Clinical evaluation and diagnostic testing for community-acquired pneumonia in adults

INTRODUCTION — Community-acquired pneumonia (CAP) is one of the most commonly diagnosed illnesses worldwide. The clinical presentation of CAP varies, ranging from mild disease characterized by limited shortness of breath and productive cough to severe disease characterized by fever, respiratory distress, and sepsis. Because of the wide spectrum of associated clinical features, CAP is a part of the differential diagnosis of most acute respiratory illnesses.

In patients with a clinically compatible syndrome, the demonstration of an infiltrate on chest imaging is generally sufficient to establish an initial working diagnosis and start empiric therapy. However, this combination of findings is ultimately nonspecific and is shared among many cardiopulmonary disorders. Thus, it is important to remain attentive to the possibility of an alternate diagnosis as a patient's course evolves.

The clinical manifestations, diagnosis, and microbiologic evaluation of CAP in immunocompetent adults are reviewed here. An general overview of CAP in adults is provided separately (see "Overview of community-acquired pneumonia in adults"). Links to detailed topics on the epidemiology, management, and prognosis of CAP are provided within the overview topic and at relevant places within the text below.

GENERAL APPROACH — The diagnosis of CAP generally requires the demonstration of an opacity on chest imaging in a patient with a clinically compatible syndrome (eg, fever, dyspnea, cough, and sputum production) [1].

●For most patients with suspected CAP, we obtain posteroanterior and lateral chest radiographs. Radiographic findings consistent with the diagnosis of CAP include lobar consolidations (image 1A-B), interstitial infiltrates (image 2A-C), and/or cavitations (image 3). Although certain radiographic features suggest specific causes of pneumonia (eg, lobar consolidations suggest infection with typical bacterial pathogens), radiographic appearance alone cannot reliably differentiate among etiologies.

●In selected cases, it is reasonable to make a clinical diagnosis without chest imaging in otherwise healthy patients with highly compatible syndromes (eg, acute onset, fever, cough, and signs of consolidation on physical examination) and lack of concern for other causes. However, clinical features alone have limited diagnostic accuracy, thus, we typically reserve this option for circumstances in which chest radiography cannot be easily obtained and the patient can be closely followed.

●When CAP is suspected based on clinical features despite a negative chest radiograph, we obtain computed tomography (CT) of the chest in selected patients. These patients include immunocompromised patients, who may not mount strong inflammatory responses and thus have negative chest radiographs, as well as patients with known exposures to epidemic pathogens that cause pneumonia (eg, severe acute respiratory syndrome coronavirus 2, Legionella spp).

●Obtaining a CT scan is also reasonable when CAP is suspected but the clinical presentation is atypical or the patient has possible alternative explanations for their syndrome (eg, chronic obstructive pulmonary disease exacerbation, pulmonary edema, atelectasis, etc). In such cases, CT scanning can help confirm or exclude the diagnosis of pneumonia [2]. Because there is no direct evidence to suggest that CT scanning improves outcomes for most patients and cost is high, we do not routinely obtain CT scans when evaluating patients for CAP. (See 'Chest imaging findings' below.)

All patients with possible CAP should be tested for coronavirus disease 2019 during the pandemic. Microbiologic testing is otherwise generally reserved for hospitalized patients and for selected outpatients in which test results could change management (eg, immunocompromised patients, patients who fail to improve with empiric therapy, during outbreaks with certain organisms such as Legionella spp, or following exposures to specific pathogens such as influenza virus) (table 1). However, the accuracy, availability, and affordability of rapid microbiologic diagnostics tests is growing, and we expect that indications for testing will broaden soon [3]. (See 'Microbiologic testing' below.)

In all cases, the initial diagnosis of CAP should be considered a working diagnosis. A substantial portion of patients initially diagnosed with CAP are ultimately found to have alternate diagnoses [4], therefore, findings that further support or refute the initial diagnoses should continue to be sought as the patient’s course progresses.

CLINICAL EVALUATION

History and physical — The history and physical examination should be directed at recognizing the clinical syndrome of CAP, evaluating its severity, assessing for complications, identifying important exposures that may indicate the cause of infection, and assessing for comorbidities that may contribute to the patient’s symptoms.

Classically, CAP is characterized by acute onset fever, cough (with or without sputum production), and shortness of breath [5-7]. In some cases, pleuritic chest pain may also be present. Less common symptoms include gastrointestinal complaints (eg, nausea, vomiting, diarrhea, abdominal pain), loss of appetite, and mental status changes. In patients with advanced age or impaired immune systems, presenting symptoms can be subtle. For example, fever may be absent in older patients and mental status changes may be the sole presenting symptom [8,9]. Tachycardia, tachypnea, hypoxemia, or increased work of breathing may be present on physical examination. Crackles (rales) and rhonchi may be heard on chest auscultation, along with other signs of consolidation (eg, tactile fremitus, egophony, dullness to percussion). As infection progresses, the dominant clinical picture may be of sepsis and/or respiratory distress.

While the clinical features described above support the diagnosis of pneumonia, no combination of symptoms and signs has been found to accurately predict the presence or absence of pneumonia [6,7,10]. As an example, the sensitivity of the combination of fever, cough, tachycardia, and crackles was less than 50 percent when chest radiograph was used as a reference standard in one retrospective review [6]. In another study evaluating >28,000 adults presenting to primary care with an acute cough attributed to a lower respiratory tract infection, independent predictors of radiographic-confirmed pneumonia included fever, tachycardia, crackles on chest auscultation, and oxygen saturation <95 percent [7]. However, the positive predictive value (PPV) of all four variables combined was estimated to be only 67 percent. Concordantly, studies comparing clinical judgement with chest radiography for the diagnosis of pneumonia have similar findings [6,11]. One cohort study of >2800 adults with acute cough estimates the PPV of clinical judgement at 57 percent [11].

Similarly, there are no signs or symptoms that reliably distinguish among the many infectious causes of pneumonia (eg, viral versus bacterial or between different bacterial causes) [12]. However, the presence of certain features may raise the index of suspicion for particular pathogens (table 2). This concept is particularly important during outbreaks, family clusters, and other known/potential exposures, where knowledge of a particular exposure may change the need for testing, alter empiric therapy selection, require quarantine, or other public health action. (See 'Important pathogens' below.)

