Annals of Internal MedicineT

In the ClinicT

Community-Acquired
Pneumonia
Prevention

C
ommunity-acquired pneumonia is an impor-
tant cause of morbidity and mortality. It can
be caused by bacteria, viruses, or fungi and
can be prevented through vaccination with pneu- Diagnosis
mococcal, influenza, and COVID-19 vaccines.
Diagnosis requires suggestive history and physical
findings in conjunction with radiographic evidence
of infiltrates. Laboratory testing can help guide Treatment
therapy. Important issues in treatment include
choosing the proper venue, timely initiation of the
appropriate antibiotic or antiviral, appropriate re- Practice Improvement
spiratory support, deescalation after negative cul-
ture results, switching to oral therapy, and short
treatment duration.

CME/MOC activity available at Annals.org.

Physician Writer doi:10.7326/AITC202204190

Michael B. Rothberg, MD, MPH
Cleveland Clinic, Cleveland,
Ohio
This article was published at Annals.org on 12 April 2022.
CME Objective: To review current evidence for prevention, diagnosis, treatment,
and practice improvement of community-acquired pneumonia.
Funding Source: American College of Physicians.
Acknowledgment: The author thanks Michael S. Niederman, MD, author of the
previous version of this In the Clinic.
Disclosures: Dr. Rothberg, ACP Contributing Author, reports grants or contracts
from the Agency for Healthcare Research and Quality and the National Institute
on Aging, consulting fees from Health Advances, participation on a data safety
monitoring board or advisory board for BMS, and stock in Moderna. All relevant
financial relationships have been mitigated. Disclosures can also be viewed at
www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M21-4473.

With the assistance of additional physician writers, the editors of Annals of

Internal Medicine develop In the Clinic using MKSAP and other resources of
the American College of Physicians. The patient information page was written by
Monica Lizarraga from the Patient and Interprofessional Partnership Initiative at
the American College of Physicians.
In the Clinic does not necessarily represent official ACP clinical policy. For ACP
clinical guidelines, please go to https://www.acponline.org/clinical_information/
guidelines/.
© 2022 American College of Physicians

1. Heron M. Deaths: Leading
Causes for 2019. National
Vital Statistics Reports.
National Center for Health
Statistics; 2021.
2. QuickStats: death rates
from influenza and
pneumonia among
persons aged ≥65 years,
by sex and age group—
National Vital Statistics
System, United States,
2018. MMWR Morb Mortal
Wkly Rep. 2020;69:1470.
[PMID: 33031359]
3. Nuorti JP, Butler JC, Farley
MM, et al. Cigarette smok-
ing and invasive pneumo-
coccal disease. Active
Bacterial Core Surveillance
Team. N Engl J Med.
2000;342:681-9. [PMID:
10706897]
4. Gupta NM, Lindenauer PK,
Yu PC, et al. Association
between alcohol use disor-
ders and outcomes of
patients hospitalized with
community-acquired pneu-
monia. JAMA Netw Open.
2019;2:e195172. [PMID:
31173120]
5. Patel MS, Patel SB,
Steinberg MB. Smoking
cessation. Ann Intern Med.
2021;174:ITC177-ITC192.
[PMID: 34904907]
6. Kobayashi M, Farrar JL,
Gierke R, et al. Use of 15-
valent pneumococcal conju-
gate vaccine and 20-valent
pneumococcal conjugate
vaccine among U.S. adults:
updated recommendations
of the Advisory Committee
on Immunization Practices
—United States, 2022.
MMWR Morb Mortal Wkly
Rep. 2022;71:109-117.
[PMID: 35085226]
7. Moberley S, Holden J,
Tatham DP, et al. Vaccines
for preventing pneumococ-
cal infection in adults.
Cochrane Database Syst
Rev. 2013:CD000422.
[PMID: 23440780]
© 2022 American College of Physicians
Community-acquired pneumonia (CAP)
can vary from a mild outpatient illness
to a more severe disease requiring hos-
pital admission or intensive care. In the
United States, CAP together with influ-
enza is the ninth leading cause of death
overall and the leading cause of death
from infectious disease (1). Outpatient
CAP is managed mainly by primary care
physicians, whereas inpatient treatment
often involves hospitalists. The key man-
agement decisions related to CAP are
Prevention
Who is at increased risk for CAP?
Persons with comorbid illness and el-
derly persons are at increased risk for
pneumonia and for having a more
complex course. The death rate from
pneumonia per 100 000 persons in-
creases rapidly with age, from 31.7
among adults aged 65 to 74 years to
377.6 among those aged 85 years or
older (2). Existing respiratory disease,
cardiovascular disease, diabetes melli-
tus, chronic liver disease, immunosup-
pression, chronic kidney disease, and
previous splenectomy are associated
with increased incidence of CAP.
Cigarette smoking and alcohol misuse
predispose to severe CAP and to bac-
teremic pneumococcal infection (3, 4).
Patients should be screened for these
conditions, should be advised to quit,
and should be offered counseling and
pharmacotherapy (5). Other common
comorbid conditions include cancer
and neurologic illnesses that predis-
pose to aspiration, such as seizures and
dementia.
Who should receive pneumococcal
vaccination, and when?
Pneumococcal vaccination is recom-
mended for all persons at increased
risk for pneumococcal infection (Table
1). For persons without high-risk condi-
tions, vaccination should be given at
age 65 years. Patients with risk factors
ITC2 In the Clinic
recognition and treatment in a timely
and effective manner, defining the app-
ropriate site of care (home, hospital, or
intensive care unit [ICU]), choosing
effective treatment based on the cause,
limiting antibiotic duration, and preven-
tion. In particular, persons who have
recently been hospitalized and received
parenteral antibiotics and those who
have previously been infected with multi-
drug-resistant organisms (MDROs) are at
increased risk for resistant infections.
should be vaccinated when the risk is
first identified.
New recommendations from the Centers
for Disease Control and Prevention
(CDC) vastly simplify the approach to vac-
cination (6). The 20-valent conjugate vac-
cine (PCV20) can be given in a single
dose to all patients with risk factors or
those aged 65 years or older. If PCV15
is administered instead, it should be fol-
lowed by the 23-valent polysaccharide
vaccine (PPSV23) 6 to 12 months later,
or as soon as 8 weeks later for persons
with cochlear implants, cerebrospinal
fluid leaks, or immunosuppressive con-
ditions. Persons who have previously
received only PPSV23 should receive a
dose of PCV15 or PCV20 at least 1 year
later.
A 2013 Cochrane review of pneumo-
coccal vaccination found that it reduces
the frequency of invasive pneumonia in
healthy, immunocompetent adults (7).
There was no evidence of effectiveness
in adults with chronic illness, but there
were very few events and CIs were
wide.
In a subsequent double-blind random-
ized controlled trial (RCT) of 84 496
adults older than 65 years, PCV13 pre-
vented both bacteremic and nonbac-
teremic vaccine strain–specific pneu-
mococcal pneumonia and vaccine-type
invasive pneumococcal pneumonia;
Annals of Internal Medicine
 Table 1. Persons Eligible for Pneumococcal Vaccination
Risk Group Characteristic or Underlying Medical Condition
Older persons Age ≥65 y
Immunocompetent persons Chronic heart disease*
Chronic lung disease†
Diabetes mellitus
Cerebrospinal fluid leak
Cochlear implant
Alcoholism
Chronic liver disease, cirrhosis
Cigarette smoking
Persons with functional or anatomic asplenia Sickle cell disease/other hemoglobinopathy
Congenital or acquired asplenia
Immunocompromised persons Congenital or acquired immunodeficiency‡
HIV infection
Chronic renal failure
Nephrotic syndrome
Leukemia
Lymphoma
Hodgkin disease
Generalized cancer
Iatrogenic immunosuppression§
Solid organ transplant
Multiple myeloma

* Includes congestive heart failure and cardiomyopathies but excludes hypertension.

