Inpatient / Infectious Disease

Community-Acquired Pneumonia (CAP)

Last Updated 10/22/2023

Admission Checklist

-- ABCs: assess for severe CAP and the need for ICU care (if need mechanical ventilation or pressor OR 3 or more of AMS, hypotension requiring resuscitation, hypothermia < 36, RR >30, O2 requirement, WBC < 4, BUN >20 platelets <100, multilobar PNA on CXR)

-- Triage: calculate CURB-65 (confusion, BUN >20, RR >30, BP <90/60, Age >65) and Pneumonia Severity Index (PSI), Shorr Score (for MRSA PNA)

‍-- Chart Check: underlying lung disease, baseline O2 needs, previous culture data, recent abx in last 90 days, abx allergies, influenza and pneumococcal vaccinations

-- Can't Miss: hypoxemia, respiratory failure, sepsis, PE, ACS

-- Admission Orders: pulse ox, supplemental O2; Labs - procalcitonin, MRSA nasal swab, consider RVP, legionella, strep urine Ag; BCx only if c/f severe PNA or sepsis (pos in <20% of inpatients with PNA); Imaging - CT if c/f PE, diagnosis uncertain, or to look for complications

-- Initial Treatment to Consider: continue, broaden, or narrow abx given in the ED based on severity and risk factors; steroids in severe CAP

HPI Intake

-- Symptom Start Date: trajectory

-- Symptoms: dyspnea, cough, sputum, fevers, chills, rigors, fatigue, myalgias, pleuritic chest pain, hemoptysis, night sweats, weight loss, diarrhea; less likely if sore throat, rhinorrhea 

-- History: prior infections, sick contacts, travel, animal exposure, occupational exposures, TB exposures, recent healthcare exposure or abx use, allergies (and reaction), smoking hx

-- MRSA Risk Factors: post-influenza, IVDU, skin pustules, cavitary PNA, nasal swab positive

Pseudomonas Risk Factors: COPD/asthma, bronchiectasis, immunosuppression, nursing home, recent healthcare exposure

-- Aspiration RFs: dementia/AMS, EtOH or other drug use, trouble swallowing, GERD, poor dentition, bedridden

To Note on Exam

-- General - appearance, oxygen needs, breathing effort, rigors, diaphoresis, AMS

-- Cardiac - rhythm, murmurs, JVP

-- Pulmonary – crackles, wheezing, reduced breath sounds, bronchophony (increased clarity and pitch of voice heard over lung), egophony (E sounds like A), tactile fremitus (vibrations more pronounced), dullness to percussion, sputum appearance

-- Extremities - edema, cap refill

-- Skin - rash (erythema multiforme), petechiae


-- Pulmonary Infectious

  • Viral - COVID, influenza, RSV, rhinovirus, adenovirus
  • Bacterial - strep pneumo, H flu, Moraxella, legionella, mycoplasma, chlamydophila pneumoniae, MRSA, pseudomonas
  • Fungal - PJP, aspergillus, histoplasmosis, coccidioides, candida

-- Pulmonary Non-Infectious - aspiration, atelectasis, PE, obstruction/plugging, COP, ILD, lung cancer, chemical or drug-induced pneumonitis, DAH

-- Cardiac - ACS, heart failure

-- Gastrointestinal - GERD, Boerhaave’s

-- Other - RA, lupus, sarcoid, vasculitis, acute chest

Assessment and Plan Dotphrase

# Community Acquired Pneumonia


-- History: pulmonary disease, immunosuppression, aspiration risk factors, healthcare exposures, recent abx, h/o MRSA or PsA, IVDU, abx allergies

-- Clinical: fever, dyspnea, cough, sputum, pleuritic chest pain

-- Exam: O2 requirement, fever, rigors, diaphoresis, AMS, crackles, reduced breath sounds, bronchophony, egophony, tactile fremitus, dullness to percussion, note appearance of mucus from productive cough, rash (erythema multiforme)

-- Data: WBC, CXR, cultures

-- Etiology/DDx: viral (COVID, flu, RSV, rhinovirus, adenovirus), bacterial (strep pneumo, moraxella, legionella, MRSA, PsA), fungal, aspiration pneumonitis, ADHF, ACS, PE



-- Imaging: CXR for all; CT if c/f PE, persistent or worsening symptoms, to assess for empyema, abscess, fungal infection; may be valuable in all immunocompromised patients; ultrasound to trend effusions

-- Labs: f/u MRSA nares; consider the added value of procalcitonin (admission and 48 hours), RVP, urine legionella, and pneumococcal antigen; BCx if severe CAP or sepsis; sputum culture often not helpful


