REVIEW

CME MOC
Anita R. Modi, MD Christopher S. Kovacs, MD
Department of Infectious Disease, Department of Infectious Disease, Cleveland Clinic;
Cleveland Clinic Assistant Professor, Cleveland Clinic Lerner College
of Medicine of Case Western Reserve University,
Cleveland, OH

Community-acquired pneumonia:
Strategies for triage and treatment
ABSTRACT hile physicians have treated pneu-
Community-acquired pneumonia significantly contributes
W monia for centuries, each stage of the
clinical decision-making process still poses
to patient morbidity and healthcare costs. As our under- challenges, from determining the most appro-
standing of this common infection grows, collaborative priate setting of care for a patient with suspect-
efforts among researchers and clinical societies provide ed pneumonia to planning follow-up after an-
new literature and updated guidelines informing its tibiotic completion. Over the years, physicians
management. This review discusses diagnostic methods, have witnessed the advent of new medical and
empiric treatment, and infection prevention strategies for respiratory therapies as well as the develop-
patients with suspected community-acquired pneumonia. ment of antibiotic resistance in the manage-
ment of this common infection.
KEY POINTS Inpatients with pneumonia fall into 2 cat-
egories: those with community-acquired pneu-
Systematically stratifying patients with suspected monia (CAP) who are admitted, and those
community-acquired pneumonia based on mortality risk who develop either hospital-acquired or venti-
can aid in designating the safest level of care for each lator-associated pneumonia while already hos-
patient. pitalized. Each patient population faces unique
organism exposures, and thus, recommended
diagnostic tests, empiric treatment regimens,
Empiric treatment should be informed by the local anti- and goals for infection prevention vary.
biogram (ie, local patterns of antibiotic resistance) with This article reviews guidelines by the In-
multidrug-resistant organism coverage added based on fectious Diseases Society of America (IDSA)
individual patient and institutional risk factors. and the American Thoracic Society (ATS)1
and interprets recent studies to address ques-
Prompt de-escalation to targeted antimicrobial therapy, tions that arise specifically in the inpatient
guided by diagnostic testing, can reduce antibiotic resis- management of CAP.
tance and antibiotic-related adverse drug reactions.
■ COMMON AND COSTLY
Appropriate clinical and radiographic follow-up after an- CAP is a significant health concern, with one
tibiotic course completion to assess for treatment failure study reporting 915,500 episodes in adults at
is a subject of ongoing debate. least 65 years of age in the United States every
year, and medical costs associated with CAP
exceeding $10 billion in 2011.2,3
The National Center for Health Statis-
tics reported 1.7 million visits to emergency
departments in the United States in 2017 in
which pneumonia was the primary discharge
diagnosis, and listed pneumonia as the cause
doi:10.3949/ccjm.87a.19067 of death for 49,157 people in 2017.4
CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 87 • NUMBER 3 MARCH 2020 145

Downloaded from www.ccjm.org on July 20, 2021. For personal use only. All other uses require permission.
 COMMUNITY-ACQUIRED PNEUMONIA