† Includes chronic obstructive pulmonary disease, emphysema, and asthma.
‡ Includes B-lymphocyte (humoral) or T-lymphocyte deficiency, complement deficiencies (particularly
C1, C2, C3, and C4 deficiencies), and phagocytic disorders (excluding chronic granulomatous disease).
§ Diseases requiring treatment with immunosuppressive drugs, including long-term systemic cortico-
steroids and radiation therapy.

however, it did not reduce all-cause co-infection (10). By limiting the spread
CAP or mortality (8). Over 4 years, it pre- of influenza, vaccination can reduce the
vented 41 cases, for a number need- incidence of CAP and protect vulnera-
ed to treat of approximately 1000. ble persons (11). The Advisory Com-
mittee on Immunization Practices (ACIP)
In an RCT of 1006 nursing home resi- 8. Bonten MJ, Huijts SM,
recommends yearly influenza vaccine Bolkenbaas M, et al.
dents in Japan, PPSV23 significantly Polysaccharide conjugate
for all persons older than 6 months. vaccine against pneumo-
reduced the frequency and mortality of coccal pneumonia in
Although patients with weakened imm-
pneumococcal pneumonia, as well as adults. N Engl J Med.
une systems, older patients, and those 2015;372:1114-25.
the frequency—but not the mortality—of [PMID: 25785969]
with chronic medical illnesses have the 9. Maruyama T, Taguchi O,
all-cause pneumonia (9). Niederman MS, et al.
highest risk for hospitalization and Efficacy of 23-valent pneu-
What is the role of influenza death, even healthy adults can be hospi- mococcal vaccine in pre-
venting pneumonia and
vaccination in the prevention talized with influenza and can spread it improving survival in nurs-
ing home residents: dou-
of CAP and its complications? to others. Details about specific influ- ble blind, randomised and
Influenza is a common cause of CAP enza vaccines were reviewed in a previ- placebo controlled trial.
BMJ. 2010;340:c1004.
and may be complicated by bacterial ous In the Clinic article (12). [PMID: 20211953]
10. Bartley PS, Deshpande A,
Yu PC, et al. Bacterial
coinfection in influenza
pneumonia: rates, patho-
Prevention... Physicians should screen for tobacco use at every visit; smokers should gens, and outcomes.
be advised to quit, be referred for counseling, and be offered pharmacologic treatment. Infect Control Hosp
Persons at risk for CAP and its complications should be offered pneumococcal and influ- Epidemiol. 2021:1-6.
[PMID: 33890558]
enza vaccination, and all patients should be encouraged to get COVID-19 vaccination. 11. Taksler GB, Rothberg MB,
Cutler DM. Association of
influenza vaccination cov-
erage in younger adults
CLINICAL BOTTOM LINE with influenza-related ill-
ness in the elderly. Clin
Infect Dis. 2015;61:1495-
503. [PMID: 26359478]
12. Uyeki TM. Influenza. Ann
Intern Med. 2021;174:
ITC161-ITC176. [PMID:
34748378]

Annals of Internal Medicine In the Clinic ITC3 © 2022 American College of Physicians

13. Rowe TA, Jump RLP,
Andersen BM, et al.
Reliability of nonlocalizing
signs and symptoms as
indicators of the presence
of infection in nursing-
home residents. Infect
Control Hosp Epidemiol.
2020:1-10. [PMID:
33292915]
14. Metlay JP, Kapoor WN,
Fine MJ. Does this patient
have community-acquired
pneumonia? Diagnosing
pneumonia by history
and physical examination.
JAMA. 1997;278:1440-5.
[PMID: 9356004]
15. Wipf JE, Lipsky BA,
Hirschmann JV, et al.
Diagnosing pneumonia
by physical examination:
relevant or relic. Arch
Intern Med.
1999;159:1082-7. [PMID:
10335685]
16. Metlay JP, Waterer GW,
Long AC, et al. Diagnosis
and treatment of adults
with community-acquired
pneumonia. An official
clinical practice guideline
of the American Thoracic
Society and Infectious
Diseases Society of
America. Am J Respir Crit
Care Med. 2019;200:e45-
e67. [PMID: 31573350]
17. Graffelman AW, le Cessie
S , Knuistingh Neven A ,
et al. Can history and
exam alone reliably pre-
dict pneumonia. J Fam
Pract. 2007;56:465-70.
[PMID: 17543257]
18. Hersh AL, King LM,
Shapiro DJ, et al.
Unnecessary antibiotic
prescribing in US ambula-
tory care settings, 2010-
2015. Clin Infect Dis.
2021;72:133-7. [PMID:
32484505]
19. Hopstaken RM, Witbraad
T, van Engelshoven JM,
et al. Inter-observer varia-
tion in the interpretation
of chest radiographs for
pneumonia in commu-
nity-acquired lower respi-
ratory tract infections. Clin
Radiol. 2004;59:743-52.
[PMID: 15262550]
20. Claessens YE, Debray MP,
Tubach F, et al. Early chest
computed tomography
scan to assist diagnosis
and guide treatment deci-
sion for suspected com-
munity-acquired pneumo-
nia. Am J Respir Crit Care
Med. 2015;192:974-82.
[PMID: 26168322]
© 2022 American College of Physicians
Diagnosis
CAP accounts for only about 5% of re-
spiratory complaints in the ambulatory
setting. History and physical examina-
tion can help suggest the presence of
pneumonia; predict the cause, which
dictates treatment; and define severity,
which determines the site of care.
How is the diagnosis made?
CAP usually presents with both respira-
tory and systemic symptoms, particu-
larly in young and immunocompetent
persons. It should be suspected when
the patient has cough, purulent spu-
tum, pleuritic chest pain, dyspnea,
fever, and chills. Among older patients
and those with chronic illness, the dis-
ease may go unrecognized because
fever may be absent or the patient may
have nonrespiratory symptoms, such as
confusion, weakness, lethargy, falling,
poor oral intake, or decompensation of
a chronic illness (for example, conges-
tive heart failure [CHF]) (13). Most
patients present acutely with symptoms
for 1 to 2 days, but symptoms may be
present for longer in older persons.
Physical findings suggestive of pneu-
monia include fever or hypothermia,
tachypnea, hypoxia, and rales or bron-
chial breath sounds on auscultation.
Unfortunately, no single symptom or
finding is sufficiently sensitive or specific
to diagnose or rule out pneumonia, and
prediction rules that combine several
findings have failed to outperform phy-
sician judgment (14). Moreover, physi-
cians' agreement on findings is often
poor, and the prognostic value of the
finding varies from one physician to the
next (15). Thus, clinical diagnosis of
pneumonia is often inaccurate, with
overall sensitivity ranging from 70% to
90% and specificity ranging from 40%
to 70% (16, 17). A chest radiograph
(CXR) can confirm the diagnosis and
identify certain complications.
When signs and symptoms from history
(cough, fever, dyspnea, pleuritic pain)
and physical examination (focal crackles,
ITC4 In the Clinic
temperature ≥38
C) were used to pre-
dict the presence of radiographic pneu-
monia in a study of 129 patients with
lower respiratory tract infection (26 with
pneumonia), no combination of signs
and symptoms was highly accurate. The
positive predictive value of each varied
from 17% to 43% (17).
When should clinicians obtain a CXR?
In general, when CAP is suspected, a
CXR should be obtained. However, for
otherwise healthy outpatients with clini-
cal features strongly suggestive of
CAP, it is reasonable to forego CXR,
bearing in mind that prescribing antibi-
otics without a CXR may contribute to
overuse, which is already a major issue
in the United States (18). It is important
to maintain a strong index of suspicion
because elderly and immunosup-
pressed patients can have radiographic
evidence of pneumonia without clinical
features.
It is especially important to obtain a
CXR if the diagnosis is questionable or
if pleural effusion, lung abscess, necrot-
izing pneumonia, or multilobar illness
is suspected. Radiographs can aid
patient management if findings of
severe illness are present (bilateral,
multilobar, or rapidly expanding infil-
trates), but patterns rarely suggest a
specific cause (for example, tuberculo-
sis, Pneumocystis jirovecii). Radiologists
frequently disagree on the presence of
an infiltrate; thus, a negative study is
more likely to rule out pneumonia than
a positive one is to rule it in (19).
However, if the patient has a convinc-
ing history and focal physical findings,
pneumonia may be present even in the
absence of a radiographic infiltrate. In
this setting, computed tomography
(CT) scans may be helpful. One single-
center study found that routine CT
scanning affected treatment in about
25% of cases (20). Another prospective
study of older patients with suspected
pneumonia found that 30% had their
Annals of Internal Medicine
 Table 2. Organisms That Commonly Cause Community-Acquired Pneumonia*
Organisms Frequency, %†
Bacteria 14.0
Streptococcus pneumoniae 5.1
Staphylococcus aureus 1.6
Pseudomonas <1
Escherichia coli 1.4
Klebsiella <1
Mycoplasma pneumoniae 1.9
Chlamydophila pneumoniae <1
Legionella pneumophila 1.4
Haemophilus influenzae <1
Other Streptococcus species <1
Mycobacterium tuberculosis <1

Viruses 27.0
Human rhinovirus 8.6
Influenza A or B 5.8
Adenovirus 1.4
Human metapneumovirus 3.9
Respiratory syncytial virus 3.0
Parainfluenza virus 3.0
Coronavirus 2.3

Fungi 1.0
Histoplasma <1
Coccidioides <1

* Adapted from reference 22.

