T he NEW ENGL A ND JOUR NA L of MEDICINE

CLINICAL PRACTICE

Patrick G. O’Malley, M.D., M.P.H., Editor

Community-Acquired Pneumonia
Thomas M. File, Jr., M.D., and Julio A. Ramirez, M.D.

This Journal feature begins with a case vignette highlighting a common clinical problem. Evidence sup-
porting various strategies is then presented, followed by a review of formal guidelines, when they exist. The
article ends with the authors’ clinical recommendations.

From the Division of Infectious

Disease, Summa Health, Akron, A 66-year-old man with underlying chronic obstructive pulmonary disease (COPD)
and the Section of Infectious presents to the emergency department with a 2-day history of fever, dyspnea,
Disease, Northeast Ohio Medical
and cough productive of green, purulent sputum. He had noted increasing dyspnea
University, Rootstown — both in
Ohio (T.M.F.); and Norton 3 days before the onset of fever. He describes one episode of acute exacerbation
Infectious Diseases Institute, of COPD that occurred 6 months earlier. The physical examination is notable for
Norton Healthcare, and the mild respira- tory distress and confusion, with disorientation to time. His
Division of Infectious Diseases,
University of Louisville — both in temperature is 38.6°C, heart rate 100 beats per minute, blood pressure 140/85
Louis- ville, KY (J.A.R.). Dr. File may mm Hg, respiratory rate 24 breaths per minute, and oxygen saturation 92% while
be contacted at he is breathing ambient air. Auscultation of the lungs reveals coarse rhonchi over
filet@summahealth.org or at
Summa Health, 75 Arch St., Suite the right midlung field. Chest radiography reveals right upper-lobe consolidation
506, Akron, OH 44304. (Fig. 1). His white-cell count is 14,000 per cubic millimeter, platelet count 159,000
N Engl J Med 2023;389:632-41.
per cubic millimeter, serum so- dium 136 mmol per liter, blood urea nitrogen 19
DOI: 10.1056/NEJMcp2303286 mg per deciliter (6.8 mmol per liter), creatinine 1.1 mg per deciliter (97.2 μmol per
Copyright © 2023 Massachusetts Medical Society. liter), and procalcitonin 5.4 ng per mil- liliter (normal range, 0.00 to 0.05). A
CME multiplex viral panel was positive for respira- tory syncytial virus. How would
at NEJM.org you further evaluate and treat this patient?

THE CLINICAL PROBLEM

Community-acquired pneumonia is an acute infection of the pulmonary

parenchyma in a patient who has acquired the infection in the community (as
distinguished from an infection acquired in a hospital). In the United States,
community-acquired pneumonia is one of the leading causes of hospitalization
and death, with approxi- mately 6 million cases reported each year.1-6 The annual
incidence of hospitaliza- tion for community-acquired pneumonia in the United
States is approximately 650 adults per 100,000 population, corresponding to
1.5 million unique hospital- izations for the disease each year.6 Factors that
increase the risk of community- acquired pneumonia include advanced age,
chronic lung disease, chronic heart dis- ease, cardiovascular disease, diabetes
mellitus, malnutrition, viral respiratory tract infections, immunocompromising
conditions, and lifestyle factors such as smoking and excessive alcohol
consumption.
The development of pneumonia is influenced by a combination of factors,
includ- ing host susceptibility, pathogen virulence, and the inoculum of
microorganisms reaching the lower airways. Respiratory pathogens must
overcome several defense mechanisms of the respiratory system before reaching
the alveoli. These defenses include mucus trapping, mucociliary clearance,
coughing, and swallowing. Pathogens can reach the alveoli by means of
microaspiration (aspiration of small amounts of oropharyngeal secretions that
often occurs during sleep), inhalation, macroaspira-
The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 T he NEW ENGL A ND JOUR NA L of MEDICINE

632 N ENGL J MED 38e;7 NEJM.ORG AUGUST 17, 2023

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 CLINICAL PRACTICE

KEY CLINICAL POINTS

COMMUNITY-ACqUIRED PNEUMONIA
• The diagnosis of community-acquired pneumonia is made on the basis of compatible
symptoms and signs, with evidence of a new infiltrate on an imaging study.
• Most outpatients with mild community-acquired pneumonia can be treated
empirically without diagnostic testing for bacteria. However, testing for SARS-CoV-2 and
influenza should be considered.
• A comprehensive approach to microbiologic testing for hospitalized patients is
recommended for determining the appropriate pathogen-directed therapy.
• The choice of antimicrobial therapy for community-acquired pneumonia varies
according to severity, coexisting conditions, and the likelihood of antimicrobial-
resistant organisms.

tion infection. The inflammatory responses explain

(aspiration of a large amount of
oropharyn- geal or upper gastrointestinal most of the host patient’s signs and symptoms
contents), or hema- togenous spread. as well as laboratory and imaging abnormalities
Microaspiration is the primary path for (Fig. 2). In some patients, the initial systemic
microorganisms into the lungs, and in- flammatory response can become
macroaspiration may lead to aspiration pneu- dysregulated and result in tissue injury and
monia.7 The alveolar macrophage is the eventual organ dys- function.9
primary defense mechanism in the lung. The Numerous microorganisms can cause
lung micro- biome may also contribute to defense commu- nity-acquired pneumonia. The bacteria
mechanisms by producing antimicrobial and viruses that are considered to be likely
molecules or compet- ing for nutrients.8 etiologic agents in all patients with community-
If pathogens overcome the alveolar defense acquired pneumo- nia are described as core
mechanisms, they will multiply and cause respiratory pathogens (Table 1).10-12 Severe
local tissue damage. Injured host cells then acute respiratory syndrome coronavirus 2
produce damage-associated molecular patterns (SARS-CoV-2) is now a predomi- nant viral
that further stimulate alveolar macrophages to pathogen in patients with community- acquired
produce cy- tokines and chemokines, pneumonia. Uncommon or infrequent causes of
triggering a local in- flammatory response. community-acquired pneumonia should be
Cytokine spillover into the bloodstream considered as likely agents in patients who
produces a systemic inflammatory response. present with risk factors for a particular
The local and systemic inflammatory patho- gen (e.g., travel or animal exposure)
responses constitute a physiologic response to lung (Table S3 in

A B

Lorem
ipsum

Figure 1. Chest Radiographs.