Laboratory evaluation

Routine blood tests — We generally obtain a complete blood count with differential and a basic metabolic panel for most patients with known or suspected CAP who are being hospitalized or who may require hospitalization based on their age, comorbidities, vital signs, or clinical appearance. Results help corroborate the diagnosis of CAP and inform the need for hospitalization and level of care (ie, need for intensive care). (See "Community-acquired pneumonia in adults: Assessing severity and determining the appropriate site of care".)

●Leukocytosis is the most common blood test abnormality with a leftward shift. Leukopenia (<4000 cells per mm³) is less common but generally connotes poorer prognosis [13,14]. Similarly, thrombocytopenia (platelet count <100,000 cells per mm³) is an uncommon finding but one that suggests poorer outcome.

●New elevations in creatinine and blood urea nitrogen also connote poor prognosis and often indicate need for hospitalization. These values along with abnormal liver function tests can be also be signs of sepsis, which mandates immediate additional evaluation and care. (See "Evaluation and management of suspected sepsis and septic shock in adults".)

Serum biomarkers — While there is much interest in using C-reactive protein (CRP) and procalcitonin to aid in the diagnosis of pneumonia and to help distinguish bacterial from viral causes of CAP [15-24], we do not find that these tests reliably add value to the initial clinical and radiographic evaluation [1,25,26]. The reported sensitivity and specificity of CRP for pneumonia both range from approximately 40 to 90 percent and vary substantially with the cutoff value used [15-19]. Similarly, the reported sensitivity for procalcitonin ranges widely from 38 to 91 percent [1]. A randomized trial of procalcitonin measurement in addition to clinical evaluation versus clinical evaluation alone for patients presenting to the emergency department with possible lower respiratory tract infection in whom clinicians were uncertain whether antibiotics were needed or not did not show any difference in antibiotic utilization or patient outcomes [25].

However, procalcitonin, in conjunction with clinical judgement, can help guide the decision to discontinue antibiotic treatment. Detailed discussion about procalcitonin in respiratory tract infections is provided separately and focuses on the utility of the assay in guiding antibiotic de-escalation in patients with established diagnoses of CAP. (See "Procalcitonin use in lower respiratory tract infections", section on 'Community-acquired pneumonia in hospitalized patients'.)

DIFFERENTIAL DIAGNOSIS — CAP is a common working diagnosis and is on the differential diagnosis of patients presenting with respiratory tract infections, patients with cough and abnormal chest imaging, and patients with sepsis. Because CAP is common and its symptoms overlap with many other cardiopulmonary disorders, it is frequently overdiagnosed [4]. In one cohort study, approximately 17 percent of patients hospitalized with CAP were ultimately found to have a different diagnosis. The most common alternate diagnoses included heart failure, malignancy, and pulmonary infarct or fibrosis [4].

Other respiratory tract infections that present similarly to CAP include, but are not limited, to influenza, coronavirus disease 2019 (COVID-19), acute bronchitis, chronic obstructive pulmonary disease exacerbations, and other respiratory tract infections. Generally, signs of consolidation on chest auscultation and the presence of an opacity on chest imaging distinguish CAP from the latter two disorders.

●Acute bronchitis – Acute bronchitis is an infection that primarily involves the large airways (bronchi) but not the alveoli or pulmonary parenchyma. Prolonged cough, typically following an upper respiratory tract infection, is the most common symptom of acute bronchitis. In contrast to pneumonia, acute bronchitis is less likely to present with abnormal vital signs (pulse >100/minute, respiratory rate >24 breaths/minute, temperature >38°C [100.4°F], or oxygen saturation <95 percent) (figure 1). Mental status changes and/or other signs of systemic infection are typically absent. Signs of consolidation (rales, egophony, or tactile fremitus) should be absent on chest auscultation. (See "Acute bronchitis in adults".)

●Influenza – Influenza is characterized by acute onset fever and cough, often accompanied by pronounced systemic symptoms including chills, rigors, myalgias, and/or arthralgias. Pneumonia complicates a minority of cases and may be primary (caused by influenza virus itself) or secondary, caused by bacterial respiratory pathogens (most often streptococci or Staphylococcus aureus). (See "Seasonal influenza in adults: Clinical manifestations and diagnosis".)

●Upper respiratory tract infection (URI) – URIs are commonly characterized by fever and cough, but in contrast to pneumonia lack signs of consolidation on chest auscultation. While the presence of rhinorrhea and pharyngitis favor the diagnosis URI over pneumonia, URIs sometimes precede or co-occur with CAP [11]. (See "The common cold in adults: Diagnosis and clinical features".)

●Coronavirus disease 2019 – The range of presenting symptoms for COVID-19 is wide, ranging from asymptomatic infection to CAP with profound acute respiratory failure. (See "COVID-19: Clinical features".)

Exacerbations of chronic pulmonary diseases are frequently triggered by viral or bacterial infection or colonization of the respiratory tract mucosa. In most cases, infection does not extend beyond the mucosa, however, pneumonia can complicate these disorders. Higher fever, adventitious breath sounds on chest auscultation (eg, focal rales/crackles or rhonchi), and impaired oxygenation suggest pneumonia and generally indicate need for chest imaging.

●Acute exacerbations of chronic obstructive pulmonary disease (AECOPD) – AECOPD is defined as an acute increase in COPD symptoms that go beyond day-to-day variation, typically characterized by increases in dyspnea, sputum volume/viscosity, and/or sputum purulence. (See "COPD exacerbations: Clinical manifestations and evaluation".)

●Acute asthma exacerbations – Cough (often worse at night), wheezing, and shortness of breath are characteristics symptoms of asthma exacerbations. In contrast to AECOPD, infectious triggers are less common. (See "Asthma in adolescents and adults: Evaluation and diagnosis".)

●Acute exacerbations of bronchiectasis – Worsening cough, sputum production, and shortness of breath also characterize bronchiectasis exacerbations. Focal crackles may also be present on physical examination. Distinguishing features from CAP include known history of bronchiectasis, chronic baseline respiratory symptoms, and/or thickened airways on chest imaging. (See "Clinical manifestations and diagnosis of bronchiectasis in adults".)