† No discernible cause is found in 62% of patients.

treatment downgraded, and another viral–bacterial co-infection. The most com-

15% had it upgraded (21). The eco-
nomic implications of routine CT scans
for pneumonia diagnosis have not
been established. However, because of
the high cost and radiation exposure, it
should be reserved for cases in which
the diagnosis remains in doubt after
CXR.
How is the cause determined?
CAP can be caused by viruses, bacte-
ria, or fungi (Table 2).
A prospective population-based sur-
veillance study conducted at 3 hospitals
in Chicago, Illinois, and 2 in Nashville,
Tennessee, before the COVID-19 pan-
demic found that despite comprehen-
sive diagnostic testing, a causative
organism was isolated in only 38% of
cases (22). Viruses were present in 27%
of patients, bacteria in 14%, and fungi
in 1%. Three percent of patients had
monly identified pathogens were human
rhinovirus, influenza, and Streptococcus
pneumoniae (pneumococcus); other
common bacterial pathogens included
Mycoplasma pneumoniae, Staphylo-
coccus aureus, Legionella, and Entero-
bacteriaceae. Control patients had few
or none of these pathogens.
History and physical examination can-
not determine the cause of pneumonia,
which requires laboratory testing. How-
ever, the history can identify risk factors 21. Prendki V, Scheffler M,
Huttner B, et al. Low-dose
for MDROs, such as hospitalization with computed tomography
for the diagnosis of pneu-
intravenous antibiotic therapy in the monia in elderly patients:
previous 90 days. This is important a prospective, interven-
tional cohort study. Eur
because treatment is usually empirical. Respir J. 2018;51. [PMID:
29650558]
History can also identify risk factors for 22. Jain S, Self WH,
less common causes of pneumonia, Wunderink RG, et al; CDC
EPIC Study Team.
such as exposure to birds (Chlamydia Community-acquired
pneumonia requiring hos-
psittaci, Cryptococcus neoformans) or pitalization among U.S.
bats (Histoplasma capsulatum) or travel adults. N Engl J Med.
2015;373:415-27.
to the southwestern United States [PMID: 26172429]

Annals of Internal Medicine In the Clinic ITC5 © 2022 American College of Physicians

23. Klompas M, Imrey PB, Yu
PC, et al. Respiratory viral
testing and antibacterial
treatment in patients hos-
pitalized with community-
acquired pneumonia.
Infect Control Hosp
Epidemiol. 2021;42:817-
25. [PMID: 33256870]
24. Schimmel JJ, Haessler S,
Imrey P, et al.
Pneumococcal urinary
antigen testing in United
States hospitals: a missed
opportunity for antimicro-
bial stewardship. Clin
Infect Dis. 2020;71:1427-
34. [PMID: 31587039]
25. Self WH, Balk RA, Grijalva
CG, et al. Procalcitonin as
a marker of etiology in
adults hospitalized with
community-acquired
pneumonia. Clin Infect
Dis. 2017;65:183-90.
[PMID: 28407054]
26. Upadhyay S, Niederman
MS. Biomarkers: what is
their benefit in the identi-
fication of infection, sever-
ity assessment, and
management of commu-
nity-acquired pneumonia.
Infect Dis Clin North Am.
2013;27:19-31. [PMID:
23398863]
27. Christ-Crain M, Stolz D,
Bingisser R, et al.
Procalcitonin guidance of
antibiotic therapy in com-
munity-acquired pneumo-
nia: a randomized trial.
Am J Respir Crit Care
Med. 2006;174:84-93.
[PMID: 16603606]
28. Vaughn VM, Flanders SA,
Snyder A, et al. Excess an-
tibiotic treatment duration
and adverse events in
patients hospitalized with
pneumonia: a multihospi-
tal cohort study. Ann
Intern Med.
2019;171:153-63. [PMID:
31284301]
© 2022 American College of Physicians
(endemic fungi, such as coccidioido-
mycosis). This information is useful
if patients do not respond to usual
therapy.
What is the role of laboratory tests?
For outpatients, no tests are needed
beyond pulse oximetry and a rapid
influenza test during influenza season
(12). For inpatients, additional testing
may be required to define disease se-
verity and identify the cause. Clinicians
should measure pulse oximetry in all
patients and arterial blood gases if car-
bon dioxide retention is suspected.
Even with extensive diagnostic testing,
a specific cause is found in fewer than
half of patients (22). Blood culture
results are positive in only about 10%
of patients with CAP, and in low-risk
patients (hospitalized without severe ill-
ness), the incidence of false-positive
results may exceed the incidence of
true-positive results, leading to poten-
tial overtreatment (16). Thus, blood cul-
tures should not be ordered routinely.
In patients with severe pneumonia and
patients suspected of being infected
with an MDRO or an unusual pathogen,
2 sets of blood cultures should be col-
lected before therapy is started, and
sputum should be collected if a good-
quality sample can be obtained. Urine
should be tested for Legionella and
pneumococcal antigens.
Serologic tests for viruses and atypical
pathogens are not useful because they
require convalescent titers in 6 to 8
weeks to identify infection. Rapid diag-
nostic tests for viral pathogens, such as
rapid antigen testing, direct fluorescent
antibody, or polymerase chain reaction
(PCR), are available and should be con-
sidered, especially during local out-
breaks (for example, during influenza
season or a COVID-19 outbreak). The
role of these tests in managing patients
with CAP and in guiding antibiotic
selection is not yet established, al-
though some studies have shown that
they may reduce unnecessary antibiotic
use and increase antiviral prescribing,
especially when results are positive for
influenza (23). Similarly, testing for
ITC6 In the Clinic
COVID-19 can inform isolation and
treatment. Other rapid diagnostic test-
ing, such as streptococcal and Legio-
nella urinary antigen testing, can re-
duce the use of overly broad empirical
antibiotic coverage and promote anti-
microbial stewardship (24).
Measurement of serum procalcitonin
has been proposed to differentiate
between bacterial and viral infections.
Levels are usually elevated with bacte-
rial and Legionella infection, but not
always with other atypical pathogens
and not with viral infections (25). In pro-
spective studies, however, the sensitiv-
ity is too low to reliably identify patients
who do not require antibiotics, so pro-
calcitonin is not recommended for this
purpose. Alternatively, among hospital-
ized patients with an initially elevated
procalcitonin level, serial measure-
ments may help determine when to
stop antibiotic therapy (26).
A randomized trial of 302 patients hos-
pitalized with CAP compared those
managed with usual care versus those
managed with an algorithm recom-
mending antibiotics and the duration of
therapy. The algorithm was based on
serial measurement of procalcitonin
levels on admission and after 6 to 24
hours, 4 days, 6 days, and 8 days. The
procalcitonin-guided group had signifi-
cantly less antibiotic use, and the dura-
tion of therapy was reduced from 12 to
5 days with similar clinical success (27).
What other disorders should
clinicians consider in patients
suspected of having CAP?
Most CAP responds to empirical antibi-
otics within 48 to 72 hours (28). If the
patient does not respond within this
window, clinicians should consider the
possibility of a resistant bacterium; a
virus; or unusual bacterial pathogens,
such as Mycobacterium tuberculosis
(which may be masked by a partial
response to empirical quinolone therapy
for CAP) or endemic fungi (histoplasmo-
sis, coccidioidomycosis, blastomycosis).
Clinicians should also consider noninfec-
tious possibilities, such as bronchiolitis
Annals of Internal Medicine
obliterans with organizing pneumonia,
pulmonary vasculitis, hypersensitivity
pneumonitis, interstitial diseases, lung
cancer, lymphangitic carcinoma, bron-
choalveolar cell carcinoma, lymphoma,
or CHF. If the patient's condition
deteriorates after an initial response to
therapy, clinicians should consider pul-
monary embolus; antibiotic-induced co-
litis; and the pneumonia complications
of empyema, meningitis, and endo-
carditis.

Diagnosis... History is valuable for defining risk factors for specific pathogens, and
physical findings help define disease severity. Clinical findings are less dramatic in el-
derly persons. Clinicians should confirm the diagnosis of CAP with a CXR, or a CT scan if
the CXR is negative and suspicion is high. Laboratory testing can also help to define se-
verity and identify complications. Diagnosing specific pathogens early can help to guide
antiviral therapy and empirical antibiotic selection. If the patient does not respond to ini-
tial therapy, a specialist should be consulted.

CLINICAL BOTTOM LINE

What is the overall approach to
treatment, and how should clinicians
determine whether a patient with
CAP requires outpatient, inpatient,
or ICU care?
Treatment decisions are based on se-
verity of illness. The first step is deter-
mining the appropriate site of care—
outpatient, hospital, or ICU—and the
second is choosing the appropriate
treatment. Hospitalization decisions
can be facilitated with the Pneumonia
Severity Index (PSI) or the British
Thoracic Society (BTS) rule (Figures 1
and 2), which predict risk for death.
Patients at high risk are generally man-
aged in the hospital. The PSI stratifies
patients into 5 categories based on
age, comorbid illness, physical exami-
nation findings, and laboratory data. In
general, patients in classes I and II are
treated as outpatients, those in class III
require careful clinical assessment to
determine the site of care, and those in
classes IV and V are admitted to the
hospital. The BTS rule, or “CURB-65,”
evaluates patients for the presence of
Confusion, blood Urea nitrogen level
above 7.0 mmol/L (>19.6 mg/dL),
Respiratory rate of 30 breaths/min or
higher, systolic Blood pressure below
Treatment
90 mm Hg or diastolic blood pressure
of 60 mm Hg or lower, and age 65
years or older. Patients meeting at least
2 of these criteria are usually hospital-
ized (16, 29, 30). For select patients, a
hospital-at-home program can substi-
tute for inpatient care, reduce costs,
and prevent readmissions (31).
A prospective study of 3181 patients
seen in 32 emergency departments
compared the PSI with the CURB-65 cri-
teria and found that both approaches
accurately identified low-risk patients.
CURB-65 was better for predicting mor-
tality in high-risk patients (29). In another
prospective study of 1651 patients,
measurement of serum procalcitonin 29. Aujesky D, Auble TE, Yealy
DM, et al. Prospective
levels supplemented data obtained by comparison of three vali-
dated prediction rules for
prognostic scoring, and patients who prognosis in community-
had low procalcitonin levels had low acquired pneumonia. Am
J Med. 2005;118:384-
mortality, regardless of PSI class or 92. [PMID: 15808136]
30. Huang DT, Weissfeld LA,
CURB-65 score (30). Kellum JA, et al; GenIMS
Investigators. Risk predic-
The PSI and CURB-65 were not devel-
tion with procalcitonin
and clinical rules in com-
oped to predict need for ICU care. munity-acquired pneumo-
nia. Ann Emerg Med.
Current guidelines recommend ICU care 2008;52:48-58.e2.
[PMID: 18342993]
if the patient requires assisted ventila- 31. Levine DM, Ouchi K,
Blanchfield B, et al.
tion, has septic shock requiring vaso- Hospital-level care at
pressors, or has at least 3 of the home for acutely ill
adults: a randomized con-
following: respiratory rate of 30 breaths/ trolled trial. Ann Intern
Med. 2020;172:77-85.
min or higher, PaO2–FiO2 ratio of 250 or [PMID: 31842232]

Annals of Internal Medicine In the Clinic ITC7 © 2022 American College of Physicians
 Figure 1. Pneumonia Severity Index for evaluating severity of illness in community-acquired
pneumonia.