Posteroanterior (Panel A) and lateral (Panel B) views show right upper-lobe infiltrate.

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 CLINICAL PRACTICE

N ENGL J MED 38e;7 NEJM.ORG AUGUST 17,

2023 633

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 T he NEW ENGL A ND JOUR NA L of MEDICINE

CAP
Severe CAP
Physiologic inflammatory
Dysregulated inflammatory response
response

Organ failure

Renal
Multiple
Cardiovascula
Local Systemic Organ organ
r failure failure
inflammato inflammato dysfuncti dysfuncti
ry ry on on
Neurologic
response response syndrome
Respirator failure
y failure
Hematologic

failure

Metabolic

failure

Hepatic failure

Clinical, Laboratory, and Imaging Correlation

American Thoracic Society and Infectious Diseases Society of
America Criteria to Define Severe CAP

Cough Major Septic shock treated with vasopressors

Fever Confusion
Sputum Criteria Respiratory failure necessitating mechanical ventilation
Chills Hypothermia
production Malaise Hypotension Minor Respiratory rate ≥30
Shortness of Anorexia Leukopenia Criteria breaths/min Confusion or
breath Pleuritic Tachycardi Thrombocytopenia disorientation or both
chest pain a Uremia Hypothermia (core temperature <36°C or 96.8°F)
Tachypnea Leukocytosi Hyperlactatemia Hypotension necessitating aggressive fluid
Dullness to s Left Increased liver resuscitation Leukopenia (white-cell count <4000
percussion Rales shift enzymes Increased cells/µl) Thrombocytopenia (platelet count
or crackles Increased C- creatinine <100,000 per µl) Uremia (blood urea nitrogen
Hypoxemia reactive protein level ≥20 mg/dl)
Pulmonary Increased Ratio of PaO2 to FIO2 ≤250
infiltrates Severe
erythrocyte Multilobar (≥2) infiltrates
CAP
sedimentation
rate One major criterion or ≥3 minor criteria

Figure 2. Pathogenesis of Community-Acquired Pneumonia (CAP) with Corresponding Clinical, Laboratory, and Imaging Abnormalities.
Pao2 denotes partial pressure of arterial oxygen, and FIo2 the fraction of inspired oxygen.

the ad- mission to the intensive care unit

Supplementary Appendix, available with the (ICU).6,15
full text of this article at NEJM.org) or in
special populations, such as
immunocompromised pa- tients (Table 1).12 STR ATEGIES AND EVIDENCE
Although community-acquired pneumonia has DIAGNOSIS AND EVALUATION
traditionally been viewed as an acute disease The diagnosis of community-acquired
of the lungs, the current understanding is that pneumonia is made on the basis of
it is a multisystem disease that can result in infiltrates shown on a
acute and long-term sequelae (Fig. 3).13,14
Community- acquired pneumonia is associated
with substan- tial long-term illness and death,
with death at
1 year occurring in approximately 30% of all
hospitalized patients and in approximately 50%
of patients whose conditions had resulted in
The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 T he NEW ENGL A ND JOUR NA L of MEDICINE

chest radiograph (or on computed tomography in a Although the procalcitonin level is typically
patient with symptoms if a chest radiograph is elevated in bacte- rial community-acquired
negative), plus supporting symptoms, signs con- sistent pneumonia, it is low in viral community-acquired
with airspace disease (e.g., rales, rhonchi, or egophony), pneumonia. Procalcito- nin levels quickly
or laboratory abnormalities result- ing from the local decline with the resolution of bacterial
and systemic inflammatory responses (Fig. 2). Testing infection, a response that can inform the
for the inflammatory biomarker procalcitonin may decision to discontinue treatment with anti-
supplement clinical judgment with regard to diagnosis microbials.16-18 However, procalcitonin levels are
and course of bacterial community-acquired not definite indicators, since false positives
pneumonia, since synthesis of procalcitonin is triggered can occur (e.g., in hemorrhagic shock or kidney
by specific cytokines in response to bacteria. injury),

