Hospital-Acquired Pneumonia

Epidemiology

Hospital-acquired pneumonia (HAP), formerly called nosocomial pneumonia, is defined as pneumonia occurring 48 hours after admission to the hospital and excluding those incubating at the time of admission. Health care–associated pneumonia (HCAP) refers to patients admitted to hospital with pneumonia that may have received IV therapy at home, wound care or specialized nursing care through a health care agency, family, or friends, or had self-administered IV medical therapy in the 30 days before pneumonia. They may have attended a hospital or hemodialysis clinic or received IV chemotherapy in the 30 days before pneumonia. They may also have been admitted to an acute care hospital for ≥2 days in the 90 days before pneumonia or resided in a nursing home or a long-term care facility. Ventilator-associated tracheobronchitis (VAT) is a localized disease of the tracheobronchial tree with clinical signs (e.g., fever, leukocytosis and purulent sputum), microbiologic information (e.g., Gram stain with bacteria and leukocytes, with either a positive semiquantitative or a quantitative sputum culture), and the absence of a new infiltrate on chest radiograph. The incidence of HAP is difficult to estimate because of overlap with other lower respiratory tract infections, such as tracheobronchitis, in mechanically ventilated patients. Given this limitation, the incidence of HAP is estimated to be between 5 and 15 cases per 1,000 hospital admissions, depending on the population being studied. A retrospective cohort analysis of 284 hospitals reported a non-ventilator associated HAP incidence of 5.5 events per 1,000 hospitalizations. Patients with tracheal intubation and/or requiring mechanical ventilation are at a 6- to 20-fold increased risk for HAP. The risk of ventilator-associated pneumonia (VAP) is estimated at 2-6 cases per 1,000 ventilator-days. The rates are much lower in nonintubated patients and in those in pediatric intensive-care units (ICUs), but higher in adult medical and surgical ICUs and burn units. When compared with matched controls, patients developing VAP have a 2- to 2.5-fold increased risk of mortality. Patients with HCAP have more severe disease, longer hospital stays, and higher mortality rates than patients with CAP. Current studies on VAT report a crude incidence rate that may vary from 2.7% to 10%; one observational study of 114 international ICUs reported an incidence rate of 10.2 cases per 1,000 ventilator-days.

Crude mortality rates for VAP range from 20% to 71%, but deaths are often due to other causes in critically ill patients. A preferred measure is attributable mortality, defined as the percentage of deaths that would not have occurred in the absence of the infection. Using matched cohort studies, an attributable mortality of up to 50% has been reported. A 2013 meta-analysis of 24 trials reported a VAP-attributable mortality of 13% in the United States. The major risk factors for mortality include:

• Severity of underlying illness
• Ineffective initial antibiotic therapy
• Infection with high-risk pathogens, such as P aeruginosa or Acinetobacter spp
• Advanced age
• Bacteremia.

Rates of secondary bacteremia range from 4% to 38%, with a median of 11%. Empyema has been reported following the development of HAP. Each episode of HAP prolongs a hospital stay by an average of 7 to 9 days, resulting in increased hospital charges. One study indicated that costs would exceed reimbursement in the vast majority of Medicare patients developing HAP.

Risk Factors and Pathogenesis

The respiratory tract is uniquely designed to prevent entry of pathogenic organisms into the lung and to eradicate these microorganisms if they manage to bypass upper airway host defenses. Included in this defense armamentarium are:

• Filtration and humidification of the upper airway
• Cough reflexes
• Mucociliary escalator, an elaborate system that transports secretions and microorganisms from the periphery of the lung to the central airways where they are expectorated or swallowed.

Organisms impact onto the mucous layer because of inertial forces whenever there is a major airway division. When they are trapped in this layer, the organisms are removed when oscillating cilia transport the mucus to the central airways. If the organisms reach the periphery of the lung, phagocytes and opsonins remove many of them and systemic humoral and cell-mediated immunity obliterates the rest. Pneumonia occurs only when this elaborate defense mechanism is overwhelmed, either because of a large aspirated inoculum or the inherent virulence of the inhaled organisms, or when the defense mechanism is breached in some manner.