Factor Points
Demographic Factors
Age 1 per year
Female sex -10
Nursing home resident +10
Comorbid Illnesses
Neoplastic disease +30
Liver disease +20
Congestive heart failure +10
Cerebrovascular disease +10
Renal disease +10
Physical Examination Findings
Altered mental status +20
Respiratory rate ≥30 breaths/min +20
Systolic blood pressure <90 mm Hg +20
Temperature <95 °F or ≥104 °F +15
Pulse ≥125 beats/min +10
Laboratory and Radiographic Findings
Arterial pH <7.35 +30
BUN level ≥30 mg/dL +20
Sodium level <130 mEq/L +20
Glucose level ≥250 mg/dL +10
Hematocrit <30% +10
PaO2 <60 mm Hg or oxygen saturation <90% +10
Pleural effusion +10

Interpretation
Score Risk Class Mortality Recommendation
<51 I 0.1% Outpatient
51–70 II 0.6% Outpatient
71–90 III 0.9% Careful assessment
91–130 IV 9.3% Inpatient
>130 V 27.0% Inpatient

BUN = blood urea nitrogen.

lower, multilobar infiltrates, confusion or Which antibiotics should be

disorientation, blood urea nitrogen level prescribed for outpatients?
of 7.1 mmol/L (20 mg/dL) or higher, leu- For patients without cardiopulmonary
kocyte count below 4  109 cells/L, pla- disease or factors that increase risk for
telet count below 100  109 cells/L, infection with MDROs, the American
temperature below 36
C, and hypoten- Thoracic Society/Infectious Diseases
sion requiring aggressive fluid resuscita- Society of America (ATS/IDSA) guide-
tion (16). lines recommend high-dose amoxicil-
lin, doxycycline, or a macrolide (azi-
Figure 2. CURB-65 score for evaluating severity thromycin, clarithromycin, or erythro-
of illness in community-acquired pneumonia.
mycin), with macrolides appropriate
Factor Points only in areas where fewer than 25% of
Confusion 1 pneumococcal isolates are resistant to
Blood Urea nitrogen level >19 mg/dL 1
Respiratory rate ≥30 breaths/min 1 macrolides (Table 3; Appendix Table,
Systolic Blood pressure <90 mm Hg or 1 available at Annals.org). In most of the
diastolic Blood pressure ≤60 mm Hg
32. Gupta V, Yu KC, Schranz Age ≥65 y 1 United States, macrolide resistance
J, et al. A multicenter
Total _____
evaluation of the US prev- now exceeds 30% (32). For outpatients
alence and regional varia-
tion in macrolide-resistant
Interpretation
with cardiopulmonary disease or fac-
S. pneumoniae in ambu-
latory and hospitalized Score Mortality Recommendation tors that increase risk for infection with
adult patients in the 0 or 1 1.5% Outpatient
United States. Open 2 9.2% Inpatient drug-resistant S pneumoniae (DRSP) or
Forum Infect Dis. 2021;8: ≥3 22% Inpatient enteric gram-negative bacteria, treat-
ofab063. [PMID:
34250183] ment should include a respiratory

© 2022 American College of Physicians ITC8 In the Clinic Annals of Internal Medicine
 Table 3. Initial Treatment Strategies for Inpatients With CAP, by Level of Severity and Risk for Drug Resistance
Level of Severity Standard Regimen Prior Respiratory Isolation Prior Respiratory Isolation Recent Hospitalization and Recent Hospitalization and
of MRSA of Pseudomonas Parenteral Antibiotics and Parenteral Antibiotics and
aeruginosa Locally Validated Risk Locally Validated Risk
Factors for MRSA Factors for P aeruginosa

Nonsevere inpatient b-Lactam þ macrolide† or Add MRSA coverage§ and Add coverage for Obtain cultures but with- Obtain cultures but
pneumonia* respiratory obtain cultures/nasal PCR P aeruginosa|| and obtain hold MRSA coverage initiate coverage for
fluoroquinolone‡ to allow deescalation or cultures to allow deescala- unless culture results are P aeruginosa only if
confirmation of need for tion or confirmation of positive culture results are positive
continued therapy need for continued If rapid nasal PCR is avail-
therapy able, add coverage if PCR
result is positive and
obtain cultures
Severe inpatient b-Lactam þ macrolide† or Add MRSA coverage§ and Add coverage for Add MRSA coverage§ and Add coverage for
pneumonia* b-lactam þ fluoroquino- obtain cultures/nasal PCR P aeruginosa|| and obtain nasal PCR and cul- P aeruginosa|| and obtain
lone‡ to allow deescalation or obtain cultures to allow tures to allow deescalation cultures to allow deescala-
confirmation of need for deescalation or confirma- or confirmation of need tion or confirmation of
continued therapy tion of need for continued for continued therapy need for continued
therapy therapy

ATS = American Thoracic Society; CAP = community-acquired pneumonia; HAP = hospital-acquired pneumonia; IDSA = Infectious
Diseases Society of America; MRSA = methicillin-resistant Staphylococcus aureus; PCR = polymerase chain reaction; VAP = ventilator-
associated pneumonia.
* As defined by the 2007 ATS/IDSA CAP severity criteria guidelines.
† Ampicillin þ sulbactam, 1.5 to 3 g every 6 hours; cefotaxime, 1 to 2 g every 8 hours; ceftriaxone, 1 to 2 g/d; or ceftaroline, 600 mg
every 12 hours, and azithromycin, 500 mg/d, or clarithromycin, 500 mg twice daily.
‡ Levofloxacin, 750 mg/d, or moxifloxacin, 400 mg/d.
§ Per the 2016 ATS/IDSA HAP/VAP guidelines: vancomycin (15 mg/kg every 12 hours, adjust on basis of levels) or linezolid (600 mg
every 12 hours).
|| Per the 2016 ATS/IDSA HAP/VAP guidelines: piperacillin–tazobactam (4.5 g every 6 hours), cefepime (2 g every 8 hours), ceftazi-
dime (2 g every 8 hours), imipenem (500 mg every 6 hours), meropenem (1 g every 8 hours), or aztreonam (2 g every 8 hours). Does
not include coverage for extended-spectrum b-lactamase–producing Enterobacteriaceae, which should be considered only on the
basis of patient or local microbiological data.

fluoroquinolone (levofloxacin or moxi- How should clinicians follow patients

floxacin) or a combination of a b-lactam
(amoxicillin–clavulanate, cefpodoxime,
or cefuroxime) with a macrolide or doxy-
cycline. If the patient has received an
antibiotic in the previous 3 months, anti-
biotics of the same class should be
avoided.
How long should outpatients continue
antibiotic treatment?
In general, it is important to keep anti-
during outpatient treatment?
Up to 10% of patients initially managed
at home do not respond to therapy
and require hospitalization (34). The
evidence base for home therapy is lim-
ited, but prudence dictates the need
for careful follow-up. Patients should
be told to measure their temperature
orally every 8 hours and to report if it 33. Dinh A, Ropers J, Duran
C, et al; Pneumonia Short
exceeds 38.3
C (101
F) or does not Treatment (PTC) Study

biotic courses as short as possible to fall below 37.2
C (99
F) after 48 hours. Group. Discontinuing
b-lactam treatment after

avoid adverse drug events and antibi- Patients should be encouraged to drink 3 days for patients with
community-acquired
otic resistance. The duration of therapy 1 to 2 quarts of liquid daily and report if pneumonia in non-critical
care wards (PTC): a dou-
should be based on the patient's clini- they cannot achieve this goal. Clini- ble-blind, randomised,
placebo-controlled, non-
cal response, severity of illness, and cians should instruct patients to report inferiority trial. Lancet.
probable pathogen. Outpatients with chest pain, severe or increasing short- 2021;397:1195-203.
[PMID: 33773631]
mild to moderate CAP can be treated ness of breath, or lethargy. 34. Tillotson G, Lodise T,
Classi P, et al. Antibiotic
for as few as 3 to 5 days if clinical treatment failure and
response is good, there has been no A follow-up visit, either in person or vir- associated outcomes
among adult patients
fever for 48 to 72 hours, and there are tually, should be arranged in 24 to 48 with community-acquired
pneumonia in the outpa-
no signs of extrapulmonary infection hours to confirm the response to ther- tient setting: a real-world
US insurance claims data-
(33). Persistent cough and sputum pro- apy. If the response is satisfactory, the base study. Open Forum
duction are not reasons to prolong an- patient should return for an examina- Infect Dis. 2020;7:
ofaa065. [PMID:
tibiotic therapy. tion in 10 to 14 days. At that time, 32195289]

