Nosocomial Pneumonia 

According to American Thoracic Society (ATS) guidelines, nosocomial pneumonia (also known as hospital-acquired pneumonia or health care–associated pneumonia) is defined as pneumonia that occurs more than 48 hours after hospital admission but that was not incubating at the time of admission. Ventilator-associated pneumonia is defined as pneumonia that occurs after 48-72 hours of endotracheal intubation. A radiographic image of nosocomial pneumonia is shown below.^[1]

Nosocomial pneumonia is the second-most-common nosocomial infection and is usually bacterial in origin. The disease adds significantly to the cost of hospital care and to the length of hospital stays.

Although most patients with nosocomial pneumonia develop fever and leukocytosis, these findings are not uniform and are not a requisite for the presumptive diagnosis of nosocomial pneumonia. Respiratory tract symptoms include an increase in respiratory rate, shortness of breath, and a productive cough.

The ATS subdivides nosocomial pneumonia into early onset (usually within the first 4 d of the hospitalization) and late onset (usually occurring after the fifth hospital day). Early onset nosocomial pneumonia tends to carry a better prognosis than does late-onset nosocomial pneumonia; the latter tends to be associated with multidrug-resistant organisms and so is characterized by higher mortality rates.^{[2, 3, 4, 5, 6]}

Go to Mycoplasma Pneumonia, Bacterial Pneumonia, and Viral Pneumonia for more complete information on this topic.

Inhalation, aspiration, and hematogenous spread are the 3 main mechanisms by which bacteria reach the lungs. The primary route through which organisms enter the lower airways is via aspiration of oropharyngeal secretions into the trachea.

Primary inhalation pneumonia develops when these organisms bypass normal respiratory defense mechanisms or when the patient inhales aerobic gram-negative organisms that colonize the upper respiratory tract or respiratory support equipment.

Aspiration pneumonia is due to the aspiration of colonized upper respiratory tract secretions.

The stomach appears to be an important reservoir of gram-negative bacilli that can ascend and colonize the respiratory tract. A prospective observational study found that patients who used acid-suppressive medications were more likely to develop hospital-acquired pneumonia than were patients who did not (5% vs 2%). Further evaluation by drug class showed that the risk for pneumonia was significantly increased with proton pump inhibitors, but not with histamine 2–blocking agents.^[7]

Hematogenously acquired infections originate from a distant source and reach the lungs via the bloodstream.

The development of nosocomial pneumonia represents an imbalance between normal host defenses and the ability of microorganisms to colonize and then invade the lower respiratory tract.

Because aerobic gram-negative bacilli are the major pathogens associated with nosocomial pneumonia, the pathophysiology of the disease relates to the destructive effect of these organisms on invaded lung tissue. Aerobic gram-negative pathogens may be divided into 2 categories. The first category includes organisms that cause necrotizing pneumonia with rapid cavitation, microabscess formation, blood-vessel invasion, and hemorrhage (eg, Pseudomonas aeruginosa).^[8] The second category consists of all other nonnecrotizing gram-negative organisms responsible for nosocomial pneumonia.

Common causes of nosocomial pneumonia

Common bacteria involved in nosocomial pneumonia include the following:

• P aeruginosa
• Klebsiella species
• Escherichia coli
• Acinetobacter species
• Staphylococcus aureus (especially methicillin-resistant S aureus [MRSA])^[9]
• Streptococcus pneumoniae
• Haemophilus influenzae

Acinetobacter species commonly colonize the respiratory tract secretions in patients in the ICU. Care must be exercised in interpretation of culture data.

Streptococcus pneumoniae should be considered in early onset hospital-acquired pneumonia. This bacterium causes up to 9% of pneumonias in elderly patients in nursing homes.

Haemophilus influenzae should also be considered in early onset hospital-acquired pneumonia.

Less-common pathogens associated with nosocomial pneumonia

The following are less-common pathogens implicated in nosocomial pneumonia:

• Serratia species
• Legionella species
• Influenza A virus
• Respiratory syncytial virus (RSV)
• Parainfluenza virus
• Adenovirus

Legionella nosocomial pneumonia occurs only in outbreaks or clusters.

Influenza A virus, RSV, parainfluenza virus, and adenovirus, may cause hospital-acquired pneumonia in the right clinical setting.

Extremely rare causes of nosocomial pneumonia

The following are rarely isolated in patients with nosocomial pneumonia:

• Enterobacter species
• Stenotrophomonas maltophilia (formerly Pseudomonas maltophilia)
• Burkholderia cepacia (formerly Pseudomonas cepacia)
• Candida species (Candida species are an uncommon cause of hospital-acquired pneumonia, and cultures positive for these organisms more often reflect colonization.)
• Oropharyngeal anaerobes (non–Bacteroides fragilis)

Although these organisms may be very uncommon causes of nosocomial pneumonia, they have been recovered in patients with ventilator-associated pneumonia.^[10] The recovery from respiratory secretions of an organism that is typically pathogenic does not prove that it is pathogenic or the cause of nosocomial pneumonia. Anaerobic organisms are not typically isolated in nosocomial pneumonia. (See Differentials in Nosocomial Pneumonia.)

