2025 ATS临床实践指南：社区获得性肺炎的诊断和管理

指南呼吸内科· 肺炎与呼吸道感染 · 2025

Table 1. Individual patient factors to consider that may strengthen or weaken recommendations.

Factors listed in this table were generated from clinical experience, observational studies, and pathophysiologic rationale, but are not supported by high-quality comparative evidence and should be integrated with clinical judgment for individual patient care.

Recommendations are not for patients with immunocompromise. See other guidelines focused on this patient population.

|Recommendation|Strength & Evidence Quality|Factors that strengthen the recommendation|Factors that weaken the recommendation| |---|---|---|---| |1. Lung ultrasound versus chest x-ray to diagnose CAP.|||| |For adults with suspected CAP, we suggest lung ultrasound is an acceptable diagnostic alternative to chest x-ray in medical centers where appropriate clinical expertise exists|Conditional Low-quality evidence|All criteria for establishing expertise met (Table 4) No availability of chest X-Ray (LUS as alternative to X-Ray) High patient risks or cost of CT scan (LUS as alternative to CT) Patient convenience and radiation exposure compared to chest x-ray & CT.|Not all criteria for establishing expertise met (Table 4) Suspicion of alternative/additional diagnoses (pulmonary embolism, malignancy) Barriers to high-quality LUS (Obesity, drains, scars, wounds, difficulty holding position)| |2. Empiric antibacterial therapy for CAP with positive respiratory virus testing|||| |For adult outpatients without co-morbidities who have clinical and imaging evidence of CAP and who test positive for a respiratory virus, we suggest not prescribing empiric antibiotics|Conditional Quality evidence|Low suspicion for bacterial co-infection (clinical history, low/normal inflammatory markers, clinical history, radiologic findings suggestive of viral etiology, viral pathogen with low prevalence of bacterial co-detection) Higher risk of harm from antibiotic exposure (History of C. difficile, severe antibiotic allergy or adverse event) Patient preference to avoid antibiotic exposure|Suspicion of bacterial co-infection (long symptom onset, "double sickening", purulent sputum, elevated inflammatory markers, radiologic findings such as consolidative infiltrate, viral pathogen with high prevalence of bacterial co-detection, exposure to Mycoplasma pneumoniae) High risk of harm if missed bacterial infection (elderly, pregnant, signs/symptoms suggestive of more severe illness) Barriers to follow-up or communication|

2. Empiric antibacterial therapy for CAP with positive respiratory virus testing

|For adult outpatients without co-morbidities who have clinical and imaging evidence of CAP and who test positive for a respiratory virus, we suggest not prescribing empiric antibiotics|Conditional Very low-quality evidence|Low suspicion for bacterial co-infection (clinical history, low/normal inflammatory markers, clinical history, radiologic findings suggestive of viral etiology, viral pathogen with low prevalence of bacterial co-detection) Higher risk of harm from antibiotic exposure (History of C. difficile, severe antibiotic allergy or adverse event) Patient preference to avoid antibiotic exposure|Suspicion of bacterial co-infection (long symptom onset, "double sickening", purulent sputum, elevated inflammatory markers, radiologic findings such as consolidative infiltrate, viral pathogen with high prevalence of bacterial co-detection, exposure to Mycoplasma pneumoniae) High risk of harm if missed bacterial infection (elderly, pregnant, signs/symptoms suggestive of more severe illness) Barriers to follow-up or communication| |---|---|---|---|

Page 58 of 119

|For adult outpatients with co-morbidities who have clinical and imaging evidence of CAP and who test positive for a respiratory virus, we suggest prescribing empiric antibiotics due to concern for bacterial-viral co-infection|Conditional Very low-quality evidence|Suspicion of bacterial co-infection (long symptom onset, "double sickening", purulent sputum, elevated or increasing inflammatory markers, radiologic findings such as consolidative infiltrate) Low likelihood that virus identified explains etiology and severity of pneumonia (ie, virus with low virulence or high-risk of co-infection) High risk of harm if missed bacterial infection -High illness severity, severe symptoms -Higher number, severe, or poorly controlled comorbidities|Low suspicion of bacterial infection (clinical history, normal inflammatory markers, radiologic findings suggestive of viral etiology) High likelihood that virus identified explains etiology and severity of pneumonia (virus with high virulence, low risk of co-infection)) Lower risk of harm if missed bacterial infection -Lower illness severity -Single, mild, or well controlled comorbidities Higher risk of harm from antibiotic exposure (History of C. difficile, antibiotic allergy/ adverse event) - Patient preference to avoid antibiotic exposure| |---|---|---|---| |For adult inpatients with clinical and imaging evidence of non-severe CAP who test positive for a respiratory virus, we suggest prescribing empiric antibiotics due to concern for bacterial-viral co-infection|Conditional Very low-quality evidence|**Same as above**|**Same as above**| |For adult inpatients with clinical and imaging evidence of severe CAP who test positive for a respiratory virus, we suggest prescribing antibiotics due to concern for bacterial-viral co-infection|Conditional Very low-quality evidence|Sepsis, severe respiratory failure, elevated or increasing inflammatory markers, Chest radiograph showing consolidation infiltrates.|Higher risk of harm from antibiotic exposure (History of C. difficile, antibiotic allergy, or antibiotic adverse event)| |3. Antibiotic duration for CAP|||| |For adult outpatients with CAP who reach clinical stability*, we suggest less than five days of antibiotics (minimum of 3 days duration), rather than five or more days of antibiotics. The duration of antibiotics should be determined based upon daily assessment of clinical stability.|Conditional Low-quality evidence|Higher risk of harm from prolonged antibiotic exposure (History of C. difficile, or an antibiotic adverse event) Patient preference to minimize antibiotic exposure|Barriers to self assessment, follow-up, or communication to ensure recovery Organism requiring longer duration (i.e., Staphylococcus aureus, Pseudomonas aeruginosa, suspected Legionella pneumophila or other intracellular microorganisms)**|

Page 59 of 119

||||Radiographic findings (high burden of disease, necrotizing process, dense consolidations) Underlying lung disease (e.g., bronchiectasis, post-obstructive pneumonia, chronic respiratory insufficiency**) Recent Hospitalization or resident in Long-term care**| |---|---|---|---| |For adult inpatients with non-severe CAP who reach clinical stability*, we suggest less than five days of antibiotics (minimum of 3 days duration), rather than five or more days of antibiotics. *The duration of antibiotics should be determined based upon daily assessment of clinical stability.|Conditional Low-quality evidence|Patient preference to minimize antibiotic exposure Resolution of inflammatory markers|Organism requiring longer duration (i.e., Staphylococcus aureus, Pseudomonas aeruginosa, suspected Legionella pneumophila or other intracellular microorganisms)** Pneumonia complication (e.g., empyema/parapneumonic effusion, abscess/necrotizing process, bacteremia, extrapulmonary infection) Underlying lung disease (e.g., bronchiectasis, post-obstructive pneumonia, chronic hypoxemia**) Pregnancy, recent antibiotics** Recent Hospitalization or resident in Long-term care**| |For adult inpatients with severe CAP who reach clinical stability, we suggest five or more days of antibiotics, rather than less than five days of antibiotics|Strong Low-quality evidence||| |4. Corticosteroids|4. Corticosteroids|4. Corticosteroids|4. Corticosteroids| |For adult inpatients with non-severe CAP, we recommend NOT administering systemic corticosteroids ***|Strong Low-quality evidence||| |For adult inpatients with severe CAP, we suggest systemic corticosteroids|Conditional Low-quality evidence|Short time interval between symptom onset and presentation.|Longer time between symptom onset and presentation|

Page 60 of 119

|||Ability to administer corticosteroids early after meeting criteria for severe CAP (within 24 hours)ICU-admittedRespiratory failure (P:F<300)Elevated inflammatory markers (ie, CRP, IL-6)|Longer time since onset of severe CAP (e.g. >72 hours)Lack of respiratory failureNormal or low inflammatory markersContraindications to corticosteroids (ie, influenza, Aspergillus, uncontrolled diabetes, recent gastrointestinal bleeding)Pregnancy**| |---|---|---|---| |**Exclusion criteria from key studies***Recommendation and factors listed are for patients without another established indication of corticosteroids.||||

|Validated definition includes either one major criterion or three or more minor criteria.| |---| |Major criteria| |Septic shock with need for vasopressors| |Respiratory failure requiring mechanical ventilation| |Minor criteria| |Respiratory rate ≥ 30 breaths/min| |$Pa_{02}$/$Fi_{02}$ ratio ≤ 250| |Multilobar infiltrates| |Confusion/disorientation| |Uremia (blood urea nitrogen level ≥ 20 mg/dl)| |Leukopenia (white blood cell count < 4,000 cells/µl)| |Thrombocytopenia (platelet count < 100,000/µl)| |Hypothermia (core temperature < 36°C)| |Hypotension requiring aggressive fluid resuscitation|

Page 61 of 119

|Table 3. Strengths of Recommendation||| |---|---|---| ||Strong Recommendation ("We recommend . . .")|Conditional Recommendation ("We suggest . . .")| |For patients|The overwhelming majority of individuals in this situation would want the recommended course of action and only a small minority would not.|The majority individuals in this situation would want the suggested course of action, but a sizeable minority would not.| |For clinicians|The overwhelming majority of individuals should receive the recommended course of action. Adherence to this recommendation according to the guideline could be used as a quality criterion or performance indicator. Formal decision aids are not likely to be needed to help individuals make decisions consistent with patient values and preferences.|Different choices will be appropriate for different patients, and you must help each patient arrive at a management decision consistent with her or his values and preferences. Decision aids may be useful to help individuals make decisions consistent with their values and preferences. Clinicians should expect to spend more time with patients when working towards a decision.| |For policy makers|The recommendation can be adapted as policy in most situations including for the use as performance indicators if supported by high- or moderate-quality evidence.|Policymaking will require substantial debates and involvement of many stakeholders. Policies are also more likely to vary between regions. Performance indicators would have to focus on the fact that adequate deliberation about the management options has taken place.|

|Table 4. Key Criteria for Establishing Expertise in Lung Ultrasound Exams|| |---|---| |Factor|Requirements| |Ultrasound Equipment|Either a cart-based or handheld ultrasound device with a low-frequency ultrasound probe that provides adequate penetration, typically 14-16cm in adults, is needed to assess for pneumonia.| |Training|Requisite training in LUS must provide background knowledge; practice in image acquisition, optimization, and interpretation; and knowledge in clinical integration. Mastery of LUS knowledge and skills through formal assessments should be demonstrated prior to use in clinical practice as recommended by specialty guidelines. [140-143]| |Imaging Protocol|A standardized protocol evaluating the superior and inferior portions of the anterior, lateral, and posterior chest wall should be utilized.[32]| |Image Archival|Dynamic ultrasound images, typically 2-4 second video loops, should be recorded, labelled per local convention, and saved in a retrievable image archival.| |Documentation|Documentation of the operator, indications, exam performed, and ultrasound findings of the pleura and lung parenchyma - including location of abnormalities using standard terminology - should be included as a report within the patient's medical record.| ||Findings from different imaging modalities shall be compared and periodic quality assurance checks of clinicians using LUS should be performed at the same level of radiology images. Discrepancies of imaging findings associated with negative outcomes should be reviewed for quality improvement.| |Patient|Patient factors that limit LUS imaging including obesity, drains, scars, wounds, and uncooperativeness, should be considered when choosing imaging modality.|

Page 62 of 119

|Table 5. Comorbidities that may warrant antibiotic therapy for outpatients with CAP who test positive for a respiratory virus. Conditions are ranked by % of committee members who would prescribe antibiotics for patients with each condition, in descending order.|| |---|---| |Comorbidity (see footnotes for further definitions and examples)|% committee members who voted this condition that may warrant antibiotics| |Greater than 50% agreement|| |Chronic pulmonary disease other than asthma|82%| |End-stage liver disease|71%| |End-stage renal disease|65%| |Cardiovascular disease|53%| |Alcoholism|53%| |Neoplastic disease|53%| |Less than 50% agreement|| |Neurological disease|47%| |Chronic liver disease|35%| |Malnutrition|35%| |Current smoker|35%| |Corticosteroid therapy* (<20mg daily or <4 weeks)|30%| |Diabetes mellitus|29%| |Chronic kidney disease|24%| |Human immunodeficiency virus* (CD4>200)|24%| |Asthma|21%| |Rheumatological diseases* (not on immunosuppressants)|18%| |Obesity (BMI>30)|12%| |*Patients with solid organ transplant on anti-rejection medications, corticosteroid therapy more than 20mg/day for 4wks, human insufficiency virus with CD4 count <200, or rheumatological diseases on immunocompromising medication should be considered immunocompromised hosts, to whom the CAP guidelines do not apply. Refer to References 6 and 7 for guidance on diagnosis and management of these patients.|| |Chronic pulmonary diseases other than asthma= chronic obstructive pulmonary disease [COPD], bronchiectasis or interstitial lung disease. End-stage liver disease = ascites, variceal hemorrhage, hepatic encephalopathy, or renal impairment. End-stage renal disease =GFR <15 mL/min lasting >3 months. Solid organ transplant recipient=not taking immunosuppressive anti-rejection medication. Cardiovascular disease =congestive heart failure, coronary artery disease, or poorly controlled hypertension. Alcoholism = recurrent or ongoing alcohol use despite inability to fulfill obligations, or despite social or interpersonal problems exacerbated by alcohol use. Neoplastic disease = not taking immunosuppressive chemotherapy. Neurological disease such as Parkinson's Disease, dementia, myasthenia gravis or amyotrophic lateral sclerosis. Chronic liver disease defined as abnormal liver function tests, coagulopathy, or other evidence of chronic liver damage lasting >3 months. Malnutrition = weight loss, BMI <18.5, reduced muscle mass, or reduced food intake or assimilation. Current smoker, including cigarettes and marijuana. Corticosteroid therapy not at immunosuppressive doses such as a cumulative dose >600 mg of prednisone. Chronic kidney disease defined as GFR 15-60 mL/min, albuminuria >30 mg / 24 hours, or other markers of kidney damage lasting >3 months. Human immunodeficiency virus (HIV) with CD4 >200 and no AIDS defining illness. Rheumatological diseases such as rheumatoid arthritis or systemic lupus erythematosus, not taking immunosuppressive medication. Obesity defined as BMI >30.||

Page 63 of 119

|Table 6. Clinical Stability Definitions* The duration of antibiotics should be determined based upon daily assessment of clinical responses.|| |---|---| |Temperature|≤37.8° C| |Heart rate|<100 beats per minute*| |Respiratory rate|<24 breaths per minute*| |Arterial oxygen saturation or partial pressure|SpO2 ≥90% or PaO2 ≥60 mmHg on room air* or baseline oxygen requirement*| |Systolic blood pressure|≥90 mmHg| |Mental status|Normal| |*All criteria were needed to be met to be considered “stable” in the Dinh study [89]. Prior 2007 guidelines and Uranga study [88] required afebrile plus no more than one sign of instability and used heart rate of ≤100 beats per minute, and respiratory rate of ≤24 breaths per minute. For el Moussaoui et al study, eligibility for 3-day duration was determined by improvement of 2 or more points on a respiratory symptom scale, temperature < 38°C and ability to take oral intake. **Neither Dinh nor Moussaoi studies included patients with chronic respiratory insufficiency. Thus this factor weakens this recommendation (See Table 1)||

Online Supplement

Diagnosis and Management of Community-acquired Pneumonia An Official American Thoracic Society Clinical Practice Guideline

Author list:

Barbara E Jones, MD MS Division of Pulmonary & Critical Care Medicine, University of Utah and Veterans Affairs Healthcare System, Salt Lake City UT. ORCID#: 0000-0002-0971-6887

Julio A Ramirez, MD, Norton Infectious Diseases Institute, Norton Healthcare and University of Louisville, Louisville, Kentucky. ORCID#: 0000-0001-7801-5374

Eyal Oren, PhD, MS, School of Public Health, San Diego State University, San Diego CA. ORCID#:

0000-0001-7817-3516

Nilam J. Soni, MD MS, Division of Pulmonary Diseases & Critical Care and Division of Hospital

Medicine, University of Texas Health San Antonio, San Antonio, Texas. ORCID#: 0000-0002-

8460-0323

Liam R. Sullivan, DO. Division of Infectious Disease, Corewell Health West, Grand Rapids, MI. ORCID #0009-0007-9035-5145

Marcos I. Restrepo, MD, MSc, PhD, South Texas Veterans Health Care System and University of Texas

Health San Antonio, San Antonio, TX. ORCID # 0000-0001-9107-3405

Daniel M Musher, MD, M. E. DeBakey Veterans Affairs Medical Center and Baylor College of Medicine,

Houston, TX, ORCID ID 0000-0002-7571-066X

Brian L. Erstad, PharmD, Department of Pharmacy Practice & Science, R. Ken Coit College of

Pharmacy, University of Arizona, Tucson, AZ, USA ORCID: 0000-0001-8909-9921

Chiagozie Pickens, MD MS Division of Pulmonary and Critical Care Medicine, Northwestern

University Feinberg School of Medicine, Chicago IL. ORCID ID #0000-0002-3400-9228.

Valerie M. Vaughn MD, MSc. Division of General Internal Medicine, University of Utah and Veterans

Affairs Healthcare System, Salt Lake City UT. ORCID#: 0000-0003-4362-7842

Scott A. Helgeson, MD, MS, Division of Pulmonary and Critical Care Medicine, Department of

Medicine, Mayo Clinic, Jacksonville, FL, USA, ORCID: 0000-0001-7590-2293

Kristina Crothers, MD, Veterans Affairs Puget Sound Healthcare System and Division of Pulmonary,

Critical Care and Sleep, Department of Medicine, University of Washington, Seattle, WA.

ORCID#: 000-001-9702-0371

Page 64 of 119

Joshua P. Metlay, MD, PhD. Division of General Internal Medicine, Massachusetts General Hospital and Harvard Medical School, Boston MA. ORCID: 0000-0003-2259-6282

Brittany D. Bissell Turpin, PharmD, PhD, Department of Pharmacy, Ephraim McDowell Regional Medical Center, Danville, KY. ORCID 0000-0002-7345-9731

Bin Cao, China-Japan Friendship Hospital; National Center for Respiratory Medicine, Beijing, PR China.