-- Oxygen: currently ***; goal SpO2 > 92%

-- Abx: 5-7 days for CAP, up to 10-14 for severe CAP, MRSA or PsA, and legionella

  • CAP - beta-lactam (ceftriaxone 1-2g IV daily, amp-sulbactam 1.5-3g IV q6 → amox-clav 875/125mg BID, cefpodoxime 200mg q12) plus azithromycin 500mg or doxycycline 100mg daily OR levofloxacin 750mg daily
  • PsA - pip-tazo 4.5g q6-8 or cefepime 2g q8 or meropenem 1g q8 for at least 7 days
  • MRSA - vancomycin or linezolid 600mg q12 for at least 7 days

Steroid: hydrocort 50 q6 or prednisone 50 PO daily if sepsis/shock or severe CAP; taper based on clinical improvement

-- Source Control: large parapneumonic effusions and any empyemas will need to be drained via thoracentesis; consider if >10mm, > half hemithorax, suspected to be causing dyspnea

-- Supportive: acetaminophen PRN, incentive spirometry, benzonatate, dextromethorphan

-- Dispo: when afebrile 48-72 hours, baseline O2 req, stable vitals

If You Remember Nothing Else

Triage by assessing for severe CAP and calculating CURB-65 and PSI. For diagnosis of bacterial pneumonia, focus on pertinent symptoms, evidence of systemic inflammation, and radiological evidence of pneumonia. For typical inpatient CAP, give ceftriaxone and azithromycin/doxy. You don't need to cover for anaerobes even if you are concerned for aspiration. If concerned for MRSA or pseudomonas, give vanc/linezolid or zosyn/cefepime. Viral etiologies are overall the most common etiology of pneumonia symptoms, but you will likely end up treating empirically with abx for at least 48 hours while collecting data and assessing response. Shoot for 5 days of treatment and re-assess after 2-3 days to see if you should extend. Consider steroids in severe CAP or sepsis/shock. Causes of persistent fevers or worsening clinical status while on abx - wrong bug (viral/fungal), wrong drug (i.e resistant), wrong process (non-infectious), no source control (effusion/empyema, abscess), not enough time (fevers usually take 2-4 days to resolve). Patients are often ready for discharge after being afebrile for 48 hours and they are close to their baseline in terms of oxygenation and hemodynamics.

Clinical Pearls

Background and Definitions

  • Pneumonia is inflammation of lung parenchyma (commonly caused by an infection) that can affect alveoli, bronchioles, and the interstitium
  • Pneumonia is the leading cause of infectious death in industrialized countries including the U.S and the most common cause of sepsis
  • Community-Acquired Pneumonia (CAP) is any infectious pneumonia contracted outside of ahospital setting
  • Historically, some PNAs were noted to be "atypical" - they had milder symptoms, dry cough, interstitial pattern on CXR, and different response to antibiotics; we now know this difference is because certain organisms preferentially exert their effect on the interstitium.
  • Confusingly, "atypical" can refer to either the PNA itself or bacteria that often cause it, including chlamydia, legionella, and mycoplasma. Note, however, that viruses are the most common cause, and that "atypical" does not imply uncommon.

Etiology and Pathophysiology

  • The most common etiology of CAP is a viral infection
  • Most common bacteria for CAP include strep pneumo, H flu, and moraxella catarrhalis
  • However, a pathogen is only detected in 38% of patients who are admitted with c/f PNA (NEJM, 2015)
  • As such, the diagnosis of bacterial CAP should focus on: 1. Evidence of Systemic Inflammation – WBC with neutrophil predominance, elevated procalcitonin, fevers, chills, fatigue, myalgias; 2. Pulmonary Signs and Symptoms – hypoxia, productive cough, dyspnea; 3. CXR Evidence of Opacities
  • Pneumonia is almost always associated with the microaspiration of airborne pathogens or oropharyngeal secretions and is very rarely transmitted via hematogenous dissemination
  • Pneumonia is often seen inpatients with failure of lung-protective mechanisms like cough reflex, mucociliary clearance, or dysfunction of alveolar macrophages
  • Pneumonia causes hypoxia via infiltration and inflammation in alveoli and interstitium which leads to decreased ventilation and V/Q mismatch; it also leads to intrapulmonary shunt (right to left) - blood continues to flow to affected areas and does not exchange oxygen
  • The expected time course of treated bacterial pneumonia: fever resolves 2-4 days, hypoxia 3-6 days, CXR clears around 1-3 months
  • Lobar pneumonia and bronchopneumonia (usually involving the lower lobes) is primarily caused by pneumococci whereas interstitial pneumonia is more commonly caused by mycoplasma and viral infections; however, a pathogen should not be assumed  just based on imaging results
  • Coxiella burnetii can cause atypical PNA - from inhalation of aerosols from secretions of livestock or ingestion of raw milk (cattle, sheep, goats)
  • Hematogenous spread to lungs is very rare, and usually when significant nidus like endocarditis and is often staph - if have PNA, likely was from airborne or aspiration
  • Aspiration pneumonitis is an important mimic of pneumonia. It happens after aspiration of aggrevating substances into airway without infection (imaging and sxs caused by lung parenchyma inflammation) – it is more abrupt and may cause low-grade fevers; It commonly will be seen in the RML/RLL if happens when upright or RUL when supine - tough to tell the difference from PNA and usually end up treating for CAP for at least 48 hours and stopping if recovers rapidly
  • If you’re concerned about aspiration PNA, think about risk factors like AMS, trouble swallowing, poor dentition - its still most commonly caused by GNRs and standard CAP organisms - anaerobes do not commonly grown in the lung
  • In retrospect, if the CXR clears quickly, it was more likely aspiration pneumonitis than bacterial pneumonia
  • Cryptogenic Organizing Pneumonia (COP) - a pneumonia-like disease and imaging appearance due to non-infectious chronic inflammation - be on lookout in those taking amiodarone or with history of rheumatoid arthritis

