Community-acquired pneumonia

Clinical problem & epidemiology

CAP is an acute infection of the lung parenchyma in a patient who has not been hospitalized or exposed to healthcare in the prior 14 days. It is among the most frequent causes of hospitalization and infectious death in adults, with mortality driven by severity at presentation and the timeliness of appropriate therapy. Streptococcus pneumoniae remains the most common identifiable bacterial pathogen, but respiratory viruses (influenza, SARS-CoV-2, RSV) now account for a large and growing fraction of cases.

• Incidence is highest at the extremes of age and in patients with chronic cardiopulmonary disease, diabetes, or immunocompromise.
• The 'HCAP' category was retired from the 2019 ATS/IDSA guidelines — broaden coverage based on validated individual risk factors, not the old label.
• Smoking is the single biggest modifiable risk factor; address it at every encounter.

Microbiology & validated risk factors for resistant pathogens

Pathogens cluster by host. S. pneumoniae, H. influenzae, and M. catarrhalis dominate typical CAP; the atypicals (Mycoplasma, Chlamydia, Legionella) are covered whenever a patient is sick enough to admit. The 2019 ATS/IDSA guideline replaced the retired 'HCAP' label with two validated, evidence-based sets of risk factors that actually predict MRSA or Pseudomonas — use these, not the old label, to decide when to broaden empiric coverage.

• Typical: S. pneumoniae, H. influenzae, M. catarrhalis.
• Atypical: Mycoplasma pneumoniae, Chlamydia pneumoniae, Legionella pneumophila.
• Validated risk factors for MRSA: prior respiratory isolation of MRSA, OR recent hospitalization + exposure to parenteral antibiotics.
• Validated risk factors for Pseudomonas: prior respiratory isolation of Pseudomonas, OR recent hospitalization + parenteral antibiotics in a patient with structurally diseased lung (bronchiectasis, severe COPD with recurrent infection).
• Injection drug use raises S. aureus risk (tricuspid valve seeding, septic emboli) — a separate clinical pattern, not captured by the two risk-factor sets above.

Clinical manifestations

The classic presentation is fever, cough, productive sputum, dyspnea, and pleuritic chest pain, with focal exam findings (crackles, consolidation). Presentation is often subtler in the elderly or immunocompromised, who may present with confusion, falls, or functional decline alone — in an older patient, CAP is a leading cause of non-specific functional decline.

• Tachypnea and tachycardia are the most sensitive physical-exam findings; their absence makes CAP less likely.
• Hypothermia, hypotension, or confusion signal severe disease (and are CURB-65 components).
• Extrapulmonary features — diarrhea, hyponatremia, hepatitis, neurologic change — suggest Legionella.

Approach to diagnosis: tiered testing

Chest imaging is required to confirm the diagnosis. Additional testing is tiered by severity: the sicker or more complex the patient, the more diagnostics are justified. Over-testing in low-severity CAP adds cost and false positives without changing management. Procalcitonin does NOT improve clinically important outcomes when used to guide initiation or duration, and should not delay antibiotics in suspected CAP.

• First tier (all patients): chest X-ray (or chest CT if X-ray is negative but suspicion is high).
• Second tier (hospitalized): blood cultures before antibiotics if possible, sputum culture if a good sample can be obtained, respiratory viral testing including influenza and SARS-CoV-2.
• Third tier (severe ICU): urinary antigens for S. pneumoniae and Legionella serogroup 1; consider bronchoscopy if atypical or opportunistic infection is suspected.
• Do NOT use procalcitonin to decide whether to start antibiotics in suspected CAP (ProCAP and related trials showed no outcome benefit). [VERIFY exact ProCAP citation]

Risk stratification & the admission decision

Severity scores objectify the risk of death, but they do not replace clinical judgment — social factors, the ability to take oral medications, and access to follow-up all affect the site of care. The PSI (Pneumonia Severity Index / PORT) is more sensitive (and more complex), weighting age and comorbidities heavily; CURB-65 is a simpler bedside score. The scoring-tools section below gives the full point tables and interpretation bands.

• CURB-65: Confusion, Urea >7 mmol/L (BUN >19 mg/dL), Respiratory rate ≥30, Blood pressure (SBP <90 or DBP ≤60), age ≥65. Score 0–1 consider outpatient; 2 admit; ≥3 consider ICU.
• PSI is preferred for borderline site-of-care decisions because it is more sensitive (fewer inappropriate outpatients).
• Clinical judgment overrides the score — a young parent unable to follow up or take oral medications may still need admission.
• Use an early-warning score (NEWS2, MEWS) to detect deterioration on the floor.