ORCID#: 0000-0001-6991-0350 Leila S Hojat, MD, MS; Case Western Reserve University, Division of Infectious Diseases and HIV.University Hospitals Cleveland Medical Center, Cleveland, OH;

James D Chalmers, MBChB, PhD, FRCPE, FERS, Respiratory Research School of Medicine, University of Dundee ORCID- 0000-0001-5514-7868

Charles S. Dela Cruz, MD, PhD, Division of Pulmonary, Allergy, Critical Care and Sleep Medicine,

University of Pittsburgh and Veterans Affairs Healthcare System, Pittsburgh, PA. ORCID# 0000-0002-5258-1797

Inessa Gendlina, MD, PhD, Division of Infectious Diseases, Department of Medicine, Montefiore

Medical Center, Albert Einstein College of Medicine, Bronx, NY.

Leila S Hojat, MD, MS; Case Western Reserve University, Division of Infectious Diseases and HIV.

University Hospitals Cleveland Medical Center, Cleveland, OH; ORCID# 0003-4364-8141

Maryrose Laguio-Vila, MD. Department of Medicine, Division of Infectious Diseases, Rochester Regional Health, Rochester, New York. ORCID#: 0000-0001-7768-2052

Stephen Y. Liang, MD, MPHS, Division of Infectious Diseases, Department of Medicine & Department of Emergency Medicine, Washington University School of Medicine, St. Louis, MO, USA.ORCID#:

0000-0002-4652-2272

Grant W. Waterer MBBS PhD, School of Medicine, University of Western Australia, ORCID#

0000-0002-7222-8018

Marilynn Paine, MPH, University Healthcare, University of Utah.

Conall Hawkins, PhD, College of Medicine, Nursing & Health Sciences, University of Galway.

Kevin Wilson, MD, Division of Allergy, Pulmonary, Critical Care, and Sleep Medicine, Department of Medicine, Boston University School of Medicine, Boston, MA, USA, ORCID: 0000-0003-4429-2263.

… on behalf of the American Thoracic Society Assembly on Pulmonary Infections and Tuberculosis.

Page 65 of 119

TABLE OF CONTENTS

Methods

Panel composition

Questions

Literature search

Evidence synthesis

Recommendations

Implications of strength of recommendations

Manuscript preparation

Peer review and approval

Updating

Funding

PICO Question #1: Lung ultrasound versus chest x-ray to diagnose CAP

Search Strategy

Flow of Information

Studies Selected

Forest Plots

Evidence Profiles

Evidence to Decision Framework

PICO Question #2: Antibacterial therapy for CAP if a test for a respiratory virus is positive Search Strategy

Flow of Information

PICO Question #3: Antibiotic duration for CAP

Search Strategy

Flow of Information

Studies Selected

Forest Plots

Evidence Profiles

PICO Question #4: Systemic corticosteroids for CAP

Search Strategy

Flow of Information

Studies Selected

Forest Plots

Evidence Profiles

Evidence to Decision Framework

Page 66 of 119

METHODS

Panel Composition

The project was proposed by one of the co-chairs (BJ) through an application to the American Thoracic Society (ATS), which subsequently invited the Infectious Diseases Society of America (IDSA) to

collaborate. The project began January 1, 2022. Cochairs BJ and JR proposed panelists based upon their expertise in the diagnosis and management of community-acquired pneumonia (CAP) and KCW

assembled a methodology team from the ATS’ Guideline Methodology Training Program (EO, BB, SL, BE). The committee was diverse with respect to gender, specialties (pulmonology, infectious disease, internal medicine, critical care, hospital medicine, emergency medicine, and evidence synthesis), level of

seniority, and geographical locations. The appointed representatives from ATS and IDSA were approved by the leadership of those societies. All panelists disclosed their conflicts of interest, which were vetted and managed according to the policies and procedures of the ATS and IDSA.

Questions

The co-chairs drafted key questions pertaining to treatment interventions in a PICO (Population,

Intervention, Comparator, and Outcome) format. The questions were discussed, revised, and finally

approved by the full committee at a virtual meeting Fall 2022. Four PICO questions were agreed upon. For each PICO question, critical and important outcomes were predetermined. An overview document was created to clarify inclusion and exclusion criteria, participants, and a priori subgroup analyses.

Literature search

The published literature was searched by a health librarian (MH) as well as reviewed by the lead

methodologist (EO) in a number of databases, including Medline/PubMed, Excerpta Medica Database

(EMBASE), and Cochrane Database of Systematic Reviews. Searching was conducted Winter and Spring

2023. The methodology team reviewed all publications retrieved from the literature searches, initially

screening based on title and/or abstract and then reviewing the full text of potentially relevant

publications. Bibliographies of selected studies, relevant systematic reviews, and articles suggested by

committee members were also reviewed. All screened article meta-fields were input into Rayyan.AI,

which was used to document and track included and excluded articles for full-text review. Randomized trials that compared performing the treatment of interest to not performing the treatment were sought first. If randomized trials were not identified, non-randomized studies that compared performing a

treatment to not performing the treatment were sought. If such studies were not found, nonrandomized studies without a control group were sought. If no direct evidence was found, indirect

evidence (e.g. population, intervention) was sought based on initial expert discussion. For one of the

PICO questions, this resulted in a shift to diagnostic interventions, with inclusion of accuracy studies that determined sensitivity and specificity of the diagnostic test.

Evidence synthesis

Findings from selected publications were extracted into an Excel spreadsheet created specifically for the project. When data were amenable to weighted pooling (i.e., meta-analysis), a random effects model

was implemented in the Cochrane Collaboration Review Manager (RevMan), version 5.4. For controlled studies, relative risk (RR) was used to report dichotomous outcomes and the mean difference (MD) was used to report continuous outcomes. The accompanying 95% confidence interval (CI) was determined.

Page 67 of 119

Statistical heterogeneity was measured using the I2 test; an I2 ≥75%, 50-75%, and 25-50% was

considered severe, moderate, and mild, respectively. Whenever heterogeneity was encountered,

sensitivity analyses were performed to identify contributing studies, reasons for the heterogeneity

sought, and subgroups analyzed. If no cause was found, we eliminated outliers and the estimates before and after elimination of outliers were both presented to the committee to inform their discussion and

judgements. Results are provided in the evidence tables. For diagnostic comparisons, a summary

receiver operator curve was constructed. The area under the curve was calculated and a bivariate model was used to find a single best estimate of sensitivity and specificity.

The Grading, Recommendations, Assessment, Development, and Evaluation (GRADE) approach was used to assess certainty in the estimated effects (i.e., the quality of evidence) for each intervention on each

outcome of interest. The methodology team created evidence profiles, which categorized the overall

certainty in the evidence into one of four levels: high, moderate, low, or very low. Each level represents the certainty in the accuracy of the estimated effects for a specific intervention. The full guideline panel reviewed the evidence profiles and provided input and feedback.

Recommendations

The methodology team distributed the completed evidence syntheses to the guideline committee by

email two weeks prior to the face-to-face meeting at the ATS conference in Washington DC, May 2023. The methodology team presented the evidence syntheses at the meeting, which were then discussed by the committee and recommendations were formulated. Decisions about whether to recommend for or against an intervention were based on the balance of desirable consequences (benefits) and undesirable consequences (burdens, adverse effects, and costs), quality of evidence, feasibility, and acceptability to patients (i.e., patient values and preferences). Guideline committee members were encouraged to

consider their non-systematic clinical observations (i.e., clinical experience) when the quality of

empirical evidence was very low.

To facilitate consensus, each recommendation was voted on using the Convergence of Opinion on

Recommendations and Evidence (CORE) approach, a modified Delphi process. Voting

percentages were calculated and rounded to the nearest multiple of five. It was decided a priori

that 1) 80% committee participation was necessary, 2) 70% agreement was necessary to make a

recommendation, 3) the strength of the recommendation would be determined by the majority among those in agreement, and 4) only those who were present for the evidence presentation could vote on

the recommendation. The methodology team and patient representatives were not voting members of guideline committee.

Evidence to Data (EtD) tables were constructed for each PICO question summarizing recommendations and providing an overview of the process.

Implications of the strength of recommendations

The strength of recommendations can be conceptualized in several ways. First, “we recommend”

conveys that the recommended course of action is the appropriate in >95% of patients, whereas a “we suggest” conveys that the recommended course of action is appropriate in >50% of patients but may not be appropriate in a sizeable minority. Second, “we recommend” conveys “just do it”, whereas ”we

suggest” conveys “slow down, think about it, discuss it”. Third, a ”we recommend” conveys that

Page 68 of 119

criticism may be warranted if the recommended course of action is not followed, whereas ”we suggest” conveys that a decision to not follow the recommended course of action may be a matter of style or

equipoise. Finally, ”we recommend” is often the basis of a performance measure, whereas ”we suggest” seldom make reasonable performance measures.

Manuscript preparation

The introduction and outline were written by the co-chairs (BJ, JR). Guideline committee members were assigned to subcommittees to create sections of the manuscript. The sections were collated and edited into a single manuscript by BJ and JR. All members of the guideline committee reviewed the manuscript; comments were addressed by the co-chairs and then incorporated into the revised manuscript. The

manuscript was redistributed to the full committee for further review. The final product was the result of collective work from the co-chairs, committee members, methodologists, and health librarian. Once the manuscript was approved by the full guideline committee, it was submitted for external peer review.

Peer review and approval

Peer review was overseen by the ATS Associate Documents Editor. The guideline was reviewed

independently by each co-sponsoring society. This included anonymous peer review by both content

experts and guideline methodology experts. Following multiple cycles of review and revision, the

guideline was reviewed and approved by the ATS Board of Directors. The IDSA chose to withdraw rather than approve the final version of the guideline.

Updating

The guideline will be reviewed by the ATS’ Pulmonary Infections and Tuberculosis Assembly within five years. If one or more questions are deemed in need of an update, or related new questions need

answered, a new task force may be approved to develop an updated guideline.

Funding

Funding was provided by both the American Thoracic Society.

Page 69 of 119

PICO question #1: Lung ultrasound versus chest x-ray to diagnose CAP

Population: Adults with suspected community-acquired pneumonia Intervention: Ultrasound, in addition to clinical judgment

Comparator: Chest x-ray, along with clinical judgment

Outcomes:

Critical

Time to appropriate diagnosis, treatment, and disposition (including emergency department length of stay)

Accuracy/Performance characteristics (sensitivity, specificity,)

Repeat visit to emergency department, clinic, or hospital/re-admission

Important

Provider experience (e.g. clinician confidence in decision-making, usability, etc.)

Us of advanced imaging

Cost

Patient satisfaction

Search strategy

(((("Community-Acquired Infections"[MeSH Terms] OR ("community acquired"[All Fields] AND

"infections"[All Fields]) OR "Community-Acquired Infections"[All Fields] OR ("community"[All Fields]

AND "acquired"[All Fields] AND "infection"[All Fields]) OR "community acquired infection"[All Fields] OR "Community-Acquired Infections"[MeSH Terms]) AND ("pneumonia"[MeSH Terms] OR "pneumonia"[All Fields] OR "pneumonias"[All Fields] OR "pneumoniae"[All Fields] OR "pneumoniae s"[All Fields])) AND

(("respiratory tract infections"[MeSH Terms] OR ("respiratory"[All Fields] AND "tract"[All Fields] AND

"infections"[All Fields]) OR "respiratory tract infections"[All Fields] ))) AND ((english[Filter]) AND

(alladult[Filter]))) AND (((("doppler ultrasound"[All Fields]) OR ("chest x ray"[All Fields])) OR ("chest

radiograph"[All Fields])) OR (("Ultrasonography, Doppler"[Mesh]) OR ("Radiography, Thoracic"[Mesh]) OR (“diagnostic imaging, lung”[Mesh])) Filters: English, Adult: 19+ years

Page 70 of 119

Page 71 of 119

Figure S1: Flow of information diagram

![graph TD\n A1[85,100 records identified from literature search] --> A2[0 duplicate records removed before screening; 14,114 abstract only]\n A1 --> A3[70,986 records screened]\n A3 --> A4[70948 removed; wrong or no comparator, outcome, design, population, or publication type]\n A3 --> A5[38 reports assessed for eligibility]\n A5 --> A6[27 studies removed -10 for incorrect comparator -7 for no CT -5 for no clinical outcomes -5 no quantitative data]\n A5 --> A7[11 studies included in review]\n](1/parseimages/figure6695adcfc261472abb436c1fe6507d61_0_2.png) Page 72 of 119

|Study|Type of Study|Location|Number of Subjects|Population|Intervention|Outcomes|Risk of Bias| |---|---|---|---|---|---|---|---| |Amatya 2018|Observational|Nepal|62|Patients with suspected CAP and CT scan in $ED^{a}$|Lung Ultrasound, CXR|Sensitivity, Specificity **|None| |Bourcier 2014|Observational|France|144|Patients with suspected CAP and CT scan in $ED^{b}$|Lung Ultrasound, CXR|Sensitivity, Specificity**|None| |Cortallero 2012|Observational|Italy|120|Patients with suspected CAP and CT scan in $ED^{c}$|Lung Ultrasound, CXR|Sensitivity, Specificity **|None| |Dhawan 2022|Observational|India|85|Patients with suspected CAP and CT scan in tertiary care hospital $ICUs^{d}$|Lung Ultrasound, CXR|Sensitivity, Specificity **|None| |Gibbons 2021|Observational|USA|110|Patients with suspected CAP and CT scan in $ED^{e}$|Lung Ultrasound, CXR|Sensitivity, Specificity **|None| |Liu 2014|Observational|China|179|Patients with suspected CAP and CT scan in $ED^{f}$|Lung Ultrasound, CXR|Sensitivity, Specificity **|None|

Page 73 of 119

|Corradi 2015|Observational|Italy|54|Patients with suspected CAP and CT scan in $ED^{g}$|Lung Ultrasound, CXR|Sensitivity, Specificity **|None| |---|---|---|---|---|---|---|---| |Fares 2015|Observational|Egypt|38|Patients with suspected CAP and CT scan in a hospital $ICU^{h}$|Lung Ultrasound, CXR|Sensitivity, Specificity **|None| |Karimi 2019|Observational|Iran|280|Patients with suspected CAP and CT scan in $ED^{i}$|Lung Ultrasound, CXR|Sensitivity, Specificity **|None| |Taghizadieh 2015|Observational|Iran|30|Patients with suspected CAP and CT scan in $ED^{j}$|Lung Ultrasound, CXR|Sensitivity, Specificity **|None| |Bitar 2018|Observational|Kuwait|82|Patients with suspected CAP and CT scan in a hospital medical-surgical $ICU^{k}$|Lung Ultrasound, CXR|Sensitivity, Specificity **|None|

ªAmatya Y, et al. Int J Emerg Med. 2018 Mar 12:11(1):8. Patients had at least three of the following signs or symptoms: temperature greater than 38 °C or history of fever, cough, dyspnea, tachypnea (respiratory rate greater than 20), or oxygen saturation lower than 92%.

<sup>b</sup>Bourcier JE, et al. Am J Emerg Med. 2014 Feb;32(2):115-8. At least three of the following items: tympanic temperature equal or higher than 38°C, cough, dyspnea, heart rate higher than 100 beats per minute, saturation of oxygen lower or equal to 92% in ambient air.

c'Cortellaro F, et al. Emerg Med J. 2012 Jan;29(1):19-23wig S, Ruiz M, Mensa J, et al. Severe community-acquired pneumonia: assessment of severity criteria. Am J Respir Crit Care Med 1998;158:1102-1108. Signs and symptoms considered as suggestive of CAP were: cough; pleuritic pain; sputum production; fever; dyspnea.

dDhawan J, et al. Indian J Crit Care Med. 2022 Aug;26(8):920-929. Clinical suspicion of pneumonia was considered when the following criteria was met: Symptoms suggestive of pneumonia (fever, cough, purulent sputum, and pleuritic chest pain), fulfilled minor criteria with at least three of the following symptoms: Respiratory rate >30 breaths/minute, PaO2/FIO2 <250, multilobar infiltrates, confusion/ disorientation, uremia [blood urea nitrogen (BUN) >20 mg/dL], leukopenia (WBC count <4,000 cells/mm3), thrombocytopenia (platelet count <100,000 cells/mm3),

hypothermia (core temperature <36°C), and hypotension requiring aggressive fluid resuscitation; fulfilled major criteria with a requirement of at least one of the following factors: Invasive mechanical ventilation and septic shock with need for vasopressors.

eGibbons RC, et al. J Emerg Med. 2021 May;60(5):615-625. Patients with one or more of the predefined signs and symptoms of COVID-19 were

eligible for enrollment. Predefined signs and symptoms included: cough, fever, dyspnea, myalgia, malaise, ageusia, anosmia, increased work of breathing, temperature $ 38 _C (100.4 F), heartrate $ 100beats/ min, respiratory rate $ 16 breaths/min, and SpO2 < 94%.

fLiu XL, et al. Emerg Med J. 2015 Jun;32(6):433-8. Signs and symptoms considered as suggestive of CAP included: cough, pleuritic pain, sputum

production, fever, dyspnea.

gCorradi F, et al. Biomed Res Int 2015:1-8. Pneumonia was clinically suspected on the basis of cough, dyspnea, body temperature >38

∘

C or

<35

∘

C, heart rate >90 beats/min, tachypnea >20 breaths/min, rales or crackles on auscultation, and abnormal oxygen saturation.

hFares Auf M-N. Med J Cairo Univ 2015;83:307–14. Pneumonia diagnosis based on suggestive history (fever, cough, sputum production,

dyspnea).General and local physical signs suggestive of pneumonia.

IKarimi E. Arch Acad Emerg Med 2019;7:e8. clinical symptoms of pneumonia such as cough, phlegm, shortness of breath, hemoptysis, and

temperature higher than 38±C.

jTaghizadieh A, et al.. Emerg 2015;3:114–6. Presence of fever, cough, pleuritic pain, sputum production, and dyspnea were considered as signs

and symptoms of CAP.

kBitar ZI, et al.. Health Sci Rep 2019;2:e102. The diagnosis of pneumonia was confirmed by a set of clinical features (clinical history and physical

examination), microbiological testing for admitted patients (blood and sputum culture, legionella and pneumococcal urinary antigen testing, and multiplex polymerase chain reaction assay for detecting Chlamydia pneumoniae, Mycoplasma pneumoniae, and respiratory tract viruses),

inflammatory markers (c-reactive protein >10 mg/L and procalcitonin ≥0.25 ng/mL), along with the presence of consolidation or opacification on a CXR or chest CT.