Clinical Presentation and Diagnosis

  • Any pulmonary disease can mimic PNA
  • You can still have viral infection AND bacterial PNA - having a positive RVP does not mean you should withhold abx
  • CURB-65 and PSI are used to calculate mortality risk and have technically not been validated for determining the ideal place of care – valuable for sick vs not sick
  • Rhinorrhea and sore throat are more commonly seen in URI and less typical for CAP
  • Sputum samples are usually not sent due to low sensitivity and yield (Strep pneumo and H Flu are tough to grow in culture); if you do send, it is considered adequate if >25 PMNs and <10 squamous epithelial cells, otherwise there is concern that the sample is saliva
  • Only send BCx if c/f severe PNA or sepsis; they are positive in <20% of inpatients with PNA
  • MRSA Nasal Swab - 98% NPV for MRSA pneumonia - if negative, can feel comfortable pulling off MRSA coverage unless severely ill (note that this was only studied on the admission MRSA swab, not if patient had c/f HAP after being admitted) (Clin Infect Dis, 2018)
  • Procalcitonin was validated in ICU settings to try and limit duration of antibiotics, thus on admission can be tough to interpret; if very high can indicate bacterial infection, and would be useful to trend; you can consider stopping abx if the level trends down to <80% of peak or is <0.25-0.50; note that it has variable sensitivity
  • For patients with a negative procalcitonin (less than 0.5) it may be reasonable to discontinue beta-lactam antibiotics but continue coverage for atypical pathogens (with azithromycin or doxycycline) because gram negatives raise it whereas atypicals are less likely to raise it
  • Note that procalcitonin may be falsely low in immunosuppressed patients
  • Legionella Urine Ag should be sent if severe CAP, or there is concomitant hyponatremia (though this association is likely overstated), diarrhea, or recent exposure
  • If immunocompromised, should consider PJP, aspergillus, histoplasma, coccidioides, candida, and CMV – note that BDG sensitive but not specific for fungal infection
  • Bacterial PNA shouldn't cause pulmonary nodules – this would be more c/w fungal infection


  • HFNC can help to reduce work of breathing and stave off intubation; it's likely better than BIPAP since patients can tolerate it for longer and it allows for easier clearance of secretions
  • In general, unless you are concerned about poor source control, resistance, or new infection, 7 days of antibiotics is almost always sufficient even in sick ICU patients with MRSA or pseudomonas
  • Only give metronidazole or clindamycin if you are concerned about lung abscess or empyema because these represent non-oxygenated areas in or around the lung where anaerobes can grow; do not treat anaerobes for run of the mill aspiration pneumonia
  • Fluoroquinolones (and less so azithromycin) can prolong QT, be wary if on other prolonging meds
  • Ciprofloxacin is less effective than levofloxacin or moxifloxacin at treating pneumococcal respiratory infections and should be avoided in PNA; reason is it has less gram-positive coverage (strep pneumo) and less lung penetration
  • Linezolid can cause serotonin syndrome when given with other medications such as SSRIs and methadone
  • Daptomycin cannot be used for pneumonia since it is inactivated by surfactant
  • Atypical bacteria need to be treated with something other than beta-lactam because they have absent or unique cell walls
  • Patients with legionella may benefit from extended courses of antibiotic (e.g., azithromycin 500 mg IV daily for 7-10 days, instead of three days)
  • "Double Pseudomonas coverage" should rarely be considered in patients with structural lung disease or those at high risk for mortality; RCTs have not seen that it helps but if you are going to cover, choose amikacin (if PsA resistant to cefepime/mero/etc is is highly likely resistant to options like fluoroquinolones) understanding there is a risk of AKI; EMCrit argues a good time to use would be septic shock 2/2 HAP/VAP in the setting of gram neg bacteremia and good underlying renal function; “In order for double-coverage to be beneficial, a chain of events must occur. The patient must truly have VAP, that VAP must be due to a gram-negative, the gram-negative must be resistant to the beta-lactam, the gram-negative must be sensitive to the second antibiotic, and broader antibiotic coverage must make a clinical difference. The likelihood of this entire sequence of events occurring is about 1-2%.”