IDSA/ATS, patients with scores of 0 or 1 can

TABLE 1
be managed as outpatients, those with scores
The CURB-65 calculator of 2 should be admitted to the hospital, and
those with scores of 3, 4, or 5 need care in the
Criteria Points intensive care unit.
C Confusion 1 An abbreviated version of this calculator,
CRB-65, allows risk-stratification of outpa-
U Urea > 7 mmol/L 1 tients without laboratory work.1
R Respiratory rate > 30 breaths per minute 1 The Pneumonia Severity Index incorpo-
rates 20 risk factors to place patients into 5
B Systolic blood pressure < 90 mm Hg 1 classes correlated with mortality risk (Table
or diastolic blood pressure < 60 mm Hg 2).5 The authors suggest outpatient manage-
65 Age ≥ 65 1 ment for those in classes I or II and inpatient
management for those in risk classes IV and V.
Patients in class III may be safely treated in an
Level of care required Total score outpatient setting with adequate support or in
Outpatient 0–1 an inpatient observation unit.
While CURB-65 may be better in busy
Inpatient 2 clinical settings, as it is a shorter risk stratifi-
Intensive care 3–5 cation scale for CAP, the Pneumonia Sever-
ity Index is preferred by the IDSA/ATS 2019
Based on information in reference 1.
guidelines as it has been more extensively
studied and validated.1
■ RISK-STRATIFICATION OF The IDSA/ATS guidelines list a separate
COMMUNITY-ACQUIRED PNEUMONIA set of major and minor criteria to define “se-
vere pneumonia” to determine which patients
The IDSA/ATS 2019 guidelines1 emphasize with suspected CAP merit intensive care.1 At
What level the importance of first determining what level least 1 of the major criteria or at least 3 of the
of patient care is needed: Is outpatient treat- minor criteria are required for the diagnosis of
of care does ment appropriate, or does the patient need severe pneumonia (Table 3).
the patient to be admitted to the hospital, or even to the The Pneumonia Patient Outcomes Re-
need: intensive care unit? Appropriate triage can search Team study, a multicenter, prospec-
prevent stresses on the patient and the health- tive controlled study of both ambulatory and
Outpatient? care system associated with under- or overes- hospitalized patients with CAP, also devised a
Inpatient? timating illness severity. Patients at high risk list of risk factors associated with death within
of death whose acuity is not fully appreciated 30 days.6 These risk factors include altered
Intensive? face inadequate support, while those admitted mental status, uremia, leukopenia, and hypox-
despite low risk of death may be unnecessarily emia. Chronic liver failure was a risk factor
subjected to the risks of the hospital setting, highlighted in this study but was not included
such as infections from healthcare-associated in the IDSA/ATS criteria.
multidrug-resistant organisms. Yet none of these scoring systems can fully
Risk calculators are routinely used to help capture all medical or psychosocial comorbid-
physicians triage their patients in everyday prac- ities that may prevent successful recovery in
tice, although they have not been specifically the outpatient setting. A retrospective chart
validated to predict the need for admission. review of more than 1,800 patients found that
CURB-65 is a simple calculator based on 45% of patients who had “low-risk” CAP by
5 risk factors first identified in 1987 (Table the Pneumonia Severity Index were never-
1).1 Patients receive 1 point each for confu- theless admitted.7 Patients with cognitive
sion, high blood urea nitrogen, high respira- impairment, coronary artery disease, diabetes
tory rate, low blood pressure, and age 65 or mellitus, pulmonary disease, multilobular ra-
older; the higher the total score, the higher diographic opacities, home oxygen therapy,
the 30-day mortality risk. According to the corticosteroid use, or use of antibiotics prior
146 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 87 • NUMBER 3 MARCH 2020

Downloaded from www.ccjm.org on July 20, 2021. For personal use only. All other uses require permission.
 MODI AND KOVACS