** - performance characteristics for CXR and LUS were calculated using CT scan results as reference standard.

Page 74 of 119

Page 75 of 119

Figure S2: Forest plots

A) Analysis #1: Ultrasound (using chest CT scan as the reference standard)

|Study|TP|FP|FN|TN|Sensitivity (95% CI)|Specificity (95% CI)|Sensitivity (95% CI)|Sensitivity (95% CI)| |---|---|---|---|---|---|---|---|---| |Amatya 2018|40|11|4|7|0.91 [0.78, 0.97]|0.39 [0.17, 0.64]||| |Bitar 2019|31|1|1|5|0.97 [0.84, 1.00]|0.83 [0.36, 1.00]||| |Bourcier 2014|23|0|0|0|1.00 [0.85, 1.00]|Not estimable||| |Corradi 2015|30|1|14|19|0.68 [0.52, 0.81]|0.95 [0.75, 1.00]||| |Cortallero 2012|25|1|1|3|0.96 [0.80, 1.00]|0.75 [0.19, 0.99]||| |Dhawan 2022|74|0|10|1|0.88 [0.79, 0.94]|1.00 [0.03, 1.00]||| |Fares 2015|28|2|6|2|0.82 [0.65, 0.93]|0.50 [0.07, 0.93]||| |Gibbons 2021|79|9|2|20|0.98 [0.91, 1.00]|0.69 [0.49, 0.85]||| |Karimi 2019|232|0|48|0|0.83 [0.78, 0.87]|Not estimable||| |Liu 2014|106|1|6|66|0.95 [0.89, 0.98]|0.99 [0.92, 1.00]||| |Taghizadieh 2015|29|1|0|0|1.00 [0.88, 1.00]|0.00 [0.00, 0.97]|||

B) Analysis #2: Chest X-Ray (using chest CT scan as the reference standard)

|Study|TP|FP|FN|TN|Sensitivity (95% CI)|Specificity (95% CI)|Sensitivity (95% CI)|Sensitivity (95% CI)|Specificity (95% CI)| |---|---|---|---|---|---|---|---|---|---| |Amatya 2018|32|9|12|9|0.73 [0.57, 0.85]|0.50 [0.26, 0.74]|||| |Bitar 2019|5|5|27|1|0.16 [0.05, 0.33]|0.17 [0.00, 0.64]|||| |Bourcier 2014|12|0|11|0|0.52 [0.31, 0.73]|Not estimable|||| |Corradi 2015|28|4|16|16|0.64 [0.48, 0.78]|0.80 [0.56, 0.94]|||| |Cortallero 2012|18|1|8|3|0.69 [0.48, 0.86]|0.75 [0.19, 0.99]|||| |Dhawan 2022|58|1|26|0|0.69 [0.58, 0.79]|0.00 [0.00, 0.97]|||| |Fares 2015|24|3|6|5|0.80 [0.61, 0.92]|0.63 [0.24, 0.91]|||| |Gibbons 2021|57|13|24|16|0.70 [0.59, 0.80]|0.55 [0.36, 0.74]|||| |Karimi 2019|263|0|17|0|0.94 [0.90, 0.96]|Not estimable|||| |Liu 2014|87|4|25|63|0.78 [0.69, 0.85]|0.94 [0.85, 0.98]|||| |Taghizadieh 2015|27|1|2|0|0.93 [0.77, 0.99]|0.00 [0.00, 0.97]||||

Page 76 of 119

Table S2: Evidence profile

Population: Adults with suspected CAP Comparison: Chest x-ray versus lung ultrasound Setting: Inpatients and outpatients

|Quality assessment|Quality assessment|Quality assessment|Quality assessment|Quality assessment|Quality assessment|Summary of Findings|Summary of Findings|Quality|Importance| |---|---|---|---|---|---|---|---|---|---| |No of studies|Design|Risk of bias|Inconsistency|Indirectness|Imprecision|Other|# patients|Effect (range)|Importance| |Sensitivity and specificity: Ultrasound (using chest CT scan as the reference standard)|||||||||| |$11^{1,2}$|Accuracy|Not Serious|Serious3|Not $serious^{4,5}$|Serious6|None|939|Sensitivity= median 95% (range 68-100%) Specificity= median 75%$^{6}$ (range 0-100%)|⊕⊕○○ LOW| |Sensitivity and specificity: Chest X-Ray (using chest CT scan as the reference standard)|||||||||| |$11^{1,2}$|Accuracy|Not Serious|Serious5|Not $serious^{4,5}$|Serious6|None|939|Sensitivity= median 70% (range 16-94%) Specificity= median 55%$^{6}$ (range 0-94%)|⊕⊕○○ LOW| |$11^{1,2}$|Accuracy|Not Serious|Serious5|Not $serious^{4,5}$|Serious6|None|939|Sensitivity= median 70% (range 16-94%) Specificity= median 55%$^{6}$ (range 0-94%)|CRITICAL LOW|

Footnotes:

1. Amatya Y, et al. Int J Emerg Med. 2018 Mar 12;11(1):8.; Bitar ZI, et al. Health Sci Rep 2019;2:e102; Bourcier JE, et al. Am J Emerg Med. 2014 Feb;32(2):115-8.; Corradi F, et al. Biomed Res Int 2015:1-8; Cortellaro F, et al. Emerg Med J. 2012 Jan;29(1):19-23; Dhawan J, et al. Indian J Crit Care Med. 2022 Aug;26(8):920-929; Fares Auf M-N. Med J Cairo Univ 2015;83:307-14; Gibbons RC, et al. J Emerg Med. 2021 May;60(5):615-625; Karimi E. Arch Acad Emerg Med 2019;7:e8; Liu XL, et al. Emerg Med J. 2015 Jun;32(6):433-8; Taghizadieh A, et al. Emerg 2015;3:114-6.

2. Testa A, et al. Crit Care. 2012 Feb 17;16(1):R30 was excluded due to being judged an outlier.

3. Inconsistency: Wide range of sensitivity and specificity estimates across studies as seen in the Forest plots.

4. Indirectness of the comparison: The question asks about ultrasound compared to chest x-ray. However, the studies are accuracy studies that compared ultrasound to a reference standard and compared chest x-ray to a reference standard (i.e., chest CT). Therefore, answering the question requires an indirect comparison with an assumption of transitivity. The committee recognized the indirect nature of the comparison but judged that it did not further reduce its certainty in the estimates beyond the inconsistency and the imprecision.

5. Indirectness of the population: The question asks about patients with suspected CAP, but most of the studies enrolled patients with suspected CAP who also required a chest CT scan (usually due to discordant results between the chest x-ray and lung ultrasound). The committee recognized the indirectness of the population but judged that it did not further reduce its certainty in the estimates beyond the inconsistency and the imprecision.

6. Imprecision: Wide confidence intervals for individual studies as seen in the Forest plots.

Page 77 of 119

Table S3: EtD framework

QUESTION

Should lung ultrasound be considered a reasonable alternative to chest x-ray in patients with suspected community-acquired pneumonia?

|POPULATION:|Patients with suspected community-acquired pneumonia| |---|---| |INTERVENTION:|Lung ultrasound| |COMPARATOR:|Chest x-ray| |SETTING:|Inpatients and outpatients|

ASSESSMENT

|Test accuracy How accurate are the tests?|| |---|---| |JUDGEMENT|RESEARCH EVIDENCE| |○ CXR is a lot more accurate ○ CXR is slightly more accurate ○ Lung US is a lot more accurate ○ Lung US is slightly more accurate ○ Chest x-ray and US are comparably accurate|TEST CHARACTERISTICS The guideline committee judged the sensitivities and specificities of lung ultrasound and chest x-ray as comparable. The estimated medians might seem quite different but, when one considers that the lung ultrasound studies were likely performed by experienced operators, the committee concluded that the accuracy of lung ultrasound is likely overestimated in the studies compared with routine clinical practice. When one accounts for this likelihood, the committee concluded that the accuracy of lung ultrasound and chest x-ray are likely comparable. Lung ultrasound Sensitivity = median 95% (range 68-100%) Specificity = median 75% (range 0-100%) Chest x-ray Sensitivity = median 70% (range 16-94%) Specificity = median 55% (range 0-94%)| |Desirable Effects How substantial are the desirable effects of making a diagnosis? “Substantial” refers to both the importance and magnitude of the desirable effects. As an example, a small improvement in a critical outcome might be considered more substantial than a large improvement in an unimportant outcome.|| |JUDGEMENT|RESEARCH EVIDENCE| |○ CXR is a lot more likely to lead to desirable effects ○ CXR is slightly more likely to lead to desirable effects ○ Lung US is a lot more likely to lead to desirable effects ○ Lung US is slightly more likely to lead to desirable effects ○ Chest x-ray and US will lead to similar desirable effects|DESIRABLE PATIENT-IMPORTANT OUTCOMES Desirable effects derive from true positive and true negative results. They include the initiation or continuation of appropriate antibiotic therapy in those you have pneumonia, the elimination of the burdens and costs of seeking alternative diagnoses in those you have pneumonia, avoiding unnecessary antibiotic therapy in those who do not have pneumonia, and promoting ongoing pursuit of the correct diagnosis in those who do not have pneumonia. For both lung ultrasound and chest x-ray, a positive test result will result in the same intervention, antibiotics, and therefore eventually the same outcomes. Thus, the committee concluded that if test accuracy is comparable (see above), downstream desirable outcomes must also be comparable. Sensitivity = true positive rate:|

Page 78 of 119

||Lung ultrasound = 95% Chest x-ray = 70% Specificity = true negative rate: Lung ultrasound = 75% Chest x-ray = 55%| |---|---| |Undesirable Effects|Undesirable Effects| |How substantial are the undesirable effects of making a diagnosis? "Substantial" refers to both the importance and magnitude of the undesirable effects. As an example, a small but important complication of diagnostic testing might be considered more substantial than a large but unimportant complication.|| |JUDGEMENT|RESEARCH EVIDENCE| |○ CXR is a lot more likely to lead to undesirable effects ○ CXR is slightly more likely to lead to undesirable effects ○ Lung US is a lot more likely to lead to undesirable effects ○ Lung US is slightly more likely to lead to undesirable effects ○ Chest x-ray and US will lead to similar undesirable effects|| |Balance of desirable and undesirable effects Does the balance between desirable and undesirable effects favor the intervention or the comparison?|| |JUDGEMENT|RESEARCH EVIDENCE The guideline committee noted that chest x-ray and lung ultrasound probably lead to similar desirable effects (foregoing or discontinuing antibiotics, additional diagnostic testing to either confirm the negative result or seek an alternative diagnosis). Thus, the committee concluded that if test accuracy is comparable (see above), downstream undesirable outcomes must also be comparable. 1 - sensitivity = false negative rate: Lung ultrasound = 5% Chest x-ray = 30% 1 - specificity = false positive rate: Lung ultrasound = 25% Chest x-ray = 45%| |Balance of desirable and undesirable effects Does the balance between desirable and undesirable effects favor the intervention or the comparison?|| |JUDGEMENT|RESEARCH EVIDENCE| |○ Favors chest x-ray ○ Probably favors chest x-ray ● Does not favor either chest x-ray or lung ultrasound ○ Favors lung ultrasound ○ Probably favors lung ultrasound ○ Varies ○ Don't know|The guideline committee noted that chest x-ray and lung ultrasound probably lead to similar desirable and undesirable effects (see above). Therefore, they concluded that the balance of effects does not favor either chest x-ray or lung ultrasound.| |Quality of evidence of test accuracy What is the committee's confidence in the above listed estimates of test accuracy (i.e., what is the quality of evidence)?|| |JUDGEMENT|RESEARCH EVIDENCE|

Page 79 of 119

|○ Very low • Low • Moderate ○ High ○ No included studies|QUALITY OF EVIDENCE The quality of evidence for both lung ultrasound and chest x-rays was low because there are accuracy studies that were downgraded due to inconsistency (there were a wide range of estimates across studies) and imprecision (the confidence intervals were wide for most studies).| |---|---| |Quality of evidence of test result/management What is the committee's confidence that the test results will lead to certain clinical actions?|| |JUDGEMENT|RESEARCH EVIDENCE| |○ Very low ○ Low • Moderate ○ High • No included studies|The committee did not evaluate published evidence regarding clinical actions that follow lung ultrasound and chest x-ray results. The committee concluded that the frequency of clinical actions following lung ultrasound and chest x-ray must be comparable if accuracy of studies is comparable (see above) since the tests lead to the same clinical actions. Clinical actions include the initiation/continuation antibiotic therapy or foregoing/discontinuing antibiotic therapy.| |Quality of evidence of management/clinical outcomes What is the committee's confidence that the clinical actions prompted by the test results will lead to certain outcomes?|| |JUDGEMENT|RESEARCH EVIDENCE| |○ Very low ○ Low • Moderate ○ High • No included studies|The committee did not evaluate published evidence regarding the clinical outcomes of antibiotic therapy in patients with CAP. However, the committee was confident that antibiotic therapy improves clinical outcomes in patients with CAP. The committee therefore concluded that clinical outcomes following lung ultrasound and chest x-ray must be comparable if the accuracy of studies is comparable (see above), since the tests lead to the same clinical actions which create those outcomes.| |Quality of evidence What is the overall certainty of the effects of the tests? This is defined as the lowest quality of evidence among the qualities of evidence of test accuracy, result/management, and management/clinical outcomes|| |JUDGEMENT|RESEARCH EVIDENCE| |● Very low ○ Low • Moderate ○ High ○ No included studies|The overall quality of evidence is very low because, even if there exists good evidence that test results effect clinical actions that improve outcomes, there is very low quality of evidence for test accuracy for both lung ultrasound and chest x-ray. ○ No included studies| |Acceptability Is the intervention acceptable to key stakeholders?|| |JUDGEMENT|RESEARCH EVIDENCE| |○ No ○ Probably no ○ Probably yes ● Yes ○ Varies ○ Don't know|Lung ultrasound and chest x-ray are non-invasive, painless, and not burdensome. Therefore, both are acceptable to most patients. This conclusion is based on the committee's non-systematic clinical observations.| |Feasibility Is the intervention feasible to implement?||

Page 80 of 119

|JUDGEMENT|RESEARCH EVIDENCE| |---|---| |o No|Both chest x-ray and lung ultrasound are available in most clinical settings. The primary limiting factor is the availability of experience operators and interpreters of lung ultrasound. This conclusion is based on the committee's non-systematic clinical observations.| |o Probably no|primary limiting factor is the availability of experience operators and interpreters of lung ultrasound. This conclusion is based on the committee's non-systematic clinical observations.| |o Yes|observations.| |o Varies|| |o Don't know||

SUMMARY OF JUDGMENTS

||JUDGEMENT|JUDGEMENT|JUDGEMENT|JUDGEMENT|JUDGEMENT|JUDGEMENT|JUDGEMENT| |---|---|---|---|---|---|---|---| |TEST ACCURACY|CXR a lot more|CXR slightly more|CXR and lung US are comparable|Lung US slightly more|Lung US a lot more|-|-| |DESIRABLE EFFECTS|CXR a lot more|CXR slightly more|CXR and lung US are comparable|Lung US slightly more|Lung US a lot more|-|-| |UNDESIRABLE EFFECTS|CXR a lot more|CXR slightly more|CXR and lung US are comparable|Lung US slightly more|Lung US a lot more|-|-| |BALANCE OF EFFECTS|Favors CXR|Probably favors CXR|Does not favor either CXR or lung US|Probably favors lung US|Favors lung US|Varies|Don't know| |QUALITY OF EVIDENCE OF TEST ACCURACY|Very low|Low|Moderate|High|No included studies|-|-| |QUALITY OF EVIDENCE OF TEST RESULT/MANAGEMENT|Very low|Low|Moderate|High|No included studies|-|-| |QUALITY OF EVIDENCE OF MANAGEMENT/OUTCOMES|Very low|Low|Moderate|High|No included studies|-|-| |QUALITY OF EVIDENCE|Very low|Low|Moderate|High|No included studies|-|-| |ACCEPTABILITY|No|Probably no|Probably yes|Yes|Varies|Don't know|-| |FEASIBILITY|No|Probably no|Probably yes|Yes|Varies|Don't know|-|

TYPE OF RECOMMENDATION

|Strong recommendation for chest x-ray|Conditional recommendation for chest x-ray|Conditional recommendation for chest x-ray or lung ultrasound|Conditional recommendation for lung ultrasound|Strong recommendation for lung ultrasound|Out|| |---|---|---|---|---|---|---| |○|○|○|○|○|○|○|

CONCLUSIONS

Recommendation

For patients with suspected community-acquired pneumonia, we suggest that lung ultrasound be considered an acceptable alternative to chest x-rays in medical centers where the appropriate

expertise exists (conditional recommendation, low quality of evidence).

Participation = 15/18 (83%)

Strong recommendation for chest x-ray = 0/15 (0%).

Conditional recommendation for chest x-ray =1/15 (6.67%).

Strong recommendation for lung ultrasound = 0/15 (0%).

Condition recommendation for lung ultrasound = 1/15 (6.67%).

Conditional recommendation for either chest x-ray or lung ultrasound = 13/15 (86.67%).

Page 81 of 119

PICO Question #2: Antibacterial therapy for CAP if a test for a respiratory virus is positive

Population: Adult CAP patients who test positive for a respiratory virus, Intervention: Antibacterial therapy,

Comparator: No antibacterial therapy.

