Community-Acquired Pneumonia.

Community-acquired pneumonia is one of the most common infections seen in emergency department patients. There is a wide spectrum of disease severity and viral pathogens are common. After a careful history and physical examination, chest radiographs may be the only diagnostic test required. The first step in management is risk stratification, using a validated clinical decision rule and serum lactate, followed by early antibiotics and fluid resuscitation when indicated. Antibiotics should be selected with attention to risk factors for multidrug-resistant respiratory pathogens. Broad use of pneumococcal vaccine in adults and children can prevent severe community-acquired pneumonia.

Key points

History, physical examination, including vital signs and saturation of peripheral oxygen, and chest radiographs results provide the essential information to clinically diagnose community-acquired pneumonia.

Careful severity assessment is a crucial step in the emergency department management of community-acquired pneumonia and should include screening for occult sepsis with a serum lactate, followed by early antibiotics and fluid resuscitation when indicated.

Risk stratification tools such as the PSI and CURB-65 should be used routinely to determine the most appropriate disposition.

Emergency department providers need to be aware of risk factors for multidrug-resistant pneumonia, limiting broad spectrum antibiotics to patients satisfying guideline-recommended criteria.

Introduction

Pneumonia is a commonly encountered respiratory infection in the emergency department (ED) that is responsible for significant morbidity and mortality in our patients. Community-acquired pneumonia (CAP) is defined as an acute lung infection involving the alveoli that occurs in a patient without recent health care exposure. CAP encompasses a clinical spectrum from walking pneumonia in an otherwise healthy patient to necrotizing or multilobar disease with septic shock. Pneumonia is the third leading reason for hospital admission, accounting for 544,000 hospitalizations from the ED annually. Despite advances in medicine, the mortality rate from CAP has remained stable over the past 4 decades. In the United States, CAP is the leading cause of sepsis and death from infection. Given the prevalence of CAP and its potential to cause severe illness, emergency providers must have a thorough understanding of this multifaceted condition and be able to take a nuanced approach to management. Emergency physicians need to recognize symptoms suggestive of CAP, order appropriate diagnostic tests, select recommended empiric antibiotics, and risk stratify the patient for proper disposition. This article provides an overview of CAP in adults and touches on drug-resistant and health care-associated disease. Opportunistic lung infections, tuberculosis, and hospital-acquired pneumonia are beyond the scope of this review.

Microbiology

The etiology and antibiotic resistance patterns of respiratory pathogens varies by geographic region and has evolved over time with the development of vaccines. In the majority of cases requiring hospitalization, no pathogen can be identified. In a 2015 US population-based surveillance study of patients admitted with CAP, only 38% of patients had a pathogen identified, and most were viral. A bacterial pathogen could be isolated in only 14% of patients. The most common pathogen was rhinovirus, followed by influenza virus, then Streptococcus pneumoniae. Mycoplasma pneumoniae and Staphylococcus aureus were the second and third most common bacterial pathogens, respectively. Other bacterial species that commonly cause CAP include Legionella pneumophila, Haemophilus influenzae, Chlamydophila pneumoniae, and Moraxella catarrhalis. 1, 4

Common viral causes include not only human rhinovirus and influenza, but also human metapneumovirus, parainfluenza virus, respiratory syncytial virus, coronavirus, adenovirus, and Middle East respiratory syndrome coronavirus. During peak influenza season, influenza may be the most common cause of CAP requiring hospitalization, although it can often be complicated by secondary bacterial infection. Fungal etiologies are generally rare in immunocompetent hosts. Coccidioidomycosis is a relatively common cause of pneumonia and pneumonitis in the Western United States that can mimic bacterial pneumonia. Other geographic endemic mycoses include Histoplasma capsulatum and Blastomyces in the Ohio and Mississippi River valleys. Opportunistic fungal pneumonias frequently seen in patients with AIDS and solid organ transplant include Pneumocystic jiroveci pneumonia, Aspergillus, Candida albicans, and Cryptococcus neoformans (Table 1 ).

Table 1

List of pneumonia pathogens according to patient population

Patient Population	Pathogens
Otherwise healthy adult, bacterial	Streptococcus pneumoniaeMycoplasma pneumoniaeStaphylococcus aureusLegionella pneumophilaHaemophilus influenzaeChlamydophila pneumoniaeMoraxella catarrhalis
Otherwise healthy adult, viral	Human rhinovirusInfluenzaHuman metapneumovirusParainfluenza virusRespiratory syncytial virusCoronavirusAdenovirus
Adults with health care exposure	Pseudomonas aeruginosaS aureusKlebsiella pneumoniaeEscherichia coli
Pediatric patients – by age Birth to 20 d 20 d to 4 mo 4 mo to 5 y School-aged children	E coliListeria monocytogenesGroup B streptococciChlamydia trachomatisStreptococcus pneumoniaeC trachomatisStreptococcus pneumoniaeM pneumoniaeRespiratory syncytial virusInfluenzaParainfluenzaAdenovirusRhinovirusC pneumoniaeM pneumoniaeStreptococcus pneumoniaeM catarrhalisH influenzaeS aureus
Unusual and opportunistic infectious etiologies	Pneumocystic jiroveciHistoplasma capsulatumBlastomycesCoccidioidomycosisAspergillusCandida albicansMucoralesCryptococcus neoformansCoxiella burnetiiMycobacterium tuberculosis