The most common means of acquiring pneumonia is via aspiration of oropharyngeal contents. As many as 45% of healthy subjects aspirate during sleep. Aspiration is more likely among patients with:

• Abnormal swallowing
• Impaired gag reflex
• Altered consciousness
• Delayed gastric emptying
• Decreased gastric motility.

Nasogastric tubes may contribute to increased oropharyngeal colonization, either by providing a conduit for organisms to migrate from the stomach or by causing erosions on mucosal surfaces, thereby exposing more binding sites to gram-negative bacilli. A crucial first step in the development of HAP is the adherence of potential pathogenic organisms to buccal mucosa (Table 6-1). This is facilitated by critical illness when increased levels of salivary proteases degrade cell-surface fibronectin.

Many patients have upper airway colonization with pathogenic bacteria, and the leading organisms associated with HAP are enteric gram-negative bacilli and
S aureus (Table 6-2). Oropharyngeal colonization with aerobic gram-negative bacilli is unusual or of short duration in healthy, non­hospitalized individuals. However, in moderately ill patients, the carriage rate increases to 16% and reaches 57% in critically ill patients. With repeat cultures, colonization rates approach 75%. Severity of illness, longer duration of hospitalization, prior or concomitant use of antibiotics, advanced age and disability, poor nutrition, intubation and major surgery have all been identified as factors associated with gram-negative oropharyngeal colonization.

Many risk factors have been identified for the development of HAP, especially in ventilated patients (Table 6-3). Endotracheal intubation breaches most of the natural barriers in the upper airway. It impairs mucociliary clearance and injures the epithelial surface, predisposing to attachment of organisms to the surface of the lower respiratory tract. The endotracheal tube may become encrusted with a biofilm containing microorganisms. This may be dislodged into the lung and serve as a source of infection. Nosocomial pneumonia is less common when patients are managed with noninvasive ventilation.

The role of the stomach as a potential reservoir for pathogens is controversial. Several studies have reported similar organisms in the stomach and trachea in patients developing HAP. However, gastric colonization preceding upper airway colonization has only been demonstrated in a minority of patients. Bacteria multiply rapidly in the presence of an ileus or impaired gastric acidification. Histamine type 2 (H₂) blockers and antacids have been identified as risk factors for HAP, presumably because they encourage an environment in which bacteria can rapidly multiply. Some authors have suggested the use of sucralfate for stress ulcer prophylaxis in the intubated patient, but reduced rates of HAP have not been consistently demonstrated.

Other less common means of acquiring pneumonia include inhalation of microorganisms, seeding from the bloodstream and reactivation of latent infection (TB). Aspiration in mechanically ventilated patients occurs around the outside of the endotracheal tube rather than through the lumen. Leakage around the endotracheal cuff can be demonstrated in most patients. Other sources of pathogens are aspiration from the stomach or the nose and paranasal sinuses. Keeping the patient in the semi­recumbent position can minimize aspiration of gastric content, but oropharyngeal aspiration is not affected by this maneuver. Outbreaks of HAP have been reported with the use of contaminated mainstream nebulizers, cascade humidifiers, manual ventilation bags, spirometers and ventilator temperature probes. Frequent changes of ventilator circuits are also associated with an increased incidence of pneumonia. Poor infection-control practices can lead to the transmission of pathogens by health care providers. Rates of infection can be reduced by hand washing, a practice too often ignored by nurses and physicians.

Enlarge 

Enlarge 

Enlarge 

Approach to Diagnosis

The diagnosis of VAP is difficult and the role of invasive testing is changing. Clinicians are often confronted with a changing clinical or radiographic setting where specific therapy is demanded. While mainly relying upon clinical judgment, clinicians have made increasing use of what has been termed “invasive testing.” Developed in 1987, several techniques have been developed to handle two problems:

• Contamination by the upper airway (by protecting the sampling fluid)
• Separation of infection from colonization (by using quantitative cultures).

However, concerns regarding diagnostic accuracy, reproducibility of results, diagnostic thresholds, non­standardized methodology and lack of clinical outcome data have made the interpretation of clinical studies difficult.