Outcomes:

Critical

Mortality (i.e. in-hospital, 28 day, 30 day, 60 day, 90 day, 180 day, <7% at 10 days)

Length of stay (i.e. hospital)

Treatment failure (i.e. decompensation, need for hospital admission, reasmission need for mechanical ventilation, need for vasopressor support, ICU transfer)

Clinical stability

Important

Antibiotic-associated adverse events (inc side effects and resistant organisms)Secondary infection

Days of antibiotics

Return to function (work, exertion, home after hospitalization) or quality of life Symptoms (i.e. total number, etc)

Cost

Search Strategy

Overall Question #1

((("antibiotic"[Title/Abstract]) OR (("Anti-Bacterial Agents"[Mesh]) OR "Anti-Bacterial Agents"

[Pharmacological Action])) AND ((("Community-Acquired Infections"[MeSH Terms] OR ("community

acquired"[All Fields] AND "infections"[All Fields]) OR "Community-Acquired Infections"[All Fields] OR

("community"[All Fields] AND "acquired"[All Fields] AND "infection"[All Fields]) OR "community acquired infection"[All Fields] OR "Community-Acquired Infections"[MeSH Terms]) AND ("pneumonia"[MeSH

Terms] OR "pneumonia"[All Fields] OR "pneumonias"[All Fields] OR "pneumoniae"[All Fields] OR

"pneumoniae s"[All Fields])) AND (("respiratory tract infections"[MeSH Terms] OR ("respiratory"[All

Fields] AND "tract"[All Fields] AND "infections"[All Fields]) OR "respiratory tract infections"[All Fields] ))))

Overall Question #1

(("communal"[All Fields] OR "communalism"[All Fields] OR "communalities"[All Fields] OR

"communality"[All Fields] OR "communally"[All Fields] OR "commune"[All Fields] OR "communes"[All

Fields] OR "community s"[All Fields] OR "communitys"[All Fields] OR "residence characteristics"[MeSH

Terms] OR ("residence"[All Fields] AND "characteristics"[All Fields]) OR "residence characteristics"[All

Page 82 of 119

Fields] OR "communities"[All Fields] OR "community"[All Fields]) AND ("acquirable"[All Fields] OR

"acquire"[All Fields] OR "acquired"[All Fields] OR "acquirement"[All Fields] OR "acquirements"[All Fields] OR "acquires"[All Fields] OR "acquiring"[All Fields]) AND ("pneumonia"[MeSH Terms] OR

"pneumonia"[All Fields] OR "pneumonias"[All Fields] OR "pneumoniae"[All Fields] OR "pneumoniae

s"[All Fields]) AND ("eur med j respir"[Journal] OR "respiratory"[All Fields]) AND ("virally"[All Fields] OR "virals"[All Fields] OR "virology"[MeSH Terms] OR "virology"[All Fields] OR "viral"[All Fields]) AND ("anti bacterial agents"[Pharmacological Action] OR "anti bacterial agents"[MeSH Terms] OR ("anti

bacterial"[All Fields] AND "agents"[All Fields]) OR "anti bacterial agents"[All Fields] OR "antibiotic"[All

Fields] OR "antibiotics"[All Fields] OR "antibiotic s"[All Fields] OR "antibiotical"[All Fields])) AND

((english[Filter]) AND (alladult[Filter]))

Indirect questions

((("antibiotic"[Title/Abstract]) OR (("Anti-Bacterial Agents"[Mesh]) OR "Anti-Bacterial Agents"

[Pharmacological Action])) AND ((("Community-Acquired Infections"[MeSH Terms] OR ("community

acquired"[All Fields] AND "infections"[All Fields]) OR "Community-Acquired Infections"[All Fields] OR

("community"[All Fields] AND "acquired"[All Fields] AND "infection"[All Fields]) OR "community acquired infection"[All Fields] OR "Community-Acquired Infections"[MeSH Terms]) AND ("pneumonia"[MeSH

Terms] OR "pneumonia"[All Fields] OR "pneumonias"[All Fields] OR "pneumoniae"[All Fields] OR

"pneumoniae s"[All Fields])) AND (("respiratory tract infections"[MeSH Terms] OR ("respiratory"[All

Fields] AND "tract"[All Fields] AND "infections"[All Fields]) OR "respiratory tract infections"[All Fields]

AND "anti bacterial agents/administration and dosage"[MeSH Terms]))))

Indirect population for Bronchitis

("antibiotic"[Title/Abstract] OR ("Anti-Bacterial Agents"[MeSH Terms] OR "Anti-Bacterial

Agents"[Pharmacological Action])) AND (("Community-Acquired Infections"[MeSH Terms] OR

("community acquired"[All Fields] AND "infections"[All Fields]) OR "Community-Acquired Infections"[All Fields] OR ("community"[All Fields] AND "acquired"[All Fields] AND "infection"[All Fields]) OR

"community acquired infection"[All Fields] OR "Community-Acquired Infections"[MeSH Terms]) AND

("bronchitis"[MeSH Terms] OR "bronchitis"[All Fields] OR "bronchitides"[All Fields]) AND ("respiratory

tract infections"[MeSH Terms] OR ("respiratory"[All Fields] AND "tract"[All Fields] AND "infections"[All

Fields]) OR "respiratory tract infections"[All Fields]))

Bronchitis

(("bronchitis"[MeSH Terms] OR "bronchitis"[All Fields] OR "bronchitides"[All Fields]) AND ("eur med j

respir"[Journal] OR "respiratory"[All Fields]) AND ("virally"[All Fields] OR "virals"[All Fields] OR

"virology"[MeSH Terms] OR "virology"[All Fields] OR "viral"[All Fields]) AND ("anti bacterial

agents"[Pharmacological Action] OR "anti bacterial agents"[MeSH Terms] OR ("anti bacterial"[All Fields] AND "agents"[All Fields]) OR "anti bacterial agents"[All Fields] OR "antibiotic"[All Fields] OR

"antibiotics"[All Fields] OR "antibiotic s"[All Fields] OR "antibiotical"[All Fields])) AND ((english[Filter])

AND (alladult[Filter]))

Indirect population for Tracheobronchitis

("antibiotic"[Title/Abstract] OR ("Anti-Bacterial Agents"[MeSH Terms] OR "Anti-Bacterial

Agents"[Pharmacological Action])) AND (("Community-Acquired Infections"[MeSH Terms] OR

Page 83 of 119

("community acquired"[All Fields] AND "infections"[All Fields]) OR "Community-Acquired Infections"[All Fields] OR ("community"[All Fields] AND "acquired"[All Fields] AND "infection"[All Fields]) OR "community acquired infection"[All Fields] OR "Community-Acquired Infections"[MeSH Terms]) AND ("tracheobronchitis"[All Fields]) AND ("respiratory tract infections"[MeSH Terms] OR ("respiratory"[All Fields] AND "tract"[All Fields] AND "infections"[All Fields]) OR "respiratory tract infections"[All Fields]))

Tracheobronchitis

("tracheobronchitis"[All Fields] AND ("eur med j respir"[Journal] OR "respiratory"[All Fields]) AND ("virally"[All Fields] OR "virals"[All Fields] OR "virology"[MeSH Terms] OR "virology"[All Fields] OR "viral"[All Fields])) AND ("anti bacterial agents"[Pharmacological Action] OR "anti bacterial agents"[MeSH Terms] OR ("anti bacterial" [All Fields] AND "agents"[All Fields]) OR "anti bacterial agents"[All Fields] OR "antibiotic"[All Fields] OR "antibiotics"[All Fields] OR "antibiotic s"[All Fields] OR "antibiotical"[All Fields])) AND ((english[Filter]) AND (alladult[Filter]))

![graph TD\n A1[5086 records identified from literature search] --> A2[11191 duplicate records removed before screening]\n A1 --> A3[3895 records screened]\n A3 --> A4[3868 records removed; wrong or no comparator, outcome, design, population, or publication type]\n A3 --> A5[27 reports sought for retrieval]\n A5 --> A6[0 reports not received]\n A5 --> A7[27 reports assessed for eligibility]\n A7 --> A8[27 reports removed for wrong study design]\n A7 --> A9[0 studies included in review]\n](1/parseimages/figure8d535fcd05fc44dba61fa77264b53765_0_4.png)

Forest Plots

None

Evidence Profile

There are no studies (randomized or non-randomized) concerning CAP as diagnosed by signs, symptoms, and imaging that compare an antibiotic to no antibiotic regimen following the identification of a respiratory viral pathogen. Therefore, an evidence profile and evidence-to-decision table were not created. The committee informed its recommendations with non-comparative evidence and non-systematic clinical observations. In published investigations, the decision to continue or discontinue antibiotics after finding a viral pathogen included other considerations such as the likelihood of bacterial coinfection depending on the specific virus identified, the difficulty in excluding concomitant bacterial infections with available diagnostic techniques, and clinical stability of the patient.

PICO question #3: Antibiotic duration for CAP

Population: Adult patients with community acquired pneumonia Intervention: Less than five days of antibiotics

Comparator: Five or more days of antibiotics

Outcomes

Critical

Mortality

Treatment success/failure CAP-related complications

Important

Duration of hospitalization Antibiotic-free days

Patient experience

Cost

Antibiotic resistance

Search Strategy

1.("Pneumonia, Bacterial"[Mesh] OR "Pneumonia"[Mesh] OR "Chlamydial Pneumonia"[Mesh] OR

"Pneumonia, Viral"[Mesh] OR "Pneumonia, Staphylococcal"[Mesh] OR "Pneumonia,

Mycoplasma"[Mesh] OR "Pneumonia, Pneumococcal"[Mesh] OR "Respiratory Tract Infections"[Mesh]) AND ("Community-Acquired Infections"[Mesh]) AND ("anti bacterial agents"[Pharmacological Action]

OR "anti bacterial agents"[MeSH Terms] OR ("anti bacterial"[All Fields] AND "agents"[All Fields])

OR "anti bacterial agents"[All Fields] OR "antibiotic"[All Fields] OR "antibiotics"[All Fields] OR "antibiotic s"[All Fields] OR "antibiotical"[All Fields] OR "antibacterial agents"[Pharmacological Action] OR "anti

bacterial agents"[MeSH Terms] OR ("anti bacterial"[All Fields] AND "agents"[All Fields]) OR "antibacterial agents"[All Fields] OR "antibacterial"[All Fields] OR "antibacterials"[All Fields] OR "antibacterially"[All

Fields])) OR "Therapeutics"[Mesh] OR "therapy" [Subheading] ) AND ("Duration of Therapy"[Mesh])

2.AND (English[Language]) AND (humans[Filter]) NOT ("infant"[mesh] OR "child"[mesh] OR

adolescent"[mesh])

3.("Pneumonia, Bacterial"[Mesh]) AND "Duration of Therapy"[Mesh]

Community acquired pneumonia treatment duration

Page 85 of 119

Page 86 of 119

4. Reference search of other identified studies

Figure S4: Flow of information diagram

![graph TD\n A1[1342 records identified from literature search] --> A2[60 duplicate records removed before screening]\n A1 --> A3[1282 records screened]\n A3 --> A4[1253 records removed for not comparing antibiotic duration]\n A3 --> A5[29 reports sought for retrieval]\n A5 --> A6[29 reports assessed for eligibility]\n A6 --> A7[4 studies included in review]\n A6 --> A8[3 removed as they did not compare antibiotics directly 21 removed because days treated were wrong 1 removed because of bronchitis treatment]\n](1/parseimages/figure9b8591388d604e7eaed7c3e850de0ba2_0_3.png) Page 87 of 119

|Study|Type of study|Location|Number of subjects (I/C)|Population|Outcomes| |---|---|---|---|---|---| |$^{1}El$ Moussaoui - 2006|RCT|Netherlands|119 (56/63)|Inpatients|Clinical cure| |$^{2}D^{I}Ignazio$ - 2004|RCT|Worldwide (Canada, Chile, India, Lithuania, Mexico, Peru, Russia, and United States)|363 (174/189)|Outpatients|Clinical cure| |$^{3}Dinh$ - 2021|RCT|France|310 (157/153)|Inpatients|Mortality, clinical cure, length of stay| |$^{4}Drehobl$ - 2005|RCT|Worldwide (United States, Canada, Argentina, Russia, India, Estonia, and Lithuania)|411 (202/209)|Outpatients|Clinical cure|

RCT = randomized controlled trial; I=intervention; C=control

2. D'Ignazio, J., Camere, M. A., Lewis, D. E., Jorgensen, D., Breen, J. D. Novel, single-dose microsphere formulation of azithromycin versus 7-day levofloxacin therapy for treatment of mild to moderate community-acquired pneumonia in adults. Antimicrob Agents Chemother; Oct 2005.

3. Dinh, A., Ropers, J., Duran, C., Davido, B., Deconinck, L., Matt, M., Senard, O., Lagrange, A., Makhloufi, S., Mellon, G., de Lastours, V., Bouchand, F., Mathieu, E., Kahn, J. E., Rouveix, E., Grenet, J., Dumoulin, J., Chinet, T., Pépin, M., Delcey, V., Diamantis, S., Benhamou, D., Vitat, V., Dombret, M. C., Renaud, B., Perronne, C., Claessens, Y. E., Labarère, J., Bedos, J. P., Aegerter, P., Crémieux, A. C. Discontinuing β-lactam treatment after 3 days for patients with community-acquired pneumonia in non-critical care wards (PTC): a double-blind, randomised, placebo-controlled, non-inferiority trial. Lancet; Mar 27 2021.

4. Drehobl, M. A., De Salvo, M. C., Lewis, D. E., Breen, J. D. Single-dose azithromycin microspheres vs clarithromycin extended release for the treatment of mild-to-moderate community-acquired pneumonia in adults. Chest; Oct 2005.

Page 88 of 119

|Study|Antibiotics (I VS C)|Intervention days|Control days| |---|---|---|---| |Inpatient|Inpatient|Inpatient|Inpatient| |$^{1}El$ Moussaoui - 2006|Patients who improved after 3 days of IV amoxicillin were randomized to placebo or 750 mg of oral amoxicillin TID|3|8| |$^{2}Dinh$ - 2021|After 72 hours of beta-lactam treatment, patients were randomized to receive placebo or 500 mg amoxicillin plus 62.5 mg of clavulanate TID|3|8| |Outpatient|Outpatient|Outpatient|Outpatient| |$^{3}D'Ignazio$ - 2004|A single 2 gm dose azithromycin microspheres vs 500 mg oral levofloxacin|1|7| ||||| |$^{4}Drehobl$ - 2005|A single 2 gm dose azithromycin microspheres vs clarithromycin|1|7|

I=intervention; C=control; TID=three times daily

1. el Moussaoui, R., de Borgie, C. A., van den Broek, P., Hustinx, W. N., Bresser, P., van den Berk, G. E., Poley, J. W., van den Berg, B., Krouwels, F. H., Bonten, M. J., Weenink, C., Bossuyt, P. M., Speelman, P., Opmeer, B. C., Prins, J. M.. Effectiveness of discontinuing antibiotic treatment after three days versus eight days in mild to moderate-severe community acquired pneumonia: randomised, double blind study. Bmj.; Jun 10 2006. 2. Dinh, A., Ropers, J., Duran, C., Davido, B., Deconinck, L., Matt, M., Senard, O., Lagrange, A., Makhloufi, S., Mellon, G., de Lastours, V., Bouchand, F., Mathieu, E., Kahn, J. E., Rouveix, E., Grenet, J., Dumoulin, J., Chinet, T., Pépin, M., Delcey, V., Diamantis, S., Benhamou, D., Vitrat, V., Dombret, M. C., Renaud, B., Perronne, C., Claessens, Y. E., Labarère, J., Bedos, J. P., Aegerter, P., Crémieux, A. C. Discontinuing β-lactam treatment after 3 days for patients with community-acquired pneumonia in non-critical care wards (PTC): a double-blind, randomised, placebo-controlled, non-inferiority trial. Lancet; Mar 27 2021. 3. D'Ignazio, J., Camere, M. A., Lewis, D. E., Jorgensen, D., Breen, J. D. Novel, single-dose microsphere formulation of azithromycin versus 7-day levofloxacin therapy for treatment of mild to moderate community-acquired Pneumonia in adults. Antimicrob Agents Chemother; Oct 2005 4. Drehobl, M. A., De Salvo, M. C., Lewis, D. E., Breen, J. D. Single-dose azithromycin microspheres vs clarithromycin extended release for the treatment of mild-to moderate community-acquired pneumonia in adults. Chest; Oct 2005.

Page 89 of 119

Figure S5: Forest Plots

A) Analysis #1: Overall all-cause mortality

![|Study or Subgroup|Less than 5 days Events|5 or more days Events|Total (95% CI)|\n|:---:|:---:|:---:|:---:|\n|Dinh - 2021|3 (152)|2 (151)|1.49 [0.25, 8.79]|\n|Total|3 (152)|2 (151)|1.49 [0.25, 8.79]|\n](1/parseimages/figure887f5d96dd5d442b890e1ab8647cbffb_0_3.png)

B) Analysis #2: Clinical cure – short follow-up (1-2 weeks)

![|Study or Subgroup|Less than 5 days Events|5 or more days Events|Total Weight|Risk Ratio M-H, Random, 95% CI|\n|:---:|:---:|:---:|:---:|:---:|\n|D'ignazio - 2005|174|189|212|0.92 [0.86, 1.00]|\n|Dinh - 2021|113|100|146|1.14 [0.99, 1.31]|\n|Dreho-bl - 2005|187|198|202|0.94 [0.90, 0.99]|\n|Moussaoui - 2006|50|56|60|0.99 [0.90, 1.10]|\n](1/parseimages/figurec76c11ba9dd24b299005a67a971aa012_0_5.png)

C) Sub-analysis #2.1: Clinical cure – short follow-up: azithromycin only

![|Study or Subgroup|Less than 5 days Events|Less than 5 days Total|5 or more days Events|5 or more days Total|Risk Ratio M-H, Random, 95% CI|\n|:---:|:---:|:---:|:---:|:---:|:---:|\n|D'ignazio - 2005|174|211|189|212|0.92 [0.86, 1.00]|\n|Dreho-bl - 2005|187|202|198|202|0.94 [0.90, 0.99]|\n|Total (95% CI)|361|413|387|414|0.94 [0.90, 0.98]|\n](1/parseimages/figurece8cf45e8ed74e1ba6f5e1f1c27806cd_0_7.png)

D) Sub-analysis #2.2: Clinical cure – short follow-up: outpatient only

![|Study or Subgroup|Less than 5 days Events|5 or more days Events|Total events|\n|:---:|:---:|:---:|:---:|\n|D'ignazio - 2005|174|211|361|\n|Drehojl - 2005|187|202|387|\n|Total (95% CI)|413|414|0.94 [0.90, 0.98]|\n](1/parseimages/figure715b349be20641e3b9559abe8566d792_0_9.png)

E) Sub-analysis #2.3: Clinical cure – short follow-up: inpatient only

![Study or Subgroup\nLess than 5 days\nEvents\nTotal\n113\n145\n100\n146\n65.3%\n1.14 [0.99, 1.31]\nM-H, Fixed, 95% CI\nDinh - 2021\nMoussaoui - 2006\n50\n199\n163\n156\nHeterogeneity: Chi* 3.54, df = 1 (P = 0.06); P = 72%\nTest for overall effect: Z = 1.65 (P = 0.10)\nRisk Ratio\nM-H, Fixed, 95% CI\nFavors [less than 5 days]\nFavors [5 or more days]\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n100\n](1/parseimages/figure85110ebeca7e4c2b8c65a1f73315bb1c_0_11.png) Page 90 of 119

F) Sub-analysis #2.4: Clinical cure – short follow-up: PSI score <71

![|Study or Subgroup|Less than 5 days Events|5 or more days Events|Total Weight|Risk Ratio M-H, Fixed, 95% CI|\n|:---:|:---:|:---:|:---:|:---:|\n|Dinh - 2021|45|50|42|1.14 [0.96, 1.34]|\n|Total (95% CI)|45|50|42|1.14 [0.96, 1.34]|\n](1/parseimages/figured02dbc91eb8a4910bbe65ce54e009c7c_0_2.png)