Microbiology: drug-resistant pathogens

The categorization of pneumonia is evolving. Until recently, pneumonia was divided into 4 categories: CAP, health care-associated pneumonia (HCAP), hospital-acquired pneumonia, and ventilator-associated pneumonia. Hospital-acquired pneumonia and ventilator-associated pneumonia are outside the scope of this article. HCAP is a category that includes patients who have been in regular contact with the health care system, including nursing home residents, patients undergoing home infusion therapy or wound care, dialysis patients, and patients hospitalized for 2 days or more in the prior 90 days. Such patients are thought to have a higher risk of pneumonia caused by multidrug-resistant (MDR) bacteria, warranting broad spectrum antibiotic coverage, similar to hospital-acquired pneumonia. Common MDR pathogens include Pseudomonas aeruginosa, methicillin-resistant S aureus (MRSA), and gram-negative Enterobacteriaceae species. This topic is currently in flux because HCAP criteria are neither sensitive nor specific for identifying patients infected with MDR organisms. Treating these patients with the same regimen as for hospital-acquired pneumonia leads to overtreatment with broad spectrum antibiotics. Excess mortality in HCAP may be largely attributable to patient comorbidities rather than drug-resistant pathogens. The Infectious Diseases Society of America (IDSA) and American Thoracic Society are removing the HCAP category from their guidelines. Forthcoming guidelines will recommend that the group formally known as HCAP be divided into 2 groups, those appropriate for limited spectrum therapy and those with 2 of 3 risk factors for MDR, who do require broad spectrum therapy. Table 2 lists risk factors for the most important MDR organisms. Although evidence on this topic remains incomplete and exact recommendations are not yet clear, it remains very important to identify ED patients who require broad spectrum empiric antibiotics.

Table 2

Risk factors for drug-resistant pneumonia pathogens

Drug-Resistant Pathogen	Risk Factors
Drug-resistant streptococcus	Age >65Beta-lactam or macrolide therapy within 3 moImmunosuppressionAlcoholismDaycare centersMedical comorbidities
Enteric gram negative	Residence in a nursing homeRecent hospitalizationRecent antibioticsCardiopulmonary diseaseSmokingUnderlying malignancy
MRSA	Age >74 yDialysisPrior MRSA infectionPrior hospitalizationRecent nursing home stayMedical comorbidities
Pseudomonas	Chronic obstructive pulmonary diseaseImmunosuppressionRecent steroid exposureHemiplegiaRecent antibiotics against gram positive organismsRecent hospitalization

Abbreviation: MRSA, methicillin-resistant Staphylococcus aureus.

Historically, penicillin was sufficient treatment for S pneumoniae, but over the past 3 decades the prevalence of drug resistant pneumococcus has increased. Alteration in the penicillin-binding protein is the main resistance mechanism. In the United States, pneumonia owing to penicillin nonsusceptible strains increased from 18% in 1991 to 35% in 2002. The clinical significance of penicillin-resistant S pneumoniae remains unclear, with mixed data on mortality impact and associated costs and durations of stay.9, 11 However, adverse outcomes are associated with high-level penicillin resistance (minimum inhibitory concentration of ≥4) and these infections require treatment with a cephalosporin or fluoroquinolone.

Macrolide-resistant pneumococcus is also a growing problem. The mechanism for high-level resistance involves methylation of a ribosomal binding site, whereas lower level resistance involves drug efflux via a membrane transporter. A study conducted in Japan, where macrolide-resistant strains exceed 90% in some areas, found that 83% of patients still had a good clinical response to azithromycin, suggesting that the clinical significance of resistance is unclear in vivo. In the United States, macrolide-resistant pneumococcus averages 27.9%, with the greatest prevalence in Louisiana and state-by-state variation as high as 33%. Macrolide resistance should be considered in high-resistance regions, particularly when there has been recent antibiotic exposure and macrolide monotherapy is being considered.

Macrolide-resistant M pneumoniae emerged around 2000 as another concerning drug-resistant CAP pathogen. Resistance is conferred from point mutations where the macrolide binds to the ribosome subunit. A study of 6 centers throughout the United States revealed an M pneumoniae macrolide resistance rate of 13.2%.

Pathophysiology

Pneumonia is an alveolar infection that occurs when the innate immune system is unable to clear a pathogen from the lower airway and alveoli. Local inflammatory factors and cytokines cause additional harm to the lung parenchyma and lead to systemic inflammation, which causes secondary symptoms such as fever, chills, and fatigue. At a histologic level, the inflammatory response causes congestion, which progresses to red and gray hepatization, and may resolve with minimal fibrosis. In terms of lung mechanics and physiology, pus in the parenchyma leads to decreased compliance and shunt, which increases the work of breathing and worsens hypoxemia and tachypnea—the most important physical examination signs of severe pneumonia for emergency physicians to focus on at the bedside.