Noninfectious illnesses that mimic or co-occur with CAP and present with pulmonary opacities and cough include:

●Heart failure with pulmonary edema – Cough accompanied by shortness of breath, particularly with exertion or when lying flat, should raise suspicion for heart failure. Physical examination findings such as an elevated jugular venous pressure, bilateral basilar crackles with resonance to percussion, S3 heart sound, displaced apical impulse, and peripheral edema should further heighten suspicion. (See "Heart failure: Clinical manifestations and diagnosis in adults".)

●Pulmonary embolism (PE) – Dyspnea, pleuritic chest pain, and hemoptysis in addition to cough are classic symptoms associated with PE. However, presenting symptoms vary and can be mild and nonspecific. History may be notable for malignancy, recent surgery, or prolonged immobilization. Physical examination findings also vary but those that support this diagnosis include tachypnea, tachycardia, and unilateral lower extremity swelling. Any suspicion for PE warrants further evaluation. (See "Clinical presentation, evaluation, and diagnosis of the nonpregnant adult with suspected acute pulmonary embolism".)

●Lung cancer – Lung cancer is an uncommon cause of acute shortness of breath and cough but should be considered in any current or prior smoker. Features that should raise suspicion for this diagnosis include a recent change in a chronic "smoker's cough," hemoptysis, and signs of focal airway obstruction on physical examination such as a unilateral wheeze or decreased breath sounds.

●Atelectasis – Atelectasis refers to collapse of a portion of the pulmonary parenchyma and can appear similar to pneumonia on chest imaging. (See "Atelectasis: Types and pathogenesis in adults".)

●Aspiration or chemical pneumonitis – Aspiration of substances that are toxic to the lower airways (eg, gastric acid, stomach contents) can lead to inflammation and opacities on chest imaging. Imaging abnormalities are typically seen in the gravity-dependent areas of the lung (eg, bases of lower lobes if upright, posterior segments of lower lobes if supine). In contrast to aspiration pneumonia, clinical improvement is often rapid following aspiration or chemical pneumonitis. (See "Aspiration pneumonia in adults", section on 'Chemical pneumonitis'.)

●E-cigarette or vaping product use associated lung injury (EVALI) – EVALI should be considered in patients with respiratory complains and recent e-cigarette use. Respiratory symptoms are similar to CAP and include fever, cough, shortness of breath, and chest pain. Concurrent gastrointestinal symptoms (eg, nausea, vomiting, diarrhea, and abdominal pain) are also common. (See "E-cigarette or vaping product use-associated lung injury (EVALI)".)

●Drug reactions – Exposure to certain drugs can cause hypersensitivity reactions or direct pulmonary toxicity. Symptoms are often similar to CAP, but the acuity of onset, pathogenesis, and radiographic appearance vary by agent. Common culprits include nitrofurantoin, amiodarone, daptomycin, methotrexate, bleomycin, and gemcitabine. (See "Drug reaction with eosinophilia and systemic symptoms (DRESS)" and "Pulmonary toxicity associated with systemic antineoplastic therapy: Clinical presentation, diagnosis, and treatment".)

●Interstitial lung diseases (ILD) – A wide variety of ILD can present similarly to CAP early in their evolution and include sarcoidosis, asbestosis, hypersensitivity pneumonitis, cryptogenic organizing pneumonia, and systemic rheumatic diseases (eg, granulomatosis with polyangiitis, rheumatoid arthritis, systemic lupus erythematosus, dermatomyositis). (See "Approach to the adult with interstitial lung disease: Clinical evaluation".)

Febrile illness and/or sepsis can also be the presenting syndrome in patients with CAP; other common causes of these syndromes include urinary tract infections, intra-abdominal infections, and endocarditis.

In all cases, monitoring the patient’s progress over time helps confirm or refute the initial diagnosis. For example, the rapid resolution of pulmonary opacities on chest imaging suggests a noninfectious process such as pulmonary edema or atelectasis. Conversely, lack of response to empiric treatment in 48 to 72 hours suggest that initial antibiotic selection was inadequate, there are complications, or that there is an alternate diagnosis. (See 'Follow-up care' below.)

CHEST IMAGING FINDINGS

Imaging techniques — Chest imaging is indicated for the majority of patients with suspected CAP to confirm the diagnosis, assess for complications (eg, parapneumonic effusion, empyema, abscess), and evaluate for alternate or concurrent diagnosis (eg, heart failure, malignancy). The presence of an opacity on chest imaging in a patient with a compatible clinical syndrome is the gold standard for diagnosis and recommended for diagnosis in the American Thoracic Society/Infectious Disease Society of America guidelines [1].

Chest radiograph (preferred) — The presence of an infiltrate on plain chest radiograph is considered the gold standard for diagnosing pneumonia when clinical and microbiologic features are supportive. For most patients with suspected CAP, obtaining posteroanterior and lateral chest radiographs is sufficient for diagnosis. Rarely, we forgo imaging in outpatients with highly compatible syndromes (eg, acute onset, fever, cough, shortness of breath, and signs of consolidation on physical examination) and lack of concern for other causes. Imaging is a necessity for hospitalized patients.

Radiographic findings consistent with the diagnosis of CAP include lobar consolidations (image 1A-B), interstitial infiltrates (image 2A-C), and/or cavitations (image 3). Although certain radiographic features suggest specific causes of pneumonia (eg, lobar consolidations suggest infection with typical bacterial pathogens), radiographic appearance alone cannot reliably differentiate among etiologies [5,27]. There is also substantial interobserver variation in the interpretation of chest radiographs in patients with possible pneumonia between different radiologists [28-30] and between emergency department physicians and radiologists [31,32].

In some cases, chest radiographs may not be sufficiently sensitive for the detection of pneumonia [2,33]. There are case reports and animal experiments favoring the hypothesis that volume depletion may produce an initially negative radiograph, which "blossoms" into infiltrates following rehydration [34]. In support of this hypothesis, one population-based cohort study of suspected CAP found that 7 percent of patients with negative initial radiographs developed changes consistent with CAP on repeat chest radiograph [35].

Thus, when clinical suspicion is high despite a negative chest radiograph, we decide to either treat empirically and/or perform a chest computed tomography (CT) depending on the patient’s severity of illness, immune status, and/or the suspected pathogen. When clinical suspicion is lower, we determine the need for additional chest imaging (eg, CT scan) or other workup based on the most likely potential causes in each individual (eg, CT pulmonary angiogram for evaluation of pulmonary embolism).