Annals of Internal Medicine In the Clinic ITC9 © 2022 American College of Physicians

35. Centers for Medicare &
Medicaid Services.
Hospital Quality Initiative
Overview. July 2008.
36. Meehan TP, Fine MJ,
Krumholz HM, et al.
Quality of care, process,
and outcomes in elderly
patients with pneumonia.
JAMA. 1997;278:2080-4.
[PMID: 9403422]
37. Kanwar M, Brar N, Khatib
R, et al. Misdiagnosis of
community-acquired
pneumonia and inappro-
priate utilization of antibi-
otics: side effects of the 4-
h antibiotic administration
rule. Chest.
2007;131:1865-9. [PMID:
17400668]
38. Horita N, Otsuka T,
Haranaga S, et al. Beta-
lactam plus macrolides or
beta-lactam alone for
community-acquired
pneumonia: a systematic
review and meta-analysis.
Respirology.
2016;21:1193-200.
[PMID: 27338144]
39. Martínez JA, Horcajada
JP, Almela M, et al.
Addition of a macrolide to
a beta-lactam-based em-
pirical antibiotic regimen
is associated with lower
in-hospital mortality for
patients with bacteremic
pneumococcal pneumo-
nia. Clin Infect Dis.
2003;36:389-95. [PMID:
12567294]
40. Lujan M, Gallego M,
Fontanals D, et al.
Prospective observational
study of bacteremic pneu-
mococcal pneumonia:
effect of discordant ther-
apy on mortality. Crit Care
Med. 2004;32:625-31.
[PMID: 15090938]
41. Yu VL, Chiou CC, Feldman
C, et al; International
Pneumococcal Study
Group. An international
prospective study of pneu-
mococcal bacteremia: cor-
relation with in vitro
resistance, antibiotics
administered, and clinical
outcome. Clin Infect Dis.
2003;37:230-7. [PMID:
12856216]
© 2022 American College of Physicians
pneumococcal and influenza vaccina-
tions should be administered if they
have not been previously. For patients
whose symptoms resolve within 5 to 7
days, a follow-up CXR is not necessary
(16). For smokers concerned about
underlying lung cancer, routine screen-
ing with low-dose CT is appropriate. If
the pneumonia is not resolving, addi-
tional imaging, laboratory testing, and
microbiological work-up are indicated.
What is the approach to antibiotic
therapy for patients hospitalized with
CAP outside the ICU?
Patients should receive initial antibiotic
therapy as soon as possible after diag-
nosis. Although therapy within 4 hours
of arrival in the emergency department
has been associated with reduced mor-
tality, an undue emphasis on early ther-
apy may lead to unnecessary use of
antibiotics and associated complica-
tions (35, 36). In 1 study, the final diag-
nosis of pneumonia in patients sus-
pected of having it in the emergency
department decreased from 75.9% to
58.9% after initiation of a program to
give patients antibiotics within 4 hours
of arrival in the emergency department
(37).
Specific therapies should be guided by
the diagnosis. Patients at risk for infec-
tion with methicillin-resistant S aureus
(MRSA), DRSP, and resistant gram-neg-
ative organisms require extended-
spectrum empirical therapy. Many risk
factors have been identified for each of
these infections, but they tend to be
only weakly associated and should not
be relied on when choosing therapy.
Importantly, the previous designation
of health care–associated pneumonia
has been abandoned because it was
not particularly associated with resist-
ant infections and led to increased use
of extended-spectrum antibiotics. The
2 consistently strong risk factors are
previous isolation of the resistant orga-
nism, especially from the respiratory
tract, and hospitalization within 90 days
with administration of antibiotics. Des-
pite multiple attempts to create models
ITC10 In the Clinic
to identify resistant organisms based
on patient risk factors, none have been
sufficiently validated for use in clinical
practice. Such validation is important
because the prevalence of resistant
organisms is low and varies among
hospitals. Recognizing this fact, the
ATS/IDSA guidelines recommend that
hospitals validate their own local risk
factors or else treat patients on the ba-
sis of a history of resistant infections or
recent hospitalization with antibiotics. If
available, rapid nasal PCR for MRSA
can be used to guide empirical ther-
apy. Patients with influenza pneumonia
should receive treatment with a neur-
aminidase inhibitor, even if they have
been sick for more than 48 hours (12).
Because co-infection is common, they
should also receive antibiotics, at least
until culture results are available. Such
patients are at increased risk for S aur-
eus infection but not for MRSA and
may therefore be treated with standard
antibiotics (10).
For patients without risk factors for re-
sistant bacteria, guidelines recommend
treatment with either an intravenous or
oral quinolone (levofloxacin, 750 mg/d,
when renal function is normal, or
moxifloxacin, 400 mg/d) or the com-
bination of a b-lactam (cefotaxime,
ceftriaxone, ampicillin–sulbactam, or
high-dose ampicillin, but not cefurox-
ime) plus a macrolide or doxycycline
(16). The addition of a macrolide to a
b-lactam has been associated with
reduced mortality, although this ben-
efit may be limited to patients with
severe pneumonia (38). Even those
with bacteremic pneumococcal pneu-
monia seem to benefit from added
macrolide coverage (39). Specific
b-lactams, such as ceftriaxone and
cefotaxime, are preferred if DRSP is
suspected because they are effective
at mean inhibitory concentrations up
to 2 mg/L (40). However, 1 study
showed increased mortality when
cefuroxime was used in patients with
bacteremic DRSP (41).
An international study of 4337 hospital-
ized patients with CAP showed that
approximately 20% had evidence of
Annals of Internal Medicine
atypical pathogen infection and that
therapy directed against these organ-
isms decreased time to clinical stability,
length of stay, and both total and CAP-
related mortality (42). Another study of
2209 hospitalized Medicare patients
with bacteremic pneumonia found that
therapy directed at atypical pathogens
reduced 30-day mortality and 30-day
readmission rate, but the benefits
occurred only with macrolides and not
with fluoroquinolones (43).
Patients who have had a resistant orga-
nism in the past should receive therapy
directed at the previously isolated or-
ganism. Therapy for suspected MRSA
includes vancomycin (15 mg/kg every
12 hours, with adjustment based on
levels) or linezolid (600 mg every 12
hours) but may be withheld if nasal
PCR results are negative. Therapy for
suspected gram-negative infections
includes piperacillin–tazobactam (4.5 g
every 6 hours), cefepime (2 g every 8
hours), ceftazidime (2 g every 8 hours),
imipenem (500 mg every 6 hours), mer-
openem (1 g every 8 hours), or aztreo-
nam (2 g every 8 hours). If culture
results are negative at 48 hours and
the patient is clinically stable, the antibi-
otic regimen may be deescalated. Al-
though there are no randomized trials
of deescalation after negative culture
results, there is a strong evidence base
from observational studies.
A study of 165 U.S. hospitals found that
among 14 170 patients who received
extended-spectrum antibiotics, only
13% had them deescalated by hospital
day 4. Deescalation seemed to be safe
and was associated with lower odds of
subsequent transfer to the ICU (adjus-
ted odds ratio, 0.38 [95% CI, 0.18 to
0.79]), shorter hospital stay, and lower
costs in propensity-matched analyses.
Importantly, hospital deescalation rates
ranged from 2% to 35%, and even hos-
pitals in the top quartile of deescalation
failed to deescalate more than 50% of
their lowest-risk patients, leaving ample
room for improvement (44).
Annals of Internal Medicine
Patients whose only risk factor is recent
hospitalization with antibiotic administra-
tion should have cultures obtained, but
antibiotics against MRSA or Pseudomonas
should be withheld unless culture results
are positive or the patient's condition
deteriorates.
Which antibiotics should be given to
patients admitted to the ICU?
Patients in the ICU should receive em-
pirical therapy with at least 2 antibiotics
(16). Clinicians should assess for risk
factors for Pseudomonas aeruginosa
and treat those without risk factors with
intravenous ceftriaxone or cefotaxime
plus either azithromycin or a respira-
tory quinolone. Patients with risk fac-
tors should be treated with an
intravenous, antipseudomonal b-lac-
tam (cefepime, piperacillin–tazobac-
tam, imipenem, meropenem) plus an
intravenous quinolone effective against
P aeruginosa (ciprofloxacin or high-
dose levofloxacin). Alternatively, an in- 42. Arnold FW, Summersgill
JT, Lajoie AS, et al;
travenous, antipseudomonal b-lactam Community-Acquired
Pneumonia Organization
combined with an aminoglycoside (CAPO) Investigators. A
worldwide perspective of
(amikacin, gentamicin, or tobramycin) atypical pathogens in
plus either an intravenous macrolide community-acquired
pneumonia. Am J Respir
(azithromycin or erythromycin) or an in- Crit Care Med.
2007;175:1086-93.
travenous antipneumococcal quino-
[PMID: 17332485]
lone (levofloxacin or moxifloxacin) 43. Metersky ML, Ma A,
Houck PM, et al.
should be used. In studies of patients Antibiotics for bacteremic
pneumonia: improved
admitted to the ICU with severe CAP, outcomes with macrolides
but not fluoroquinolones.
mortality was reduced when combi-
Chest. 2007;131:466-73.
nation therapy was used; monother- [PMID: 17296649]
apy, even with a quinolone, was not as 44. Deshpande A, Richter SS,
Haessler S, et al. De-esca-
effective. lation of empiric antibiot-
ics following negative
cultures in hospitalized
A 2012 meta-analysis of 28 observatio- patients with pneumonia:
rates and outcomes. Clin
nal studies involving nearly 10 000 crit- Infect Dis. 2021;72:1314-
22. [PMID: 32129438]
ically ill patients found that macrolide 45. Sligl WI, Asadi L, Eurich
DT, et al. Macrolides and
use (generally in a combination regi- mortality in critically ill
men) was associated with an 18% patients with community-
reduction in mortality compared with
acquired pneumonia: a
systematic review and
nonmacrolide regimens and that a meta-analysis. Crit Care
Med. 2014;42:420-32.
b-lactam–macrolide combination had a [PMID: 24158175]
46. Baddour LM, Yu VL,
trend toward reduced mortality com- Klugman KP, et al;
pared with a b-lactam–quinolone regi- International
Pneumococcal Study
men (45). In patients with bacteremic Group. Combination anti-
biotic therapy lowers mor-
pneumococcal pneumonia and critical tality among severely ill
patients with pneumococ-
illness, studies have found that mortality cal bacteremia. Am J
was lower with combination therapy Respir Crit Care Med.
2004;170:440-4. [PMID:
than with monotherapy (46). 15184200]
In the Clinic ITC11 © 2022 American College of Physicians
 What are the other components of nolone monotherapy. Patients who
ICU care for CAP? have responded to a b-lactam–macro-
Hydration should be ensured and sup- lide combination can be continued on
plemental oxygen should be given to macrolide monotherapy unless culture
maintain oxygen saturation above 90%. results justify dual therapy. Quinolones
Chest physiotherapy should be consid- have excellent oral bioavailability. For
ered. Intubation and mechanical venti- patients with a working gastrointestinal
lation are required in patients with tract, there is little added benefit to in-
oxygen saturation below 90% on maxi- travenous quinolones (50).
47. Girou E, Schortgen F,
Delclaux C, et al. mal mask oxygen, inability to clear
Association of noninvasive
secretions, inability to protect the air- To facilitate a switch to oral therapy,
ventilation with nosoco-
mial infections and sur- way, or hypercarbia. If the patient has hospitals should consider using a
vival in critically ill
patients. JAMA. only hypoxemia or hypercarbia and is standing order set supplemented by
2000;284:2361-7. [PMID:
alert and cooperative, noninvasive pos- antibiotic stewardship. Such programs
11066187]
48. Stern A, Skalsky K, Avni T, itive-pressure ventilation should be have been shown to reduce the num-
et al. Corticosteroids for
pneumonia. Cochrane considered. This therapy is associated ber of days patients receive intrave-
Database Syst Rev.
with fewer complications than endotra- nous therapy and shorten hospital stay
2017;12:CD007720.
[PMID: 29236286] cheal intubation, including ventilator- (51). Gains seem to be maintained
49. Horby P, Lim WS, even after stewardship efforts are
Emberson JR, et al; associated pneumonia (47). Studies of
RECOVERY Collaborative
steroids in CAP tend to be small and reduced, and the program can save
Group. Dexamethasone in
hospitalized patients with heterogeneous in nature (48). Many money (52).
Covid-19. N Engl J Med.
2021;384:693-704. are subject to bias and report problem- What is the role of nondrug therapies?
[PMID: 32678530]
50. Belforti RK, Lagu T, atic outcomes, such as length of stay. In outpatients, nondrug therapy con-
Haessler S, et al.
Association between ini-
Therefore, routine use of systemic corti- sists of oral hydration. For hospitalized
tial route of fluoroquino- costeroids is not recommended, but patients, nondrug therapies include in-
lone administration and
outcomes in patients hos- patients with refractory septic shock travenous hydration and oxygen for hy-
pitalized for community-
acquired pneumonia. Clin
(16) or COVID-19 requiring ventilator poxemia. Chest physiotherapy has not
Infect Dis. 2016;63:1-9. support may benefit from steroids (49). been widely studied but can improve
[PMID: 27048748]
51. Fishbane S, Niederman
When can clinicians transition outcomes in patients with pneumonia
MS, Daly C, et al. The
impact of standardized hospitalized patients from who have more than 30 mL of sputum
order sets and intensive
clinical case management intravenous to oral antibiotics? per day and impaired clearance of
on outcomes in commu-
nity-acquired pneumonia. Switching to oral antibiotics is indicated secretions (53).
Arch Intern Med.
2007;167:1664-9. [PMID: once cough, sputum production, and A 2013 meta-analysis of 6 randomized
17698690]
52. Ciarkowski CE, Timbrook
dyspnea improve; the patient is afe- trials involving 434 patients evaluated
TT, Kukhareva PV, et al. A brile on 2 occasions 8 hours apart; and the following 4 types of chest physio-
pathway for community-
acquired pneumonia with they are able to take oral medications. therapy: conventional chest physiother-
rapid conversion to oral
therapy improves health This switch can be made as early as 24 apy, active cycle breathing, osteopathic
care value. Open Forum manipulation, and positive expiratory
Infect Dis. 2020;7:
to 48 hours after admission and can be
ofaa497. [PMID: done safely even if pneumococcal bac- pressure. No method reduced mortal-
33269294]
53. Graham WG, Bradley DA. teremia has been documented. Longer ity, but osteopathic manipulation and
Efficacy of chest physio-
therapy and intermittent durations of therapy are usually positive expiratory pressure reduced
positive-pressure breath-
needed for patients infected with P aer- the duration of hospital stay by 2.02
ing in the resolution of
pneumonia. N Engl J uginosa or S aureus and for those with and 1.4 days, respectively (54).
Med. 1978;299:624-7.
[PMID: 355879] extrapulmonary complications, such as In severely ill patients, nondrug therapy
54. Yang M, Yan Y, Yin X, et
al. Chest physiotherapy empyema or meningitis, but should be can include noninvasive ventilatory
for pneumonia in adults.
Cochrane Database Syst
individualized to specific patient situa- support and mechanical ventilation for
Rev. 2013:CD006338. tions. An oral regimen that covers all those with respiratory failure. One
[PMID: 23450568]
55. Stefan MS, Priya A, Pekow organisms isolated in blood or sputum observational study found that nonin-
PS, et al. The comparative
effectiveness of noninva- cultures and corresponds to the intra- vasive ventilation was associated with
sive and invasive ventila-
tion in patients with
venous therapy should be selected. lower mortality, but only for patients
pneumonia. J Crit Care. For some patients, this means a b- with comorbid chronic obstructive pul-
2018;43:190-6. [PMID:
28915393] lactam–macrolide combination or qui- monary disease (COPD) or CHF (55).