CAP affects patients of all ages across the spectrum of health, with certain organisms having a predilection for specific patient subgroups. Any condition that causes decreased mucociliary clearance and cough, like cigarette smoking, puts patients at increased risk, as do conditions that lead to aspiration such as cerebral vascular accidents, esophageal disorders, and neuromuscular disorders. Old age and dehydration affect how pneumonia manifests and can make recognition more difficult. Underlying cardiopulmonary disease or structural lung disease can also delay recognition.

Epidemiology

Pneumonia carries the highest mortality of any infectious disease. Lower respiratory tract infections are the most common infectious cause of death in the world, with 3.5 million deaths annually worldwide (World Health Organization). In the United States, influenza and pneumonia are listed together as the 9th leading cause of death. The highest rates are among elderly adults aged 65 to 79 years. Outcomes in patients requiring hospitalization for pneumonia are poor: the 30-day mortality is 10% to 12% and the readmission rate is 18%. Interestingly, mortality after a CAP hospitalization remains increased at 1 year and 5 years. Patients with chronic obstructive pulmonary disease, diabetes mellitus, renal failure, congestive heart failure, coronary artery disease, and liver disease have an increased incidence of CAP. Fortunately, the use of the pneumococcal conjugate vaccine may be responsible for up to a 35% decrease in the incidence of pneumonia.

Differential diagnosis

In day-to-day practice, the main differential diagnosis in an immunocompetent ambulatory patient presenting with acute cough illness is CAP versus viral bronchitis. Correctly distinguishing between the two depends in large part on elements of the history and physical examination, particularly age, upper respiratory infection symptoms, pulse oximetry, and lung sounds. In EM practice, chest radiographs also plays a central role. CAP is distinguished by the presence of an alveolar infiltrate on chest radiographs, whereas those with bronchitis will have a normal chest radiographs (with the possible exception of peribronchial thickening). Accurately distinguishing pneumonia from acute bronchitis is one of the most important ways that emergency physicians can improve antibiotic stewardship.

In the case of acute cough with a negative radiographs, possible diagnoses include viral bronchitis, asthma, gastroesophageal reflux disease, postnasal drip, sinusitis, or medication side effect, particularly an angiotensin-converting enzyme inhibitor–induced cough (Table 3 ). When an infiltrate is present on radiographs, important noninfectious causes to consider include pulmonary edema, pulmonary embolism with pulmonary infarction, lung cancer, alveolar hemorrhage (arteriovenous malformation, Goodpasture’s syndrome, Wegener’s granulomatosis), bronchiectasis, cryptogenic organizing pneumonia, acute eosinophilic pneumonia, interstitial lung diseases, vasculitis, cocaine-induced lung injury, pulmonary contusion, drug reaction, and high altitude pulmonary edema, among others.

Table 3

Differential diagnosis of noninfectious causes of an infiltrate on chest radiographs, and differential diagnosis of cause of a normal chest radiographs in the setting of acute cough illness

Chest Radiograph Findings	Causes
Abnormal chest radiograph, infectious	Refer to micro table
Abnormal chest radiograph, noninfectious	Cardiogenic pulmonary edemaBronchiectasisPulmonary infarctionArteriovenous malformationInterstitial lung diseaseCryptogenic organizing pneumoniaAcute eosinophilic pneumoniaPneumonitisVasculitisCocaine-induced lung injuryPulmonary contusionDrug reactionHigh altitude pulmonary edemaLung cancer
Normal chest radiograph	BronchitisAsthmaGastroesophageal reflux diseaseUpper respiratory tract infectionMedication side-effect

When the chest radiograph is abnormal, in addition to the common bacterial and viral infectious etiologies, emergency physicians need to keep in mind less common infectious causes of pneumonia. Immunocompetent hosts may develop infections from endemic mycoses such as histoplasmosis, blastomycosis, or coccidiomycosis. Septic emboli should be considered in patients with multiple sites of infection or history of injection drug use. Immunocompromised hosts, particularly patients infected with the human immunodeficiency virus (HIV), are at risk for opportunistic lung infections, including from fungal pathogens, such as P jiroveci (P jiroveci pneumonia), Aspergillus, C albicans, Mucormycosis, C neoformans, as well as Mycobacterium tuberculosis and Mycobacterium avium complex. Tuberculosis and P jiroveci pneumonia are remarkably common and easily misdiagnosed causes of pneumonia in HIV-infected patients with CD4 counts of less than 500 cells/mm3. When evaluating pneumonia, emergency physicians should consider the possibility that the patient has HIV or is otherwise immunocompromised.

