytosis or leukopenia, and (d) purulent tracheal secretions. However, nosocomial pneumonia may be a subtle illness, and the usual clinical criteria for the diagnosis may lead to false-negative rates as high as 66%. The explanation for such potential misdiagnosis includes confusion of atelectasis and pulmonary edema with infectious infiltrate on chest x-ray films, lack of febrile response to infection from the critically ill (especially the elderly), inability to mount an effective WBC response, and poor-quality sputum samples.
The most convincing argument for accurate diagnosis before treatment is the potential for development of superinfection and increased mortality with antimicrobial-resistant organisms. Furthermore, empiric antimicrobial treatment compromises efforts to document infection and recover a causative organism. Also, there is a risk for drug reactions, drug-drug interactions, and Clostridium difficile infection, which can lead to increased morbidity and mortality. Finally, treating all suspected cases of pneumonia in the ICU with a long course of empiric antimicrobials is an unnecessary and inefficient use of financial resources. Streamlining antibiotic therapy—that is, the practice of converting a broad-spectrum regimen to a more specific regimen on the second or third day of therapy—can help reduce the risk for superinfection and emergence of resistance.
When pneumonia is suspected in the ICU, the physician’s quest to obtain a “good” sputum sample (i.e., one with 25 WBCs per high-power field) should be relentless. Sputum examined by Gram’s stain for bacteria, potassium hydrox-ide for fungi, Ziehl-Neelsen stain for acid-fast bacilli, and, when appropriate, silver methenamine stain for P. carinii may be all that is needed, along with culture results to establish the cause of pneumonia. Often, sputum cannot be expectorated by the critically ill patient because of extreme weakness, poor cough, the presence of an endotracheal or tracheostomy tube, or unresponsiveness. Thus, the nasotracheal suction catheter is often used to obtain a sputum sample. When introduced through an endotracheal tube, the aspiration catheter is capable of obtaining samples representative of true lower respiratory secretion. Often, the critically ill patient cannot tolerate deep suctioning (e.g., because of face or head trauma or induction of oxygen desaturation), and a more invasive technique for obtaining a sample of lower respiratory tract secretion is needed. Transtracheal aspiration is rarely used today because the procedure is not commonly taught, and it is risky when performed by an unskilled operator. Furthermore, the procedure cannot be carried out in an intubated patient.
Percutaneous needle aspiration can be accurate in the diagnosis of a peripheral cavitary lesion or an anaerobic lung abscess. However, it provides a small inoculum of a small sampling area (and is thus prone to yield a high false-negative rate), is contraindicated in the mechanically ventilated patient, and has a high rate of complications, such as pneumothorax and hemorrhage.
Bronchial washings have been of little use in the diagnosis of pneumonia in the ICU. Transbronchial biopsy is helpful in the diagnosis of a central mass lesion, but its diagnostic utility is unproved with pneumonic infiltrates, and it also is associated with the problems of limited sample area and risk for pneumothorax. Bronchoalveolar lavage of samples of more than 1 million alveoli has been shown to be safe and is the method of choice at many institutions. Although some investigators believe that the rate of contamination of bronchoalveolar lavage cultures with colonizing bacteria is high, bronchoalveolar lavage sampling is attractive because it is easily performed at bedside, allows immediate results (Gram’s stain and cell count), and has a low element of risk (only a rare episode of hemoptysis or pneumothorax). Protected specimen bronchoscopy (protected brush) allows collection of lower respiratory specimens protected from contamination by secretions that may pool on the outside of the bronchoscope. With much lower cutoffs for the number of colony-forming units per milliliter to indicate infection (103 for protected brush versus 105 for expectorated sputum), excellent results can be obtained with essentially no false-negatives, suggesting a very high sensitivity. Disadvantages to the use of protected specimen bronchoscopy include reduced accuracy after prior use of antimicrobials, delay before obtaining culture results, concern that protected specimen bronchoscopy samples only a small area of lung, cost, and concern that many centers are not yet equipped to perform these procedures. In general, the role of bronchoscopy in the diagnosis of pneumonia in the ICU patient remains controversial among pulmonologists.
Finally, open lung biopsy, the most invasive procedure, has long been regarded as the gold standard in the diagnosis of pneumonia. Open lung biopsy has been found to have 97% accuracy and a complication rate of 9.6%. Furthermore, the results frequently lead to antimicrobial changes. Despite the accuracy of this procedure, it cannot be performed quickly and easily in most patients (especially ventilated ICU patients), and the complication rate of 10% is too high to justify its use on a routine basis. Open lung biopsy should be reserved for patients in whom other methods of specimen recovery have failed.