G) Sub-analysis #2.5: Clinical cure - short follow-up: PSI score >70

![|Study or Subgroup|Less than 5 days Events|5 or more days Events|Total Weight|Risk Ratio M-H, Fixed, 95% CI|\n|:---:|:---:|:---:|:---:|:---:|\n|Dinh - 2021|68|95|100.0%|1.15 [0.94, 1.41]|\n|Total (95% CI)|68|58|100.0%|1.15 [0.94, 1.41]|\n](1/parseimages/figure9149e773f4ba4373bc026e51074ae3a9_0_4.png)

H) Sub-analysis #2.6: Clinical cure - short follow-up: PSI score <91

![|Study or Subgroup|Less than 5 days Events|5 or more days Events|Total Weight|Risk Ratio M-H, Fixed, 95% CI|\n|:---:|:---:|:---:|:---:|:---:|\n|Dinh - 2021|74|88|100.0%|1.13 [0.97, 1.31]|\n|Total (95% CI)|74|65|100.0%|1.13 [0.97, 1.31]|\n](1/parseimages/figure7aa5894a50e04535aa74ada4052244f0_0_6.png)

I) Sub-analysis #2.7: Clinical cure – short follow-up: PSI score >90

![Study or Subgroup\nLess than 5 days\nEvents\nTotal\n0.1\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100%\n100\n](1/parseimages/figuree215a960ee2b4a7980898522637e441c_0_8.png)

J) Analysis #3: Clinical cure – long follow-up (3-4 weeks)

![Study or Subgroup\nLess than 5 days\nEvents\nTotal\n164\n211\n182\n212\n29.1%\n19.8%\n19.9%\n19.0%\n19.1%\n19.2%\n19.3%\n19.4%\n19.5%\n19.6%\n19.7%\n19.8%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9%\n19.9\n](1/parseimages/figure7fdb987088ec476fb94e8446cf484592_0_10.png)

K) Sub-analysis #3.1: Clinical cure – long follow-up: azithromycin only

Page 91 of 119

|Study or Subgroup|Less than 5 days|Less than 5 days|5 or more days|5 or more days||Risk Ratio|Risk Ratio|Risk Ratio| |---|---|---|---|---|---|---|---|---| |Study or Subgroup|Events|Total|Events|Total|Weight|M-H, Random, 95% CI|M-H, Random, 95% CI|Risk Ratio| |D'ignazio - 2005|164|211|182|212|49.3%|0.91 [0.83, 0.99]||| |Drehobl - 2005|175|202|172|209|50.7%|1.05 [0.97, 1.14]||| |Total (95% CI)||413||421|100.0%|0.98 [0.84, 1.13]||| |Total events|339||354|||||| |Heterogeneity: Tau* = 0.01; Chi* = 5.83, df = 1 (P = 0.02); P = 83%||||||||| |Test for overall effect: Z = 0.30 (P = 0.76)|||||||||

L) Sub-analysis #3.2: Clinical cure – long follow-up: outpatient only

|Study or Subgroup|Less than 5 days|Less than 5 days|5 or more days|5 or more days||Risk Ratio|Risk Ratio|Risk Ratio| |---|---|---|---|---|---|---|---|---| |Study or Subgroup|Events|Total|Events|Total|Weight|M-H, Random, 95% CI|M-H, Random, 95% CI|Risk Ratio| |D'ignazio - 2005|164|211|182|212|49.3%|0.91 [0.83, 0.99]||| |Drehobl - 2005|175|202|172|209|50.7%|1.05 [0.97, 1.14]||| |Total (95% CI)||413||421|100.0%|0.98 [0.84, 1.13]||| |Total events|339||354|||||| |Heterogeneity: Tau* = 0.01; Chi* = 5.83, df = 1 (P = 0.02); P* = 83%||||||||| |Test for overall effect: Z = 0.30 (P = 0.76)|||||||||

M) Sub-analysis #3.3: Clinical cure – long follow-up: inpatient only

|Study or Subgroup|Less than 5 days|Less than 5 days|5 or more days|5 or more days|Risk Ratio|Risk Ratio|Risk Ratio| |---|---|---|---|---|---|---|---| |Study or Subgroup|Events|Total|Events|Total|Weight|M-H, Random, 95% CI|Risk Ratio| |Dinh - 2021|105|141|107|141|49.5%|0.98 [0.86, 1.12]|| |Moussaoui - 2006|47|52|49|56|50.5%|1.03 [0.90, 1.18]|| |Total (95% CI)||193||197|100.0%|1.01 [0.92, 1.11]|| |Total events|152||156||||| |Heterogeneity: Tau* = 0.00; Chi* = 0.33, df = 1 (P = 0.57); P* = 0%|||||||| |Test for overall effect: Z = 0.15 (P = 0.88)||||||||

N) Analysis #4: Hospital length of stay

|Study or Subgroup|Less than 5 days|Less than 5 days|5 or more days|5 or more days|Mean Difference|Mean Difference| |---|---|---|---|---|---|---| |Study or Subgroup|Mean|SD|Total|Mean|SD|Mean Difference| |Dinh - 2021|6|3.7|152|6.3|151|-0.30 [-1.13, 0.53]| |Total (95% CI)||152|||151|100.0%| |Heterogeneity: Not applicable||||||-100 -50| |Test for overall effect: Z = 0.71 (P = 0.48)||||||Favors [less than 5 days]|

Page 92 of 119

Table S6: Evidence Profiles

Table S6.1

Population: Adult outpatients and inpatients with CAP who reach clinical stability Comparison: Less than five days of antibiotics versus five or more days of antibiotics Setting: All patients (outpatients and inpatients)

CI: confidence interval; RCT: randomized controlled trial; RR: risk ratio

Footnotes:

2. Drehobl, M. A., De Salvo, M. C., Lewis, D. E., Breen, J. D.. Single-dose azithromycin microspheres vs clarithromycin extended release for the treatment of mild-to-moderate community-acquired pneumonia in adults. Chest; Oct 2005.

3. Dinh, A., Ropers, J., Duran, C., Davido, B., Deconinck, L., Matt, M., Senard, O., Lagrange, A., Makhloufi, S., Mellon, G., de Lastours, V., Bouchand, F., Mathieu, E., Kahn, J. E., Rouveix, E., Grenet, J., Dumoulin, J., Chinet, T., Pépin, M., Delcey, V., Diamantis, S., Benhamou, D., Vitrat, V., Dombret, M. C., Renaud, B., Perronne, C., Claessens, Y. E., Labarère, J., Bedos, J. P., Aegerter, P., Crémieux, A. C., Discontinuing β-lactam treatment after 3 days for patients with community-acquired pneumonia in non-critical care wards (PTC): a double-blind, randomised, placebo-controlled, non-inferiority trial. Lancet; Mar 27 2021.

4. D'Ignazio, J., Camere, M. A., Lewis, D. E., Jorgensen, D., Breen, J. D. Novel, single-dose microsphere formulation of azithromycin versus 7-day levofloxacin therapy for treatment of mild to moderate community-acquired Pneumonia in adults. Antimicrob Agents Chemother; Oct 2005.

5. Risk of bias: Studies with minor protocol violations judged insufficient to warrant downgrading.

6. Inconsistency: Significant heterogeneity (defined as p-het <0.05 or 1$^{2}$ > 50%).

7. Imprecision: Wide confidence intervals, the ends of the confidence interval are likely to lead to different courses of action.

Page 93 of 119

Table S6.2

Population: Adult outpatients with CAP who reach clinical stability

Comparison: Less than five days of antibiotics versus five or more days of antibiotics

Setting: Outpatients

|Certainty assessment|Certainty assessment|Certainty assessment|Certainty assessment|Certainty assessment|Certainty assessment|Certainty assessment|Certainty assessment|Effect|Effect|Effect|Certainty|Importance| |---|---|---|---|---|---|---|---|---|---|---|---|---| |№ of studies|Study design|Risk of bias|Inconsistency|Indirectness|Imprecision|Other considerations|less than 5 days|5 or more days|Relative (95% CI)|Absolute (95% CI)|Certainty|Importance| |Clinical cure rate at 1-2 weeks||||||||||||| |$2^{1,2}$|RCT|Not $serious^{3}$|Not serious|Not serious|Not serious|none|361/413 (87.4%)|387/414 (93.5%)|RR-0.94 (0.90-to 0.98)RR 0.96 (0.91 to 1.01)|56 fewer per-1,000 (from 93 fewer-to 19 fewer)45 fewer per 1,000 (from 87 fewer to 10 more)|⊕○○○ Moderate|CRITICAL| |Clinical cure rate at 3-4 weeks||||||||||||| |$2^{1,2}$|RCT|Not $serious^{3}$|$Serious^{4}$|Not serious|$Serious^{5}$|none|339/413 (82.1%)|354/421 (84.1%)|RR 0.98 (0.84 to 1.13)|17 fewer per 1,000 (from 135 fewer to 109 more)|⊕⊕○○ Low|CRITICAL|

Cl: confidence interval; RCT: randomized controlled trial; RR: risk ratio

Footnotes:

1. Drehobl, M. A., De Salvo, M. C., Lewis, D. E., Breen, J. D.. Single-dose azithromycin microspheres vs clarithromycin extended release for the treatment of mild-to-moderate community-acquired pneumonia in adults. Chest; Oct 2005.

2. D'Ignazio, J., Camere, M. A., Lewis, D. E., Jorgensen, D., Breen, J. D. Novel, single-dose microsphere formulation of azithromycin versus 7-day levofloxacin therapy for treatment of mild to moderate community-acquired Pneumonia in adults. Antimicrob Agents Chemother; Oct 2005.

3. Risk of bias: Studies with minor protocol violations judged insufficient to warrant downgrading.

4. Inconsistency: Significant heterogeneity (defined as p-het <0.05 or $I^2 > 50$%).

5. Imprecision: Wide confidence intervals, the ends of the confidence interval are likely to lead to different courses of action.

Page 94 of 119

Table S6.3

Population: Adult inpatients with CAP who reach clinical stability

Comparison: Less than five days of antibiotics versus five or more days of antibiotics

Setting: Inpatients

|Certainty assessment|Certainty assessment|Certainty assessment|Certainty assessment|Certainty assessment|Certainty assessment|Certainty assessment|№ of patients|№ of patients|Effect|Effect|Certainty|Importance| |---|---|---|---|---|---|---|---|---|---|---|---|---| |№ of studies|Study design|Risk of bias|Inconsistency|Indirectness|Imprecision|Other considerations|less than 5 days|5 or more days|Relative (95% CI)|Absolute (95% CI)|Certainty|Importance| |Overall mortality|Overall mortality|Overall mortality|Overall mortality|Overall mortality|Overall mortality|Overall mortality|Overall mortality|Overall mortality|Overall mortality|Overall mortality|Overall mortality|Overall mortality| |$1^{1}$|RCT|Not $serious^{2}$|Not serious|Not serious|$Serious^{3}$|none|17/535 (3.2%)|18/371 (4.9%)|RR 0.63 (0.27 to 1.49)|18 fewer per 1,000 (from 35 fewer to 24 more)|⊕○○○ Moderate|CRITICAL| |Clinical cure rate at 1-2 weeks||||||||||||| |$2^{1,4}$|RCT|Not $serious^{2}$|$Serious^{5}$|Not serious|$Serious^{3}$|none|163/199 (81.9%)|156/206 (75.7%)|RR 1.06 (0.90 to 1.24)|45 more per 1,000 (from 76 fewer to 182 more)|⊕⊕○○ Low|CRITICAL| |Clinical cure rate at 3-4 weeks||||||||||||| |$2^{1,4}$|RCT|Not $serious^{2}$|Not serious|Not serious|$Serious^{3}$|none|152/193 (78.8%)|156/197 (79.2%)|RR 1.01 (0.92 to 1.11)|8 more per 1,000 (from 63 fewer to 87 more)|⊕○○○ Moderate|CRITICAL| |Hospital length of stay||||||||||||| |$1^{1}$|RCT|Not serious|Not serious|Not serious|$Serious^{3}$|none|172|171|-|MD 0.35 lower (1.17 lower to 0.47 higher)|⊕○○○ Moderate|IMPORTANT|

Cl: confidence interval; RR: risk ratio; MD: mean difference.

Footnotes:

1. Dinh, A., Ropers, J., Duran, C., Davido, B., Deconinck, L., Matt, M., Senard, O., Lagrange, A., Makhloufi, S., Mellon, G., de Lastours, V., Bouchand, F., Mathieu, E., Kahn, J. E., Rouveix, E., Grenet, J., Dumoulin, J., Chinet, T., Pépin, M., Delcey, V., Diamantis, S., Benhamou, D., Vitrat, V., Dombret, M. C., Renaud, B., Perronne, C., Claessens, Y. E., Labarère, J., Bedos, J. P., Aegerter, P., Crémieux, A. C., Discontinuing β-lactam treatment after 3 days for patients with community-acquired pneumonia in non-critical care wards (PTC): a double-blind, randomised, placebo-controlled, non-inferiority trial. Lancet; Mar 27 2021.

2. Risk of bias: Minor protocol violations, judged not severe enough to warrant downgrading.

3. Imprecision: wide confidence intervals, the ends of the confidence interval are likely to lead to different courses of action.

4. el Moussaoui, R., de Borgie, C. A., van den Broek, P., Hustinx, W. N., Bresser, P., van den Berk, G. E., Poley, J. W., van den Berg, B., Krouwels, F. H., Bonten, M. J., Weenink, C., Bossuyt, P. M., Speelman, P., Opmeer, B. C., Prins, J. M.. Effectiveness of discontinuing antibiotic treatment after three days versus eight days in mild to moderate-severe community acquired pneumonia: randomised, double blind study. Bmj; Jun 10 2006.

5. Inconsistency: Significant heterogeneity (defined as p-het <0.05 or I$^{2}$ > 50%).

PICO Question #4: Systemic corticosteroids for CAP

Population: Hospitalized adult CAP patients Intervention: Corticosteroids

Comparator: No corticosteroids

Outcomes:

Critical

Mortality

Treatment/clinical failure Clinical stability

Adverse drug events

Important

Symptoms

Disability or return to independence/function Length of stay

Antibiotic days

Search strategy

admission"[All Fields])) AND (((("Community-Acquired Infections"[MeSH Terms] OR ("community

acquired"[All Fields] AND "infections"[All Fields]) OR "Community-Acquired Infections"[All Fields] OR

("community"[All Fields] AND "acquired"[All Fields] AND "infection"[All Fields]) OR "community acquired infection"[All Fields] OR "Community-Acquired Infections"[MeSH Terms]) AND ("pneumonia"[MeSH

Terms] OR "pneumonia"[All Fields] OR "pneumonias"[All Fields] OR "pneumoniae"[All Fields] OR

"pneumoniae s"[All Fields])) AND (("respiratory tract infections"[MeSH Terms] OR ("respiratory"[All

Fields] AND "tract"[All Fields] AND "infections"[All Fields]) OR "respiratory tract infections"[All Fields] ))) AND ((english[Filter]) AND (alladult[Filter])))) AND ((("Adrenal Cortex Hormones"[Mesh])) OR

"Steroids/therapeutic use"[Mesh] OR Corticosteroid*) Filters: Adult: 19+ years, English Sort by: Most

Recent

Page 95 of 119

Page 96 of 119

![92 records identified from literature search\n\n0 duplicate records removed before screening\n\n67 records removed*\n-38 for study type\n-28 for drug\n-15 for publication type\n-6 for outcome\n-2 for population\n\n25 reports assessed for eligibility\n\n10 studies excluded\n-4 for study type\n-2 for intervention\n-2 for pneumonia type\n-1 for no clinical outcomes\n-1 unavailable\n\nsome studies removed for multiple reasons\n](1/parseimages/figure89c6c4ad1b5c4901b18eac5d76f1d66a_0_2.png) Page 97 of 119

|Study|Type of Study|Location|Number of Subjects|Population|Intervention|Outcome s|Risk of Bias| |---|---|---|---|---|---|---|---| |Blum 2015|RCT|Switzerland|785|Hospital admission with $CAP^{a}$|Prednisone 50 mg daily for 7 days|Mortality Clinical stability Adverse drug events Length of stay Antibiotic duration|None| |Confalonie ri 2005|RCT|Italy|46|ICU admission with severe $CAP^{b,c}$|Hydrocortisone 200 mg IV bolus followed by 10 mg/hour for 7 days|Mortality Adverse drug events Length of stay|Serious s| |Dequin 2023|RCT|France|800|ICU admission with severe $CAP^{d}$|Hydrocortisone 200 mg per day for 4 days with predefined criteria to administer for a total of 8 or 14 days with gradual taper|Mortality Adverse drug events|Serious s| |Fernandez 2011|RCT|Spain|45|Hospital admission with $CAP^{e}$|Methylprednisole 200 mg IV followed by 20 mg/6 h for 3 days, then 20 mg/12 h for 3 days, then 20 mg/day for 3 days|Mortality Length of stay|Serious s| |Fitzgerald 2022|RCT|Multination al|79|Hospital admission with CAP and new pleural effusion|Dexamethasone 4 mg IV every 12 hours for 48 hours|Clinical stability Adverse drug events|Very Serious s|

Page 98 of 119

|||||||Length of stay Antibiotic duration|| |---|---|---|---|---|---|---|---| |Marik 1993|RCT|South Africa|30|ICU admission with CAPf|Hydrocortisone 10 mg/kg|Mortality Length of stay|Very Serious| |Meduri 2022|RCT|USA|584|ICU admission with severe CAPg|Methylprednisone 40 mg IV, then 40 mg/day days 1-7, 20 mg/day days 8-14, 12 mg/day days 15-17, and 4 mg/day days 18-20|Mortality Adverse drug events Length of stay|Serious s| |Meijvis 2011|RCT|Netherlands|304|Hospital admission with CAPh|Dexamethasone 5 mg IV followed by 5 mg daily for 3 days|Mortality Length of stay Antibiotic duration|None| |Mikami 2007|RCT|Japan|31|Hospital admission with CAPi|Prednisolone 40 mg IV daily for 3 days|Length of stay Antibiotic duration|Very Serious s| |Nafae 2013|RCT|Egypt|80|Hospital admission with CAPh|Hydrocortisone 200 mg IV bolus followed by 10 mg/hour for 7 days|Mortality Length of stay Antibiotic duration|Serious s| |Sabry 2011|RCT|Egypt|80|ICU admission with CAPg|Hydrocortisone 200 mg IV bolus followed by 12.5 mg/hour for 7 days|Mortality Length of stay Antibiotic duration|Serious s| |Snijders 2010|RCT|Netherlands|213|Hospital admission with CAPi|Prednisone 40 mg daily for 7 days|Mortality Treatment failure Clinical stability|Serious s|

Page 99 of 119

|||||||Adverse drug events Length of stay|| |---|---|---|---|---|---|---|---| |Torres 2015|RCT|Spain|120|Hospital admission with severe CAPc,h|Methylprednisolo ne 0.5 mg/kg IV every 12 hours for 5 days|Mortality Treatment failure Clinical stability Length of stay|Serious s| |Wagner 1956|RCT|USA|113|Hospital admission with confirmed pneumococc al pneumonia|Hydrocortisone PO 80 mg once, then 60 mg every 6 hours for 3 doses, then 40 mg every 6 hours for 4 doses, then 20 mg every 6 hours for 4 doses, then 10 mg every 6 hours for 4 doses, then 10 mg every 12 hours for 2 doses|Mortality|Very Serious s| |Witterman s 2021|RCT|Netherlands|412|Hospital admission with CAPk|Dexamethasone 6 mg PO daily for 4 days|Mortality Length of stay|Serious s|

ªNiederman MS, Mandell LA, Anzueto A, et al. Guidelines for the management of adults with community-acquired pneumonia. Diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med 2001;163:1730–54.