The initial diagnosis of VAP depends upon clinical suspicion and the presence of new or progressive radiographic infiltrates. Clinical suspicion is triggered by the presence of some combination of fever, purulent tracheobronchial secretions and leukocytosis. This combination of findings is sensitive in predicting the presence of pneumonia in a group of patients pre­selected for suspected VAP but does not perform as well among unselected patients. If all three criteria are required along with a new or progressive lung infiltrate, the specificity for the diagnosis of VAP drops below 50%.

The incidence of pneumonia in immunocompetent patients with a normal chest radiograph and a compatible clinical presentation (a common finding in immunocompromised patients with Pneumocystis jirovecii pneumonia) is unknown. The usual radiographic abnormalities are new or worsening infiltrates or air bronchograms. The sensitivity ranges from 50% to 78% for new or worsening infiltrates and 58% to 83% for air bronchograms. The inter-rater reliability of chest radiographs is low, as there is only marginal reproducibility between readers. These findings suggest that a combination of clinical criteria and radiographic abnormalities is a useful screening tool for the presence of VAP, but other investigations may be required to confirm the diagnosis. Many noninfectious disorders may be associated with pulmonary infiltrates and fever, including:

• HF
• Atelectasis
• ARDS
• Drug reactions
• PE.

Other investigations that are routinely performed are blood cultures and thoracentesis of pleural fluid if a sample is obtainable. Blood cultures will be positive in 8% to 20% of all patients with HAP. Among patients with severe HAP, other sources of infection besides the lung are found in up to 50% of cases. Identifying the causative organism and choosing appropriate therapy quickly is essential in HAP and VAP; therefore, the introduction of commercial multiplex polymerase chain reaction (mPCR) platforms which are able to detect a large spectrum of bacterial, viral and fungal pathogens in sputum or bronchoalveolar lavage (BAL) samples more quickly did much to improve diagnosis. Serologic studies searching for organisms, such as M pneumoniae, C pneumoniae and Legionella spp, are rarely performed since the results of these investigations often are unavailable in time to affect clinical outcomes.

The clinical pulmonary infection score (CPIS) combines clinical, radiographic, physiologic (PaO₂/fraction of inspired oxygen) and microbiologic data into a single numerical result in order to improve diagnostic specificity. When the CPIS exceeds 6, a good correlation with positive quantitative culture findings of bronchoscopic and nonbronchoscopic BAL specimens is seen. These findings have not always been reproducible, but the accuracy of the CPIS score is improved if a Gram stain of a deep respiratory tract culture is added to the evaluation.

Microscopic examination of endotracheal aspirates is the simplest noninvasive means of obtaining respiratory secretions from mechanically ventilated patients. It is an attractive option since it is readily performed at the bedside and requires few special skills. Using the standard criteria developed for assessment of Gram’s stain will produce a valid representation of the infectious organisms. Qualitative endo­tracheal aspirates usually allow one to identify organisms found by invasive tests, suggesting high sensitivity, but they frequently identify other nonpathogenic organisms, reducing the positive predictive value of the test. Sterile endotracheal aspirate cultures make the diagnosis of VAP unlikely.

The results of quantitative endotracheal aspirate cultures vary with the bacterial load, duration of ventilation and prior antimicrobial administration. The sensitivity of endotracheal aspirates ranges widely from 38% to 100%, while specificity ranges from 14% to 100%. Antibody coating of bacteria (a systemic response to infection) and the presence of elastin fibers (a means of detecting lung parenchymal destruction) are neither sensitive nor specific for the diagnosis of VAP.