Supportive measures in the treatment of pneumonia in the ICU include physiotherapy and postural drainage, percussion, and tracheal suctioning to mobilize purulent secretions. These measures may improve gas exchange in the patient with thick, copious secretions. Other supportive measures include intermittent positive pressure breathing, supplemental oxygen therapy, adequate analgesics, and antipyretics. Intermittent positive pressure breathing is used to increase lung inflation in many types of lung disease, including pneumonia. However, its utility in pneumonia is unproved, and its routine use is not recommended in the treatment of pneumonia. Oxygen therapy is critical in patients with pneumonia and documented hypoxemia. Indiscriminate use of oxygen and oxygen toxicity must be avoided. Codeine or parenteral meperidine for analgesia is sometimes needed to allow deeper breathing and coughing. Finally, if antipyretics are used, they should be given around the clock; sporadic use can increase the patient’s discomfort by causing periods of heavy sweating.
The first step in choosing an antimicrobial agent is to obtain a thorough history and physical examination in conjunction with a Gram’s-stained smear of respiratory secretions. In patients ill enough to require ICU admission, bronchoscopy or protected brush is often necessary to determine an etiologic agent. If a good expectorated sputum sample can be obtained and a presumptive diagnosis made on the basis of a Gram’s-stained smear, further diagnostic study is usually not pursued. The treatment of gram-negative pneumonia depends in part on the antimicrobial sensitivity patterns of a particular hospital. Treatment of gram-negative bacillary pneumonia in the extremely ill patient requires initial antipseudomonal coverage. Ticarcillin or piperacillin plus an aminoglycoside is the commonly used regimen for the ICU patient. Ceftazidime, cefepime, aztreonam, and imipenem plus cilastatin have been used successfully as monotherapy, again depending on known susceptibility patterns. When staphylococci are suspected in a Gram’s-stained smear, the incidence of methicillin-resistant staphylococci becomes a critical issue. Suspicion of methicillin-resistant staphylococci requires initial therapy with vancomycin. When H. influenzae or M. catarrhalis is suspected, initial therapy with a second- or third-generation cephalosporin is begun. When sputum reveals an abundance of polymorphonuclear leukocytes but no staining organisms, Legionella must be considered and initial therapy with erythromycin begun.
Recently, the Infectious Disease Society of America published guidelines for the treatment of community-acquired pneumonia. For treatment of the patient hospitalized in an ICU, the panel recommended erythromycin, azithromycin, or a fluoroquinolone plus cefotaxime, ceftriaxone, or a b-lactamase inhibitor.
Table 7-2 lists the most common causes of pneumonia in ICU patients and the treatments of choice. (S.L.B.)
Table 7-2. Drug regimens of choice for etiologic agents in pneumonia
American Thoracic Society. Hospital-acquired pneumonia in adults: diagnosis, assessment of severity, initial antimicrobial therapy, and preventive strategies. A consensus statement. Am Rev Respir Crit Care Med 1996;153:1711.
Defines severe, hospital-acquired pneumonia requiring admission to ICU. The role of bronchoscopy in diagnosis of pneumonia in ICU remains controversial among pulmonologists.
Andrews CP, et al. Diagnosis of nosocomial bacterial pneumonia in acute, diffuse lung injury. Chest 1981;80:3.
The diagnosis of bacterial pneumonia in the setting of acute respiratory distress syndrome is particularly difficult.
Ashbaugh DG, Petty TL. Sepsis complicating the acute respiratory distress syndrome. Surg Gynecol Obstet 1972;135:865.
Mortality is high in patients with acute respiratory distress syndrome and concurrent pulmonary infection.
Baigelman W, et al. Bacteriologic assessment of the lower respiratory tract in intubated patients. Crit Care Med 1986;14:864.
Routine nasal suctioning compared well with flexible fiberoptic bronchoscopy.
Bartlett JG, et al. Bacteriology of hospital-acquired pneumonia. Arch Intern Med 1986;146:868.
Pathogens causing nosocomial pneumonia include, in order of decreasing prevalence, gram-negative bacilli, anaerobic bacteria, S. aureus, and S. pneumoniae.
Bartlett JG, et al. Community-acquired pneumonia in adults. Clin Infect Dis 1998; 26:811.
Infectious Disease Society of America guidelines on the treatment of community-acquired pneumonia. For patients hospitalized in an ICU, the panel recommends erythromycin, azithromycin, or a fluoroquinolone plus cefotaxime, ceftriaxone, or a b-lactamase inhibitor.