Computed tomography (CT) scan — High-resolution CT is more sensitive for the detection of pneumonia than chest radiograph [2,36,37]. CT scanning can be helpful to better characterize pneumonia and identify complications [36,38]. This is particularly true for immunocompromised patients who are at risk for infection with broad array of pathogens. The enhanced sensitivity and specificity of CT scan can help distinguish among causes (eg, invasive fungal infections, pneumocystis pneumonia, bacterial pathogens) [39]. (See "Approach to the immunocompromised patient with fever and pulmonary infiltrates".)

We also obtain chest CT for selected patients in whom CAP is suspected based on clinical features despite a negative chest radiograph to help confirm the diagnosis and assess for other causes of symptoms. (See 'Chest radiograph (preferred)' above.)

Chest CT can also be helpful to rule out pneumonia in patients with equivocal clinical syndromes and nondiagnostic chest radiographs (opacities present but not clear if due to pneumonia versus pulmonary edema, atelectasis, contusion, chronic lung disease, or other etiologies). This is most common in patients with multiple comorbidities presenting with nonspecific syndromes. A chest CT in such cases can help rule out pneumonia in up to 50 percent of cases [2].  

Ultrasound and other studies — There are increasing data on the use of lung ultrasound to diagnose pneumonia, particularly in unstable patients in the emergency department or intensive care unit in whom it is difficult to obtain good-quality chest radiographs. In three large meta-analyses, the sensitivity of lung ultrasound was approximately 80 to 90 percent and the specificity approximately 70 to 90 percent [40-42]. Diagnostic performance can vary, however, depending on the ultrasonographer's level of experience [43]. It is therefore reasonable for those with experience performing lung ultrasounds to use this modality when a chest radiograph is not likely to be a good-quality study.

DETERMINING SITE OF CARE — For patients with a working diagnosis of CAP, the next steps in management are defining the severity of illness and determining the most appropriate site of care. Determining the severity of illness is based on clinical judgement and can be supplemented by use of severity scores (algorithm 1). (See "Community-acquired pneumonia in adults: Assessing severity and determining the appropriate site of care".) (Related Pathway(s): Community-acquired pneumonia: Determining the appropriate site of care for adults.)

The triage of patients with known or suspected coronavirus disease 2019 is discussed separately. (See "COVID-19: Evaluation of adults with acute illness in the outpatient setting" and "COVID-19: Infection prevention for persons with SARS-CoV-2 infection".)

MICROBIOLOGIC TESTING

Approach to testing — The benefit of obtaining a microbiologic diagnosis should be balanced against the time and cost associated with an extensive evaluation in each patient. Apart from testing all patients for COVID-19 during the pandemic, we generally take a tiered approach to microbiologic evaluation based on CAP severity and the site of care (table 1).

●For most patients with mild CAP being treated in the ambulatory setting, microbiologic testing is not needed apart from COVID-19 testing during the pandemic. Testing for influenza can also be considered when community incidence is high and when results would change management (eg, those who meet criteria for antiviral treatment) [44,45]. (See "Seasonal influenza in adults: Clinical manifestations and diagnosis".)

Diagnostic testing for outpatients with mild CAP is otherwise generally unnecessary since empiric antibiotic therapy is generally successful, and knowledge of the infecting pathogen does not usually improve outcomes [46].

●For most patients with moderate CAP admitted to the general medical ward, we obtain the following:

•Blood cultures

•Sputum Gram stain and culture

•Urinary antigen testing for Streptococcus pneumoniae

•Testing for Legionella spp (polymerase chain reaction [PCR] when available, urinary antigen test as an alternate)

•COVID-19 testing

•Testing for respiratory viruses during respiratory virus season (PCR preferred, particularly for influenza)

•Rapid nasal PCR or culture for methicillin-resistant Staphylococcus aureus (MRSA) in patients with risk factors for MRSA, severe disease, or a biphasic illness (viral respiratory syndrome followed by new deterioration)

•A screening test for human immunodeficiency virus (HIV) infection

Blood and urine cultures should ideally be performed before the start of empiric antibiotics. To help improve the diagnostic yield of sputum, patients should rinse their mouth and cough deeply before expectoration. The sample should be inoculated on culture media as soon as possible.

●For most hospitalized patients with severe CAP, including those admitted to the intensive care unit, we obtain the tests bulleted above. In addition, for patients with severe disease who fail to respond to empiric therapy, we obtain bronchoscopic specimens for microbiologic testing when feasible, weighing the benefits of obtaining a microbiologic diagnosis against the risks of the procedure (eg, need for intubation, bleeding, bronchospasm, pneumothorax) on a case-by-case basis. When pursuing bronchoscopy, we usually send specimens for aerobic culture, Legionella culture, fungal stain and culture, pneumocystis pneumonia stain, or PCR if there is concern for impaired immunity, and test for respiratory viruses using multiplex PCR panels.

In all cases, we modify this approach based on epidemiologic exposures, patient risk factors/clinical features, public health needs, and the likelihood that the test results would help refine the differential diagnosis and/or change management. As examples:

●For patients with known or probable exposures to epidemic pathogens such as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), Middle East Respiratory Syndrome (MERS), Legionella, or Avian influenza viruses, we ensure that our evaluation includes tests for these pathogens. (See 'Important pathogens' below.)

●For patients with cavitary pneumonia, we may include testing for tuberculosis, fungal pathogens, and Nocardia.

●For immunocompromised patients, we broaden our differential to include opportunistic pathogens such as Pneumocystis jirovecii, fungal pathogens, parasites, and less common viral pathogens such as cytomegalovirus. The approach to diagnostic testing varies based on the type and degree of immunosuppression and other patient-specific factors. (See "Approach to the immunocompromised patient with fever and pulmonary infiltrates" and "Epidemiology of pulmonary infections in immunocompromised patients".)

