Lower Respiratory Tract Infections

Community-Acquired Pneumonia

Pneumonia, and CAP in particular, is a leading cause of death in the United States, particularly in the elderly. This is in part due to the nonspecific presentation and lack of a practical diagnostic gold standard to confirm CAP; therefore the diagnosis frequently relies upon symptom assessment and clinical gestalt. Several critical decisions must be made early in the course of CAP: first, whether CAP is present; second, if the patient has risk factors for multidrug-resistant organisms (MDROs); third, which clinical setting (outpatient, inpatient, or intensive care unit [ICU]) is appropriate for management; and fourth, which antimicrobial agents and what duration of therapy are most appropriate. The Infectious Disease Society of America and the American Thoracic Society have issued joint guidelines on the management and diagnosis of CAP, which integrate evidence-based approaches to these dilemmas and have been proven to reduce mortality.

Diagnosis is based on respiratory and infectious symptoms such as cough, fever, and malaise associated with radiographic infiltrate suggestive of pneumonia. Although it is possible for a chest radiograph to be clear in pneumonia, this is highly unusual, and typically further imaging (e.g., computed tomography) is not required to establish a clinically significant pneumonia.

Other testing can support the diagnosis or help with risk stratification, but normal results do not rule out pneumonia. For example, sputum microbiology, although helpful if positive, is negative in greater than half of patients with pneumonia. Renal function, leukocyte count, and hematocrit are all useful in risk stratification schemes, such as the Pneumonia Severity Index, but do little for establishing or refuting a diagnosis. Novel biomarkers, such as procalcitonin, have shown promise in study settings, but practical implementation trials to date have shown no known diagnostic benefit, and biomarkers are currently not part of any guideline-driven CAP treatment protocols.

The next decision is where to place the patient for clinical management; the outpatient setting is appropriate for the mildly to moderately ill, the inpatient setting is needed for the moderately to severely ill, and the intensive care setting for the severely ill. Tools like the Pneumonia Severity Index (a complex score typically requiring a calculator to use, but reasonably accurate) or the CURB-65 (a simple, but less accurate, score) can be used to assist such risk stratification, but ultimately, this is a clinical decision based on the patient’s appearance, baseline level of health, and risk for progression of disease.

Once the patient’s risk factors are categorized and a working diagnosis of pneumonia is established, the next step is to obtain additional diagnostic tests based on the most likely etiologies and risk factors for the patient. This is shown in Table 9.1 .

Once diagnostics are underway, appropriate empiric treatment should be initiated based on the patient’s clinical situation. Empiric therapy is targeted at common organisms, knowing that the most commonly isolated pathogen in CAP is Streptococcus pneumoniae , with other typical organisms, including Haemophilus influenzae, Moraxella catarrhalis , and atypical organisms such as Legionella spp. and Mycoplasma pneumoniae . Viral causes of illness, such as influenza and parainfluenza, are also common, but given the propensity for bacterial superinfections to occur in the critically ill, antibacterial therapy is often given when empirically treating a severe viral pneumonia. Of note, H. influenzae is thus named because of being mistakenly attributed as the cause of influenza in the 1890s due to this organism frequently being isolated from influenza patients during a pandemic.

Conversely, it should be noted that viruses can mimic bacterial pneumonia and, particularly during influenza season, testing for influenza and other viral pathogens should be performed in hospitalized patients, both for accurate diagnosis of viral, bacterial, or coinfections causing pneumonia and in order to initiate infection prevention strategies to limit the risk of nosocomial spread of viral infections.

CAP patients may be at risk for infections with MDROs in certain settings. Patients who were recently hospitalized, on antibiotics, or chronically ill with immune or pulmonary disease or chronic illness may lead to colonization with organisms such as Pseudomonas aeruginosa , notorious for developing resistance to first-line antimicrobial agents.

For common organisms (not the potential MDROs), empiric treatment is targeted at these organisms. For mild cases, a macrolide antibiotic or, as an alternative, doxycycline may be appropriate. As patients become more severely ill, options diverge to either be fluoroquinolone based or a combination of a β-lactam and macrolide, as this provides common and atypical bacteria coverage, and dual therapy carries potential morbidity and mortality benefits for those with severe S. pneumoniae pneumonia. In the ICU setting, a combination of a β-lactam (or cephalosporin) plus a macrolide or fluoroquinolone antibiotic is appropriate empiric coverage ( Fig. 9.1 ). In patient at risk for MDRO, coverage for specific organisms (methicillin-resistant Staphylococcus aureus [MRSA], resistant Pseudomonas , etc.) may be indicated.

Generally speaking, factors that increase the suspicion for MRSA include Gram-positive cocci on the sputum Gram stain, known intravenous (IV) drug abuse, recent influenza infection, prior recent antimicrobial exposure, or if patient is on dialysis. In these situations, vancomycin or linezolid may need to be added to the regimen for empiric MRSA coverage. Of note, daptomycin should not be used for pulmonary infections because it is inactivated by surfactant and therefore cannot adequately penetrate the lungs to treat pneumonia.

Viral causes of CAP are common, with influenza being the most concerning. Some strains of influenza can be treated with antiviral medications such as oseltamivir, but meta-analytic reports have not consistently shown a benefit with this treatment, particularly if started >48 hours after symptom onset. Nonetheless, antiviral treatment remains part of most practitioners’ approach to suspected or confirmed influenza.