Fiberoptic bronchoscopy (FOB) allows direct sampling of the lower respiratory tract. However, the bronchoscope itself will be contaminated by organisms found in the upper airway. Sampling is usually performed in the distal airways where contamination is less likely. BAL involves the sequential instillation and aspiration of a large volume of saline through the distal port of the bronchoscope while it is wedged in a peripheral airway. BAL samples approximately 1 million alveoli if a standard 120-mL aliquot of saline is introduced into a pulmonary sub­segment. The recovered material is examined for the presence of inflammatory cells, intracellular organisms and the percentage of squamous epithelial cells (SECs). SECs signify upper airway contamination, while the presence of intracellular organisms in ≥2% of phagocytic cells indicates true bacterial infection. Quantitative cultures are performed and the presence of organisms in concentrations of 10⁴ colony-forming units (CFUs)/mL is considered to be sufficient to confirm the diagnosis of pneumonia and identify the causative pathogen. The sensitivity of quantitative BAL fluid cultures ranges from 42% to 93%, with a mean of 73%. This variability depends upon prior antibiotic treatment, type of study population and the reference test used. Previous antibiotic therapy reduces sensitivity.

Protected specimen brush (PSB) involves a double-catheter brush system with telescoping cannulas and a distally occluding wax plug. The bronchoscope is inserted into a peripheral segment and the catheter is advanced beyond the tip of the bronchoscope. The inner cannula is advanced to eject the wax plug, and purulent secretions are sampled. The brush is retracted into the inner cannula, and the inner cannula is retracted into the outer cannula. Once the bronchoscope is removed, the brush can be advanced, cut with sterile wire cutters and placed in 1 mL sterile saline. Quantitative cultures are performed, and the presence of organisms in concentrations of 10³ CFUs/mL is considered to be sufficient to confirm the diagnosis of pneumonia and identify the causative pathogen.

The sensitivity for PSB ranges from 33% to 100%, with a median of 67%, and the specificity ranges from 50% to 100%, with a median of 95%. Overall, PSB appears to be more specific than sensitive in diagnosing VAP. One of the difficulties associated with quantitative invasive techniques is the reproducibility of the test, especially around the diagnostic thresholds. The test properties vary directly with the cutoff points chosen, and some studies have indicated that the results vary when multiple samples are obtained.

Other techniques have been developed because of the inconvenience, expense and necessity of operator expertise and potential side effects of diagnostic testing using fiberoptic bronchoscopy (FOB). Blinded bronchial sampling (BBS) involves blindly wedging a catheter into a distal bronchus and then aspirating secretions but without instilling fluid. Mini-BAL usually employs a single-sheathed 50-cm sterile plugging telescoping catheter with the installation of 20 to 150 mL of lavage fluid. In some instances, an unprotected catheter is used. The blinded protected specimen brush (BPSB) incorporates a sterile brush, which is protected from contamination.

The sensitivity of BBS ranges from 74% to 97%, that of mini-BAL from 63% to 100%, and that of BPSB from 58% to 86%. The specificity ranges from 74% to 100% for BBS, from 66% to 96% for mini-BAL, and from 71% to 100% for BPSB. In general, these ranges are similar to those reported for BAL and PSB. Side effects for blinded techniques appear to be minimal and, at worst, may be similar to those seen with FOB.

Several studies have indicated that management decisions based on the results of invasive tests as compared with empiric therapy or decisions based on the results of endotracheal aspirates lead to more frequent antibiotic changes. There are conflicting data as to whether invasive testing reduces mortality. A randomized controlled trial comparing the utility of bronchoscopic lavage with quantitative culture of BAL fluid or endotracheal aspirate with nonquantitative culture of the aspirate identified no significant difference in the 28-day mortality rate (the primary outcome), the rate of targeted therapy, days alive without antibiotics, hospital or ICU length of stay, or maximum organ dysfunction scores. There are advantages of diagnostic testing strategies, depending on the population studied. In patients with late-onset VAP, invasive diagnostic testing strategies can decrease antibiotic utilization and cost. It is recommended that a lower respiratory protected-specimen quantitative culture be collected prior to antibiotic administration in appropriate patients, but collection should not delay initiation of therapy.

It is important to recognize that choice of sampling methods and diagnosis vary considerably depending on expertise and availability of specialized equipment. Factors that contribute to this variability include patient populations, physician expertise and preferences, as well as institutional standards. Cost effectiveness and impact of various diagnostic approaches remain controversial and no gold standard for diagnosis of VAP has been definitively established.