634 N ENGL J MED 38e;7 NEJM.ORG AUGUST 17, 2023

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 CLINICAL PRACTICE

and some bacteria (e.g., mycoplasma) may cause Table 1. Respiratory Pathogens in Community-Acquired Pneumonia (CAP).*
pneumonia in patients with normal procalcitonin levels.
If access to chest radiography is limited, a diagnosis Pathogen Group Pathogen
can be suggested by the findings from a comprehensive Common or core
examination, including evidence of lung consolidation. Gram-positive Streptococcus pneumoniae, methicillin-
Community-acquired pneu- monia is considered to be bacteria susceptible Staphylococcus aureus, Strep.
severe if there are manifestations of organ pyogenes, other streptococci
dysfunction or organ failure. The American Thoracic Gram-negative Hemophilus influenzae, Moraxella
Society and Infec- tious Diseases Society of America bacteria catarrhalis, Enterobacteriaceae (e.g.,
Klebsiella pneu- moniae)
(ATS–IDSA) criteria for defining severe community-
Atypical bacteria Legionella pneumophila, Mycoplasma pneumoniae,
acquired pneumonia are shown in Figure 2.1
Chlamydophila pneumoniae
Respiratory viruses Influenza virus, SARS-CoV-2, respiratory
SITE OF CARE
syncytial virus, parainfluenza virus,
The human meta- pneumovirus,
decision regarding the site of care rhinoviruses, common human
depends on many variables, including severity coronaviruses
of illness, associated disease, presence of Uncommon or infrequent
hypoxemia, ade- quacy of home support, and Gram-positive Methicillin-resistant Staph. aureus, nocardia
probability of ad- herence to treatment. The bacteria spe- cies, Rhodococcus equi
severity of a patient’s illness is primarily a Gram-negative Enterobacteriaceae, including extended-
determination based on clinical judgment, bacteria spectrum beta-lactamases or
carbapenem-resistant
which can be supplemented by the use of enterobacteriaceae; nonfermenting
severity scores. The most commonly used ba-
severity scores are the Pneumonia Severity cilli (e.g., pseudomonas or
Index (PSI) and CURB-65, a score that acinetobacter);
Francisella tularensis
incorpo- rates confusion, urea, respiratory rate,
Atypical bacteria Chlamydia psittaci, Coxiella burnetii
blood pres- sure, and age ≥65 years.19,20
Mycobacteria Mycobacterium tuberculosis, nontuberculous
Calculations for determining PSI scores are mycobacteria
provided in the Sup- plementary Appendix.
Viruses Cytomegalovirus, herpes simplex, varicella
The CURB-65 scale rang- es from 0 to 5; scores zoster, MERS-CoV
are calculated by assigning 1 point each for the
Fungi Pneumocystis jirovecii, aspergillus species,
presence of new-onset con- fusion, blood urea muco- rales species, histoplasma
nitrogen level greater than 19 mg per deciliter, * Risk factors associated withspecies,
specificcryptococ-
pathogens cus species,
are shown in Table
respiratory rate greater than 30 breaths per S3. MERS- CoV denotes Middle East respiratory syndrome coronavirus,
minute, systolic blood pressure less than 90 and SARS-CoV-2 severe acute respiratory syndrome coronavirus 2.
mm Hg or diastolic blood pressure less than
60 mm Hg, and age of 65 years or older. accurate, and cost-effective methods of obtaining
Outpatient treatment is recommended for a results for most patients at the point of
patient with a CURB-65 score of 0 or 1, a short service. However, molecular diagnostic
hospital stay or close observation should be techniques that combine sensitivity, specificity,
considered for a patient with a score of 2, and and a rapid turn- around time are becoming
hospitaliza- tion is recommended for a patient increasingly avail- able.21-23 The coronavirus
with a score of 3 to 5. Indications for ICU caredisease 2019 (Covid-19) pandemic has
are based on further criteria, including the useillustrated the etiologic impor- tance of
of mechanical ventilation and the presence of respiratory viruses that are primarily
shock (Fig. 2). Severity score thresholds have identified by molecular testing.
not been defined for the treatment of patients Microbiologic testing for bacterial causes is
who are immuno- compromised; thresholds generally not recommended for most patients
for admission should be based on clinical who are treated in ambulatory settings, since
judgment.12 empirical antibiotic therapy is largely successful.
However, testing for viruses (e.g., SARS-CoV-2
MICROBIOLOGIC TESTING and influ- enza) should be considered, since the
The identification of the causative agents of results can affect the choice of therapy.
com- munity-acquired pneumonia was limited Establishing an
in the past because there were no rapid, easily
performed,
The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 CLINICAL PRACTICE

N ENGL J MED 38e;7 NEJM.ORG AUGUST 17, 2023 635

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 T he NEW ENGL A ND JOUR NA L of MEDICINE