Making a microbiologic diagnosis also allows for directed therapy, which has been shown to improve outcomes [47-49], and can also limit antibiotic overuse and prevent antimicrobial resistance and unnecessary complications, such as Clostridioides (formerly Clostridium) difficile infections. Microbiologic test results are also critical for defining the local epidemiology and antimicrobial resistance patterns, which guide the care of other patients. In addition, a negative evaluation can help support a decision to narrow or discontinue antibiotic treatment. When defining the scope of our microbiologic evaluation, we also take the certainty of the diagnosis of CAP into consideration. Because a substantial portion of patients hospitalized with an initial clinical diagnosis of CAP are ultimately found to have alternate diagnoses [4], pursuing a comprehensive microbiologic evaluation can help reach the final diagnosis.

Our approach is similar to that of the American Thoracic Society (ATS)/Infectious Diseases Society of America (IDSA) guidelines, which does not routinely recommend microbiologic testing for outpatients with CAP (outside of the pandemic) [1]. However, the ATS/IDSA is more conservative in its approach for hospitalized patients.

Blood cultures, sputum Gram stain and culture, and MRSA nasal testing are recommended for hospitalized patients with severe CAP, concern for infection with MRSA or Pseudomonas, or recent hospitalization and/or receipt of IV antibiotics.

The ATS/IDSA recommends reserving testing for Legionella spp and urine pneumococcal antigen testing for patients with severe CAP, compromised immune systems, or epidemiologic exposures.

For other hospitalized patients, the ATS/IDSA neither recommends for nor against obtaining sputum Gram stain and culture but states that the decision should be individualized based on clinical presentation, local epidemiology, and antimicrobial stewardship processes. The differences in approaches may be in part due to a difference in vantage point; the ATS/IDSA guidelines address patients with known CAP [1], whereas this topic addresses the evaluation of an undifferentiated patient in whom CAP is suspected.

Overall, the utility of pursuing a microbiologic evaluation has been debated in the literature, with some studies indicating that diagnostic yield from testing is low and rarely changes management, while other studies suggest that establishing a diagnosis provides diagnostic clarity, can help tailor the treatment strategy, and provides valuable data at the population level on prevailing pathogens and antibiotic-resistance profiles [50,51]. In practice, most hospitalized patients are treated empirically without a microbiologic diagnosis [52-54].

Important pathogens — Certain pathogens are critical to detect because they represent important epidemiologic challenges, serious conditions that require treatment that differs from standard empiric regimens, and/or require special measures to prevent transmission from infected patients to staff and other patients. These organisms include:

●Severe acute respiratory syndrome coronavirus 2 – The epidemiology of SARS-CoV-2, the virus that causes COVID-19, is discussed separately. (See "COVID-19: Epidemiology, virology, and prevention".)

●Community-acquired methicillin-resistant S. aureus (CA-MRSA) – Because CA-MRSA infections can be severe and some but not all empiric treatment regimens cover CA-MRSA, it is important to make the diagnosis. Features that raise suspicion for CA-MRSA include concurrent or prior influenza infection [55], necrotizing pneumonia, or other rapidly progressive pneumonia, particular occurring in young adults. The presence of gram-positive cocci on a high-quality sputum Gram stain and/or positive nasal MRSA test further raise suspicion. The diagnosis is confirmed by culture with susceptibility testing. (See "Methicillin-resistant Staphylococcus aureus (MRSA) in adults: Epidemiology", section on 'Community-associated MRSA infection'.)

●Legionella spp – Detection of Legionella has important public health implications and can inform treatment selection. While there are no distinguishing clinical features, known outbreaks or family clusters (often related to contaminated water sources) should raise suspicion. Recent travel is a risk factor. The diagnosis can be made by urine antigen testing (limited to serogroup 1), PCR, or culture on special media. (See "Clinical manifestations and diagnosis of Legionella infection" and "Treatment and prevention of Legionella infection".)

●Influenza and other respiratory viruses – Influenza and other respiratory viruses are important to recognize because of the need for appropriate infection control in hospitalized patients, for public health reporting purposes, and for rapid treatment with antiviral agents in the cases of influenza and SARS-CoV-2. During influenza outbreaks, outpatients with compatible syndromes can be diagnosed on clinical grounds alone. Real-time reverse-transcriptase PCR is the test of choice to diagnose influenza but requires a laboratory and a laboratory technician. Rapid point-of-care diagnostic tests can be done in an emergency department or office without a laboratory technician; these tests show a sensitivity of only about 50 to 60 percent, but specificity is >95 percent, so a positive test helps establish the diagnosis but a negative result does not rule out influenza [56,57].

The need to know the influenza type depends partly on whether multiple influenza strains with different resistance patterns are circulating. For example, the majority of seasonal influenza H1N1 during the 2008 to 2009 season were resistant to oseltamivir, but almost all isolates of H1N1 influenza since the 2009 pandemic have been susceptible to oseltamivir [58,59]. (See "Seasonal influenza in nonpregnant adults: Treatment" and "Seasonal influenza in adults: Clinical manifestations and diagnosis" and "Antiviral drugs for influenza: Pharmacology and resistance".)

For severe and sporadic cases of influenza-like illness, there is more urgency to make a specific diagnosis due to concern for avian influenza A H5N1, avian influenza A H7N9, or other emerging strains. In such situations, PCR for influenza is appropriate [56]. Reagents to test for novel influenza strains when they first emerge may be available only in public health laboratories; local hospital laboratory supervisors and infection control practitioners will know the specimen referral process. (See "Seasonal influenza in adults: Clinical manifestations and diagnosis" and "Avian influenza: Epidemiology and transmission" and "Avian influenza: Clinical manifestations and diagnosis".)

●Zoonotic pathogens – Several pathogens have emerged from animal sources to cause outbreaks of respiratory disease in humans (eg, H1N1 pandemic influenza, H5N1 avian influenza, H7N9 avian influenza, SARS-CoV-2, MERS-CoV). Other zoonotic pathogens that can cause respiratory disease in humans include hantaviruses (eg, Sin Nombre virus) and Yersinia pestis; such pathogens should be considered in patients with relevant travel or exposure history [60,61].

●Agents of bioterrorism – Agents of bioterrorism that can cause a pneumonic syndrome include Bacillus anthracis (inhalational anthrax), Y. pestis (pneumonic plague), Francisella tularensis (tularemia), Coxiella burnetii (Q fever), Legionella spp, influenza virus, hantavirus, and ricin (table 3) [62]. (See "Identifying and managing casualties of biological terrorism".)