ᵇNiederman MS, Bass JB Jr, Campbell GD, et al. Guidelines for the initial management of adults with community-acquired pneumonia: diagnosis, assessment of severity, and initial antimicrobial therapy. American Thoracic Society. Medical Section of the American Lung Association. Am Rev Respir Dis 1993;148:1418-1426.

`c`Ewig S, Ruiz M, Mensa J, et al. Severe community-acquired pneumonia: assessment of severity criteria. Am J Respir Crit Care Med 1998;158:1102–1108.

`dDiagnosis of Community-Acquired Pneumonia (CAP)` suggested by at least two of the following: cough, purulent sputum, chest pain, dyspnea + Focal shadowing/infiltrate on chest X-ray or CT-scan + one of the following: Pneumonia Severity Index (PSI) > 130 (Fine class V), Patient placed on mechanical ventilation (invasive or not) for acute respiratory failure, with a PEEP level of 5 cm of water or more, Patient treated by high-flow oxygen therapy with a FiO2 of 50% or more and a PaO2:FiO2 ratio lower than 300, or Patient treated by oxygen therapy with a partial rebreathing-mask with a reservoir bag,

provided that the PaO2 is less than 180 mmHg for oxygen flow 6 L/min, 210 mmHg for oxygen flow 7

L/min, 240 mmHg for 8 L/min, 270 mmHg for 9 L/min, or 300 mHg for 10 L/min or more

ePneumonia based on presence of a lung radiographic opacity and at least two of the following

conditions: fever (>38.5°C), purulent expectoration, pleuritic chest pain, or leukocytosis (white blood cell count of >10,000/mm3) + with extensive radiographic consolidations (affecting at least two lobes) and

respiratory failure (ratio of partial O2 pressure to the fraction of inspired O2, <300)

fBritish Thoracic Society Research Committee. Community-acquired pneumonia in adults in British

Hospitals in 1982-1983: a survey of aetiology, mortality, prognostic factors and outcome. Q J Med

1987;62:195-220

gMandell LA, Wunderink RG, Anzueto A, et al. Infectious diseases society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults.

2007;Clin Infect Dis 44(Suppl 2):S27-72

hFine MJ, Singer DE, Hanusa BH, et al. Validation of a pneumonia prognostic index using the

MedisGroups Comparative Hospital Database. Am J Med 993;94:153–59.

IClinical signs and symptoms of lower respiratory tract infections + Radiographic abnormalities

consistent with infection neither preexisting nor caused by any other previous conditions

jClinical symptoms suggestive of CAP: cough (with or without sputum), fever (.38.58C), pleuritic chest

pain, or dyspnea + new consolidations on chest radiograph.

kNew opacities on chest radiography, and two of the following signs and symptoms: cough, production

of sputum, temperature >38.0°C or <36.0°C, abnormalities at auscultation consistent with pneumonia,

C-reactive protein (CRP) >15 mg/L, white blood cell count >10×109 or <4×109 cell/L, or >10% of bands in leukocyte differentiation

Page 100 of 119

Page 101 of 119

Figure S8: Forest plots

A) Analysis #1: Mortality in CAP

![Study or Subgroup Events Corticosteroids Events Total Events Total Weight M-H, Random, 95% CI M-H, Random, 95% CI\n1.1.2 RCT Only Blum 2015 Blum 2015 Confalonieri 2005 Dequin 2023 25 400 3 47 395 24.6% Fernandez-Serrano S2011 1 23 1 22 1 24\nMarik 1993 1 14 3 16 17 27 39 1.24\nMeduri 2022 47 286 50 277 38.8% 0.91 [0.63, 1.31]\nMeijis 2011 9 151 11 153 7.7% 0.83 [0.35, 1.94]\nNafae 2013 4 60 6 20 4.2% 0.33 [0.07, 0.71]\nSabry 2011 2 40 6 10 49 1.47 0.83 [0.05, 3.15]\nTorres 2015 6 61 9 59 6.0% 0.64 [0.24, 1.70]\nWagner 1956 1 52 1 61 0.81 [1.17, 0.08, 18.30]\nWittermans 2021 4 203 7 198 3.8% 0.56 [0.17, 1.87]\nSubtotal (95% CI) 1809 1766 100.0% 0.68 [0.53, 0.86]\nTotal events 111 161 0.68 [0.53, 0.86]\nTotal (95% CI) 1809 1766 100.0% 0.68 [0.53, 0.86]\nTotal events 111 161 0.68 [0.53, 0.86]\nTHeogeneity: Tau* = 0.00; Chi* = 12.22, df = 12 (P = 0.43); Pi* = 2% Test for overall effect: Z = 3.20 (P = 0.001)\nTHeogeneity: Tau* = 0.00; Chi* = 12.22, df = 12 (P = 0.43); Pi* = 2% Test for overall effect: Z = 3.20 (P = 0.001)\nTest for subgroup differences: Not applicable\n#B) Analysis #2: Mortality in Severe CAP\n](1/parseimages/figure81d666e8a98f4fb69d018c4acaea4a01_0_3.png)

B) Analysis #2: Mortality in Severe CAP

![Study or Subgroup Events Corticosteroids Control Weight Risk Ratio M-H, Random, 95% CI 1.11.4 Severe CAP Only Confalonieri 2005 0 23 4 7 23 2.2% 0.07 [0.00, 1.10] Dequin 2023 25 400 47 395 33.5% 0.53 [0.33, 0.84] Marik 1993 1 14 3 16 3.6% 0.38 [0.04, 3.26] Meduri 2022 47 286 50 277 39.9% 0.91 [0.63, 1.31] Sabry 2011 2 40 6 40 6.6% 0.33 [0.07, 1.55] Torres 2015 6 81 9 59 14.2% 0.64 [0.24, 1.70] Subtotal (95% CI) 824 810 100.0% 0.62 [0.41, 0.94] Test for overall effect: Z = 2.24 (P = 0.02) Total (95% CI) 824 810 100.0% 0.001 0.1 1 10 1000 Test for overall effect: Z = 2.24 (P = 0.02) Test for subgroup differences: Not applicable\n](1/parseimages/figure4475d086900940d6a14c220abc2782b6_0_5.png) Page 102 of 119

C) Analysis #3: Mortality in CAP Patients Admitted to Ward

![Study or Subgroup Events Corticosteroids Events Total Events Total Weightes Risk Ratio M-H, Random, 95% CI 1.1.1.5 Ward Patients Only Meijis 2011 Wittermans 2021 4 203 7 198 33.0 Subtotal (95% CI) Total events Heterogeneity: Tau* = 0.00; Chi* = 0.28, df= 1 (P = 0.60); I* = 0% Test for overall effect: Z = 0.90 (P = 0.37)\n\nTotal events Heterogeneity: Tau* = 0.00; Chi* = 0.28, df= 1 (P = 0.60); I* = 0% Test for overall effect: Z = 0.90 (P = 0.37)\n\nTotal (95% CI) Total events 13 354 354 351 100.0% R=300 0.001 0.001 0.1 10 10000\n####T.1.1.5 Ward Patients Only\nMeijis 2011 9 151 11 153 67.0% R=300 0.83 [0.35, 1.94]\nWittermans 2021 4 203 7 198 33.0% 0.56 [0.17, 1.87]\nSubtotal (95% CI) Total events 13 351 100.0% I* = 0% Test for overall effect: Z = 0.90 (P = 0.37)\n\nTotal events Heterogeneity: Tau* = 0.00; Chi* = 0.28, df= 1 (P = 0.60); I* = 0% Test for overall effect: Z = 0.90 (P = 0.37)\n\n####T.1.1.5 Ward Patients Only\nMeijis 2011 9 151 11 153 67.0% R=300 0.83 [0.35, 1.94]\nWittermans 2021 4 203 7 198 33.0% 0.56 [0.17, 1.87]\nSubtotal (95% CI) Total events 13 351 100.0% I* = 0% Test for overall effect: Z = 0.90 (P = 0.37)\n\nTotal events\n1.1.1.5 Ward Patients Only\nMeijis 2011 9 151 11 153 67.0% R=300 0.83 [0.35, 1.94]\nMeijis 2011 9 151 11 153 67.0% R=300 0.83 [0.35, 1.94]\n####T.1.1.5 Ward Patients Only\nStudy or Subgroup differences: Not applicable\n\nCorticosteroids\nControl\n(1)\n](1/parseimages/figure01e81822fca84ef0803cbb789fc71793_0_2.png)

D) Analysis #4: Treatment Failure in CAP

![Study or Subgroup Events Corticosteroids Control\nControl\n| 1.2.1 RCTs Only\n| Snijders 2010\n| Torres 2015\n| Subtotal (95% CI)\n| 1.78\n| 61\n| 165\n| 42\n| 166\n| 157\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n| 42\n\n](1/parseimages/figure0b01b37684b64fa0b6264e859cb7de08_0_4.png)

E) Analysis #5: Clinical Stability in CAP

![Study or Subgroup Mean Mean SD Total Mean SD Total Weight Weight Mean Difference IV, Random, 95% CI 1.3.1 RCTs Only Blum 2015 2.97 0.67 392 4.47 0.74 393 28.3% -1.50 [-1.60, -1.40] Fitzgerald 2022 1.78 0.7 51 1.28 1.11 28 27.4% 0.50 [S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+S+\n](1/parseimages/figurefd2efb2c239a44099943fc6785b64c74_0_6.png) Page 103 of 119

F) Analysis #6: Length of Stay in CAP

|Study or Subgroup|Corticosteroids|Corticosteroids|Corticosteroids|Control|Control|Control|Mean Difference IV, Random, 95% CI||| |---|---|---|---|---|---|---|---|---|---| |Study or Subgroup|Mean|SD|Total|Mean|SD|Total|Weight|Mean Difference IV, Random, 95% CI|Mean Difference IV, Random, 95% CI| |1.11.2 RCTs Only|||||||||| |Blum 2015|6.33|0.74|392|7.33|0.74|393|18.7%|-1.00 [-1.10, -0.90]|| |Confalonieri 2005|22.25|11.15|23|29.25|17.89|23|0.5%|-7.00 [-15.62, 1.62]|| |Fernandez-Serrano S2011|10.6|3.16|23|13|7.13|22|3.0%|-2.40 [-5.65, 0.85]|| |Fitzgerald 2022|7|3.81|51|6.17|2.34|28|9.8%|0.83 [-0.53, 2.19]|| |Marik 1993|4.3|3.8|14|4.6|5.9|16|2.6%|-0.30 [-3.81, 3.21]|| |Meduri 2022|7.67|5.96|297|9|8.2|287|11.2%|-1.33 [-2.50, -0.16]|| |Meijysis 2011|6.83|2.99|151|8.07|4.71|153|13.5%|-1.24 [-2.13, -0.35]|| |Mikami 2007|11.3|5.5|15|15.5|10.7|16|1.0%|-4.20 [-10.14, 1.74]|| |Nafae 2013|9.27|2.4|60|16.5|2.24|20|11.3%|-7.23 [-8.38, -6.08]|| |Snijders 2010|10|12|104|10.6|12.8|109|2.9%|-0.60 [-3.93, 2.73]|| |Torres 2015|10.83|4.94|61|11.17|5.32|59|7.0%|-0.34 [-2.18, 1.50]|| |Wittermans 2021|4.5|0.75|203|5|0.6|198|18.6%|-0.50 [-0.63, -0.37]|| |Subtotal (95% CI)|10.83|4.94|61|11.17|5.32|1324|100.0%|-1.53 [-2.14, -0.91]|| |Heterogeneity: Tau* = 0.52; Chi* = 163.82, df = 11 (P < 0.00001); P* = 93%|||||||||| |Total (95% CI)|1394||||||1324|100.0%|-1.53 [-2.14, -0.91]| |Heterogeneity: Tau* = 0.52; Chi* = 163.82, df = 11 (P < 0.00001); P* = 93%|||||||||| |Test for overall effect: Z = 4.89 (P < 0.00001)|||||||||| |Test for overall effect: Z = 4.89 (P < 0.00001)||||||||||

G) Analysis #7: Length of Stay in Severe CAP

![Study or Subgroup Mean SD Total Control Mean SD Total Weight Mean Difference IV, Random, 95% CI 1.2.3 Severe CAP\nConfalonieri 2005\nMarik 1993\n4.3\n3.8\n14\n4.6\n5.9\n6.8\n7.6\n8.0\n9\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n1.2%\n\n](1/parseimages/figurefddcabebc80a455f82f99c529d6e6adc_0_4.png) Page 104 of 119

H) Analysis #8: Length of Stay in CAP Patients Admitted to Ward

![Study or Subgroup Mean SD Total Mean 51 6.17 2.34 28 20.3% Mean Difference IV, Random, 95% CI 4.12 Ward Patients Only Fitzgerald 2022 7 3.81 51 6.17 2.34 28 20.3% Mean 2011 6.83 2.99 151 8.07 4.71 153 30.4% Mean 2012 2.35 Mikami 2007 11.3 5.5 15 15.5 10.7 16 1.8% 4.20 [-10.14, 1.74]Wittermans 2021 4.5 0.75 203 5 0.6 198 47.5% -0.50 [-0.63,-0.37]Subtotal (95% CI) 420 395 100.0% -0.52 [-1.33, 0.28]THERogeneity: Tau* = 0.35; Chi* = 7.85, df= 3 (P = 0.05); P* = 62%Test for overall effect: Z = 1.27 (P = 0.21)Heterogeneity: Tau* = 0.35; Chi* = 7.85, df= 3 (P = 0.05); P* = 62%Test for overall effect: Z = 1.27 (P = 0.21)Test for subgroup differences: Not applicable\n[1.1.1 RCTs Only\n[2.1.1 RCTs Only]\n[3.1.1 RCTs Only]\n[4.1.1 RCTs Only]\n[5.1.1 RCTs Only]\n[6.1.1 RCTs Only]\n[7.1.1 RCTs Only]\n[8.1.1 RCTs Only]\n[9.1.1 RCTs Only]\n[10.1.1 RCTs Only]\n[11.1.1 RCTs Only]\n[12.1.1 RCTs Only]\n[13.1.1 RCTs Only]\n[14.1.1 RCTs Only]\n[15.1.1 RCTs Only]\n[16.1.1 RCTs Only]\n[17.1.1 RCTs Only]\n[18.1.1 RCTs Only]\n[19.1.1 RCTs Only]\n[10.1.1 RCTs Only]\n[11.1.1 RCTs Only]\n[12.1.1 RCTs Only]\n[13.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RCTs Only 2.0\n[14.1.1 RC\n](1/parseimages/figure1e021ce90f58401d8b0ab9bbb2e457cc_0_2.png)

I) Analysis #9: Antibiotic Duration in CAP

![Study or Subgroup Mean SD Control Mean Control Control\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q\n](1/parseimages/figure6785030d856a4d97a931f8a1fab9093d_0_4.png)

J) Analysis #10: Antibiotic Duration in CAP Patients Admitted to Ward

![Study or Subgroup Mean SD Total Mean SD Total Mean SD Total Weight Name Difference IV, Random, 95% CI 1.5.2 Ward Patients Only\nFitzgerald 2022\n22.4 15.4 51 20.4 13.8 28 14.5%\nMean Difference IV, Random, 95% CI 1.5.2 Ward Patients Only\nFitzgerald 2022\n21.5 15.4 51 20.4 13.8 28 14.5%\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitz�gerald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.24 Ward Patients Only\nFitzgald 2022\n20.00 [-46.3, 63]\nMean Difference IV, Random, 95% CI 1.5.2\n](1/parseimages/figure4af10da1ed24458eb0f96d9ce16e7cda_0_6.png) Page 105 of 119

K) Analysis #11: Adverse Events in CAP

|Study or Subgroup|Corticosteroids Events|Control Total|Events|Total|Weight|Risk Ratio M-H, Random, 95% CI|Risk Ratio M-H, Random, 95% CI| |---|---|---|---|---|---|---|---| |1.6.1 RCTs Only|||||||| |Blum 2015|96|392|61|393|19.1%|1.58 [1.18, 2.11]|M-H, Random, 95% CI| |Confalonieri 2005|18|23|6|23|9.7%|3.00 [1.46, 6.17]|| |Dequin 2023|63|400|88|395|19.0%|0.71 [0.53, 0.95]|| |Fitzgerald 2022|25|51|10|28|12.4%|1.37 [0.78, 2.43]|| |Meduri 2022|167|297|162|287|22.2%|1.00 [0.86, 1.15]|| |Snijders 2010|41|104|37|109|17.5%|1.16 [0.82, 1.65]|| |Subtotal (95% CI)|1267|1267||1235|100.0%|1.21 [0.90, 1.62]|| |Total events|410||364||||| |Heterogeneity: Tau² = 0.10; Chi² = 24.53, df = 5 (P = 0.0002); P = 80%|||||||| |Test for overall effect: Z = 1.29 (P = 0.20)|||||||| |Total events|410||364||||| |Heterogeneity: Tau² = 0.10; Chi² = 24.53, df = 5 (P = 0.0002); P = 80%|||||||| |Test for overall effect: Z = 1.29 (P = 0.20)|||||||| |Test for subgroup differences: Not applicable||||||||