MRSA polymerase-chain-reaction (PCR) assay

Patient with CAP
can be useful, since a negative result can allow
for discontinuation of anti-MRSA therapy.24 A
more comprehensive microbiologic workup
Short-Term
Outcomes or Long-Term Outcomes or should be performed on the basis of
Acute Sequelae of CAP Post-Acute Sequelae of CAP epidemiologic exposures as well as individual
(outcomes within 30 days after CAP) (outcomes months or years after
CAP)
patient characteristics such as
immunosuppression.12
Pulmonary dysfunction Development of new or
TREATMENT
(e.g., persistent shortness worsening pulmonary disease
of breath, hypoxemia) (e.g., chronic bronchitis, Empirical
emphysema) antimicrobial therapy targets common
Cardiovascular pathogens on the basis of risk factors.1
dysfunction (e.g., Development of new or Antiviral therapy for influenza or SARS-CoV-2
arrhythmias, myocardial worsening cardiovascular
infarction, heart failure, disease infection should be administered according to
stroke) (e.g., stroke, myocardial clinical fac- tors, the results of diagnostic tests,
infarction, arrhythmia, heart or both.1,25 Therapy should be administered as
Musculoskeletal failure)
dysfunction (e.g., soon as pos- sible after community-acquired
generalized weakness Development of new or pneumonia is di- agnosed. Therapy for patients
affecting activities of daily worsening neurologic disease who are immuno- compromised is beyond the
living) (e.g., memory loss, confusion,
depression, dementia)
scope of this article and has been described
Neurologic dysfunction elsewhere.12
(e.g., memory loss, confusion) Development of new or
worsening musculoskeletal,
Dysfunction of other organs
Ambulatory Patients
metabolic, or kidney disease
(e.g., kidney, liver) For most patients who are younger than 65
Pneumonia and other infections years of age, otherwise healthy, and have not
Deterioration of coexisting
conditions (e.g., chronic
recently received treatment with antibiotics,
Rehospitalization
obstructive pulmonary recent ATS– IDSA guidelines recommend one
disease, heart failure, Death of the follow- ing three oral medication
diabetes, kidney disease) (approximately 30 to 35% of options: amoxicillin (1 g three times daily),
patients 1 yr after
Rehospitalizati hospitalization for CAP)
doxycycline (100 mg twice daily), or a
macrolide (azithromycin at a dose of 500 mg
on day 1, then 250 mg daily, or clarithro- mycin
Figure 3. Potential Clinical Outcomes of CAP.
at a dose of 500 mg twice daily [extended-
release, 1000 mg daily]). Macrolides should be
considered only in areas where pneumococcal
resistance to macrolides is less than 25% —
etiologic diagnosis of community-acquired pneu- which excludes the United States, where
monia in hospitalized patients is important resistance ex- ceeds 30%.
for several reasons, including the appropriate For
patients who have taken antibiotics
selec- tion of an antibiotic for use against a
with- in the past 3 months, have serious
specific pathogen, promote good antimicrobial
coexisting conditions (e.g., chronic heart, lung,
steward- ship, and allow identification of
kidney, or liver disease; diabetes mellitus; or
pathogens associ- ated with notifiable diseases
alcohol depen- dence), or who are smokers,
such as SARS-CoV-2 infection or legionnaires’
amoxicillin–clav- ulanate (875 mg orally twice
disease. Available rec- ommended tests include
daily [extended-re- lease, 2 g twice daily]) and
sputum Gram’s staining and culture, blood
either a macrolide (preferred) or doxycycline
cultures, urine immunochro- matographic
are recommended. Pa- tients who cannot take
analysis for Streptococcus pneumoniae and
beta-lactam agents owing to hypersensitivity or
Legionella pneumophila serogroup 1, and mo-
adverse effects can instead be treated with a
lecular techniques such as multiplex assays
respiratory fluoroquinolone (levofloxacin at a
that include SARS-CoV-2. In addition, if there
dose of 750 mg daily or moxi- floxacin at a
is a risk of methicillin-resistant Staphylococcus
dose of 400 mg daily) or one of two more
aureus (MRSA) infection, obtaining a nasal
recently approved agents, lefamulin or
swab for a
omadacycline.26,27
The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 T he NEW ENGL A ND JOUR NA L of MEDICINE

636 N ENGL J MED 38e;7 NEJM.ORG AUGUST 17, 2023

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 CLINICAL PRACTICE

Hospitalized patient with CAP and respiratory failure necessitating mechanical ventilation or septic shock treated with vasopressors

No Yes

Presence of risk factor for pseudomonas as cause of CAP?

No Yes

Presence of risk factors for MRSA? Presence of risk factors for MRSA?

No Yes No Yes

Group 1 Group 2 Group 3 Group 4

Empirical therapy for Empirical therapy Empirical therapy for core Empirical therapy for core
core pathogens for core pathogens plus MRSA pathogens plus pseudomonas pathogens plus pseudomonas
and MRSA

Ampicillin–sulbactam Beta-lactam + atypical coverage Piperacillin–tazobactam Piperacillin–tazobactam

or ceftriaxone or or alternative regimen as in or cefepime or or cefepime or
cefotaxime group 1 ceftazidime or ceftazidime or
+ + imipenem or imipenem or
Azithromycin or clarithromycin Vancomycin or linezolid meropenem meropenem
or doxycycline + +
Azithromycin or Azithromycin or
clarithromycin or doxycycline clarithromycin or
Alternative regimens: or levofloxacin or doxycycline or levofloxacin
levofloxacin or moxifloxacin moxifloxacin or moxifloxacin
or +
omadacycline or lefamulin Vancomycin or linezolid

Figure 4. Empirical Therapy for Hospitalized Patients with CAP.

In the presence of severe CAP with respiratory failure necessitating mechanical ventilation or shock treated with
vasopressors, initial therapy against methicillin-resistant Staphylococcus aureus (MRSA) and pseudomonas can be
considered, pending assessment for risk factors and subsequent microbiologic test results. Strong risk factors for
pseudomonas include known colonization or previous infection and gram-negative bacilli on Gram’s staining;
weak risk factors include receipt of intravenous (IV) antibiotics in the previous
3 months, bronchiectasis, and frequent exacerbations of chronic obstructive pulmonary disease necessitating
glucocorticoid therapy or antibiotic use. Strong risk factors for MRSA include known colonization or previous
infection and gram-positive cocci in clusters on Gram’s staining; weak risk factors include receipt of IV antibiotics
in the previous 3 months, recent influenza-like illness, cavitary infil- trate or empyema, and end-stage renal disease.
In the presence of any strong risk factors, initiation of empirical therapy targeting MRSA or Pseudomonas aeruginosa is
recommended. However, in patients with weak risk factors, the decision to initiate empirical therapy for multidrug-
resistant pathogens should be based on clinical judgment and individual assessment. For patients in group 1 who
are receiv- ing care in the intensive care unit, combination therapy is recommended with a beta-lactam plus
macrolide or a beta-lactam plus fluo- roquinolone. The selection of an antipseudomonal antibiotic should be made
on the basis of the susceptibilities of previous isolates or hospital antibiograms (or both), if available. Empirical
therapy with two agents may be needed if the local prevalence of drug-resistant strains is high or in patients with a
history of multidrug-resistant infection. The combination of piperacillin–tazobactam and vancomy- cin has been
associated with acute kidney injury; we generally avoid this combination if possible. Recommended therapies and
doses for patients with normal renal function are ampicillin–sulbactam (3 g IV every 6 hours), ceftriaxone (1 to 2 g
IV daily), cefotaxime (1 to 2 g IV every 8 hours), azithromycin (500 mg IV or orally daily), clarithromycin (500 mg
twice daily) or clarithromycin XL (two 500-mg tablets once daily), doxycycline (100 mg orally or IV twice daily),
levofloxacin (750 mg IV or orally daily), moxifloxacin (400 mg IV or orally daily), omadacycline (200-mg IV loading
dose on day 1 followed by 100 mg IV daily, or 300 mg orally twice daily on day 1 then
300 mg daily), lefamulin (150 mg IV every 12 hours or 600 mg orally every 12 hours), vancomycin (15 to 20 mg per
kilogram of body weight IV every 8 to 12 hours or a loading dose of 20 to 35 mg per kilogram IV not to exceed 3000
mg for severe CAP; subsequent dose amounts should be based on area-under-the-curve values), linezolid (600 mg IV
or orally twice daily), piperacillin–tazobactam (4.5 g IV every 6 hours), cefepime (2 g IV every 8 hours), ceftazidime
(2 g IV every 8 hours), imipenem (500 mg IV every 6 hours), and meropenem (1 g IV every 8 hours).