Diagnostic tests

Blood cultures — We typically obtain two sets of blood cultures in patients hospitalized with CAP at the time the initial working diagnosis is established. To enhance diagnostic yield, blood cultures should ideally be obtained prior to empiric antibiotic administration. Although the value of obtaining blood cultures for all hospitalized patients is debated [1], results both help direct patient care and inform important public health programs.

●For patients, a positive culture with a likely pathogen (eg, S. pneumoniae) provides a definitive microbiologic diagnosis and allows for directed therapy, which has been shown to improve outcomes [47-49] and minimize potential adverse effects from unnecessary antibiotic exposure. In addition, blood cultures also help identify complications (eg, bacteremia) or alternate diagnoses (eg, endocarditis), which might have otherwise gone undetected.

●For many health care institutions, blood culture data is the primary source for microbiologic data used to inform institutional antibiograms and antimicrobial stewardship efforts. At the national level, isolates identified serve as an important resource for tracking resistance patterns of S. pneumoniae and other common pathogens (eg, the United States Centers for Disease Control and Prevention surveillance network is reliant on blood culture data [63]). These are the data used for evaluating efficacy of current S. pneumoniae vaccines and the serotypes needed for inclusion in future vaccines [64-66].

The major downsides of obtaining blood cultures include their low diagnostic yield and the low degree of certainty that results improve clinical outcomes. The rate of positive blood cultures ranges from approximately 4.7 to 16 percent among patients hospitalized with CAP [67-75]. In one retrospective cohort study in the United States evaluating >130,000 patients with pneumonia, blood cultures were positive in 4.7 percent (though an uncertain percentage were obtained while on antibiotics); the likelihood of a positive culture rose with the severity of illness [74]. S. pneumoniae is the most common blood culture isolate (approximately 33 percent), followed by S. aureus and Pseudomonas aeruginosa. Approximately two-thirds of isolates were resistant to first-line CAP treatment regimens. While the degree to which culture results lead to changes in management that improve outcomes is not well studied, available data suggest that the number might be low [69,72]. In one study, with a 10 percent rate of false-positive cultures, hospital stays were prolonged due to a perceived need for vancomycin to treat S. aureus, when the laboratory calls to report gram-positive cocci, which were actually coagulase-negative staphylococci that had yet to be identified.

Sputum Gram stain and culture — We generally obtain a sputum Gram stain and culture for most hospitalized patients with CAP because results can help direct therapy and reduce antibiotic overuse. To enhance diagnostic yield, specimens should ideally be from a deep cough, obtained before antibiotics, and processed within two hours of collection [76].

Culture results should be interpreted based on the following:

●Quantitation of growth (heavy, moderate, or light)

●Clinical correlation

●Correlation with the Gram stain

●Quality of the specimen

True pathogens are generally present in moderate or heavy amounts by Gram stain and culture. However, certain pathogens are regarded as significant regardless of concentration, including Legionella spp, B. anthracis, Mycobacterium tuberculosis, Mycoplasma pneumoniae, Chlamydia pneumoniae, and Chlamydia psittaci; these respiratory pathogens are virtually never colonizers but always represent disease.

Antibiotics may alter the yield of any subsequent culture of respiratory secretions [77]. Specimens collected after antibiotics are given are more likely to grow S. aureus or gram-negative bacilli (GNB), which usually represent early airway colonizers.

When S. pneumoniae or Haemophilus influenzae are the etiologic agent, false-negative cultures may occur because of their fastidious growth requirements and the lack of selective media. By contrast, S. aureus and GNB are relatively rare pulmonary pathogens that are easily grown in respiratory secretions because they are hardy and easily recognized by growth on selective media. The failure to grow these organisms is strong evidence against their presence.

A "good-quality" sputum sample is one with moderate or many polymorphonuclear leukocytes but few (or no) squamous epithelial cells on Gram stain. This is discussed in detail separately. (See "Sputum cultures for the evaluation of bacterial pneumonia", section on 'Expectorated sputum'.)

The utility of sputum Gram stain and culture is debated. Rates of pathogen detection and impact on clinical outcomes vary among studies [50,52,78-83]. One meta-analysis evaluating 24 studies including >4500 patients detected bacterial pathogens in 73 percent of good-quality sputum Gram stains (95% credible interval [CrI] 26-96 percent); detection rates dropped to 36 percent (95% CrI 22-53 percent) when including lower quality Grain stains as well [50]. The diagnostic yield was highest for S. pneumoniae (sensitivity 69 percent [95% CrI 56-80 percent]; specificity 91 percent [95% CrI 83-96 percent]) and H. influenzae (sensitivity 76 percent [95% CrI 60-87 percent]; specificity of 97 percent [95% CrI 91–99 percent]). Other studies have found detection rates <10 percent [52,78,79]. Differences among studies may be due to variations in specimen quality, timeliness of specimen processing, clinical microbiology laboratory capabilities, and whether patients were exposed to antibiotics before sampling. The widespread use of broad-spectrum empiric treatments before Gram stain and culture return also likely blunts capacity to assess the specific impact of Gram stain and culture guided therapy.  

Urine pneumococcal antigen testing — We generally send the urine S. pneumoniae antigen test in most hospitalized patients with pneumonia because the test is easy to obtain, turnaround time is rapid, and results are less affected by antibiotic administration when compared with Gram stain and culture. As with other microbiologic tests, when positive, results help direct therapy and reduce antibiotic overuse. The urine S. pneumoniae antigen assay has a sensitivity of >70 percent and a specificity of approximately 98 percent. The diagnostic approach for pneumococcal pneumonia, including the performance characteristics of this assay, are discussed separately. (See "Pneumococcal pneumonia in patients requiring hospitalization", section on 'Diagnosis'.)

Testing for Legionella — We also test for infection with Legionella spp in most hospitalized patients. The main testing options for Legionella infection include nucleic acid detection (eg, PCR), urine antigen tests, and culture (table 4).

●When testing for Legionella in patients with pneumonia, we prefer to use PCR on a lower respiratory tract sample (eg, sputum or bronchoalveolar lavage specimen) because PCR has high diagnostic accuracy and detects all Legionella species and serogroups.