Diagnosis: History and Physical Examination

Although the diagnostic criteria for CAP seem relatively straightforward, making the correct diagnosis can be difficult. A thoughtful history and physical examination with close attention to the actual respiratory rate and core temperature, as well as careful interpretation of chest radiographs, are required. This caution is especially true in the elderly. The clinical diagnosis of CAP is made on the basis of respiratory symptoms such as cough, sputum production, dyspnea, chest pain, signs of fever, and hypoxemia, as well as an infiltrate on chest imaging. Additional symptoms may include myalgia, fatigue, abdominal pain, and headache, making CAP difficult to distinguish from viral infection, particularly influenza, based on history alone. Possible chest examination findings include dullness to percussion, decreased breath sounds, and inspiratory crackles.

Because patients in the ambulatory setting with symptoms suggestive of a respiratory infection have a prevalence of pneumonia of only about 5%, it is possible to rule out CAP without a chest radiograph. Nonelderly patients without any abnormal vital signs and without a focally abnormal auscultatory examination have a probability of CAP (abnormal chest radiographs) of less than 1%. Ruling out CAP without an radiographs should not be attempted, however, in elderly patients, because many will not mount typical signs and symptoms. As many as 30% of elderly patients with CAP are afebrile in the ED and preexisting oxygenation problems and lung disease can further complicate the picture.

Fig. 1 outlines a simplified diagnostic approach for ambulatory patients with cough and a suspicion for CAP. Patients who are not elderly, with normal vital signs and normal auscultatory examination, can generally be given a diagnosis of bronchitis and discharged home without antibiotics. If vital signs or examination are abnormal, or the patient is elderly, a chest radiograph should be obtained. If the film is normal, the correct diagnosis is usually bronchitis. Rarely, very early in the disease course, or in severely dehydrated patients, the chest radiographs will be normal in CAP, but this is an uncommon issue in ambulatory patients less than 65 years old. If the chest radiographs confirms the diagnosis of CAP, the next step is risk stratification (see Severity Assessment and Clinical Decision Rules). In many patients who are determined to be at low risk for mortality by a clinical decision rule, no further diagnostic testing is required before discharge. The only important remaining step in such patients is proper oral antibiotic selection (see Fig. 1).

Fig. 1

Summary of the diagnostic approach to acute cough illness in ambulatory patients. This shows the central role of the chest radiographs and, in cases of community-acquired pneumonia (CAP), severity assessment (typically with a clinical decision rule). Abnl, abnormal; abx, antibiotics; ausc., auscultation; cx, culture; CXR, chest radiograph; ICU, intensive care unit; inpt, inpatient; IV, intravenous; outpt, outpatient; tx, treatment; VS, vital signs.

Diagnosis: Imaging

A chest radiograph is the crucial diagnostic test for CAP in adults. Physical examination alone is less sensitive and specific than a chest radiograph. Infiltrates may be subtle, so a study that includes both the posteroanterior and lateral views is preferred. Opacities typically develop within 12 hours. The most common findings include peribronchial nodules, silhouette sign, parapneumonic effusions, and ground glass opacities.26, 27 More severe pneumonia is characterized by multilobar involvement, cavitation, and bilateral pleural effusions. However, the radiographic appearance does not necessarily correspond with clinical severity and infiltrates may evolve independent of clinical improvement or worsening.

Although etiology cannot be reliably predicted by chest radiograph appearance, the following classic associations between radiographic findings and etiology are described in the literature. S pneumoniae typically appears as alveolar or lobar pneumonia. The lower lobe and multilobar involvement is frequently seen. Bilateral disease and interstitial infiltrates are found in 50% of cases. Mycoplasma typically produces reticulonodular opacities or patchy consolidations. Mycoplasma targets bronchial epithelium and, therefore, can cause bronchial wall thickening (peribronchial cuffing) in central bronchi like respiratory viruses. Similarly, Chlamydia pneumoniae classically shows patchy consolidations or reticular opacities. Viral pneumonitis/pneumonia can be radiographically indistinguishable from bacterial CAP, but ground glass opacities are the most common findings after a normal chest radiographs. Finally, aspiration pneumonia commonly affects the posterior and inferior segments of the lung and may cavitate if subacute or chronic.

The interpretation of a chest radiographs, however, is an imperfect practice. Not only do these radiographic patterns poorly predict etiology, but radiologists may miss infiltrates in up to 15% of cases. The timing of CAP presentation may also affect radiographic appearance. The chest radiographs may seems to be normal in the first few hours of S pneumoniae, in patients with HIV with early P jirovecii infection, or in the setting of severe dehydration, as is common in patients from skilled nursing facilities. In such cases, a clinical diagnosis of pneumonia should be considered even if the chest radiograph seems to be normal, and it is reasonable to treat for pneumonia and repeat the radiograph in 1 to 2 days.

Computed tomography (CT) is more sensitive than plain radiographs for detecting pneumonia. Although the clinical significance of an infiltrate seen on CT but not on chest radiographs is debated, a recent prospective surveillance study showed that illness severity, pathogens, and clinical outcomes were comparable between patients with CAP with an abnormal chest radiographs and those diagnosed by CT scan without an obvious infiltrate on radiographs. If clinical suspicion for pneumonia remains high despite a negative chest radiograph, advanced imaging should be considered and empiric treatment initiated. CT scans may be better at visualizing certain areas of the lung, such as the upper lobes and lingula, and at elucidating interstitial infiltrates as seen with atypical pathogens. CT scanning is useful to further characterize necrotizing infection, multilobar disease, empyema, and pleural involvement. It can also help to differentiate CAP from tuberculosis or lung cancer, which can be difficult to distinguish on chest radiographs alone.