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 CLINICAL PRACTICE

N ENGL J MED 38e;7 NEJM.ORG AUGUST 17, 2023 637

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 T he NEW ENGL A ND JOUR NA L of MEDICINE

Hospitalized Patients
Start empirical therapy for bacterial causes of CAP in patient hospitalized with CAP
The choice of the appropriate antibiotic agent
for treatment of a patient who has been
admitted to the hospital is based on the
presence of risk fac- tors for MRSA or
Microbiologic 1 Identification of a viral
Workup pathogen pseudomonas (or both), as shown in Figure 4.
(viral respiratory panel multiplex In patients admitted to a general ward
Yes No
Continue without risk factors for MRSA or
for bacterial
therapy
CAP pseudomonas, combination therapy with a
2 Identification of a bacterial pathogen
beta- lactam plus a macrolide or doxycycline or
(sputum mono- therapy with a fluoroquinolone is
culture or PCR, blood cultures, urinary recommended (see group 1 in Fig. 4).
Continue
No Yes
for bacterial
therapy
Although data from ran- domized trials are
CAP lacking, many observational studies have
suggested that macrolide combina- tion
Radiographic 3
regimens are associated with better clinical
Radiographic findings compatible with a
Findings bacterial outcomes in patients with severe community-
pathogen (lobar consolidation, dense alveolar acquired pneumonia, possibly owing to the
No Yes
Continue im- munomodulatory effects of macrolides.28-31
therapy
for bacterial
CAP If risk factors for MRSA, pseudomonas, or
other gram- negative pathogens not covered by
the standard community-acquired pneumonia
Inflammatory 4 White-cell count compatible with a bacterial
Markers
regimens out- lined above are present,
pathogen
coverage should be ex- panded (see groups 2,
No Yes
Continue 3, and 4 in Fig. 4).
for bacterial
therapy
CAP
Patients
with severe community-acquired
5 CRP compatible with a bacterial pneu- monia who are admitted to the ICU are
pathogen
(>150 mg/liter) more likely to be at risk for resistant pathogens,
Continue includ- ing MRSA and pseudomonas.1,32,33 The
No Yes
for bacterial
therapy establish- ment of an etiologic diagnosis is
CAP
important in the treatment of these patients.
6 PCT at the time of admission and 24–48 hr Evidence to guide appropriate therapy in
after admission compatible with a bacterial patients with severe com- munity-acquired
pathogen (>0.25 ng/ml) pneumonia is limited, but com- mon practice is
to administer anti-MRSA therapy and
Continue
No Yes
for bacterial
therapy antipseudomonas therapy to patients in the
CAP ICU who have shock that is being treated with
vasopressors or respiratory failure that necessi-
Clinical 7
tates mechanical ventilation, pending results of
Evidence of clinical
Finding deterioration cultures and PCR tests (group 4 in Fig. 4).34
24–48 hr after admission Therapies that modify the host response,
No Yes
Continue such as dexamethasone, interleukin-6
for bacterial
therapy
CAP inhibitors, and kinase inhibitors, have been
established for pa- tients with community-
Viral CAP without evidence of bacterial coinfection
acquired pneumonia due to SARS-CoV-2
Discontinuation of antibiotic therapy for bacterial CAP infection.25 The use of glucocorti- coids in the
treatment of other causes of commu- nity-
Figure 5. Discontinuation of Antibacterial Therapy in Viral Pneumonia acquired pneumonia is evolving, with recent
without Evidence of Bacterial Coinfection. evidence showing a benefit of survival among
CRP denotes C-reactive protein, PCR polymerase chain reaction, and patients with severe community-acquired
PCT procalcitonin. pneu- monia (i.e., patients who had been
admitted to the ICU and had received
mechanical ventilation) and patients at high
risk for respiratory failure who had been
treated with hydrocortisone at a dose of 200
The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 T he NEW ENGL A ND JOUR NA L of MEDICINE

mg daily initially, followed by a taper.35

Glucocorticoid therapy should be avoided in
pa- tients with influenza or aspergillus
pneumonia.

638 N ENGL J MED 38e;7 NEJM.ORG AUGUST 17, 2023

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 CLINICAL PRACTICE

DE-ESCALATION OF ANTIMICROBIAL THERAPY associated with hospitalization.