© 2022 American College of Physicians ITC12 In the Clinic Annals of Internal Medicine
When should a consultation be
requested for hospitalized patients,
and which types of specialists or
subspecialists should be consulted?
An infectious disease consultation is
appropriate if there are questions
about initial antibiotic therapy or when
the patient does not respond to initial
therapy. Questions about appropriate
site of care or the need for vasopres-
sors or ventilatory support are appro-
priate indications for critical care
consultation. Pulmonary or thoracic
surgical consultation is appropriate for
placement of a chest tube if a compli-
cated parapneumonic effusion or em-
pyema is found on thoracentesis
because early therapy can reduce hos-
pital stay and avoid complications.
Cardiology consultation may be need-
ed in cases of ischemia or CHF.
In a study of 170 patients with pneumo-
coccal pneumonia, 19.4% had at least 1
major cardiac event, including 12 with
acute myocardial infarction, 8 with new-
onset atrial fibrillation or ventricular
tachycardia, and 13 with newly diag-
nosed or worsening heart failure without
other cardiac complications. Patients
with cardiac events had significantly
higher mortality (27.3% vs. 8.8%) (56).
When can patients be discharged from
the hospital, and how long should
antibiotics be continued?
Patients can be discharged once they are
clinically stable (temperature ≤37.8
C,
heart rate <100 beats/min, respiratory
rate <24 breaths/min, oxygen saturation
≥90%, systolic blood pressure ≥90 mm
Hg, and normal mental status) (57).
Therapy may be continued after dis-
charge, but the total duration should not
exceed 5 to 7 days. In some cases, 3
days may be sufficient (33). Excess treat-
ment is common, particularly after dis-
charge, and is associated with increased
incidence of adverse drug events (28). If
switching to oral therapy, it is not neces-
sary to observe the response in the
Annals of Internal Medicine
hospital. In one study, clinically stable
patients were observed on oral ther-
apy before discharge, and no deterio-
ration occurred (58). Another study
compared patients who remained in
the hospital for 1 day after the switch
with those discharged on the same
day and found no differences in mor-
tality or 14-day readmission rate (59).
Patients may be discharged on intra-
venous antibiotics as long as they are
clinically stable; consultation with an
infectious disease specialist can en-
56. Musher DM, Rueda AM,
sure that the duration and route of Kaka AS, et al. The associ-
ation between pneumo-
outpatient therapy are appropriate coccal pneumonia and
acute cardiac events. Clin
(60). Programs directed by infectious Infect Dis. 2007;45:158-
disease specialists seem to produce 65. [PMID: 17578773]
57. Mandell LA, Wunderink
better outcomes at lower costs (61). RG, Anzueto A, et al;
Infectious Diseases
What are the indications for follow- Society of America.
Infectious Diseases
up CXR after discharge? Society of America/
American Thoracic Society
consensus guidelines on
Routine CXR before discharge is the management of com-
unnecessary, but patients who do not
munity-acquired pneumo-
nia in adults. Clin Infect
achieve clinical stability and those who Dis. 2007;44 Suppl 2:
S27-72. [PMID:
deteriorate despite therapy require 17278083]
58. Rhew DC, Hackner D,
an aggressive evaluation, including Henderson L, et al. The
CXR. As with outpatient pneumonia, if
clinical benefit of in-hos-
pital observation in ‘low-
the patient has a good clinical res-
risk’ pneumonia patients
after conversion from par-
ponse to therapy, CXR need not be enteral to oral antimicro-
bial therapy. Chest.
repeated. 1998;113:142-6. [PMID:
9440581]
59. Nathan RV, Rhew DC,
How can patients prevent recurrent Murray C, et al. In-hospi-
tal observation after anti-
CAP? biotic switch in
pneumonia: a national
Patients should receive pneumococcal evaluation. Am J Med.
2006;119:512.e1-7.
and influenza vaccinations; avoid [PMID: 16750965]
60. Sharma R, Loomis W,
smoking; and optimize treatment of Brown RB. Impact of man-
datory inpatient infectious
comorbid illnesses, such as CHF and
disease consultation on
COPD. They should be evaluated for outpatient parenteral anti-
biotic therapy. Am J Med
medical conditions that could predis- Sci. 2005;330:60-4.
[PMID: 16103785]
pose them to recurrent infection. One 61. Shah A, Petrak R,
study found that 6% of patients with Fliegelman R, et al.
Infectious diseases spe-
CAP had a new comorbid condition, cialty intervention is asso-
ciated with better
including diabetes mellitus, cancer, outcomes among pri-
COPD, and HIV infection (62). If pneu-
vately insured individuals
receiving outpatient par-
monia recurs in the same location, the enteral antimicrobial ther-
apy. Clin Infect Dis.
possibility of bronchiectasis, aspirated 2019;68:1160-5. [PMID:
30247512]
foreign body, or endobronchial ob- 62. Falguera M, Martín M,
Ruiz-González A, et al.
struction should be considered. Re- Community-acquired
current pneumonia or pneumonia with pneumonia as the initial
manifestation of serious
an unusual pathogen may signal underlying diseases. Am
J Med. 2005;118:378-
immunodeficiency. 83. [PMID: 15808135]
In the Clinic ITC13 © 2022 American College of Physicians
 In a 2014 review of 12 studies, the 30-day caused the readmission only 17.9% to
readmission rate for patients with CAP var- 29.4% of the time; other common causes
ied from 16.8% to 20.1% (63). Pneumonia were exacerbations of CHF or COPD.