Berk SL, Verghese A. Emerging pathogens in nosocomial pneumonia. Eur J Clin Microbiol Infect Dis 1989;8:11.
Gram-negative bacilli have become the most common etiologic agents in nosocomial pneumonia, but some gram-positive cocci, such as enterococci, group B streptococci, staphylococci, and pneumococci, have taken on new significance.
Bryan CS, Reynolds KL. Bacteremic nosocomial pneumonia. Analysis of 172 episodes from a single metropolitan area. Am Rev Respir Dis 1984;129:668.
Bacteremic nosocomial pneumonia causes a 58% mortality rate.
Bryant LR, et al. Misdiagnosis of pneumonia in patients needing mechanical respiration. Arch Surg 1973;106:286.
Bacterial pneumonia is often overdiagnosed in the ICU setting.
Celis R, et al. Nosocomial pneumonia. A multivariate analysis of risk and prognosis. Chest 1988;93:319.
Identification of predisposing factors to nosocomial pneumonia (“high risk” microorganisms, bilateral pneumonia, respiratory failure, inappropriate antibiotics, age over 60 years, and presence of an ultimately or rapidly fatal underlying disease) may improve prognosis.
Chastre J, et al. Prospective evaluation of the protected specimen brush for the diagnosis of pulmonary infections in ventilated patients. Am Rev Respir Dis 1984; 130:924.
Quantitative cultures obtained from protected specimen brush bronchoscopy are useful in the diagnosis and treatment of pulmonary infections in ventilated patients.
Chastre J, et al. Diagnosis of nosocomial bacterial pneumonia in intubated patients undergoing ventilation: comparison of the usefulness of bronchoalveolar lavage and the protected specimen brush. Am J Med 1988;85:499.
Quantitative cultures from protected brush specimen were sensitive and specific in the diagnosis of ICU pneumonia.
Craven DE, et al. Risk factors for pneumonia and fatality in patients receiving continuous mechanical ventilation. Am Rev Respir Dis 1986;133:792.
The presence of an intracranial pressure monitor, treatment with cimetidine, hospitalization during fall-winter seasons, and ventilator circuit changes every 24 hours were found to be significant risk factors for nosocomial pneumonia in the ventilated patient.
Craven DE, et al. Nosocomial infection and fatality in medical and surgical intensive care unit patients. Arch Intern Med 1988;148:1161.
Identifies nine variables significantly associated with fatality in nosocomial infection.
Ekenna O, et al. Isolation of b-hemolytic streptococci from the respiratory tract: serotypic distribution and clinical significance. Am J Med Sci 1988;295:94.
Group B streptococci were found to cause pneumonia in patients with mean age of 68.1 years. There was a 34% incidence of S. aureus coinfection.
Fagan JY, et al. Detection of nosocomial lung infection in ventilated patients. Use of a protected specimen brush and quantitative culture techniques in 147 patients. Am Rev Respir Dis 1988;138:110.
Protected specimen brush bronchoscopy was used in evaluating patients with pulmonary infiltrates and purulent tracheal secretions, the majority of whom did not have bacterial pneumonia.
Fagan JY, et al. Nosocomial pneumonia in patients receiving continuous mechanical ventilation. Prospective analysis of 52 episodes with use of a protected specimen brush and quantitative culture techniques. Am Rev Respir Dis 1989;139:877.
P. aeruginosa and S. aureus were involved in 33% of ventilator-associated pneumonias studied by protected specimen brush bronchoscopy.
Fang GD, et al. New and emerging etiologies for community-acquired pneumonia with implications for therapy. A prospective multicenter study of 359 cases. Medicine 1990;69:307.
Prospective study of community-acquired pneumonia defines etiologic agents well. Mortality is determined by etiologic agent.
Gaussorgues P, et al. Comparison of nonbronchoscopic bronchoalveolar lavage to open lung biopsy for the bacteriologic diagnosis of pulmonary infections in mechanically ventilated patients. Intesive Care Med 1989;15:94.
A cuffed catheter blindly guided through an endotracheal tube can be used for bronchoalveolar lavage in the diagnosis of pneumonia.
Graybill JR, et al. Nosocomial pneumonia: a continuing major problem. Am Rev Respir Dis 1973;108:1130.
Gram-negative nosocomial pneumonia, often following oropharyngeal colonization, is a common and frequently catastrophic event with high morbidity and mortality rates despite new antimicrobial agents.
Hitt CM, et al. Streamlining antimicrobial therapy for lower respiratory infections. Clin Infect Dis 1997;24(Suppl 2):S231.