If the etiologic cause of community-acquired Communication and coordination with the pa- tient’s
pneu- monia has been identified by means of primary care clinician for early outpatient
reliable microbiologic methods and there is
no labora- tory or epidemiologic evidence of
coinfection, treatment regimens should be
simplified and di- rected toward that
pathogen.36,37 If a causative pathogen is not
identified, empirical treatment should be
continued, provided the patient’s symp- toms are
abating. If a screening nasal swab for MRSA is
negative, empirical anti-MRSA therapy can
usually be discontinued.
In
patients in whom viral community-acquired
pneumonia is suspected owing to the
identifica- tion of a virus (including SARS-CoV-
2) by means of a molecular test and in whom
there is no evi- dence of concurrent bacterial
infection or clini- cal deterioration,
antibacterial treatment can be discontinued
(Fig. 5). Most patients have some clinical
improvement within 48 to 72 hours after the
start of antibacterial treatment. Intravenous
antibiotic regimens can be transitioned to oral
regimens with a similar spectrum activity as
the patient’s condition improves.38,39

DURATION OF THERAPY
Typically,
patients continue to receive
treatment until they have been afebrile and in
a clinically stable condition for at least 48 hours.
Treatment should usually continue for a
minimum of 5 days; however, 3 days may be an
adequate treatment duration for certain
patients whose condition is completely stable.40-
42
Extended courses of therapy may be indicated
for patients with immunocom- promising
conditions, infections caused by cer- tain
pathogens (e.g., P. aeruginosa), or complications
such as empyema. Serial procalcitonin
thresholds as an adjunct to clinical judgment
may help guide the discontinuation of
antibiotic therapy.17,18

HOSPITAL DISCHARGE AND FOLLOW-UP

Hospital discharge is appropriate when the
pa- tient is in a clinically stable condition, is
able to take oral medication, and has a safe
environ- ment for continued care; overnight
observation after a switch to oral therapy is
not necessary. Early discharge based on
clinical stability and criteria for the switch to
oral therapy is encour- aged to reduce
unnecessary hospital costs and risks
The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 CLINICAL PRACTICE

follow-up is encouraged to reduce pandemic and suggest selective microbiologic

the likelihood of readmission to the testing. However, we now advocate for a more
hospital.43 A follow-up chest comprehensive approach to microbiologic
radiograph is indicated in only a testing for all hospitalized patients with
minority of pa- tients, such as those community- acquired pneumonia, including
at risk for lung cancer on the basis of testing for SARS- CoV-2 infection.
age, smoking history, or persistence
of symptoms.1,44
The roleAREA ofS OF
the UNCERTAINT
lung microbiome
Y in
community- acquired pneumonia is an area of
PREVENTION
ongoing re- search.8 Further understanding of
Smoking and excessive alcohol the lung microbiome may provide information
consumption regarding inflammatory response and
should be addressed. susceptibility to spe- cific pathogens.
In addition, vaccines against Microbiologic diagnosis with the use of
influenza, Covid-19, and Strep. rapid multiplexed molecular platforms is a
pneumoniae should be administered swiftly ad- vancing technology.22 Further
according to current Advisory studies are needed to determine the clinical
Committee on Immunization Practices effect and cost–benefit ratio of these rapid
recommen- dations.45 molecular tests.
Although
ATS–IDSA guidelines recommend
The recommendations we describe monotherapy with GUIDELINES
amoxicillin as a first-line
align with the most recent ATS–IDSA op- tion for ambulatory patients at low risk, we
guidelines.1 We agree with the often add a macrolide that targets atypical
recommendation that the addition of pathogens, since such pathogens are relatively
anaero- bic coverage for suspected common and not readily identified in low-risk
aspiration pneumonia should not be patients, and treating such patients may hasten
routine practice unless there is evi- recovery.46-48
dence of a lung abscess or empyema. There is an observed association between com-
The current guidelines were
published before the Covid-19

N ENGL J MED 38e;7 NEJM.ORG AUGUST 17, 2023 639

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 T he NEW ENGL A ND JOUR NA L of MEDICINE

munity-acquired pneumonia and an increased concerned about secondary bacterial

risk of cardiovascular disease.49,50 Further infection, particularly given the elevated
stud- ies are needed to gain a better procalcitonin lev- el. In the absence of known
understanding of this relationship and to risk factors for MRSA or pseudomonas, we
develop interventions aimed at reducing would initiate treatment in the emergency
cardiovascular risk as well as the risk of other department with intravenous azithromycin
sequelae of community-acquired pneumonia. and ceftriaxone. If testing proved negative for
atypical bacteria, we would discon- tinue
CONCLUSIONS AND azithromycin therapy. We would discharge him
RECOMMENDATIONS with continued oral therapy (e.g., amoxicil- lin–
clavulanate if no bacterial pathogen was
The patient described in the vignette was identified); if his condition had reached
clini- cally stable, had a CURB-65 score of 2, clinical stability in 48 to 72 hours, he should
and had only one minor criterion for severe complete a 5-day course of the medication.
community- acquired pneumonia (i.e., Outpatient follow-up should be scheduled a
confusion); therefore, he should be admitted to week after discharge.
a general ward. Although a viral pathogen was
identified, we would be Disclosure forms provided by the authors are available with
the full text of this article at NEJM.org.