to presentation had increased odds of hospi-

TABLE 2
talization.
Clinical judgment should be applied to the Pneumonia Severity Index calculator
results of any of these calculators to appropri- and associated risk classes
ately triage patients with pneumonia.
Risk factor Points
■ DIAGNOSIS
OF COMMUNITY-ACQUIRED PNEUMONIA Demographics
Men Age (years)
Imaging Women Age (years) – 10
After triaging a patient with suspected CAP
Nursing home resident +10
to the safest level of care, several radiographic
and laboratory methods can be used to verify
Comorbidities
the diagnosis and identify the organism most
likely responsible for the ongoing infection. Neoplasm +30
Chest radiographs with demonstrable infil- Liver disease +20
trates are required to diagnose CAP and to Heart failure +10
distinguish it from upper respiratory tract in- Stroke +10
fection.1 Renal failure +10
Different organisms can be associated with
characteristic infiltrate patterns, which often Physical examination findings
manifest within 12 hours of symptom onset: Altered mental status +20
Focal nonsegmental or lobar pneumo-
Respiratory rate > 30 breaths per minute +20
nia (Figure 1). Typical bacterial pneumonias
caused by organisms such as Streptococcus Systolic blood pressure < 90 mm Hg +20
pneumoniae tend to manifest with an airspace Temperature < 95°F or > 104°F +15
opacity in 1 segment or lobe, though antibi- Heart rate > 125 beats per minute +10
otic use can alter their pathophysiology to cre-
ate a patchy, multilobular opacity pattern. Laboratory and radiographic findings
Multifocal bronchopneumonia or lobar Arterial pH < 7.35 +30
pneumonia. Bronchopneumonias, similarly Blood urea nitrogen > 30 mg/dL +20
characterized by a patchy pattern, are most Sodium < 130 mmol/L +20
commonly caused by Staphylococcus aureus, Glucose > 250 mg/dL +10
Haemophilus influenzae, and fungi.8 Hematocrit < 30% +10
Focal or diffuse “interstitial” pneumo-
Partial pressure of arterial oxygen < 60 mm Hg +10
nia (Figure 2). Atypical bacterial organisms
including Legionella pneumophila, Mycoplasma Pleural effusion +10
pneumoniae, and Chlamydophila pneumoniae
frequently involve the lung bases in a diffuse, Risk class Total points
bilateral, reticulonodular pattern, but can I < 51
start as isolated lobar opacities on chest radi- II 51–70
ography.9 Viral organisms are associated with III 71–90
diffuse, bilateral lung involvement as well. IV 91–130
Early radiographic identification of pulmo-
V > 130
nary complications, such as pleural effusions
From Fine MJ, Auble TE, Yealy DM, et al. A prediction rule to identify low-risk patients with com-
or cavitating lesions, can provide more clues munity-acquired pneumonia. N Engl J Med 1997; 336(4):243–250. Copyright 1997, Massachusetts
Medical Society. Reprinted with permission from the Massachusetts Medical Society..
to the causative organism and allow for timely
intervention.9
How accurate is chest radiography?
The utility of chest radiographs in diagnos- curacy in diagnosing viral pneumonia, 67% in
ing CAP is ultimately subject to interobserver diagnosing bacterial pneumonia, and no sta-
variability, with some studies citing 65% ac- tistical reliability for differentiating bacterial
CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 87 • NUMBER 3 MARCH 2020 147

Downloaded from www.ccjm.org on July 20, 2021. For personal use only. All other uses require permission.
 COMMUNITY-ACQUIRED PNEUMONIA

TABLE 3 TABLE 4
Severe pneumonia: Indications for blood culture
Infectious Diseases Society of America testing in suspected
and American Thoracic Society criteria community-acquired pneumonia
Major criteria Intensive care unit admission

Respiratory distress requiring mechanical ventilation Cavitary infiltrates

Septic shock Leukopenia

Minor criteria Active alcohol abuse

Confusion Chronic liver failure

Respiratory rate > 30 breaths per minute Asplenia (anatomic or functional)

Blood urea nitrogen > 7 mmol/L Positive pneumococcal urine antigen test

Leukopenia resulting from infection Pleural effusion

Thrombocytopenia Based on information in reference 1.

Hypothermia
Hypotension requiring aggressive fluids tibiotic regimens.1 Pretreatment Gram stain
and culture in patients able to adequately
PaO2 / FiO2 < 250 expectorate a good-quality specimen or endo-
Multilobar infiltrates tracheal aspirate in intubated patients should
be collected. Patients fulfilling criteria for se-
Having at least 1 major criterion or at least 3 minor criteria suggests vere pneumonia as defined by the IDSA/ATS
the need for intensive care. guidelines merit blood and sputum cultures as
well as urinary antigen tests for L pneumophila
From Mandell LA, Wunderink RG, Anzueto A, et al; Infectious Diseases Society of America; American
Thoracic Society. Infectious Diseases Society of America/American Thoracic Society consensus guide-
and S pneumoniae (Table 4).1
lines on the management of community-acquired pneumonia in adults. Clin Infect Dis 2007; 44(suppl Overall, active surveillance of more than
2):S27–S72, by permission of Oxford University Press.
2,200 patients with CAP requiring hospital-
ization noted that 38% of blood and sputum
cultures, nasopharyngeal and oropharyngeal
from nonbacterial pneumonias.10 A Swedish swabs, and urinary antigens yielded a causative
retrospective chart review of 103 outpatients organism.12 Viral organisms accounted for 25%
with suspected CAP noted that just 88% of of these cases and bacterial organisms account-
patients with high clinical concern for CAP ed for 14%; 5% of patients with viral pneumo-
demonstrated radiographic evidence of infec- nias were coinfected with either another respi-
tion.11 ratory virus or a bacterial organism.
Microbiology Procalcitonin testing
A thorough social history should be gathered Procalcitonin testing can help differentiate vi-
for every patient with suspected CAP to screen ral from bacterial pathogens in patients admit-
for potential occupational, travel, or endemic ted for CAP, preventing the use of unnecessary
exposures. This will guide microbiologic testing antibiotics and allowing prompt de-escalation
and empiric antibiotic treatment.1 For example, of empiric therapy more effectively than clini-
patients presenting during flu season or with cal judgment alone.13 While any infectious
known exposures to poultry in areas of prior in- pneumonia can precipitate elevations of this
fluenza outbreaks should be screened for influ- serum biomarker, typical bacteria tend to re-
enza A and B with a nasopharyngeal swab. sult in higher procalcitonin levels than atypi-
Isolating a specific organism in outpatients cal bacteria or viruses.14 Cytokines, associated
with CAP may not be necessary but is recom- with bacterial infections, enhance procalcito-
mended to guide de-escalation of empiric an- nin release, while interferons, associated with
148 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 87 • NUMBER 3 MARCH 2020