Risk factors

As previously mentioned, the stomach appears to be an important reservoir of gram-negative bacilli that can ascend and colonize the respiratory tract. A prospective observational study found that patients who used acid-suppressive medications were more likely to develop hospital-acquired pneumonia than were patients who did not (5% vs 2%).

Further evaluation by drug class showed that the risk for pneumonia was significantly increased with proton pump inhibitors, but not with histamine 2–blocking agents.^[7]

Incidence in the United States

Nosocomial pneumonia is the second-most-common nosocomial infection in the United States and is usually bacterial in nature. It is one of the most common diagnoses made in medical and surgical ICUs and is common in patients undergoing mechanical ventilation. Nosocomial pneumonia also occurs in patients in the general hospital wards who are not receiving mechanical ventilation.

International incidence

The international incidence and prevalence of nosocomial pneumonia is similar to that in the United States, with comparable rates of responsible microorganisms.

Race and sex predilections

Nosocomial pneumonia has no race or sex predilection

Age predilection

Nosocomial pneumonia is most common in elderly patients; however, patients of any age may be affected.

Morbidity and mortality in nosocomial pneumonia

Intubation and ventilatory support bypass the normal host defense mechanisms, predisposing patients with nosocomial pneumonia to infection.

In addition, nosocomial pneumonia that develops in ICU patients is associated with high morbidity and mortality rates, because these patients are already typically critically ill.

Compromised cardiac and lung function may further decrease the cardiopulmonary reserve of pneumonia, accounting for the high mortality and morbidity rates associated with nosocomial pneumonia.

Barotrauma may decrease an already compromised lung function and alter chest radiographic appearances.

As previously mentioned, early onset nosocomial pneumonia tends to carry a better prognosis than does late-onset nosocomial pneumonia. The latter tends to be associated with multidrug-resistant organisms, meaning that it is associated with higher mortality rates.^{[2, 3, 4, 5, 6]}

The prognosis in patients with nosocomial pneumonia depends on preexisting underlying cardiopulmonary function and host defenses.

As previously noted, early onset nosocomial pneumonia tends to carry a better prognosis than does late-onset nosocomial pneumonia. The latter tends to be associated with multidrug-resistant organisms and is therefore associated with higher mortality rates.^{[2, 3, 4, 5, 6]}

All patients with presumed nosocomial pneumonia should undergo testing to rule out conditions that mimic nosocomial pneumonia. The presumptive diagnosis of nosocomial pneumonia is difficult, because the diagnosis does not depend on the presence of fever, and leukocytosis is unhelpful. A summary of management strategies is available through a practice guideline provided by the ATS.^[11]

The list of differential diagnoses represents the main clinical problem in nosocomial pneumonia. For example, any disorder that results in leukocytosis with variable degrees of left shift may be included in the differential diagnoses. An inflammatory or infectious process can cause fever; therefore, do not regard this symptom as an indication of an infectious disease process. Many conditions other than nosocomial pneumonia can cause pulmonary infiltrates, which can be seen on chest radiographs.

Consider all of these differential diagnoses carefully before settling on a diagnosis and embarking on a course of antimicrobial therapy.

The most common causes of infiltrates in ventilated patients with fever and/or leukocytosis include the following conditions:

• Congestive heart failure
• Pulmonary embolus or infarction
• Acute respiratory distress syndrome (ARDS)
• Pulmonary drug reactions
• Collagen vascular diseases

Collagen-vascular diseases include systemic lupus erythematosus (SLE), bronchiolitis obliterans-organizing pneumonia (BOOP), interstitial lung disease, bronchogenic carcinomas, and metastatic carcinomas.

ARDS is usually readily diagnosable according to the microatelectatic changes on the chest radiograph and the progressive and severe hypoxemia indicated by arterial blood gas (ABG) levels. (Little or no fever may accompany these symptoms.)

In most cases, the diagnosis of nosocomial pneumonia is clinical, although it is supported by appropriate cultures; these can include semiquantitative cultures from bronchoalveolar lavage (BAL) samples.

The definitive diagnosis of nosocomial pneumonia rests on tissue biopsy, which is rarely performed. Therefore, the clinician is forced to grapple with various findings that are not specific to nosocomial pneumonia. (See Differentials in Nosocomial Pneumonia.)