REFERENCES Amer- ican

1. Metlay JP, Waterer GW, Long AC, et Thoracic Society
10. Jain S, Self WH, Wunderink RG, et al.
al. Diagnosis and treatment of adults workshop report.
Community-acquired pneumonia requir- ing
with community-acquired pneumonia: Ann Am Thorac Soc
hospitalization among U.S. adults. N Engl J Med
an of- ficial clinical practice guideline of 2021;18:1087-97.
2015;373:415-27.
the American Thoracic Society and 11. Gadsby NJ, Musher DM. The microbi- al
Infectious Diseases Society of America. etiology of community-acquired pneu- monia in
Am J Respir Crit Care Med adults: from classical bacteriol- ogy to host
2019;200(7):e45-e67. transcriptional signatures. Clin Microbiol Rev
2. Agency for Healthcare Research and 2022;35(4):e0001522.
Quality. Community-acquired 12. Ramirez JA, Musher DM, Evans SE, et al.
pneumonia clinical decision support Treatment of community-acquired pneu- monia in
implementation toolkit. January 2018 immunocompromised adults: a consensus
(https://www.ahrq statement regarding initial strategies. Chest
.gov/hai/tools/ambulatory-care/cap 2020;158:1896-911.
-toolkit/index.html). 13. Dela Cruz CS, Wunderink RG, Chris- tiani DC,
3. McDermott KW, Roemer M. Most et al. Future research directions in pneumonia.
fre- quent principal diagnoses for Am J Respir Crit Care Med 2018;198:256-63.
inpatient stays in U.S. Hospitals, 2018. 14. Feldman C, Anderson R. Pneumonia as a
Rockville, MD: Agency for Healthcare systemic illness. Curr Opin Pulm Med 2018;24:237-
Research and Quality, July 2021 43.
(https://hcup-us.ahrq 15. Cavallazzi R, Furmanek S, Arnold FW, et al. The
.gov/reports/statbriefs/sb277-Top burden of community-acquired pneumonia
-Reasons-Hospital-Stays-2018.pdf ). requiring admission to ICU in the United States.
4. Murphy SL, Kochanek KD, Xu J, Arias Chest 2020;158:1008- 16.
E. Mortality in the United States, 2020. 16. Gilbert DN. Procalcitonin as a bio- marker in
NCHS Data Brief 2021;427:1-8. respiratory tract infection. Clin Infect Dis
5. Musher DM, Thorner AR. Community- 2011;52:Suppl 4:S346-S350.
acquired pneumonia. N Engl J Med 17. Schuetz P, Wirz Y, Sager R, et al. Pro-
2014; 371:1619-28. calcitonin to initiate or discontinue anti- biotics in
6. Ramirez JA, Wiemken TL, Peyrani P, acute respiratory tract infec- tions. Cochrane
et al. Adults hospitalized with Database Syst Rev 2017; 10:CD007498.
pneumonia in the United States: 18. Schuetz P, Beishuizen A, Broyles M, et al.
incidence, epidemi- ology, and Procalcitonin (PCT)-guided antibiotic
mortality. Clin Infect Dis 2017; 65:1806- stewardship: an international experts consensus
12. on optimized clinical use. Clin Chem Lab Med
7. Mandell LA, Niederman MS. Aspira- 2019;57:1308-18.
tion pneumonia. N Engl J Med 19. Fine MJ, Auble TE, Yealy DM, et al. A
2019;380: 651-63. prediction rule to identify low-risk pa- tients with
8. Faner R, Sibila O, Agustí A, et al. The community-acquired pneumo- nia. N Engl J Med
microbiome in respiratory medicine: 1997;336:243-50.
cur- rent challenges and future
perspectives. Eur Respir J
2017;49:1602086.
9. Dela Cruz CS, Evans SE, Restrepo MI,
et al. Understanding the host in the
man- agement of pneumonia: an official

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 T he NEW ENGL A ND JOUR NA L of MEDICINE

Quimioter 2022;35: Suppl (https://covid19treatment

20. Lim WS, van der Eerden MM, Laing R, et al. 1:15-20. guidelines.nih.gov/).
Defining community acquired pneu- monia 24. Parente DM, Cunha CB, 26. File TM Jr, Goldberg L, Das A, et al.
severity on presentation to hospi- tal: an Mylonakis E, Timbrook TT. Efficacy and safety of intravenous-to-oral
international derivation and vali- dation study. The clinical utility of meth- lefamulin, a pleuromutilin antibiotic, for
Thorax 2003;58:377-82. icillin-resistant the treatment of community-acquired
21. File TM Jr. New diagnostic tests for staphylococcus aureus bacterial pneumonia: the phase III lefam-
pneumonia: what is their role in clinical practice? (MRSA) nasal screening to ulin evaluation against pneumonia (LEAP
Clin Chest Med 2011;32:417-30. rule out MRSA pneumonia: a I) trial. Clin Infect Dis 2019;69:1856-67.
22. Hanson KE, Azar MM, Banerjee R, et al. diagnostic meta-analysis 27. Stets R, Popescu M, Gonong JR, et al.
Molecular testing for acute respiratory tract with antimicrobial Omadacycline for community-acquired
infections: clinical and diagnostic stewardship implica- tions. pneumonia. N Engl J Med
recommendations from the IDSA’s diag- nostics Clin Infect Dis 2018;67:1-7. 2019;380:517- 27.
committee. Clin Infect Dis 2020; 71:2744-51. 25. National Institutes of 28. Metersky ML, Ma A, Houck PM, Brat-
23. Burillo A, Candel FJ, Canut-Blasco A. Value of Health. Corona- virus disease zler DW. Antibiotics for bacteremic
syndromic panels in the manage- ment of severe 2019 (COVID-19) treatment pneu-
community-acquired pneumonia. Rev Esp guidelines. 2023