Downloaded from www.ccjm.org on July 20, 2021. For personal use only. All other uses require permission.
 MODI AND KOVACS

viral infections, inhibit procalcitonin release.

This biomarker is not perfect, however, and
will not be elevated in up to 23% of typical
bacterial infections.14
For this reason, procalcitonin should not
replace clinical judgment in guiding the deci-
sion to initiate antimicrobial therapy for pa-
tients with suspected CAP but can be used in
conjunction with clinical judgment to de-es-
calate therapy. In patients whose clinical his-
tories suggest alternative causes of respiratory
distress or improvement with concomitantly
administered therapies such as diuresis, a neg-
ative procalcitonin can help guide cessation
of antibiotics. On the other hand, in patients
with polymerase chain reaction-proven influ-
enza, an elevated procalcitonin can suggest
continuation of antibiotics to treat bacterial
superinfection.

■ MANAGEMENT
OF COMMUNITY-ACQUIRED PNEUMONIA
Antibiotic therapy
The selection of antibiotics before a causative
pathogen is identified should be informed by
the patient’s risk factors and degree of illness
(Table 5, Table 6).1
Patients on a medical floor should be
Figure 1. Focal lobar pneumonia.
started on either a respiratory fluoroquino-
lone or a combination of a beta-lactam plus
a macrolide; intensive care patients should
receive a beta-lactam plus either a macro-
lide or a respiratory fluoroquinolone. Doxy-
cycline can be used as an alternative to the
macrolide or respiratory fluoroquinolone to
cover atypical organisms such as Chlamydia
pneumoniae, Legionella pneumophila, and My-
coplasma pneumoniae in patients with pro-
longed QTc. In penicillin-allergic patients,
aztreonam should be used in combination
with an aminoglycoside and a respiratory
fluoroquinolone.
Patients who may have been exposed to in-
fluenza or who have a history of injection drug
use or structural lung disease or who have a
lung abscess, cavitary infiltrates, or endobron-
chial obstruction also merit coverage against
community-acquired methicillin-resistant S
aureus (MRSA) with vancomycin or linezolid.
Those with confirmed or suspected influenza Fgure 2. Diffuse interstitial pneumonia.
A presenting within 48 hours of symptom
CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 87 • NUMBER 3 MARCH 2020 149

Downloaded from www.ccjm.org on July 20, 2021. For personal use only. All other uses require permission.
 COMMUNITY-ACQUIRED PNEUMONIA

TABLE 5 TABLE 6

Common organisms Initial antibiotic therapy

in community-acquired pneumonia for community-acquired pneumonia
Outpatient care Outpatients without comorbiditiesa
Amoxicillin
Streptococcus pneumoniae Or doxycycline
Mycoplasma pneumoniae Or a macrolide

Haemophilus influenzae Outpatients with comorbidities

Combination therapy:
Chlamydophila pneumoniae Amoxicillin/clavulanate or a cephalosporin
Plus a macrolide or doxycycline
Respiratory virus (influenza A and B, adenovirus,
Or monotherapy with a fluoroquinolone
respiratory syncytial virus, parainfluenza)
Patients on a medical floor
Inpatient (non-intensive care) A fluoroquinolone
S pneumoniae Or a combination of a beta-lactam plus a macrolide
M pneumoniae Intensive care patients
A beta-lactam
C pneumoniae Plus either a macrolide or a fluoroquinolone
H influenzae Add coverage as needed for:
Legionella species Methicillin-resistant Staphylococcus aureus (MRSA)
Pseudomonas aeruginosa
Aspiration-related oral flora Influenza A
Respiratory viruses a
Comorbidities include heart, lung, liver, or renal disease,
diabetes mellitus, alcoholism, malignancy, and asplenia
Inpatient (intensive care)
Use clinical S pneumoniae
Based on information from reference 1.