Treatment... The most important clinical decisions in the treatment of CAP include
determining the site of care, selecting antibiotic therapy, delivering supportive care, and
determining the need for ventilatory support. Antibiotic therapy differs by site of care.
However, all patients should receive timely empirical therapy directed at pneumococ-
cus, atypical pathogens, and other organisms depending on risk factors. The PSI and
the CURB-65 score aid decisions about hospital admission. Patients should be managed
in the ICU if they require ventilatory or vasopressor support or close observation.
Consultation should occur in cases of severe disease and when patients do not respond
to initial therapy or have complications. Broad-spectrum empirical antibiotics should be
deescalated promptly after negative culture results, and patients can be transitioned to
oral antibiotics and discharged once they are clinically stable. Routine follow-up CXR is
unnecessary if the patient is responding well. Patients should be offered pneumococcal,
COVID-19, and influenza vaccinations and should be encouraged to avoid smoking.

CLINICAL BOTTOM LINE

Practice Improvement
What measures do stakeholders use to process measures represents good
measure the quality of care? care, it may not affect these outcomes
(64), which are influenced by various
Process measures for pneumonia qual-
medical and social conditions.
ity of care, including measurement of
oxygenation, prompt initiation of appro- What do professional organizations
priate antibiotics, drawing of blood cul- recommend with regard to
63. Prescott HC, Sjoding MW, tures before antibiotic administration, prevention and treatment?
Iwashyna TJ. Diagnoses of
early and late readmis-
providing smoking cessation counsel- The ATS and IDSA issue joint guide-
sions after hospitalization ing, and administration of influenza and lines on the treatment of CAP (16). The
for pneumonia. A system-
atic review. Ann Am pneumococcal vaccine, are no longer American College of Physicians has
Thorac Soc.
2014;11:1091-100.
collected or publicly reported by the issued a guideline on appropriate use
[PMID: 25079245] Centers for Medicare & Medicaid of short-course antibiotics, which in-
64. Werner RM, Bradlow ET.
Relationship between Services (CMS). Pneumonia is not one cludes treatment of CAP (65). The CDC
Medicare's Hospital
Compare performance of the conditions for which CMS publishes the ACIP recommendations
measures and mortality requires core measures. Instead, CMS for influenza, COVID-19, and pneumo-
rates. JAMA.
2006;296:2694-702. reports risk-standardized mortality, re- coccal vaccination and Core Elements
[PMID: 17164455]
65. Lee RA, Centor RM, admission, and excess days in acute of Antibiotic Stewardship (66). The rec-
Humphrey LL, et al;
Scientific Medical Policy
care for patients with pneumonia. ommendations in this article reflect
Committee of the Although attention to the previous these guidelines.
American College of
Physicians. Appropriate
use of short-course antibi-
otics in common infec-
tions: best practice advice
from the American
College of Physicians. Ann
Intern Med.
2021;174:822-7. [PMID:
33819054]
66. Centers for Disease
Control and Prevention.
Core Elements of Hospital
Antibiotic Stewardship
Programs. Accessed at
www.cdc.gov/antibiotic-
use/core-elements/
hospital.html on 10
January 2022.

© 2022 American College of Physicians ITC14 In the Clinic Annals of Internal Medicine
In the Clinic Patient Information

Tool Kit
https://medlineplus.gov/pneumonia.html
https://medlineplus.gov/languages/
pneumonia.html
Information and handouts in English and
other languages from the National
Institutes of Health's MedlinePlus.
Community-Acquired
Pneumonia www.nhlbi.nih.gov/health/pneumonia
www.nhlbi.nih.gov/health-topics/espanol/
neumonia
Information in English and Spanish from
the National Heart, Lung, and Blood
Institute.

www.cdc.gov/pneumococcal/index.html

In the Clinic
www.cdc.gov/pneumococcal/index-sp.
html
Information on pneumococcal disease in
English and Spanish from the Centers for
Disease Control and Prevention.
Information for Health Professionals
www.atsjournals.org/doi/full/10.1164/
rccm.201908-1581ST
2019 clinical practice guideline on diagnosis
and treatment of adults with community-
acquired pneumonia from the American
Thoracic Society and the Infectious
Diseases Society of America.

www.cdc.gov/vaccines/vpd/pneumo/hcp/
index.html
Pneumococcal vaccination information and
resources from the Centers for Disease
Control and Prevention.

www.cdc.gov/pneumococcal/clinicians/
index.html
Information on pneumococcal disease in
English and Spanish from the Centers for
Disease Control and Prevention.

Annals of Internal Medicine In the Clinic ITC15 © 2022 American College of Physicians
 In the Clinic
WHAT YOU SHOULD KNOW Annals of Internal Medicine
ABOUT PNEUMONIA
What Is Pneumonia?
Pneumonia is a serious infection of the lungs.
Community-acquired pneumonia is when you
develop pneumonia outside a hospital or nurs-
ing home. Pneumonia can be caused by bacte-
ria, viruses, or fungi and can range from mild to
severe. It is important that pneumonia be identi-
fied and treated quickly.

What Are the Symptoms?