Guidelines do not support the use of ultrasound examination for the diagnosis of pneumonia in adults when other imaging modalities are available. However, a metaanalysis concluded that, in the hands of experienced operators, ultrasound examination may have a sensitivity and specificity as high as 94% and 96%, respectively. Ultrasound examination may offer an ideal alternative diagnostic modality in pediatric patients and critically ill patients in whom it is difficult to obtain a 2-view film (Fig. 2 ).

Fig. 2

Chest radiograph demonstrating lobar pneumonia.

Diagnosis: Additional Testing

When history, physical examination, and imaging studies confirm the diagnosis of CAP, additional etiologic testing may or may not be required depending on the patient’s clinical status and disposition. This area holds considerable controversy in emergency medicine. Identification of the causative organism uses resources and is costly; even when it reveals the pathogen, etiologic testing only rarely leads to a change in treatment. In contrast, cases in which an organism is identified may allow for deescalation from broad to more narrow spectrum treatment during the hospitalization, thereby decreasing cost of care and preventing adverse effects. Etiologic testing is also important from a public health standpoint, allowing for epidemiologic monitoring of drug susceptibility and the incidence of specific pathogens.

The IDSA recommends directing testing toward patients with the highest expected yield. For outpatients, etiologic testing, with the exception of a rapid diagnostic tests for influenza, is rarely indicated. If obtained before antibiotic administration, blood cultures grow a pathogen in 5% to 14% of CAP cases, depending on disease severity. The sensitivity of blood cultures is halved if antibiotics have already been administered. The most common isolate is S pneumoniae. Although pneumococcal infections rarely require a change in therapy based on isolate susceptibility, these isolates do allow surveillance for penicillin and macrolide resistance. Among hospitalized patients, blood cultures should always be drawn in those with severe CAP requiring admission to the intensive care unit (ICU). In patients admitted to the floor, blood cultures are optional, but should be performed in patients with alcohol abuse, liver disease, leukopenia, effusion, or asplenia.

Patients with severe CAP should also have a respiratory specimen obtained for culture and Gram stain. Sputum should be obtained within 12 hours of initiation of antibiotics. The yield of respiratory cultures is significantly higher from endotracheal aspirates or bronchoscopic sampling and should be sent on all intubated patients with CAP. Absence of S aureus or gram-negative rods in a respiratory specimen suggests these pathogens are not the cause of illness and can allow the antibiotic regimen to be narrowed early in the hospitalization.

Although use of polymerase chain reaction has increased the detection of viral respiratory pathogens in CAP, the role of these tests, other than for influenza, remains unclear. Influenza testing is recommended for admitted patients when local influenza activity is high. It is also recommended that patients with severe CAP have urinary antigen tests for S pneumoniae as well as Legionella pneumophilia if Legionella serogroup 1 is suspected. S pneumoniae urinary antigen tests remain positive for 3 days after initiating therapy. Local health department notification is required if Legionella is detected because this may indicate an outbreak. In general, more etiologic testing should be done for patients with a history of alcohol abuse, liver disease, lung disease, leukopenia, cavitary infiltrates, asplenia, pleural effusion, and recent travel.

Serum lactate is a widely recommended screening test for severe sepsis. When elevated in the setting of infection, the lactate level independently predicts mortality. A prospective observational study comparing CURB-65 and serum lactate in 1641 patient with CAP showed that lactate better predicted 28-day mortality, hospitalization, and ICU admission. Lactate should routinely be drawn along with blood cultures as part of the workup for patients being hospitalized with severe CAP to help guide resuscitation, treatment, and disposition.

Testing for acute phase reactants has a growing role in evaluation of CAP and other infectious diseases, particularly in pediatric populations. Procalcitonin is an acute phase reactant that has a very low circulating level (<0.15 ng/mL) normally, but increases with inflammatory diseases, particularly in response to bacterial toxins. The procalcitonin level may be helpful in distinguishing bacterial from viral infections, although the data are not consistent.3, 15 Procalcitonin may also serve as a marker of disease severity; if low or decreasing, it may allow for the deescalation or termination of antibiotics in the inpatient setting. In a metaanalysis of 14 randomized trials, the use of procalcitonin was associated with a decrease in the duration of antibiotic therapy from 8 to 4 days, without a change in mortality. Another metaanalysis showed lower mortality among critically ill patients who were allowed to have procalcitonin-guided cessation of antibiotics. Other biomarkers currently under investigation to differentiate viral from bacterial illness include C-reactive protein and cortisol. Although biomarkers may have a future role in risk-stratification of CAP in the ED setting, ED studies are limited at this time (Fig. 3 ).