  • POCUS is good option to trend effusions
  • If an effusion is large or loculated you will likely need to drain via thoracentesis; empyema may need a pigtail catheter to continuously drain
  • A Parapneumonic effusion develop in the pleural space adjacent to a PNA; they are seen in 20-57% of inpatients with PNA and 5-10% progress to empyema; it usually takes ~300cc of fluid to be seen on CXR; effusions are “complicated” if fluid has a positive GS or culture, or is loculated
  • An empyema is purulence in pleural space – these will always require drainage
  • Light’s criteria is used to assess for an exudative vs transudative effusion; ratio of the protein in effusion/serum > 0.5, ratio of LDH effusion/serum > 0.6, or if LDH significantly elevated in general; you can also look for low pH, low glucose as other suggestions that it is an exudative effusion, but not part of light’s criteria
  • In general, pleural effusions alone are rarely the cause of dyspnea or hypoxia unless it takes up > half of a hemithorax - you should be thinking about other etiologies for symptoms
  • If the patient does not improve with antibiotics or they keep getting recurrent pneumonia - think recurrent aspiration, underlying lung dx (COPD, bronchiectasis), post-obstructive (malignancy), immunodeficiency, or COP – a CT may help clarify
  • Mycoplasma can cause erythema multiforme and autoimmune hemolytic anemia (AIHA) from IgM cold agglutinin

Trials and Literature

Treatment Guidelines and Trials

  • CAP Treatment Guidelines 2019 (Am J Respir Crit Care Med, 2019)
  • CAP Review Articles (NEJM, 2014) (NEJM, 2023)
  • CDC-EPIC Study (NEJM, 2015) - viruses cause 23% of CAP, strep pneumo 5%, only 38% of cases identify ANY organism
  • In patients with severe PNA treated on the floor, switching to PO abx after 3 days IV abx vs 7 days of IV abx showed similar 28-day mortality mortality (4% vs 6%) and reduced length of stay by almost 2 days (BMJ 2006)
  • It’s okay to discharge patients on the day you switched to PO abx - they don’t need to be watched in the hospital (Am J Med 2006)
  • Duration of antibiotic treatment in CAP - 5 Days of abx (physician’s choice) for hospitalized patients with CAP was non-inferior to longer courses; however mostly used fluoroquinolones (80%) which is not great and not very generalizable (was done in a region of Spain) AND only randomized after patients had already been getting better by 5 days (JAMA, 2016)
  • Patients with severe CAP admitted to the ICU benefit from steroids with improved 28-day mortality (6.2% vs. 11.9%) (NEJM, 2023); note that guidelines do not recommend steroids in CAP, despite some evidence that they may improve time to clinical stability (Lancet, 2015)
  • CAP-START Trial - beta-lactam monotherapy was non-inferior (90-day mortality) to a combo of beta-lactam and macrolide or fluoroquinolone alone; notable that was done in an area with low rates of atypicals (NEJM, 2015)

Procalcitonin and Pneumonia

Exam in Pneumonia

  • Does This Patient Have a Pleural Effusion (Rational Clinical Exam, JAMA)

Other Resources

Blogs and Chapters

  • EMCrit Blog Post on PsA "Double Coverage" - argues double coverage for GNR is very unlikely to help, and the antibiotics most likely to help (amikacin) have a chance of harm via AKI; a good time to use would be septic shock 2/2 VAP iso gram neg bacteremia and good underlying renal function; “In order for double-coverage to be beneficial, a chain of events must occur. The patient must truly have VAP, that VAP must be due to a gram-negative, the gram-negative must be resistant to the beta-lactam, the gram-negative must be sensitive to the second antibiotic, and broader antibiotic coverage must make a clinical difference. The likelihood of this entire sequence of events occurring is about 1-2%.”
  • EMCrit Blog Post on Azithromycin and QTc - argues that azithromycin may minimally prolong QTc but is unlikely to be associated with an increased risk of torsades
  • Internet Book of Critical Care (IBCC) - Severe CAP

Differentials and Schema