judgment Staphylococcus aureus reliable negative predictive value and can also
when triaging Legionella species be used to de-escalate empiric antimicrobial
therapy.16
patients with Gram-negative bacilli Should microbiologic evaluation fail to
community- H influenzae identify a causative organism, the patient’s
acquired individual risk factors as listed above must be
From Mandell LA, Wunderink RG, Anzueto A, et al; Infectious Diseases
Society of America; American Thoracic Society. Infectious Diseases Soci- considered in de-escalating therapy to a final
pneumonia ety of America/American Thoracic Society consensus guidelines on the
regimen with coverage for MRSA, Pseudomo-
management of community-acquired pneumonia in adults. Clin Infect
Dis 2007; 44(suppl 2):S27–S72, by permission of Oxford University Press. nas aeruginosa, or atypical pathogens as indi-
cated. Pseudomonal pneumonia has been asso-
ciated with higher risk of mortality and relapse
onset or with severe illness should be treated than pneumonia caused by other pathogens.
with oseltamivir.1
Corticosteroids as adjunctive therapy
If an organism is identified by culture,
The use of adjunctive corticosteroids for CAP
polymerase chain reaction, or serology, the management has been widely contested. The
empiric antibiotic regimen should be tailored IDSA/ATS guidelines recommend against cor-
to this organism. MRSA nares screening can ticosteroid use for adjunctive treatment of CAP
be reliably used to guide empiric and targeted except in patients with refractory septic shock.1
antimicrobial regimens; patients started on
vancomycin or linezolid based on the above- Later management
stated risk factors can be safely de-escalated Patients who are hemodynamically stable,
on the basis of a negative nasal swab.15 The can ingest medications safely, and have a nor-
pneumococcal urinary antigen has a similarly mal gastrointestinal tract can be discharged
150 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 87 • NUMBER 3 MARCH 2020

Downloaded from www.ccjm.org on July 20, 2021. For personal use only. All other uses require permission.
 MODI AND KOVACS

on oral therapy without waiting to observe atypical pathogens which may not be covered
the clinical response. Antibiotics should be by the ongoing antibiotic regimen. Pulmonol-
given for at least 5 days, though longer dura- ogy consultation is also indicated for patients
tions may be needed in immunocompromised with complications of pneumonia such as em-
patients or in those with pulmonary or extra- pyema that require procedural intervention.
pulmonary complications.1
An infectious disease consultation may be ■ TAKE-HOME POINTS
beneficial if long-term intravenous antibiotic
therapy is anticipated or if the patient progres- • CAP continues to contribute to patient
sively deteriorates on guideline-based antimi- morbidity and mortality as well as health-
crobial therapy. care costs.
Pulmonary consultation may be needed • Professional societies have released col-
for bronchoscopy to obtain deep respiratory laborative guidelines to streamline prac-
samples, especially if the patient is clinically tice patterns and provide evidence-based
worsening and the causative pathogen re- protocols for the diagnosis, treatment, and
mains unidentified. We acknowledge that the prevention of this common infection.
yield of bronchoscopy and bronchoalveolar • Further research is needed to delineate ap-
lavage samples is reduced with longer dura- propriate strategies to de-escalate antibiot-
tions of antibiotic therapy, yet believe that in ics in the absence of a causative organism,
the context of clinical worsening in spite of define the dose and duration of adjunctive
antibiotics, bronchoalveolar lavage may help steroid use, and clarify patient follow-up
successfully identify multidrug-resistant or after discharge from the hospital. ■