Fig. 3

Ultrasound image showing air bronchograms suggestive of pneumonia.

Severity assessment and clinical decision rules

Owing to the very wide spectrum of disease severity in CAP, it is strongly recommended that emergency physicians routinely use a structured, clinical decision rule to risk stratify patients with CAP. A number of rules have been developed and validated, all of which are used to determine the intensity of diagnostic testing and optimal safe disposition of the patient—to home, inpatient ward, or inpatient ICU. Besides helping to identify the sickest subset of patients, another important benefit of clinical decision rules is that they promote safe discharge of low-risk patients who might otherwise be admitted to the hospital unnecessarily.

The IDSA divides the signs of severe CAP into major and minor criteria. Approximately 10% of patients hospitalized with CAP will require an ICU stay. It is absolutely crucial that emergency physicians recognize the signs of severe disease to ensure appropriate disposition. The minor criteria include:

Respiratory rate of greater than 30 breaths per minute;

Pao 2/Fio 2 of less than 250;

Multilobar infiltrates;

Confusion;

Uremia;

Leukopenia;

Thrombocytopenia;

Hypothermia; and

Hypotension.

The clinically less helpful major criteria are mechanical ventilation and septic shock requiring vasopressors. ICU admission is recommended with any major criteria or 3 or more minor criteria. The minor signs represent a valuable guide to identifying subtle organ system dysfunction and hypoperfusion. A careful assessment for these subtle signs of poor organ perfusion is the key to effective care of patients with CAP. Critical early interventions include prompt antibiotic treatment, monitoring serial blood gases and serum lactate levels, and frequent reassessment.

The 2 most widely validated and used CAP clinical decision rules are the Pneumonia Severity Index (PSI) and CURB-65. The PSI was developed in 1997 by the Pneumonia Patient Outcomes Research Team in an effort to predict short-term CAP mortality (Table 4 ). It is intended for immunocompetent adults based on data available at presentation. The higher the score, the higher the risk of death or eventual admission to the ICU. In practice, the score helps emergency providers to determine an appropriate disposition destination. PSI class IV and V patients should be hospitalized, with class V usually requiring ICU admission. Class III patients may be appropriate for 23-hour observation, 1 or 2 doses of intravenous antibiotics, and hydration. Classes I and II patients can usually be safely managed as outpatients. The PSI has been externally validated in several large trials and a randomized control trial at 19 hospitals confirmed that using it safely reduces low risk admissions.

Table 4

Clinical decision rules

Clinical Decision Rule	Factors	Points	Score and Stratification
Pneumonia Severity Index (PSI)	MaleAge >50Nursing home residentNeoplastic diseaseCongestive heart failureCerebrovascular diseaseRenal diseaseLiver diseaseAltered mental statusHR ≥125 beats/minRR ≥30 breaths/minSystolic BP <90 mm HgTemperature <35°C or ≥40°CArterial pH <7.35BUN ≥30 mg/dLSodium <130 mmol/LGlucose ≥250 mg/dLHematocrit <30%Partial pressure arterial O2 <60 mm Hg or O2 sat <90%Pleural effusion	AgeAge + 10303010101020201020201530202010101010	Point totalIf all are absent ≤70 low risk71–90 low risk91–130 moderate>130 high risk	Risk classIIIIIIIVV	MortalityClass I: 0.1%–0.4%Class II: 0.6%–0.7%Class III: 0.9%–2.8%Class IV: 8.2%–12.5%Class V: 27.1%–31.1%	SuggestedDispositionI, II: outpatientIII: observationIV, V: hospitalize
British Thoracic Society (BTS) modifiedOrCURB-65	Confusion/orientationBUN >20 mg/dL (7 mmol/L)RR ≥30 breaths/minLow BP: <90 SBP, ≤60 DBPAge ≥65 y	11111	Score total0–5	Mortality (30-d)0 factors: 0.7%1 factors: 2.1%2 factors: 9.2%3 factors: 14.5%4 factors: 40%5 factors: 57%		Suggested Disposition0–1: treat as outpatients2: admit to floor≥3: ICU

Abbreviations: BP, blood pressure; BUN, blood urea nitrogen; DBP, diastolic blood pressure; HR, heart rate; ICU, intensive care unit; RR, respiratory rate; SBP, systolic blood pressure.