Physical findings in nosocomial pneumonia relate to the pneumonia’s distribution in the chest. Physically, lobar lesions caused by nosocomial pneumonia mimic those caused by any other type of pneumonia (eg, rales in the location of the pneumonic process).

In most cases, neither consolidation nor pleural effusions are features of nosocomial pneumonia. The presence of either should prompt consideration of an alternate diagnosis.

A white blood cell (WBC) count is usually suggested, but it does not yield a specific finding.

The WBC count may be normal or elevated in cases of nosocomial pneumonia or conditions that mimic the disease. A left shift reflects the stress that the patient is being subjected to and neither rules out nor confirms infection. The degree of left shift indicates the degree of stress in the host.

Neither leukocytosis nor a normal WBC count favors the diagnosis of nosocomial pneumonia over the diseases that mimic nosocomial pneumonia, as these can produce a similar elevation.

Obtain blood cultures as early as possible to retrospectively diagnose infection with hematogenous pathogens.

In bacteremic nosocomial pneumonia, blood culture results are frequently positive if obtained early in the disease process and if the patient is not already receiving antimicrobial therapy.

Obtain serial chest radiographs to assist in evaluating the progress of the pneumonia and the efficacy of appropriate antimicrobial therapy. (See the image below.)

Radiographs may also be useful in distinguishing various mimics from actual nosocomial pneumonia. In these patients, computed tomography (CT) scanning or spiral CT scanning may be useful.

Go to Imaging Atypical Pneumonia and Imaging Typical Pneumonia for more complete information on these topics.

These techniques have variable sensitivities and specificities, although there are accepted criteria for semiquantitative cultures to improve the diagnostic reliability of bronchoscopically derived cultures.

In one study, a bronchoscopic bacteriologic strategy was shown to reduce the short-term mortality risk in patients.

Histologic study of lung tissue reveals either necrotizing or nonnecrotizing pneumonia, depending on the pathogen. P aeruginosa produces a necrotizing pneumonia with vessel invasion, local hemorrhage, and microabscess formation.^[8] Other aerobic gram-negative bacilli produce a polymorphonuclear response at the site of invasion, but microabscess formation and vessel invasion are absent.

Electrocardiograms (ECGs) and ventilation-perfusion scans should eliminate pneumonia mimics. ECGs, cardiac enzymes, and Swan-Ganz readings may rule out left ventricular failure caused by exacerbation of heart failure or new myocardial infarction.

Obtain other tests that are related to the possible underlying causes of the pulmonary infiltrates; for example, if lupus pneumonitis is suspected, ask the patient about a history of SLE pneumonitis. Afterward, serologic tests should be performed to assess for SLE.

Tests such as ABG studies are merely used to assess the degree of severity of lung dysfunction but are not useful in diagnosing nosocomial pneumonia. Obtain ABGs to help diagnose a diffusion defect related to interstitial lung diseases.

Multiple organisms as a cause of nosocomial pneumonia

Multiple pathogens obtained by nontissue biopsy culture methods are proof of lower airway colonization.

Most patients with nosocomial pneumonia require ventilatory support at some point; the majority of patients need supplemental oxygen therapy as well. Patients with nosocomial pneumonia should be transferred to the ICU if they are on the general medical wards and cannot be maintained without ventilatory support.

Antimicrobial therapy

Ordinarily, nosocomial pneumonia is treated for 14 days. If the patient indeed has nosocomial pneumonia and the appropriate antimicrobial therapy is administered, chest radiography shows significant improvement in the pulmonary infiltrate during the 2 weeks of treatment. Pulmonary infiltrates that are unchanged after a 2-week course of therapy suggest that the infiltrates may not be infectious in origin. Start a diagnostic workup to consider other infectious diseases that do not respond to antibiotics (eg, herpesvirus type 1 [HSV-1] pneumonitis) or noninfectious diseases (eg, bronchogenic carcinomas).

Before empiric antimicrobial therapy is initiated, an attempt should be made to rule out mimics of nosocomial pneumonia. If mimics of nosocomial pneumonia can be excluded with a reasonable degree of certainty, then empiric therapy for nosocomial pneumonia is appropriate. (See Differentials in Nosocomial Pneumonia.)