640 N ENGL J MED 38e;7 NEJM.ORG AUGUST 17, 2023

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 CLINICAL PRACTICE

monia: improved outcomes with macro-

lides but not fluoroquinolones. Chest 2007; pathogen directed antibiotic treatment Post-discharge services for different
131:466-73. and empirical broad spectrum antibiotic diag- nosis than index hospitalization
29. Martínez JA, Horcajada JP, Almela M, treatment in patients with community predict decreased 30-day readmissions
et al. Addition of a macrolide to a beta- ac- quired pneumonia: a prospective among Medicare beneficiaries. J Gen
lactam-based empirical antibiotic regi- ran- domised study. Thorax Intern Med 2019;34:1766-74.
men is associated with lower in-hospital 2005;60:672-8. 44. US Preventive Services Task Force,
mortality for patients with bacteremic 37. Viasus D, Simonetti AF, Garcia-Vidal Krist AH, Davidson KW, et al. Screening
pneumococcal pneumonia. Clin Infect C, Niubó J, Dorca J, Carratalà J. Impact of for lung cancer: US Preventive Services
Dis 2003;36:389-95. antibiotic de-escalation on clinical out- task force recommendation statement.
30. Restrepo MI, Mortensen EM, Waterer comes in community-acquired pneumo- JAMA 2021;325:962-70.
GW, Wunderink RG, Coalson JJ, Anzueto coccal pneumonia. J Antimicrob Chemoth- 45. Murthy N, Wodi AP, Bernstein H,
A. Impact of macrolide therapy on mortal- er 2017;72:547-53. McNally V, Cineas S, Ault K. Advisory
ity for patients with severe sepsis due to 38. Ramirez JA, Srinath L, Ahkee S, committee on immunization practices
pneumonia. Eur Respir J 2009;33:153-9. Huang A, Raff MJ. Early switch from in- recommended immunization schedule
31. Martin-Loeches I, Lisboa T, Rodri- travenous to oral cephalosporins in the for adults aged 19 years or older —
guez A, et al. Combination antibiotic treatment of hospitalized patients with United States, 2022. MMWR Morb
therapy with macrolides improves surviv- community-acquired pneumonia. Arch In- Mortal Wkly Rep 2022;71:229-33.
al in intubated patients with community- tern Med 1995;155:1273-6. 46. File TM Jr, Marrie TJ. Does empiric
acquired pneumonia. Intensive Care Med 39. Aliberti S, Zanaboni AM, Wiemken T, therapy for atypical pathogens improve
2010;36:612-20. et al. Criteria for clinical stability in hos- outcomes for patients with CAP? Infect
32. Niederman MS, Torres A. Severe pitalised patients with community-acquired Dis Clin North Am 2013;27:99-114.
com- munity-acquired pneumonia. Eur pneumonia. Eur Respir J 2013;42:742-9. 47. File TM Jr, Eckburg PB, Talbot GH,
Respir Rev 2022;31:220123. 40. Dimopoulos G, Matthaiou DK, Kara- Llorens L, Friedland HD. Macrolide
33. Sibila O, Restrepo MI, Anzueto A. georgopoulos DE, Grammatikos AP, thera- py for community-acquired
What is the best antimicrobial Athanassa Z, Falagas ME. Short- versus pneumonia due to atypical pathogens:
treatment for severe community- long-course antibacterial therapy for outcome as- sessment at an early time
acquired pneumo- nia (including the community-acquired pneumonia: a point. Int J An- timicrob Agents
role of steroids and statins and other meta- analysis. Drugs 2008;68:1841- 2017;50:247-51.
immunomodulatory agents). Infect Dis 54. 48. Shames JM, George RB, Holliday WB,
Clin North Am 2013; 27:133-47. 41. el Moussaoui R, de Borgie CAJM, van Rasch JR, Mogabgab WJ. Comparison of
34. Haessler S, Guo N, Deshpande A, et den Broek P, et al. Effectiveness of antibiotics in the treatment of
al. Etiology, treatments, and outcomes discon- tinuing antibiotic treatment mycoplas- mal pneumonia. Arch Intern
of patients with severe community- after three days versus eight days in Med 1970; 125:680-4.
acquired pneumonia in a large U.S. mild to moder- ate-severe community 49. Perry TW, Pugh MJV, Waterer GW, et
sample. Crit Care Med 2022;50:1063- acquired pneumo- nia: randomised, al. Incidence of cardiovascular events
71. double blind study. BMJ 2006;332:1355. af- ter hospital admission for
35. Dequin P-F, Meziani F, Quenot J-P, et 42. Dinh A, Ropers J, Duran C, et al. Dis- pneumonia. Am J Med 2011;124:244-
al. Hydrocortisone in severe community- continuing β-lactam treatment after 3 days 51.
acquired pneumonia. N Engl J Med for patients with community-acquired 50. Menéndez R, Méndez R, Aldás I, et al.
2023; 388:1931-41. pneumonia in non-critical care wards Community-acquired pneumonia patients
36. van der Eerden MM, Vlaspolder F, (PTC): a double-blind, randomised, at risk for early and long-term cardiovas-
de Graaff CS, et al. Comparison place- bo-controlled, non-inferiority cular events are identified by cardiac
between trial. Lan- cet 2021;397:1195-203. bio- markers. Chest 2019;156:1080-91.
43. Tak HJ, Chen LW, Wilson FA, et al. Copyright © 2023 Massachusetts Medical Society.

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.
 CLINICAL PRACTICE

N ENGL J MED 38e;7 NEJM.ORG AUGUST 17, 2023 641

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org by Kevin Chavarria on August 17, 2025.
Copyright © 2023 Massachusetts Medical Society. All rights reserved, including those for text and data mining, AI training, and similar technologies.