The CURB-65 score combines just 5 variables to determine disease severity, placing more emphasis on physiologic parameters. It is much easier to calculate than the PSI (see Table 4). One point each is assigned for confusion, blood urea nitrogen of 20 mg/dL or greater, respiratory rate of 30 or greater, blood pressure less than 90 mm Hg or diastolic 60 mm Hg or greater, and age greater than 65. Patients with a score of 2 or higher should be admitted to the hospital. The CURB-65 is not as well-validated as the PSI and when the 2 measures are compared, the PSI with its 20 variables boasts a slightly higher discriminatory power for mortality and classifies a slightly higher percentage of patients as low risk, and in that sense has greater usefulness.24, 41

Other pneumonia clinical decision rules that are less-widely used in the United States include SMART-COP, A-DROP, and CAP-PIRO. The Australian SMART-COP score assigns points for low systolic blood pressure, multilobar infiltrates, low albumin, tachypnea, tachycardia, confusion, hypoxemia, and acidemia. A score of 3 or more points identified 92% of patients who would later receive vasopressor or ventilator support. The A-DROP scoring system, developed by the Japanese Respiratory Society, uses the variables age (males ≥70, females ≥75), dehydration (blood urea nitrogen >210 mg/mL), respiratory failure (SaO2 ≤90%), confusion, and hypotension. Similarities between these clinical decision rules are not surprising. The CURB-65 seems to have emerged as the preferred score in the ED because of its simplicity, although the PSI still has a role because it is better at identifying patients who seem to be “on the fence” but can actually be safely discharged.

In addition to clinical decision rules, assessment of nonmeasurable factors is important when determining disposition. Before patient discharge, emergency physicians must consider the likelihood of direct complications of the pneumonia, such as hypoxemia or pleural effusion, exacerbation of underlying disease, the patient’s ability to take oral medication, and availability of a caregiver. Discharge is appropriate when the respiratory rate is less than 24 breaths per minute, the saturation of peripheral oxygen is greater than 90%, mental status is normal, and the patient can tolerate oral intake. Finally, psychosocial factors and patient preference must be considered. Outpatient care of CAP costs 25 times less than hospitalization and patients resume normal activity faster, so providers should encourage discharge when it is deemed clinically and socially safe.

Treatment

Antibiotic therapy for CAP should be directed at the most common pathogens with consideration of local resistance patterns and patient disposition. Once the diagnosis is made, antibiotics should be administered as soon as possible. There is no longer a CMS quality metric regarding time to treatment, but a widely accepted target is within 6 hours of presentation.1, 44 Patients who exhibit signs of sepsis should receive antibiotics within 1 hour.

If available, local empiric antibiotic treatment guidelines that reflect the hospital antibiogram should be followed. In nonsevere cases, the goal is to reliably cover S pneumoniae and atypical bacterial pathogens. The decision to provide broader coverage is based on health care exposure risk factures, a history of structural lung disease, or other specific conditions (eg, known MRSA colonization).

For outpatients without coexisting illnesses or recent antibiotic use, the IDSA recommends a macrolide (azithromycin, clarithromycin) or doxycycline.1, 24 If coexisting illness is present or the patient has recently used antibiotics, a respiratory fluoroquinolone (moxifloxacin, levofloxacin) is recommended; a beta-lactam (eg, amoxicillin) plus macrolide may also be used. Risk factors for drug-resistant Streptococcus pneumoniae include comorbidities such as heart disease, liver disease, renal disease, diabetes, alcoholism, immunosuppression, or antimicrobial use within the previous 3 months (see Table 2). For these patients, a respiratory fluoroquinolone should be prescribed or a beta-lactam plus a macrolide. In 1 cohort of patients treated for CAP in the ED, approximately one-half had 1 risk factor for drug-resistant S pneumonia, raising concern for potential overuse of fluoroquinolones. In areas such as Louisiana with a large percentage (48%) of macrolide-resistant S pneumoniae (defined as a minimum inhibitory concentration of ≥16), a respiratory fluoroquinolone or beta-lactam plus macrolide can be used. See Tables 5 for treatment guidelines.

Table 5

Treatment of adults with community-acquired pneumonia

Patient Characteristics	Regimen
Outpatient: previously healthy	MacrolideDoxycycline
Outpatient: with comorbidities (heart, lung renal disease, diabetes, alcoholism) or recent use of antibiotics concerning for drug-resistant Streptococcus pneumoniae	Respiratory fluoroquinoloneBeta-lactam plus macrolide
Outpatient: macrolide-resistance streptococcus areas (>25% of infection)	Respiratory fluoroquinoloneBeta-lactam plus macrolide
Inpatient: floor	Respiratory fluoroquinoloneBeta-lactam plus macrolide
Inpatient: intensive care unit	Beta-lactam plus azithromycin or respiratory fluoroquinolonePenicillin allergic: respiratory fluoroquinolone plus aztreonam
Inpatient: Pseudomonas	Antipseudomonal beta-lactam such as piperacillin-tazobactam, cefepime, meropenem, imipenem plus ciprofloxacin or levofloxacin
Inpatient: methicillin-resistant Staphylococcus aureus	Vancomycin or linezolid

For hospitalized patients, the IDSA guidelines recommends a beta-lactam plus a macrolide (eg, ceftriaxone plus azithromycin). There is increasing evidence that patients do better with a combination of antibiotics rather than fluoroquinolone monotherapy, possibly related to immunomodulation. During high local influenza activity, hospitalized patients generally should also be treated with oseltamivir. Droplet and contact precautions should be used when influenza is suspected.