L) Analysis #12: Adverse Events in Severe CAP

![Study or Subgroup Events\nCorticosteroids Events\nTakY = 0.10; ChiP = 24.53, df = 5 (P = 0.0002); P = 80% Test for overall effect: Z = 1.29 (P = 0.20) - 1.40; ChiP = 14.20, df = 2 (P = 0.0002); P = 80% Total events 410 - 1.40; ChiP = 24.53, df = 5 (P = 0.0002); P = 80% Test for overall effect: Z = 0.45 (P = 0.20) - 1.40; ChiP = 24.53, df = 5 (P = 0.0002); P = 80% Test for overall effect: Z = 0.45 (P = 0.20) - 1.40; ChiP = 24.53, df = 5 (P = 0.0002); P = 80% Test for overall effect: Z = 0.45 (P = 0.20) - 1.40; ChiP = 24.53, df = 5 (P = 0.0002); P = 80% Test for subgroup differences: Not applicable\n](1/parseimages/figure2f11adfbb04241bb995ccf68d1754921_0_4.png)

M) Analysis #13: Hyperglycemia in CAP

![Study or Subgroup Events Control Events Total Events Total Weight M-H, Random, 95% CI Cl M-H, Random, 95% CI Blumzard 2022\n13.1.2 Severe CAP Definonier 2005\n18.2\n19.3\n12.3\n14.2\n15.3\n16.2\n17.2\n18.3\n19.4\n19.5\n19.6\n19.7\n19.8\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.9\n19.\n](1/parseimages/figure63428ec21da44673bb791fd6e5859443_0_6.png) Page 106 of 119

N) Analysis #14: Gastrointestinal Bleeding in CAP

|Study or Subgroup|Corticosteroids|Corticosteroids|Control|Control|Risk Ratio|Risk Ratio|Risk Ratio| |---|---|---|---|---|---|---|---| |Study or Subgroup|Events|Total|Events|Total|Weight|M-H, Random, 95% CI|Risk Ratio| |Blum 2015|3|392|4|393|14.9%|0.75 [0.17, 3.34]|| |Confalonieri 2005|1|23|1|23|4.5%|1.00 [0.07, 15.04]|| |Dequin 2023|9|400|13|395|47.2%|0.68 [0.30, 1.58]|| |Meduri 2022|6|297|4|287|21.1%|1.45 [0.41, 5.08]|| |Sabry 2011|2|40|2|40|9.1%|1.00 [0.15, 6.76]|| |Torres 2015|0|61|1|59|3.3%|0.32 [0.01, 7.76]|| |Total (95% CI)||1213||1197|100.0%|0.83 [0.47, 1.48]|| |Total events|21||25||||| |Heterogeneity: $Tau^{2}$ = 0.00; $Chi^{2}$ = 1.37, df = 5 (P = 0.93); $F^{2}$ = 0% Test for overall effect: Z = 0.62 (P = 0.54)||||||||

O) Analysis #15: Neuropsychiatric Events in CAP

|Study or Subgroup|Corticosteroids|Corticosteroids|Control|Control|Risk Ratio|Risk Ratio|Risk Ratio| |---|---|---|---|---|---|---|---| |Study or Subgroup|Events|Total|Events|Total|Weight|M-H, Random, 95% CI|Risk Ratio| |Blum 2015|5|392|2|393|14.4%|2.51 [0.49, 12.84]|| |Fitzgerald 2022|2|51|0|28|4.3%|2.79 [0.14, 56.13]|| |Meduri 2022|4|297|3|287|17.3%|1.29 [0.29, 5.71]|| |Snijders 2010|4|104|3|109|17.7%|1.40 [0.32, 6.09]|| |Torres 2015|1|61|0|59|3.8%|2.90 [0.12, 69.87]|| |Wittermans 2021|10|209|7|293|42.6%|2.00 [0.77, 5.18]|| |Total (95% CI)|1114|1114|1169|1169|100.0%|1.85 [1.00, 3.44]|| |Total events|26|26|15|15|||| |Heterogeneity: $Tau^{2}$ = 0.00; $Chi^{2}$ = 0.68, df = 5 (P = 0.98); $P^{2}$ = 0% Test for overall effect: $Z = 1.95$ (P = 0.05)||||||||

P) Analysis #16: Nosocomial Infection in CAP

|Study or Subgroup|Corticosteroids Events|Control Events|Control Events|Weight|Risk Ratio M-H, Random, 95% CI|Risk Ratio M-H, Random, 95% CI| |---|---|---|---|---|---|---| |Blum 2015|13|392|14|393|11.3%|0.93 [0.44, 1.95]| |Confalonieri 2005|0|23|4|23|0.8%|0.11 [0.01, 1.95]| |Dequin 2023|39|400|44|395|31.4%|0.88 [0.58, 1.32]| |Fitzgerald 2022|2|51|3|28|2.2%|0.37 [0.06, 2.06]| |Meduri 2022|59|297|56|287|43.4%|1.02 [0.73, 1.41]| |Meijvis 2011|7|151|5|153|5.1%|1.42 [0.46, 4.37]| |Snijders 2010|10|104|4|109|5.1%|2.62 [0.85, 8.09]| |Torres 2015|1|61|0|59|0.7%|2.90 [0.12, 69.87]| |Total (95% CI)||1479||1447|100.0%|0.99 [0.76, 1.29]| |Total events|131|||130||| |Heterogeneity: $Tau^{2}$ = 0.01; $Chi^{2}$ = 7.61, df = 7 (P = 0.37); $F^{2}$ = 8%||||||| |Test for overall effect: Z = 0.06 (P = 0.95)|||||||

Page 107 of 119

Table S8: Evidence profiles

Table S8.1

Population: Adults with CAP Comparison: Systemic corticosteroids versus no systemic corticosteroids Setting: All patients (outpatients and inpatients)

|Quality assessment|Quality assessment|Quality assessment|Quality assessment|Quality assessment|Quality assessment|Quality assessment|Summary of Findings|Summary of Findings|Summary of Findings|Summary of Findings|Quality|Importance| |---|---|---|---|---|---|---|---|---|---|---|---|---| |No of studies|Design|Risk of bias|Inconsistency|Indirectness|Imprecision|Other|Steroids|Placebo|Relative Effect (95% CI)|Absolute Effect (95% CI)|Quality|Importance| |Mortality|Mortality|Mortality|Mortality|Mortality|Mortality|Mortality|Mortality|Mortality|Mortality|Mortality|Mortality|Mortality| |131-13|RCT|Serious14|Not serious|Not serious|Not serious|None|111/1809 (6.1%)|161/1766 (9.1%)|RR 0.68 (0.53 to 0.86)|29.12 per 1000 (42.77 to 12.74)|⊕○○○ Moderate|CRITICAL| |Treatment Failure|Treatment Failure|Treatment Failure|Treatment Failure|Treatment Failure|Treatment Failure|Treatment Failure|Treatment Failure|Treatment Failure|Treatment Failure|Treatment Failure|Treatment Failure|Treatment Failure| |210-11|RCT|Not serious|Serious15|Not serious|Serious16|None|42/165 (25.5%)|42/168 (25.0%)|RR 0.83 (0.25 to 2.80)|52.53 per 1000 (231.75 to - 556.20)|⊕⊕○○ Low|CRITICAL| |Clinical Stability|Clinical Stability|Clinical Stability|Clinical Stability|Clinical Stability|Clinical Stability|Clinical Stability|Clinical Stability|Clinical Stability|Clinical Stability|Clinical Stability|Clinical Stability|Clinical Stability| |41,10-11,17|RCT|Serious14|Serious15|Not serious|Serious18|None|682|654|-|MD-0.45 (-1.77 to 0.86)|⊕○○○ Very Low|CRITICAL| |Adverse Events|Adverse Events|Adverse Events|Adverse Events|Adverse Events|Adverse Events|Adverse Events|Adverse Events|Adverse Events|Adverse Events|Adverse Events|Adverse Events|Adverse Events| |61,2-3,6,10,17|RCT|Not serious|Serious15|Not serious|Serious18|None|410/1267 (32.4%)|364/1235 (29.5%)|RR 1.21 (0.90 to 1.62)|-61.95 per 1000 (29.5 to - 182.9)|⊕⊕○○ Low|CRITICAL| |Length of Stay|Length of Stay|Length of Stay|Length of Stay|Length of Stay|Length of Stay|Length of Stay|Length of Stay|Length of Stay|Length of Stay|Length of Stay|Length of Stay|Length of Stay| |121,2,4-8,10,11,13-17|RCT|Serious14|Serious15|Not serious18|Not serious18|None|1394|1324|-|MD-1.53 (-2.14 to - 0.91)|⊕⊕○○ Low|IMPORTANT| |Antibiotic Duration|Antibiotic Duration|Antibiotic Duration|Antibiotic Duration|Antibiotic Duration|Antibiotic Duration|Antibiotic Duration|Antibiotic Duration|Antibiotic Duration|Antibiotic Duration|Antibiotic Duration|Antibiotic Duration|Antibiotic Duration| |51,7-8,17-19|RCT|Not serious|Serious15|Not serious18,20|Serious16|None|669|610|-|MD-2.01 (-4.46 to 0.45)|⊕○○○ Very Low|IMPORTANT| |Footnotes:|Footnotes:|Footnotes:|Footnotes:|Footnotes:|Footnotes:|Footnotes:|Footnotes:|Footnotes:|Footnotes:|Footnotes:|Footnotes:|Footnotes:|

Cl: confidence interval; RCT: randomized controlled trial; RR: risk ratio; MD: mean difference.

Footnotes:

$^{1}$ Blum CA, Nigro N, Briel M, et al. Adjunct prednisone therapy for patients with community-acquired pneumonia: a multicentre, double-blind, randomised, placebo-controlled trial. Lancet. 2015;385(9977):1511-8. doi: 10.1016/S0140-6736(14)62447-8.

Am J Respir Crit Care Med. 2005;171(3):242-8. doi: 10.1164/rccm.200406-808OC. Epub 2004 Nov 19. PMID: 15557131.

3. Dequin PF, Meziani F, Quenot JP, et al. Hydrocortisone in Severe Community-Acquired Pneumonia. N Engl J Med. 2023. doi: 10.1056/NEJMoa2215145. Epub haïd of print. PMID: 36942789.

4. Fernández-Serrano S, Dorca J, Garcia-Vidal C, et al. Effect of corticosteroids on the clinical course of community-acquired pneumonia: a randomized controlled trial. Crit Care. 2011;15(2):R96. doi: 10.1186/cc10103.

* Marik P, Kraus P, Sribante J, Havlik I, Lipman J, Johnson DW. Hydrocortisone and tumor necrosis factor in severe community-acquired pneumonia. A randomized controlled study. Chest. 1993;104(2):389-92. doi: 10.1378/chest.104.2.389.

6 Meduri GU, Shih MC, Bridges, L. et al. Low-dose methylprednisolone treatment in critically ill patients with severe community-acquired pneumonia. Intensive Care Med. 2022;48(8):1009-1023. doi: 10.1007/s00134-022-06684-3.

7 Meijvis SC, Hardeman H, Remmelts HH, et al. Dexamethasone and length of hospital stay in patients with community-acquired pneumonia: a randomised, double-blind, placebo-controlled trial. Lancet. 2011;377(9782):2023-30. doi:10.1016/50140-6736(11)60607-7.

$^{8}$ Nafae RM, Ragab MI, Amany FM, Rashed SB. Adjuvant role of corticosteroids in the treatment of community-acquired pneumonia. Egyptian Journal of Chest Diseases and Tuberculosis. 2013;62(3):439-445. doi: 10.1016/j.ejdt.2013.03.009.

$^{9}$ Sabry N, Omar E. Corticosteroids and ICU Course of Community Acquired Pneumonia in Egyptian Settings. Pharmacology & Pharmacy. 2011;2(7):23-81. doi: 10.4236/pp.2011.2209.

10 Snijders D, Daniels JM, de Graaff CS, van der Werf TS, Boersma WG. Efficacy of corticosteroids in community-acquired pneumonia: a randomized double-blinded clinical trial. Am J Respir Crit Care Med. 2010;181(9):975-82. doi: 10.1164/rccm.200905-0808OC.

11 Torres A, Sibila O, Ferrer M, Polverino E, et al. Effect of corticosteroids on treatment failure among hospitalized patients with severe communityacquired pneumonia and high inflammatory response: a randomized clinical trial. JAMA. 2015;313(7):677-86. doi: 10.1001/jama.2015.88.

12 Wagner HN Jr, Bennett IL Jr, Lasagna L, Cluff LE, Rosenthal MB, Mirick GS. The effect of hydrocortisone upon the course of pneumococcal pneumonia treated with penicillin. Bull Johns Hopkins Hosp. 1956;98(3):197-215.

13 Wittermans E, Vestjens SMT, Spoorenberg SMC, et al. Adjunctive treatment with oral dexamethasone in non-ICU patients hospitalised with community-acquired pneumonia: a randomised clinical trial. Eur Respir J. 2021;58(2):2002535. doi: 10.1183/13993003.02535-2020.

14 Risk of bias: Several studies have either an unclear risk of bias or a high risk of bias due to outcome reported differing from protocol, protocol violations, early study termination, limited information regarding randomization or enrollment procedures, etc.

15 Inconsistency: Large heterogeneity (p<0.05 or I2 >50%)

16 Imprecision: Wide confidence intervals, defined as the ends of the confidence intervals leading to different courses of action.

17 Fitzgerald DB, Waterer GW, Budgeon C, et al. Steroid Therapy and Outcome of Parapneumonic Pleural Effusions (STOPPE): A Pilot Randomized Clinical Trial. Am J Respir Crit Care Med. 2022;205(9):1093-1101. doi: 10.1164/rccm.202107-1600OC.

18 Indirectness of the population: Fitzgerald enrolled patients with a pleural effusion who received antibiotics for a longer duration; therefore, it is likely that the population was more severely ill. However. The committee concluded that this did not alter its confidence in the estimated effects because it was only one of many studies.

19 Mikami K, Suzuki M, Kitagawa H, et al. Efficacy of corticosteroids in the treatment of community-acquired pneumonia requiring hospitalization. Lung. 2007;185(5):249-255. doi: 10.1007/s00408-007-9020-3.

20 Indirectness of outcomes: Varied definition of outcomes between studies.

Page 108 of 119

Page 109 of 119

Table S8.2

Population: Adult inpatients with severe CAP

Comparison: Systemic corticosteroids versus no systemic corticosteroids

Setting: Intensive care units

Cl: confidence interval. CO: cohort study. MD: mean difference. OBS: RCT: randomized controlled trial. RR: risk ratio.

Footnotes:

$^{1}$ Confalonieri M, Urbino R, Potena A, et al Hydrocortisone infusion for severe community-acquired pneumonia: a preliminary randomized study. Am J Respir Crit Care Med. 2005;171(3):242-8. doi: 10.1164/rccm.200406-808OC. Epub 2004 Nov 19. PMID: 15557131.

$^{2}$ Dequin PF, Meziani F, Quenot JP, et al. Hydrocortisone in Severe Community-Acquired Pneumonia. N Engl J Med. 2023. doi: 10.1056/NEJMoa2215145. Epub ahead of print. PMID: 36942789.

$^{3}$Marik P, Kraus P, Sribante J, Havlik I, Lipman J, Johnson DW. Hydrocortisone and tumor necrosis factor in severe community-acquired pneumonia. A randomized controlled study. Chest. 1993;104(2):389-92. doi: 10.1378/chest.104.2.389.

$^{4}$ Sabry N, Omar E. Corticosteroids and ICU Course of Community Acquired Pneumonia in Egyptian Settings. Pharmacology & Pharmacy. 2011;2(2):73-81. doi: 10.4236/pp.2011.22009.

$^{5}$ Meduri GU, Shih MC, Bridges L, et al. Low-dose methylprednisolone treatment in critically ill patients with severe community-acquired pneumonia. Intensive Care Med. 2022;48(8):1009-1023. doi: 10.1007/s00134-022-06684-3.

$^{6}$ Torres A, Sibila O, Ferrer M, Polverino E, et al. Effect of corticosteroids on treatment failure among hospitalized patients with severe community-acquired pneumonia and high inflammatory response: a randomized clinical trial. JAMA. 2015;313(7):677-86. doi: 10.1001/jama.2015.88.

$^{7}$ Risk of bias: Several studies have either an unclear risk of bias or a high risk of bias due to outcome reported differing from protocol, protocol violations, early study termination, limited information regarding randomization or enrollment procedures, etc.

$^{8}$ Indirectness: Meduri et al. and Torres et al. included a minority of patients who may have had non-severe CAP. However, the committee concluded that this did not alter its confidence in the estimated effects.

$^{9}$ Inconsistency: Large heterogeneity (p<0.05, I2 >50%)

$^{10}$ Imprecision: Wide confidence intervals, defined as the ends of the confidence intervals leading to different courses of action.

Page 110 of 119

Table S8.3

Population: Adult inpatients with non-severe CAP

Comparison: Systemic corticosteroids versus no systemic corticosteroids

Setting: Medical wards

Cl: confidence interval. CO: cohort study. MD: mean difference. OBS: RCT: randomized controlled trial. RR: risk ratio.

Footnotes:

$^{1}$ Meijvis SC, Hardeman H, Remmelts HH, et al. Dexamethasone and length of hospital stay in patients with community-acquired pneumonia: a randomised, double-blind, placebo-controlled trial. Lancet. 2011;377(9782):2023-30. doi: 10.1016/S0140-6736(11)60607-7.

$^{2}$ Wittermans E, Vestjens SMT, Spoorenberg SMC, et al. Adjunctive treatment with oral dexamethasone in non-ICU patients hospitalised with community-acquired pneumonia: a randomised clinical trial. Eur Respir J. 2021;58(2):2002535. doi: 10.1183/13993003.02535-2020.

$^{3}$ Risk of bias: Several studies have either an unclear risk of bias or a high risk of bias due to outcome reported differing from protocol, protocol violations, early study termination, limited information regarding randomization or enrollment procedures, etc.

$^{4}$ Indirectness of the population: Fitzgerald enrolled patients with a pleural effusion who received antibiotics for a longer duration; therefore, it is likely that the population was more severely ill. Meijvis et al. and Wittermans et al. included a minority of patients who may have had non-severe CAP. However, the committee concluded that this did not alter its confidence in the estimated effects because it was only one of many studies.

CAP. However, the committee concluded that this did not alter its confidence in the estimated effects because it was only one of many studies.

$^{5}$Imprecision: Wide confidence intervals, defined as the ends of the confidence intervals leading to different courses of action.

$^{6}$ Fitzgerald DB, Waterer GW, Budgeon C, et al. Steroid Therapy and Outcome of Parapneumonic Pleural Effusions (STOPPE): A Pilot Randomized Clinical Trial. Am J Respir Crit Care Med. 2022;205(9):1093-1101. doi: 10.1164/rccm.202107-16000C.