Treatment: empiric therapy by site of care

Empiric therapy follows the site of care and host risk. The principles: cover the core pathogens, add atypicals for inpatient therapy, and add MRSA/Pseudomonas coverage only for the validated risk factors above. The regimen table below summarizes the common scenarios; de-escalate the moment cultures and susceptibilities return.

• Outpatient (healthy): amoxicillin high-dose, doxycycline, or a respiratory fluoroquinolone.
• Outpatient (comorbidities): amoxicillin/clavulanate plus a macrolide or doxycycline; or a respiratory fluoroquinolone alone.
• Inpatient (non-severe): an antipneumococcal beta-lactam (ceftriaxone, cefotaxime, ampicillin-sulbactam, or ertapenem) PLUS a macrolide; or a respiratory fluoroquinolone.
• Severe (ICU): beta-lactam PLUS macrolide (azithromycin) — fluoroquinolone is an acceptable alternative; add vancomycin/linezolid AND an antipseudomonal beta-lactam when validated risk factors are present. Avoid fluoroquinolone monotherapy in ICU disease.

Duration, oral step-down & treatment failure

Shorter is better. Patients who improve can complete therapy orally, and most courses run ≥5 days. Transition to oral therapy does not require repeat blood cultures, a documented normal temperature, or a normal WBC — only clinical stability and the ability to absorb oral drugs. Failure to improve by day 3 should prompt structured reassessment rather than reflexive broadening. Corticosteroids reduce mortality in severe CAP with septic shock; they should be avoided in influenza or undrained infection.

• Minimum 5 days of therapy, provided the patient is afebrile for 48–72 hours and clinically stable.
• Switch IV→PO when the patient is hemodynamically stable, improving, and able to take oral medications — high-bioavailability oral agents (fluoroquinolones, linezolid, doxycycline, metronidazole, fluconazole) achieve serum levels close to IV.
• Causes of failure by day 3: wrong pathogen or resistant organism, complication (empyema, abscess, metastatic infection), wrong diagnosis (PE, heart failure, atypical presentation), or inadequate source control.
• Corticosteroids (e.g., hydrocortisone) reduce mortality in severe CAP with septic shock; avoid in influenza or active undrained infection. [VERIFY regimen and point estimate — CAPE COD II, JAMA Intern Med 2023]

Aspiration pneumonitis vs aspiration pneumonia

Not every infiltrate after a witnessed aspiration event is an infection. Aspiration pneumonitis is a chemical lung injury from sterile gastric contents (typically after a large-volume aspiration in an altered patient) and may not require antibiotics — supportive care plus observation is often sufficient. Aspiration pneumonia is a bacterial infection that develops subacutely, classically in the dependent lobes, with anaerobic and mixed flora. Distinguishing the two prevents unnecessary broad-spectrum antibiotics.

• Aspiration pneumonitis: acute onset after a witnessed aspiration, chemical injury; consider supportive care first and add antibiotics only if the infiltrate persists or fever/purulence develops over 24–48 h.
• Aspiration pneumonia: subacute presentation in a patient with aspiration risk (stroke, seizure, dysphagia, substance use), dependent-lobe infiltrate (RLL, RML, posterior LUL).
• Routine anaerobic coverage is NOT needed for simple CAP; add anaerobic coverage (clindamycin or metronidazole, or a beta-lactam/beta-lactamase inhibitor) only for true aspiration pneumonia, lung abscess, or empyema.

Guidelines

• Diagnosis and Treatment of Adults with Community-Acquired Pneumonia (ATS/IDSA)Metlay JP, et al. Am J Respir Crit Care Med. 2019;200(7):e45–e67The current definitive US guideline — severity, diagnostics, empiric therapy, and the validated MRSA/Pseudomonas risk factors. [VERIFY no 2024–2025 replacement update]
• Recommended Adult Immunization Schedule — United States (pneumococcal)ACIP / CDC MMWR [VERIFY most recent schedule year]Current PCV20 / PCV15→PPSV23 recommendations for adults.

Reviews

• Community-Acquired PneumoniaWunderink RG, Waterer G. N Engl J MedA concise NEJM clinical review of diagnosis and management.

Key trials

• CAPE COD II — Hydrocortisone in severe community-acquired pneumoniaJAMA Intern Med. 2023 [VERIFY exact authors, volume, and point estimate]Randomized trial of hydrocortisone in severe CAP with septic shock; mortality benefit. The current basis for steroid use in severe CAP.
• Procalcitonin-guided antibiotic therapy in community-acquired pneumonia (ProCAP)[VERIFY exact citation — ProCAP / procalcitonin CAP trial]Procalcitonin guidance did not improve clinically important outcomes in CAP — do not use to decide initiation.

Books & reference

• Principles and Practice of Infectious Diseases, 9e — Community-Acquired Pneumonia chapterMandell, Bennett, DolinComprehensive reference for pathogens and management.