A final consideration for CAP is aspiration pneumonia; in patients with known dysphagia or considered at risk for aspiration, the organisms involved may include more oral anaerobes than those typically micro-aspirated in CAP. In such a situation, it is reasonable to add anaerobic coverage such as amoxicillin–clavulanate or clindamycin to the empiric coverage. This is controversial, as much of the radiographic and clinical presentation of aspiration may be due to chemical pneumonitis from stomach contents, as opposed to true infection, and thus the need for any antimicrobial therapy is a clinical judgement call.

Duration of antimicrobial therapy for any of the described CAP infections is typically 5 to 7 days, depending on clinical improvement. Often, patients can be de-escalated or have antimicrobials stopped altogether if a single etiologic agent is identified and can be targeted or a better alternative diagnosis becomes evident.

Health Care–Associated Pneumonia and Ventilator-Associated Pneumonia

HAP and ventilator-associated pneumonia (VAP) are clinically differentiated from CAP based on the risk factors noted within their names. A pneumonia is considered “health care associated” if symptom onset is ≥48 hours after hospital admission, and VAP is similarly defined by onset ≥48 hours after intubation (noting that there is a difference between clinical definitions and the reported infection control definitions used). Similar to other pneumonias, the pathogenesis of HAP is thought to be due to micro-aspiration of oropharyngeal flora. However, due to the more vulnerable population seen in the hospital, their risk for aspiration is presumed to be higher, and their oral flora is anticipated to change to reflect the higher proportion of MDROs in the hospital setting relatively quickly. Thus a patient is considered at risk for HAP after being in the hospital after 48 hours, though the longer a patient is hospitalized, the greater their risk for MDRO becomes.

Among the hospital flora, the most common Gram-negative organisms of concern are Escherichia coli, Klebsiella pneumoniae, Enterobacter spp. , Pseudomonas, and Acinetobacter . The main Gram-positive bacteria of concern is MRSA. The risk factors for MDRO in this population are prolonged hospitalization (>4 days), recent use of (>4 days) IV antimicrobials (within the last 90 days), presence of acute respiratory distress syndrome (ARDS) or shock, need for renal replacement therapy, or a high prevalence of MDROs in the community. For the purposes of risk stratification, a high prevalence is considered 10% to 20% of S. aureus isolates being MRSA or 10% of Gram-negative rods demonstrating multidrug resistance. Many institutions now internally report “antibiograms,” which list the rates of bacterial resistance to make clinicians aware of the prevalence of antimicrobial resistance and help guide empiric antimicrobial drug choices. If local resistance is unknown, it should be considered high until proven otherwise.

Viral and fungal causes of HAP and VAP are less common, though influenza can cause both HAP and VAP if infection control adherence is suboptimal. Cytomegalovirus (CMV) and Epstein–Barr virus (EBV) are rare causes of HAP or VAP and more commonly represent reactivation of the virus due to severe illness and thus do not represent invasive disease or require antiviral therapy. Another common point of confusion in this population is whether Candida species noted on Gram stain or bronchoalveolar lavage are pathogens. Once again, pneumonia caused by Candida species is rare and most often represents colonization of the airway with this organism, not invasive pathogenic infection. It has been questioned whether Candida pneumonia exists as a clinical entity at all, but if it does occur, it is only in the most severely immune compromised hosts.

The clinical diagnosis of HAP/VAP, similar to CAP, is primarily clinical. Infiltrates are usually seen on chest radiography, but other signs of infection such as fever, purulent sputum, hypoxia, and leukocytosis are not as consistent and require interpretation in each individual case. Sputum cultures have low sensitivity in HAP, but negative sputum is considerably more useful for ruling out VAP, likely reflecting the ease of obtaining a higher-quality specimen than can be obtained via an endotracheal tube. Molecular tests can be used, such as multiplex polymerase chain reaction (PCR), but these should be considered for de-escalation purposes based on identification of an etiologic agent, not as the basis for withholding or administering antibiotics.

In patients without MDRO risk factors, piperacillin–tazobactam, cefepime, or levofloxacin are appropriate empiric regimens. If an MDRO is suspected, empiric coverage options include piperacillin–tazobactam, cefepime, ceftazidime, imipenem, meropenem, or aztreonam for Gram-negative coverage. Double coverage for Pseudomonas aeruginosa can be used for high-risk patients by adding an aminoglycoside for additional coverage for highly resistant organisms. Further use of double antimicrobial coverage should depend on local “antibiogram” data. The need for empiric coverage of Gram-positive MDROs is identical to that of CAP. As noted earlier, in patients at risk for MDRO pneumonia, vancomycin or linezolid are appropriate first-line choices for MRSA coverage.

Duration of therapy is typically 7 days for HAP and VAP. Longer regimens used to be used for select organisms such as Acinetobacter and Pseudomonas , but in the most recent guidelines, the recommended duration was shortened to only 14 days. De-escalation should occur when susceptibility and culture results are available, or 48 hours after cultures are obtained if there is no growth. Typical antimicrobial de-escalation removes MDRO coverage in the absence of a demonstrated MDRO at that time and sets a definitive end date for the treatment program.

Biomarkers play a limited role in HAP and VAP. Procalcitonin, probably the most studied of these, is not recommended to use as a basis for initiating or withholding treatment. Current guidelines advise considering serial procalcitonins for de-escalation for patients beyond the 7-day treatment time frame; thus in most cases, procalcitonin should not affect therapy. Other biomarkers have less data to advise their proper use.