●If PCR is not available or if sputum cannot be obtained, urine antigen testing is an acceptable alternative, especially in regions such as the United States where the prevalence of Legionella pneumophila serogroup 1 is high. The main advantages of the urinary antigen test are its rapid turnaround time and high specificity. A key disadvantage of the urine Legionella assay is that it does not detect all L. pneumophila serotypes or other Legionella species.

Because the urinary antigen test only detects L. pneumophila serotype 1, we generally send PCR or culture on a lower respiratory tract sample when urine antigen assays are negative and Legionella infection is still suspected (eg, patients with severe bilateral interstitial pneumonia, particularly if immunocompromised). For patients with positive Legionella urinary antigen tests or PCR, we generally also obtain a culture from the lower respiratory tract for epidemiologic purposes (eg, comparison with isolates from other patients or potential sources). Testing for Legionella infection is discussed in detail separately. (See "Clinical manifestations and diagnosis of Legionella infection", section on 'Diagnosis'.)

Testing for respiratory viruses

Coronavirus disease 2019 testing — During the pandemic, we test all patients with known or suspected CAP for COVID-19, regardless of treatment setting. Testing for COVID-19 is discussed in detail separately. (See "COVID-19: Diagnosis".)

Other respiratory viruses — During respiratory virus season (eg, late fall to early spring in the northern hemisphere), we also test for a broad array of respiratory viruses (eg, influenza, adenovirus, parainfluenza, respiratory syncytial virus, human metapneumovirus, and rhinovirus) in all hospitalized patients. Apart from influenza and SARS-CoV-2, we generally do not test for respiratory viruses in outpatients.

When testing for influenza, PCR is preferred over rapid antigen testing. For other respiratory viruses, the types of assays available (eg, direct PCR, multiplex PCR, antigen-detection, serology, culture) vary among institutions. In some cases, multiplex PCR panels that test for a wide array of viral and bacterial pathogens are used. Care should be taken when interpreting results. While respiratory viruses are increasingly recognized as a cause of CAP, they can also serve as cofactors for bacterial infection and can be asymptomatic colonizers. Thus, a positive result does not exclude bacterial coinfection. Similarly, negative results do not exclude a viral pathogen, particularly if only obtaining an upper respiratory tract specimen from a patient with lower respiratory tract disease [84,85]. The overall impact of viral testing on patient outcomes and limiting antibiotic use is thus far limited [86,87].

Performance characteristics of individual respiratory viral diagnostic assays are discussed separately. (See "Parainfluenza viruses in adults" and "Respiratory syncytial virus infection: Clinical features and diagnosis in infants and children" and "Human metapneumovirus infections" and "Diagnosis, treatment, and prevention of adenovirus infection", section on 'Pneumonia' and "Seasonal influenza in adults: Clinical manifestations and diagnosis".)

Our approach is similar to that of the ATS/IDSA guidelines, which do not recommend testing for respiratory viruses (apart from influenza and SARS-CoV-2) in outpatients [88]. For hospitalized patients, the guidelines reserve testing for respiratory viruses (apart from influenza and SARS-CoV-2) for patients with severe CAP and immunocompromised patients.

Differentiating viral from bacterial CAP — There is no single clinical feature, radiographic finding, or test that can reliably distinguish viral from bacterial pneumonia (figure 2). However, in combination, the following findings are highly suggestive of viral pneumonia:

●A positive viral test (eg a positive influenza PCR)

●A low procalcitonin level (typically <0.25 ng/mL)

●Lack of microbiologic evidence of bacterial infection (eg, negative blood and sputum cultures and/or urine antigen testing)

●Lack of radiographic findings that suggest bacterial infection (eg, dense lobar consolidations, alveolar consolidations with air bronchograms)

Because viruses are estimated to cause 30 to 50 percent of CAP cases [89,90], when these findings occur in combination in a patient who is stable or improving, it is reasonable to stop antibiotics early.

Our approach to using procalcitonin to guide antibiotic discontinuation in patients with CAP is reflected in this algorithm (algorithm 2) and discussed in detail separately. (See "Procalcitonin use in lower respiratory tract infections", section on 'Community-acquired pneumonia in hospitalized patients'.)

Multiplex molecular assays — Molecular assays that detect multiple respiratory pathogens including both bacteria and viruses from a single respiratory tract sample are being increasingly used [91]. How these assays are used in practice varies widely among institutions. In some institutions where multiplex panels are available on site, these assays are used routinely for initial diagnosis. In other institutions, use of these panels is reserved for patients with severe pneumonia, immunocompromised patients, or for those who fail to respond to initial empiric therapy.

The most commonly available multiplex assays include the BioFire FilmArray Pneumonia Panel (PN panel) and Pneumonia Plus Panel (PNplus panel), which are US Food and Drug Administration-cleared for use on lower respiratory tract specimens (sputum and bronchoalveolar lavage fluid). Each detects a broad range of bacterial and viral pathogens (including C. pneumoniae, M. pneumoniae, Bordetella pertussis, influenza, respiratory syncytial virus, parainfluenza, adenovirus, coronavirus, metapneumovirus, and others) and certain antimicrobial resistance markers. Several studies suggest that results from these multiplex panels are largely concordant with conventional culture [92-94], though further study is needed to confirm concordance.

While use of multiplex panels increases the likelihood of detecting a micro-organism in a respiratory tract sample [95-97], the predictive value of these results is not clear. As an example, the detection of a viral pathogen does not rule out the possibility of bacterial coinfection (both bacteria and viruses are identified in approximately 5 to 10 percent of CAP with an identified etiology) [89]. Similarly, some viral and bacterial pathogens can colonize the airways; their detection does not definitively indicate infection. In sum, while these assays are promising, the true impact of multiplex molecular assays on patient outcomes is yet to be substantiated by data.

FOLLOW-UP CARE

Monitoring response — Because a substantial portion of patients initially diagnosed with CAP will ultimately be found to have an alternate diagnosis, close follow-up is important for both outpatient and hospitalized patients (algorithm 3).

Most patients will show some signs of improvement within 48 to 72 hours of starting treatment. Subjectively, we look for improvement in cough, sputum production, dyspnea, and chest pain. Objectively, we assess for resolution of fever and normalization of heart rate, respiratory rate, oxygenation, and white blood cell count. Generally, patients demonstrate some clinical improvement within 48 to 72 hours.