• Fever or chills
• A cough that produces a lot of mucus
• Chest pain that is worse with deep breathing
• Shortness of breath
• Feeling tired and weak
• Confusion
What Are the Risk Factors?
• Being age 65 years or older
• Having other health conditions, like diabetes or
lung, heart, liver, or kidney disease
•Drinking alcohol
•Smoking cigarettes
•Having the flu or COVID-19
•Having a weakened immune system
Pneumonia can be prevented by quitting smoking
and by receiving vaccines. The pneumococcal,
influenza, and COVID-19 vaccines have all been
shown to prevent pneumonia and its
complications.
How Is It Treated?
• Treatment depends on how severe your pneu-
monia is. Your doctor will determine whether you
can be treated at home or in the hospital.
• Most patients can be treated at home. Some who
are very ill or have a risk for complications might
need to stay in the hospital. If you have to stay in
the hospital, your doctor will monitor your heart
and breathing rates, and you might be given IV
fluids or medicine.
• If your pneumonia is caused by bacteria, your
doctor will prescribe antibiotics. Symptoms usu-
ally start to go away within 2 to 3 days of starting
medicine. It is important to finish all of your anti-
biotics, even if you are feeling better.
• Follow up with your doctor 1 to 2 days after start-
ing antibiotics to make sure you are responding

How Is It Diagnosed?
• Your doctor will take a history; check your vital
signs, including your oxygen level; and perform a
physical examination.
• You might be given a flu test or a COVID-19 test.
• You might have other tests. A chest x-ray may be
helpful to confirm the diagnosis. You may need
to have a chest CT scan. Tests of the sputum (the
mucus you produce when coughing) or urine
may help your doctor learn what type of bacteria
is causing your pneumonia. Blood tests may help
determine the severity of the infection.
Patient Information
well.
• Drink plenty of fluids and stay hydrated.
• Get plenty of rest. Feeling tired and coughing
may last for a month or longer.
Questions for My Doctor
• Should I be treated at home or in the hospital?
• What medicine do I need to take?
• What can I do to help relieve my symptoms?
• When should I have a follow-up visit?
• How can I prevent another episode of
pneumonia?

For More Information

American Lung Association
www.lung.org/lung-health-diseases/lung-disease-lookup/
pneumonia
MedlinePlus
https://medlineplus.gov/pneumonia.html

© 2022 American College of Physicians ITC16 In the Clinic Annals of Internal Medicine
Appendix Table. Drug Treatments for CAP
Agent Mechanism of Action Dosage Benefits Adverse Effects Notes

Antibiotics for
community-acquired
MRSA
Linezolid (Zyvox) Bacteriostatic; binds to 600 mg PO or IV every Penetrates well into the Myelosuppression, par- Drug interactions may
50S ribosomal sub- 12 h lung and is active ticularly thrombocyto- lead to serotonin syn-
unit to inhibit bacte- against MRSA penia, vomiting, drome
rial protein synthesis diarrhea, seizures, Do not use with tricy-
hypoglycemia clic antidepressants,
monoamine oxidase
inhibitors, or selective
serotonin reuptake
inhibitors
Monitor blood
pressure
Clindamycin* Bacteriostatic; binds to 600 mg PO every 8 h Can inhibit toxin pro- Diarrhea, esophagitis, Can cause
(Cleocin) 50S ribosomal sub- duction by MRSA hypersensitivity Clostridioides difficile–
unit to inhibit bacte- associated diarrhea
rial protein synthesis Caution with asthma,
severe hepatic
disease
Vancomycin Bactericidal; inhibits 15–20 mg/kg IV every Active against MRSA, Ototoxicity, nephrotox- Avoid rapid infusion
(Vancocin) cell wall and RNA 8–12 h with extensive clinical icity, neutropenia, (causes histamine
synthesis experience nausea, hypokalemia release)
Is not therapy for Avoid optimal extrava-
methicillin-suscepti- sation
ble Staphylococcus Monitor trough con-
aureus centrations
In chronic kidney dis-
ease, individualize
dose

Antipseudomonal
b-lactams
Piperacillin– Bactericidal; interferes Piperacillin–tazobac- Active against pneu- Anaphylaxis, rash, nau- Only use for patients
tazobactam with peptidoglycan tam: 3.375 mg IV mococci and sea, vomiting, diar- with pseudomonal
Cefepime cross-linking and pre- every 4–6 h Pseudomonas rhea, phlebitis, risk factors, although
Imipenem vents formation of the Cefepime: 1–2 g IV aeruginosa seizures (high doses), generally active
Meropenem bacterial cell wall every 12 h hypokalemia (high against DRSP
Imipenem: 1 g IV every doses), elevated liver Can dose daily if
8h enzymes, prolonged patient has renal
Meropenem: 1 g IV prothrombin time insufficiency
every 8 h (especially if patient is
using coumadin)
Seizure potential
greater with imipe-
nem than
meropenem

Cephalosporins
Cefuroxime Bactericidal; interferes Cefuroxime: 500 mg Active against pneumo- Anaphylaxis, rash, Not to be used alone
Cefpodoxime with peptidoglycan PO twice daily cocci and Haemophilus nausea, vomiting, in CAP
Ceftriaxone cross-linking and pre- Cefpodoxime: 400 mg influenzae, including diarrhea, elevated Combine with a macro-
Cefotaxime vents formation of the PO twice daily b-lactamase–producing liver function test lide
bacterial cell wall Ceftriaxone: 1–2 g organisms results, interstitial ne- Although cefuroxime
every 12–24 h (usually phritis, altered coagu- can be used as oral
every 24 h) lation, pseudomem- therapy, it should not
Cefotaxime: 1 g every branous colitis be used IV because it
8h is not as active
against DRSP as other
cephalosporins

Continued on following page

Annals of Internal Medicine In the Clinic © 2022 American College of Physicians

Appendix Table—Continued
Agent Mechanism of Action Dosage Benefits Adverse Effects Notes

Glycylcycline
Tigecycline* Bacteriostatic; binds to 100 mg IV initially, then Can be used for CAP,
(Tygacil) 30S ribosomal sub- 50 mg IV every 12 h
unit to inhibit bacte- Infuse over 30–60 min
rial protein synthesis Penetrates well into res-
Macrolides
Azithromycin Bacteriostatic; binds to Azithromycin: 500 mg Covers pneumococcus,
Clarithromycin 50S ribosomal sub- IV or PO on day 1, fol-
unit and inhibits bacte- lowed by 500 mg IV
rial protein synthesis or PO for 7–10 d for
hospitalized patients
(250 mg on days 2–5
for outpatients)
Azithromycin in the
microspheres oral
extended-release for-
mulation: 2 g PO on
day 1 without follow-
up dosing for out-
patients
Clarithromycin: 500
mg PO twice daily,
or 1000 mg/d PO
(extended-release
preparation) for
outpatients
Vomiting, diarrhea, Avoid with pregnancy
but only when there hepatotoxicity, pan- With severe hepatic
are no other options creatitis, anemia disease, use 25 mg
Increase in all-cause for maintenance
piratory secretions; mortality infusion
active against pneu-
mococcus and atypi-
cal pathogens, but
not P aeruginosa
Nausea, vomiting, diar- Use as monotherapy
atypical pathogens, rhea, QT prolonga- only in patients with-
and H influenzae tion, dyspepsia out cardiopulmonary
(clarithromycin) disease or modifying
factors; otherwise,
combine with a
b-lactam
Erythromycin is less ex-
pensive but is not rec-
ommended because
of the need for more
frequent dosing,
more intestinal upset,
and no coverage of
H influenzae

Penicillins
Amoxicillin– Bactericidal; interferes Amoxicillin–clavula- Active against pneu- Anaphylaxis, rash, Do not use alone in
clavulanate with peptidoglycan nate: 875 mg PO mococci and b-lacta- nausea, vomiting, di- CAP
Ampicillin cross-linking and pre- twice daily mase–producing arrhea, phlebitis, seiz- Combine with a
Ampicillin– vents formation of the Ampicillin: 500–1000 H influenzae ures (high doses), macrolide
sulbactam bacterial cell wall mg PO 3 times daily High doses (1 g 3 times hypokalemia (high
Ampicillin–sulbactam: daily) active against doses), elevated liver
1–2 g IV every 6 h DRSP enzymes, prolonged
prothrombin time
(especially if patient is
using warfarin)

Quinolones
Ciprofloxacin Bactericidal; interferes Ciprofloxacin: 400 mg Active against P aerugi- Seizures, hypersensitiv- Only use ciprofloxacin
Levofloxacin with bacterial DNA IV every 8 h nosa, atypical patho- ity, photosensitivity, in severe CAP
Moxifloxacin gyrase Levofloxacin: 500–750 gens, and tendon rupture, nau- Not always reliable
Kills bacteria in a mg/d IV or PO H influenzae sea, vomiting, diar- against pneumococci,
concentration- Moxifloxacin: 400 Levofloxacin and moxi- rhea, QT and should be com-
dependent fashion mg/d IV or PO floxacin are the “res- prolongation bined with other
piratory quinolones,” agents if DRSP is pos-
with activity against sible
DRSP, H influenzae, If used in severe CAP,
and atypical do not use as
pathogens monotherapy

Tetracyclines
Doxycycline Bacteriostatic; binds to 100 mg IV or PO twice Active against key Nausea, vomiting, diar- Not always fully reli-
30S ribosomal sub- daily bacterial and atypical rhea, photosensitivity able against
unit and interferes pathogens pneumococci
with bacterial protein
synthesis

CAP = community-acquired pneumonia; DRSP = drug-resistant Streptococcus pneumoniae; IV = intravenously; MRSA = methicillin-
resistant Staphylococcus aureus; PO = orally.
* Black box warning.

© 2022 American College of Physicians In the Clinic Annals of Internal Medicine