The precise pathogen that causes a given case of nosocomial pneumonia is usually unknown. Therefore, empiric antimicrobial therapy is the only practical approach. Delaying therapy until the pathogen is identified is not recommended. For empiric coverage of nosocomial pneumonia, monotherapy is as effective as combination therapy for early nosocomial pneumonia.^[2]

For proven pseudomonal nosocomial pneumonia, double drug coverage with a high degree of antipseudomonal activity and a low resistance potential should be used. Optimal combinations include meropenem or doripenem plus either levofloxacin or aztreonam. Alternately, antipseudomonal penicillin (eg, piperacillin) in combination with levofloxacin, meropenem, aminoglycoside, or aztreonam, may provide equal efficacy.^[12]

A course of antimicrobials for ventilator-associated tracheitis that lasts longer than 7 days does not protect against progression to hospital-acquired pneumonia or ventilator-associated pneumonia. Prolonged antimicrobial use is associated with an increased risk of subsequent multidrug resistance acquisition.^[13]

Principles of appropriate empiric antibiotic coverage in nosocomial pneumonias

Direct empiric coverage against common nosocomial pathogens P aeruginosa, Klebsiella species, E coli,and MRSA. Coverage against P aeruginosa also covers other nosocomial pneumonia pathogens.^[8]

Enterobacter species usually do not cause nosocomial pneumonia. S maltophilia and B cepacia are common colonizers of respiratory secretions, but they rarely, if ever, cause nosocomial pneumonia in most hosts. However, they are potential pathogens in patients with bronchiectasis or cystic fibrosis.

Oropharyngeal anaerobes are unimportant from a therapeutic standpoint, as they are not typically isolated in nosocomial pneumonia.

Empiric monotherapy versus combination therapy

The optimal empiric monotherapy for nosocomial pneumonia consists of ceftriaxone, ertapenem, levofloxacin, or moxifloxacin. Monotherapy may be acceptable in patients with early onset hospital-acquired pneumonia. Avoid monotherapy with ciprofloxacin, ceftazidime, or imipenem, as they are likely to induce resistance potential.^[14]

Late-onset hospital-acquired pneumonia, ventilator-associated pneumonia, and health care–associated pneumonia require combination therapy using an antipseudomonal cephalosporin, beta lactam, or carbapenem plus an antipseudomonal fluoroquinolone or aminoglycoside plus an agent such as linezolid or vancomycin to cover MRSA.^[15] The FDA warns against the concurrent use of linezolid with serotonergic psychiatric drugs, unless indicated for life-threatening or urgent conditions. Linezolid may increase serotonin CNS levels as a result of MAO-A inhibition, increasing the risk of serotonin syndrome.^[16]

A retrospective analysis of 320 patients with documented MRSA bacteremia initially treated with vancomycin from January 2005 to April 2010 revealed high failure rates.^[17] The study suggests that optimizing the use of vancomycin by targeting higher trough values of 15-20 mg/L and area under the curve at 24 hour (AUC24h) to minimum inhibitory concentration (MIC) ratios to 400 or more in selected patients should be considered.

Optimal combination regimens for proven P aeruginosa nosocomial pneumonia include (1) piperacillin/tazobactam plus amikacin or (2) meropenem plus levofloxacin, aztreonam, or amikacin.^[12] Avoid using ciprofloxacin, ceftazidime, gentamicin, or imipenem in combination regimens, as combination therapy does not eliminate the resistance potential of these antibiotics.

When selecting an aminoglycoside for a combination therapy regimen, amikacin once daily is preferred to gentamicin or tobramycin to avoid resistance problems. When selecting a quinolone in a combination therapy regimen, use levofloxacin, which has very good anti– P aeruginosa activity (equal or better than ciprofloxacin at a dose of 750 mg).^{[2, 3, 4, 5, 6]}

Consult an infectious disease specialist to assess the microbiology of the specimens obtained from the patient, to rule out the mimics of nosocomial pneumonia, and to administer empiric or specific empiric antimicrobial therapy.

Consult a pulmonologist to help with mechanical ventilation (often required in patients with nosocomial pneumonia).

Other consultations include the following:

• Rheumatologist (if the patient appears to have lupus or SLE pneumonitis)
• Cardiologist (if the patient has heart failure)
• Oncologist (for possible pulmonary infiltrates caused by a lymphangitic spread of a malignancy)

Most patients with nosocomial pneumonia are intubated, are instructed to receive nothing by mouth (NPO), and are limited to bed rest.

In a study of intensive care patients by Grau et al, total parenteral nutrition supplemented by alanine-glutamine was associated with a lower rate of infectious complications and superior glycemic control.^[18]

Beds that permit some degree of patient turning may decrease the risk of nosocomial pneumonia in at-risk patients.

Failure to successfully wean the patient from the respirator (possibly because of a lack of cardiopulmonary function or a superimposed process [eg, HSV-1 pneumonitis]) is a common problem following intubation for nosocomial pneumonia.

HSV-1 pneumonitis develops in intubated patients who have unchanging or persistent pulmonary infiltrates after 2 weeks of antimicrobial therapy. These patients usually have low-grade fevers with variable degrees of leukocytosis. Demonstrating HSV-1 in samples of respiratory secretions may establish the diagnosis.

Start treatment with acyclovir in patients diagnosed with HSV-1 infection; acyclovir decreases hypoxemia and subsequently permits weaning of the patient from the respirator.