Both failure to improve and very rapid response should raise suspicion for diagnoses other than CAP. Reasons for clinical failure can generally be categorized as either progression of initial infection (eg, overwhelming infection despite appropriate antibiotics, infection with a drug resistant pathogen) or development of complications (eg, lung abscess, empyema, heart failure, other cardiac events).

By contrast, rapid resolution of pulmonary infiltrates on chest imaging also suggests an alternate, and usually noninfectious etiology. Pulmonary infiltrates in CAP are primarily caused by the accumulation of white blood cells in the alveolar space and typically take weeks to resolve.

These concepts are discussed further elsewhere. (See "Overview of community-acquired pneumonia in adults", section on 'Complications and prognosis' and "Nonresolving pneumonia".)

Follow-up imaging — Most patients with prompt clinical resolution after treatment do not require a follow-up chest radiograph as radiographic response lags behind clinical response [98]. However, follow-up clinic visits are good opportunities to review the patient's risk for lung cancer based on age, smoking history, and recent imaging findings (algorithm 4).

This approach is similar to that outlined by the American Thoracic Society (ATS)/Infectious Diseases Society of America, which recommend not obtaining a follow-up chest radiograph in patients whose symptoms have resolved within five to seven days [1].

Most available studies have evaluated whether short-term follow-up imaging detects previously undiagnosed lung cancer [99-102]. The overall diagnostic yield is low in these studies (<5 percent), and the benefit appears greatest in those with risk factors for lung cancer (ie, males aged >50 with substantial smoking history) who otherwise meet criteria for screening via computed tomography. We therefore only recommend follow-up imaging for patients who independently meet criteria for lung cancer screening. (See "Screening for lung cancer".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Community-acquired pneumonia in adults".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and “Beyond the Basics.” The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topic (see "Patient education: Pneumonia in adults (The Basics)")

●Beyond the Basics topic (see "Patient education: Pneumonia in adults (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●General approach − The diagnosis of community-acquired pneumonia (CAP) generally requires the demonstration of an opacity on chest imaging in a patient with a clinically compatible syndrome (eg, fever, dyspnea, cough, and sputum production). (See 'General approach' above.)

•For most patients with suspected CAP, we obtain posteroanterior and lateral chest radiographs alone. In selected cases, if clinical suspicion for CAP is high but chest imaging is negative, we obtain a chest computed tomography (CT). Similarly, chest CT can help rule out pneumonia in patients with only moderate suspicion for pneumonia but nondiagnostic chest radiographs.

•Rarely, we make the diagnosis of CAP based on clinical features alone. We typically reserve this option for outpatients who have a highly compatible syndrome (eg, fever, dyspnea, cough, tachycardia, crackles on chest auscultation, and oxygen saturation <95 percent) who lack concern for other causes, or when chest radiography is not available. Because clinical features alone have limited diagnostic accuracy, we generally follow these patients closely to ensure they are improving with treatment.

●Chest imaging − Radiographic findings consistent with the diagnosis of CAP include lobar consolidations (image 1A-B), interstitial infiltrates (image 2A-C), and/or cavitations (image 3). Although certain radiographic features suggest specific causes of pneumonia (eg, lobar consolidations suggest infection with typical bacterial pathogens), radiographic appearance alone cannot reliably differentiate among etiologies (figure 2). (See 'Chest imaging findings' above.)

●Initial working diagnosis − Once an initial working diagnosis is made, the next step is to determine the most appropriate site of care. Typically, the decision is made based on clinical judgement and can be supplemented by clinical prediction rules (algorithm 1). (See 'Determining site of care' above.)

●Microbiologic testing − The need for microbiologic testing varies based on the severity of illness and clinical suspicion for specific pathogens (eg, outbreak pathogens, methicillin-resistant Staphylococcus aureus, etc). Generally, we take a tiered approach to testing based on site of care (table 1). (See 'Microbiologic testing' above.)

•For most patients with mild CAP being treated in the ambulatory setting, microbiologic testing is not needed apart from COVID-19 testing during the pandemic. Testing for influenza can also be considered when community incidence is high and when results would change management (eg, those who meet criteria for antiviral treatment). (See "Seasonal influenza in adults: Clinical manifestations and diagnosis".)

•For most patients with moderate CAP admitted to the general medical ward, we obtain the following:

-Blood cultures (ideally before antibiotics)

-Sputum Gram stain and culture

-Urinary antigen testing for Streptococcus pneumoniae

-Testing for Legionella spp (polymerase chain reaction [PCR] when available, urinary antigen test as an alternate)

-Testing for COVID-19 (reverse-transcriptase PCR preferred)

-Testing for other respiratory viruses during respiratory virus season (PCR preferred, particularly for influenza)

-Rapid nasal PCR or culture for MRSA (if risk factors for MRSA, severe disease, or a biphasic illness [viral respiratory syndrome followed by deterioration])

-Screening test for HIV

•For most hospitalized patients with severe CAP, including those admitted to the ICU, we obtain the tests bulleted above. In addition, we obtain bronchoscopic specimens for microbiologic testing when feasible, weighing the benefits of obtaining a microbiologic diagnosis against the risks of the procedure (eg, need for intubation, bleeding, bronchospasm, pneumothorax) on a case-by-case basis.

•In all cases, we modify this approach based on epidemiologic exposures, patient risk factors/clinical features, public health needs, and the likelihood that the test results would help refine the differential diagnosis and/or change management.

●Differential diagnosis − A substantial portion of patients initially diagnosed with CAP are ultimately found to have alternate diagnoses. Therefore, findings that further support or refute the initial diagnoses should continue to be sought as the patient’s course progresses (algorithm 3). The most common alternate diagnoses included heart failure, malignancy, pulmonary infarct, and fibrosis. (See 'Differential diagnosis' above.)

●Clinical course − Most patients with CAP will show some signs of improvement within 48 to 72 hours of starting antibiotics. Failure to respond within this time period should raise suspicion for progression of infection with a drug resistant pathogen, development of complications (eg, lung abscess, empyema heart failure, other cardiac events), or an incorrect diagnosis. (See 'Follow-up care' above.)

ACKNOWLEDGMENT — UpToDate gratefully acknowledges John G Bartlett, MD (deceased), who contributed as author on earlier versions of this topic and was a founding Editor-in-Chief for UpToDate in Infectious Diseases.