■ REFERENCES 9. Vilar J, Domingo ML, Soto C, Cogollos J. Radiology of

bacterial pneumonia. Eur J Radiol 2004; 51(2):102–113.
1. Metlay JP, Waterer GW, Long AC, et al. Diagnosis and doi:10.1016/j.ejrad.2004.03.010
treatment of adults with community-acquired pneumo- 10. Katz DS, Leung AN. Radiology of pneumonia. Clin Chest
nia. An official clinical practice guideline of the Ameri- Med 1999; 20(3):549–562. pmid:10516903
can Thoracic Society and Infectious Diseases Society of 11. Moberg AB, Taléus U, Garvin P, Fransson SG, Falk M.
America. Am J Respir Crit Care Med 2019; 200(7):e45– Community-acquired pneumonia in primary care: clinical
e67. doi:10.1164/rccm.201908-1581ST assessment and the usability of chest radiography. Scand
2. Jackson ML, Neuzil KM, Thompson WW, et al. The J Prim Health Care 2016; 34(1):21–27.
burden of community-acquired pneumonia in seniors: doi:10.3109/02813432.2015.1132889
results of a population-based study. Clin Infect Dis 2004; 12. Jain S, Self WH, Wunderink RG, et al; CDC EPIC Study
39(11):1642–1650. doi:10.1086/425615 Team. Community-acquired pneumonia requiring
3. Pfuntner A, Wier LM, Steiner C. Costs for hospital stays hospitalization among US adults. N Engl J Med 2015;
in the United States, 2011. HCUP Statistical Brief #168. 373(5):415–427. doi:10.1056/NEJMoa1500245
December 2013. Agency for Healthcare Research and 13. Schuetz P, Christ-Crain M, Thomann R, et al; ProHOSP Study
Quality, Rockville, MD. www.hcup-us.ahrq.gov/reports/ Group. Effect of procalcitonin-based guidelines vs standard
statbriefs/sb168-Hospital-Costs-United-States-2011.pdf. guidelines on antibiotic use in lower respiratory tract infec-
Accessed February 6, 2020. tions: the ProHOSP randomized controlled trial. JAMA 2009;
4. Centers for Disease Control and Prevention (CDC). Na- 302(10):1059–1066. doi:10.1001/jama.2009.1297
tional Center for Health Statistics. Pneumonia. cdc.gov/ 14. Self WH, Balk RA, Grijalva CG, et al. Procalcitonin as a
nchs/fastats/pneumonia.htm. Accessed February 6, 2020. marker of etiology in adults hospitalized with communi-
5. Fine MJ, Auble TE, Yealy DM, et al. A prediction rule ty-acquired pneumonia. Clin Infect Dis 2017; 65(2):183–
to identify low-risk patients with community-acquired 190. doi:10.1093/cid/cix317
pneumonia. N Engl J Med 1997; 336(4):243–250. 15. Parente DM, Cunha CB, Mylonakis E, Timbrook TT. The
doi:10.1056/NEJM199701233360402 clinical utility of methicillin-resistant Staphylococcus
6. Mortensen EM, Coley CM, Singer DE, et al. Causes of aureus (MRSA) nasal screening to rule out MRSA pneu-
death for patients with community-acquired pneu- monia: a diagnostic meta-analysis with antimicrobial
monia: results from the Pneumonia Patient Outcomes stewardship implications. Clin Infect Dis 2018; 67(1):1–7.
Research Team cohort study. Arch Intern Med 2002; doi:10.1093/cid/ciy024
162(9):1059–1064. doi:10.1001/archinte.162.9.1059 16. West DM, McCauley LM, Sorensen JS, Jephson AR,
7. Labarere J, Stone RA, Scott Obrosky D, et al. Factors Dean NC. Pneumococcal urinary antigen test use in
associated with the hospitalization of low-risk patients diagnosis and treatment of pneumonia in seven Utah
with community-acquired pneumonia in a cluster- hospitals. ERJ Open Res 2016; 2(4). pii:00011-2016.
randomized trial. J Gen Intern Med 2006; 21(7):745–752. doi:10.1183/23120541.00011-2016
doi:10.1111/j.1525-1497.2006.00510.x
8. Franqut T. Imaging of community-acquired pneumonia. Address: Anita R. Modi, MD, Department of Infectious Dis-
J Thorac Imaging 2018; 33(5):282–294. ease, G21, Cleveland Clinic, 9500 Euclid Avenue, Cleveland, OH
doi:10.1097/RTI.0000000000000347 44106; [email protected]

CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 87 • NUMBER 3 MARCH 2020 151

Downloaded from www.ccjm.org on July 20, 2021. For personal use only. All other uses require permission.