The authors emphasize an approach to antibiotic therapy in lower respiratory infection in which a broad-spectrum empiric regimen is switched to a narrower-spectrum antibiotic based on culture results and other data.
Johanson G, et al. Bacteriologic diagnosis of nosocomial pneumonia following prolonged mechanical ventilation. Am Rev Respir Dis 1988;137:259.
Bronchoalveolar lavage provides the best reflection of the pulmonary bacterial burden in intubated patients.
Johanson WG. Ventilator-associated pneumonia. Light at the end of the tunnel? Chest 1990;97:1027.
Concise overview of methods for diagnosing ventilator-associated pneumonia.
Johanson WR, et al. Nosocomial respiratory infections with gram-negative bacilli. The significance of colonization of the respiratory tract. Ann Intern Med 1972;77:701.
Colonization of the respiratory tract with gram-negative bacilli is extremely common in the ICU patient, and respiratory tract infection occurs in 23% of such colonized patients.
Joshi N, Localio AR, Hamory BH. A predictive risk index for nosocomial pneumonia in the intensive care unit. Am J Med 1991;93:135.
Factors that lead to high risk for pneumonia in the ICU include endotracheal intubation, upper abdominal thoracic surgery, bronchoscopy, and presence of a nasogastric tube.
Kappstein I, et al. Incidence of pneumonia in mechanically ventilated patients treated with sucralfate or cimetidine as prophylaxis for stress bleeding: bacterial colonization of the stomach. Am J Med 1991;91(Suppl 2A):125S–131S.
Increased gastric pH may lead to a higher incidence of retrograde colonization of the oropharynx from the stomach with Enterobacteriaceae.
Karnad A, Alvarez S, Berk SL. Pneumonia caused by gram-negative bacilli. Am J Med 1985;79:63.
In pneumonia caused by gram-negative bacilli, associated bacteremia is most commonly seen with P. aeruginosa or Serratia marcescens, and outcome is poor.
Meduri GU, Baselski V. The role of bronchoalveolar lavage in diagnosing nonopportunistic bacterial pneumonia. Chest 1991;100:179.
Bronchoalveolar lavage is a simple and relatively safe technique with an expanding role in the diagnosis of nonopportunistic pulmonary infection.
Miller KS, Sahn SA. Chest tubes: indications, technique, management and complications. Chest 1987;91:259.
Current standards in critical care often require invasive procedures such as bronchoscopy and pulmonary biopsy. The potential for pneumothorax is significant, and those performing invasive procedures should be comfortable with chest tube insertion.
Rello J, et al. Severe community-acquired pneumonia in the elderly: epidemiology and prognosis. Clin Infect Dis 1996;23:723.
S. pneumoniae is the most common organism—seen in about half of all cases—to cause life-threatening pneumonia in the elderly.
Salata RA, et al. Diagnosis of nosocomial pneumonia in intubated intensive care unit patients. Am Rev Respir Dis 1987;135:426.
In intubated ICU patients, serial examination of endotracheal aspirates for elastin fibers, graded Gram’s stain, and bacterial colony counts are useful in the diagnosis of pneumonia.
Sanderson PJ. The sources of pneumonia in ICU patients. Infect Control 1986;7:104.
Describes rates of colonization and pneumonia in ICUs in Great Britain.
Scheld WM, Mandell GL. Nosocomial pneumonia: pathogenesis and recent advances in diagnosis and therapy. Rev Infect Dis 1991;13(Suppl 9):S743.
Pneumonia is the leading cause of death from nosocomial infections, with a mortality rate of 20% to 50%.
Seindenfeld JJ, et al. Incidence, site, and outcome of infections in patients with the adult respiratory distress syndrome. Am Rev Respir Dis 1986;134:12.
Bacterial pathogens, the majority of them gram-negative organisms, account for most infections in patients with adult respiratory distress syndrome, and mortality rate can reach as high as 70% to 80%.
Stevens RM, et al. Pneumonia in an intensive care unit. A 30-month experience. Arch Intern Med 1974;134:106.
Established the high incidence and mortality rate of ICU pneumonia.
Winterbauer RH, et al. The use of quantitative cultures and antibody coating of bacteria to diagnose bacterial pneumonia by fiberoptic bronchoscopy. Am Rev Respir Dis 1983;128:98.
Quantitative cultures and immunofluorescent demonstration of antibody-coated bacteria are used to differentiate colonizing from infecting bacteria in lower respiratory tract secretions obtained by fiberoptic bronchoscopy.
Young LS. Treatment of respiratory infections in the patient at risk. Am J Med 1984;76(Suppl 5A):61.
Includes review of treatment response among patients with gram-negative pneumonia.