The risk for MDR pathogens must be considered before selecting a treatment regimen for hospitalized patients (see Table 2). In those with 2 or more MDR risk factors such as medical comorbidities, recent hospitalization, or recent antibiotics, immunosuppression coverage for P aeruginosa is recommended. Pseudomonas, which is invariably MDR, may cause 1% to 8% of severe CAP cases, and is associated with a case fatality rate of 50% to 100%. To cover MDR organisms, an antipseudomonal cephalosporin (eg, cefepime, ceftazidime), carbapenem (eg, meropenem, imipenem), or antipseudomonal penicillin (eg, piperacillin-tazobactam) plus an antipseudomonal fluoroquinolone is recommended. MRSA coverage with vancomycin or linezolid should be added in patients with suspected recent or coinfection with influenza, chronic glucocorticoid use, or other risk factors for MRSA (see Table 2).18, 46

The recommended duration for CAP therapy is 5 to 7 days. Evidence suggests there is no difference in outcomes when treatment duration is 7 days or less compared with 8 days or more. Exceptions include S aureus lobar pneumonia, which may require extended treatment for 2 weeks, and S aureus bacteremia, which generally requires 4 weeks of intravenous treatment. Atypical infections are also an exception; if M pneumoniae or C pneumoniae are known to be the causative organism, 10 to 14 days is recommended. For Legionella, 14 to 21 days of therapy is recommended. For hospitalized patients, an early switch from intravenous to oral antibiotics does not compromise outcome and decreases the duration of stay.

In addition to antimicrobials and symptomatic care, other CAP treatment adjuncts should be considered, corticosteroids being the most important and widely debated. Steroids may attenuate the inflammatory response, reduce the frequency of acute respiratory distress syndrome, and decrease the length of illness. A systematic review and metaanalysis suggested that steroids reduce the need for mechanical ventilation and rate of acute respiratory distress syndrome by 5% (estimated number needed to treat of 20). Despite this evidence in favor of steroids, there are many high-quality studies showing no benefit. The case for adjunctive steroids is stronger in severe CAP. Steroids should be trialed in patients with vasopressor-dependent shock and selectively in patients with severe CAP and evidence of inadequate cortisol response. Early physical therapy is another important treatment adjunct in hospitalized patients. A single-center retrospective study showed an association between physical therapy for 30 minutes or more and lower 30-day readmission rate.

Long-Term Host Effects

Discharged patients should be informed about the usual course of illness in CAP. In previously healthy adults with pneumococcal pneumonia, fevers typically resolve within 3 days of initiating antibiotics. One week after presentation, 80% of patients will still have fatigue and 50% will have dyspnea. Overall, patients improve clinically much faster than radiographs clear. Patients hospitalized for CAP have a 1-year mortality rate 2.5 times greater than controls and mortality remains elevated for 2 years, even in patients with no comorbidities.55, 56, 57 Excess cardiovascular risk has been observed for 5 to 10 years after infection. This phenomenon is likely due to the fact that systemic inflammation destabilizes coronary plaques and produces a procoagulant effect.58, 59 Primary care providers should ensure that after a pneumonia diagnosis patients receive aspirin and statin if they are eligible.60, 61

Prevention

In the primary care setting, the most important measure to reduce a patient’s risk of CAP is to encourage smoking cessation, because tobacco use interferes with immune system and lung function. Children should be vaccinated against S pneumoniae and H influenza with PCV13 and HIB. All healthy adults over the age of 65 should receive vaccination against S pneumoniae. The PCV13 should be given to adults 65 and older who have not previously received a dose and the PPSV23 should be given at least 1 year later. This regimen is especially important for patients with chronic obstructive pulmonary disease in whom pneumococcal vaccine has been shown to reduce the likelihood of CAP (needed to treat = 21) and exacerbations of chronic obstructive pulmonary disease. Influenza vaccines should be given yearly to all eligible patients. Finally, frequent hand washing should be encouraged to patients and providers alike, primarily to prevent spread of respiratory viruses.

Summary

History, physical examination, including vital signs and saturation of peripheral oxygen, and chest radiographs results provide the essential information to clinically diagnose CAP.

CAP is caused by both bacterial and viral pathogens.

It is essential to query the patient’s past medical history for risk factors that predispose to drug-resistant pneumonia.

The concept of HCAP is changing; ED providers need to be aware of risk factors for MDR pneumonia, limiting broad spectrum antibiotics to patients satisfying guideline-recommended criteria.

In severe CAP, ED providers should collect blood cultures before administering antibiotics and sputum cultures when applicable, although in most cases etiologic testing does not reveal the causative pathogen.

Careful severity assessment is a crucial step in ED CAP management and should include screening for occult sepsis with a serum lactate, followed by early antibiotics and fluid resuscitation when indicated.

Risk stratification tools such as the PSI and CURB-65 should be used routinely to determine the most appropriate disposition for a patient.

Emergency providers should be familiar with the latest guidelines for antimicrobial treatment for both outpatient and inpatient CAP, which will continue to change as resistance patterns in respiratory pathogens evolve.

Vaccination must be encouraged to continue to prevent respiratory infections in children and adults.