$^{7}$Mikami K, Suzuki M, Kitagawa H, et al. Efficacy of corticosteroids in the treatment of community-acquired pneumonia requiring hospitalization. Lung. 2007;185(5):249-255. doi: 10.1007/s00408-007-9020-3.

$^{8}$Indirectness of outcomes: Varied definition of outcomes between studies.

Page 111 of 119

Table S9: EtD frameworks

A) Analysis #1: Patients with Severe CAP

QUESTION

|Should hospitalized patients with severe community-acquired pneumonia receive systemic corticosteroids rather than no systemic corticosteroids?|| |---|---| |POPULATION:|Hospitalized patients with SEVERE community-acquired pneumonia| |INTERVENTION:|Systemic corticosteroids| |COMPARISON:|No systemic corticosteroids| |SETTING:|Inpatients|

ASSESSMENT

|Desirable Effects How substantial are the desirable effects? "Substantial" refers to both the importance of the outcomes and the magnitude of effect. As an example, a small improvement in a critical outcome might be more substantial than a large improvement in an unimportant outcome.|| |---|---| |JUDGEMENT|RESEARCH EVIDENCE| |○ Trivial ○ Small ● Moderate ○ Large ○ Varies ○ Don't know The magnitude of the effect is small but mortality is a very important outcome to most patients. Beneficial effects are seen in heterogeneous populations of patients with CAP. These effects appear to be driven by|DESIRABLE EFFECTS Decreased mortality: CAP patients- 13 RCTs, 3575 patients, 6.1% versus 9.1% (NNT 34), RR 0.68, 95% CI 0.53-0.86. Severe CAP- 6 RCTs, 1634 patients, 9.8% versus 15.1% (NNT 17), RR 0.62, 95% CI 0.41-0.94. Non-severe CAP- 3 RCTs, 705 patients, 3.7% versus 5.1% (NNT 72) RR 0.73, 95% CI 0.36 to 1.46. Decreased length of stay: CAP patients- 12 RCTs, 3403 patients, MD -1.11 days, 95% CI -1.66 to -0.55 days. Severe CAP- 4 RCTs, 780 patients, MD -1.06 days, 95% CI -2.01 to -0.12 days. Non-severe CAP- 5 RCTs, 1556 patients, MD -0.61 days, 95% CI -1.36 to 0.14 days. RCT= randomized controlled trial, NNT= number needed to treat, RR = risk ratio, CI = confidence interval, MD = mean difference.|

Page 112 of 119

|patients with severe CAP.|| |---|---| |Undesirable Effects How substantial are the undesirable effects? "Substantial" refers to both the importance of the outcomes and the magnitude of effect. As an example, a small increase in a critical adverse outcome might be more substantial than a large increase in an unimportant adverse outcome.|| |JUDGEMENT|RESEARCH EVIDENCE| |○ Trivial ● Small ○ Moderate ○ Large ○ Varies ○ Don't know The magnitude of the effect is small and is of average importance since it is treatable and reversible upon discontinuation.|UNDESIRABLE EFFECTS Increased hyperglycemia: CAP patients- 7 RCTs, 2476 patients, 14.5% versus 9% (NNH 18), RR 1.71, 95% 1.21-2.40. Severe CAP- no research evidence. Non-severe CAP- no research evidence. RCT= randomized controlled trial, NNH = number needed to harm, RR = risk ratio, CI = confidence interval.| |Certainty of evidence What is the committee's confidence in the accuracy of the above listed estimates (i.e., what is the quality of evidence)?|| |JUDGEMENT|RESEARCH EVIDENCE| |○ Very low ● Low ○ Moderate ○ High ○ No included studies|QUALITY OF EVIDENCE Overall quality of evidence is the lowest quality of evidence among the critical outcomes. For the critical outcome of adverse effects, there is low-quality evidence because there are RCTs downgraded for inconsistency and imprecision.| |Balance of effects Does the balance between desirable and undesirable effects favor the intervention or the comparison?|| |JUDGEMENT|RESEARCH EVIDENCE|

Page 113 of 119

|o Favors the comparison o Probably favors the comparison o Does not favor either the intervention or the comparison o Probably favors the intervention • Favors the intervention o Varies o Don't know|The desirable effects (decreased mortality + decreased length of stay) were judged to outweigh the undesirable effects (increased hyperglycemia), thereby favoring systemic corticosteroids in patients with severe CAP.| |---|---| |Acceptability Is the intervention acceptable to key stakeholders?|| |JUDGEMENT|RESEARCH EVIDENCE| |o No o Probably no o Probably yes • Yes o Varies o Don't know|The intervention is most often delivered intravenously or via the gastrointestinal tract, which is acceptable to most patients. This is based on the committee's non-systematic clinical observations.| |Feasibility Is the intervention feasible to implement?|| |JUDGEMENT|RESEARCH EVIDENCE| |o No o Probably no o Probably yes • Yes o Varies o Don't know|The intervention is most often delivered intravenously or via the gastrointestinal tract, which is feasible in all hospitals. This is based on the committee's non-systematic clinical observations.|

Page 114 of 119

SUMMARY OF JUDGEMENTS

||JUDGEMENT|JUDGEMENT|JUDGEMENT|JUDGEMENT|JUDGEMENT|JUDGEMENT|JUDGEMENT| |---|---|---|---|---|---|---|---| |DESIRABLE EFFECTS|Trivial|Small|Moderate|Large||Varies|Don't know| |UNDESIRABLE EFFECTS|Trivial|Small|Moderate|Large||Varies|Don't know| |BALANCE OF EFFECTS|Favors the compariso n|Probably favors the compariso n|Does not favor either the intervention or the comparison|Probably favors the interventio n|Favors the interventio n|Varies|Don't know| |CERTAINTY OF EVIDENCE|Very low|Low|Moderate|High|||No include d studies| |ACCEPTABILITY Y|No|Probably no|Probably yes|Yes||Varies|Don't know| |FEASIBILITY|No|Probably no|Probably yes|Yes||Varies|Don't know|

TYPE OF RECOMMENDATION

|Strong recommendation against the intervention|Conditional recommendation against the intervention|Conditional recommendation for either the intervention or the comparison|Conditional recommendation for the intervention|Strong recommendation for the intervention|Strong recommendation for the intervention| |---|---|---|---|---|---| |○|○|○|○|○|○|

CONCLUSIONS

Recommendation

For hospitalized patients with severe community-acquired pneumonia, we suggest systemic corticosteroids (conditional recommendation, low quality of evidence).

Participation = 16/18 (89%)

Strong recommendation for systemic corticosteroids = 1/16 (6.25%).

Conditional recommendation for systemic corticosteroids =14/16 (87.50%).

Strong recommendations against systemic corticosteroids = 0/16 (0%).

Condition recommendation against systemic corticosteroids = 1/16 (6.25%).

B) Analysis #2: Patients Admitted to Ward

Page 115 of 119

QUESTION

|Should hospitalized patients with severe community-acquired pneumonia received systemic corticosteroids rather than no systemic corticosteroids?|| |---|---| |POPULATION:|Hospitalized patients with NON-SEVERE community-acquired pneumonia| |INTERVENTION:|Systemic corticosteroids| |COMPARISON:|No systemic corticosteroids| |SETTING:|Inpatients| |||

ASSESSMENT

|Desirable Effects How substantial are the desirable effects? "Substantial" refers to both the importance of the outcomes and the magnitude of effect. As an example, a small improvement in a critical outcome might be more substantial than a large improvement in an unimportant outcome.|| |---|---| |JUDGEMENT|RESEARCH EVIDENCE| |● Trivial (if any) o Small o Moderate o Large o Varies o Don't know Beneficial effects are seen in heterogeneous populations of patients with CAP, but the effects appear to be driven by patients with severe CAP. There is no evidence of benefit in patients with non-severe CAP.|DESIRABLE EFFECTS Decreased mortality: CAP patients- 13 RCTs, 3575 patients, 6.1% versus 9.1% (NNT 33), RR 0.68, 95% CI 0.53-0.86. Severe CAP- 6 RCTs, 1634 patients, 9.8% versus 15.1% (NNT 19), RR 0.62, 95% CI 0.41-0.94. Non-severe CAP- 3 RCTs, 705 patients, 3.7% versus 5.1% (NNT 72) RR 0.73, 95% CI 0.36 to 1.46. Decreased length of stay: CAP patients- 12 RCTs, 3403 patients, MD -1.11 days, 95% CI -1.66 to -0.55 days. Severe CAP- 4 RCTs, 780 patients, MD -1.06 days, 95% CI -2.01 to -0.12 days. Non-severe CAP - 5 RCTs, 1556 patients, MD -0.61 days, 95% CI -1.36 to 0.14 days. RCT= randomized controlled trial, NNT= number needed to treat, RR = risk ratio, CI = confidence interval, MD = mean difference.| |Undesirable Effects How substantial are the undesirable effects? "Substantial" refers to both the importance of the outcomes and the magnitude of effect. As an example, a small increase in a critical adverse outcome might be more substantial than a large increase in an unimportant adverse outcome.||

Page 116 of 119

|JUDGEMENT|RESEARCH EVIDENCE| |---|---| |○ Trivial ● Small ○ Moderate ○ Large ○ Varies ○ Don't know The magnitude of the effect is small and it is average importance since it is treatable and reversible upon discontinuation.|UNDESIRABLE EFFECTS Increased hyperglycemia: CAP patients- 7 RCTs, 2476 patients, 14.5% versus 9% (NNH 18), RR 1.71, 95% 1.21-2.40. Severe CAP- no research evidence. Non-severe CAP- no research evidence. RCT= randomized controlled trial, NNH = number needed to harm, RR = risk ratio, CI = confidence interval.| |Certainty of evidence What is the committee's confidence in the accuracy of the above listed estimates (i.e., what is the quality of evidence)?|| |JUDGEMENT|RESEARCH EVIDENCE| |○ Very low ● Low ○ Moderate ○ High ○ No included studies|QUALITY OF EVIDENCE The overall quality of evidence is determined by the lowest quality of evidence among critical outcomes. For the critical outcome of mortality, there is low-quality evidence because there are RCTs downgraded for risk of bias and imprecision.| |Balance of effects Does the balance between desirable and undesirable effects favor the intervention or comparison?|| |JUDGEMENT|RESEARCH EVIDENCE| |● Favors the comparison ○ Probably favors the comparison ○ Does not favor either the intervention or the comparison ○ Probably favors|The undesirable effects (increased hyperglycemia) outweigh the desirable effects (there might be no desirable effects in non-severe CAP), thereby favoring no systemic corticosteroids in patients with non-severe CAP.|

Page 117 of 119

|the intervention o Favors the intervention o Varies o Don't know|| |---|---| |Acceptability Is the intervention acceptable to key stakeholders?|| |JUDGEMENT|RESEARCH EVIDENCE| |o No o Probably no o Probably yes ● Yes o Varies o Don't know|The intervention is most often delivered intravenously or via the gastrointestinal tract, which is acceptable to most patients. This is based on the committee's non-systematic clinical observations.| |Feasibility Is the intervention feasible to implement?|| |JUDGEMENT|RESEARCH EVIDENCE| |o No o Probably no o Probably yes ● Yes o Varies o Don't know|The intervention is most often delivered intravenously or via the gastrointestinal tract, which is feasible in all hospitals. This is based on the committee's non-systematic clinical observations.|

Page 118 of 119

SUMMARY OF JUDGEMENTS

||JUDGEMENT|JUDGEMENT|JUDGEMENT|JUDGEMENT|JUDGEMENT|JUDGEMENT|JUDGEMENT| |---|---|---|---|---|---|---|---| |DESIRABLE EFFECTS|Trivial|Small|Moderate|Large||Varies|Don't know| |UNDESIRABLE EFFECTS|Trivial|Small|Moderate|Large||Varies|Don't know| |BALANCE OF EFFECTS|Favors the comparison|Probably favors the comparison|Does not favor either the intervention or the comparison|Probably favors the intervention|Favors the intervention|Varies|Don't know| |CERTAINTY OF EVIDENCE|Very low|Low|Moderate|High|||No included studies| |ACCEPTABILITY|No|Probably no|Probably yes|Yes||Varies|Don't know| |FEASIBILITY|No|Probably no|Probably yes|Yes||Varies|Don't know|

TYPE OF RECOMMENDATION

CONCLUSIONS

Recommendation

For hospitalized patients with non-severe community-acquired pneumonia, we recommend NOT administering systemic corticosteroids (strong recommendation, low quality of evidence). Remark: The recommendation is strong despite the very low certainty of effects because the intent is to avoid harm due to unnecessary systemic corticosteroids.

Participation = 16/18 (89%) Strong recommendations for systemic corticosteroids = 0/16 (0%). Condition recommendation for systemic corticosteroids = 0/16 (0%).

Strong recommendation against systemic corticosteroids = 13/16 (81.25%).

Conditional recommendation against systemic corticosteroids =3/16 (8.75%).

Page 119 of 119

表 1
Recommendation	Strength & Evidence Quality	Factors that strengthen the recommendation	Factors that weaken the recommendation
1. Lung ultrasound versus chest x-ray to diagnose CAP.
For adults with suspected CAP, we suggest lung ultrasound is an acceptable diagnostic alternative to chest x-ray in medical centers where appropriate clinical expertise exists	Conditional Low-quality evidence	All criteria for establishing expertise met (Table 4) No availability of chest X-Ray (LUS as alternative to X-Ray) High patient risks or cost of CT scan (LUS as alternative to CT) Patient convenience and radiation exposure compared to chest x-ray & CT.	Not all criteria for establishing expertise met (Table 4) Suspicion of alternative/additional diagnoses (pulmonary embolism, malignancy) Barriers to high-quality LUS (Obesity, drains, scars, wounds, difficulty holding position)
2. Empiric antibacterial therapy for CAP with positive respiratory virus testing
For adult outpatients without co-morbidities who have clinical and imaging evidence of CAP and who test positive for a respiratory virus, we suggest not prescribing empiric antibiotics	Conditional Quality evidence	Low suspicion for bacterial co-infection (clinical history, low/normal inflammatory markers, clinical history, radiologic findings suggestive of viral etiology, viral pathogen with low prevalence of bacterial co-detection) Higher risk of harm from antibiotic exposure (History of C. difficile, severe antibiotic allergy or adverse event) Patient preference to avoid antibiotic exposure	Suspicion of bacterial co-infection (long symptom onset, "double sickening", purulent sputum, elevated inflammatory markers, radiologic findings such as consolidative infiltrate, viral pathogen with high prevalence of bacterial co-detection, exposure to Mycoplasma pneumoniae) High risk of harm if missed bacterial infection (elderly, pregnant, signs/symptoms suggestive of more severe illness) Barriers to follow-up or communication

表 2
For adult outpatients with co-morbidities who have clinical and imaging evidence of CAP and who test positive for a respiratory virus, we suggest prescribing empiric antibiotics due to concern for bacterial-viral co-infection	Conditional Very low-quality evidence	Suspicion of bacterial co-infection (long symptom onset, "double sickening", purulent sputum, elevated or increasing inflammatory markers, radiologic findings such as consolidative infiltrate) Low likelihood that virus identified explains etiology and severity of pneumonia (ie, virus with low virulence or high-risk of co-infection) High risk of harm if missed bacterial infection -High illness severity, severe symptoms -Higher number, severe, or poorly controlled comorbidities	Low suspicion of bacterial infection (clinical history, normal inflammatory markers, radiologic findings suggestive of viral etiology) High likelihood that virus identified explains etiology and severity of pneumonia (virus with high virulence, low risk of co-infection)) Lower risk of harm if missed bacterial infection -Lower illness severity -Single, mild, or well controlled comorbidities Higher risk of harm from antibiotic exposure (History of C. difficile, antibiotic allergy/ adverse event) - Patient preference to avoid antibiotic exposure
For adult inpatients with clinical and imaging evidence of non-severe CAP who test positive for a respiratory virus, we suggest prescribing empiric antibiotics due to concern for bacterial-viral co-infection	Conditional Very low-quality evidence	Same as above	Same as above
For adult inpatients with clinical and imaging evidence of severe CAP who test positive for a respiratory virus, we suggest prescribing antibiotics due to concern for bacterial-viral co-infection	Conditional Very low-quality evidence	Sepsis, severe respiratory failure, elevated or increasing inflammatory markers, Chest radiograph showing consolidation infiltrates.	Higher risk of harm from antibiotic exposure (History of C. difficile, antibiotic allergy, or antibiotic adverse event)
3. Antibiotic duration for CAP
For adult outpatients with CAP who reach clinical stability*, we suggest less than five days of antibiotics (minimum of 3 days duration), rather than five or more days of antibiotics. The duration of antibiotics should be determined based upon daily assessment of clinical stability.	Conditional Low-quality evidence	Higher risk of harm from prolonged antibiotic exposure (History of C. difficile, or an antibiotic adverse event) Patient preference to minimize antibiotic exposure	Barriers to self assessment, follow-up, or communication to ensure recovery Organism requiring longer duration (i.e., Staphylococcus aureus, Pseudomonas aeruginosa, suspected Legionella pneumophila or other intracellular microorganisms)**

表 3
			Radiographic findings (high burden of disease, necrotizing process, dense consolidations) Underlying lung disease (e.g., bronchiectasis, post-obstructive pneumonia, chronic respiratory insufficiency) Recent Hospitalization or resident in Long-term care
For adult inpatients with non-severe CAP who reach clinical stability, we suggest less than five days of antibiotics (minimum of 3 days duration), rather than five or more days of antibiotics. The duration of antibiotics should be determined based upon daily assessment of clinical stability.	Conditional Low-quality evidence	Patient preference to minimize antibiotic exposure Resolution of inflammatory markers	Organism requiring longer duration (i.e., Staphylococcus aureus, Pseudomonas aeruginosa, suspected Legionella pneumophila or other intracellular microorganisms) Pneumonia complication (e.g., empyema/parapneumonic effusion, abscess/necrotizing process, bacteremia, extrapulmonary infection) Underlying lung disease (e.g., bronchiectasis, post-obstructive pneumonia, chronic hypoxemia) Pregnancy, recent antibiotics Recent Hospitalization or resident in Long-term care
For adult inpatients with severe CAP who reach clinical stability, we suggest five or more days of antibiotics, rather than less than five days of antibiotics	Strong Low-quality evidence
4. Corticosteroids	4. Corticosteroids	4. Corticosteroids	4. Corticosteroids
For adult inpatients with non-severe CAP, we recommend NOT administering systemic corticosteroids ***	Strong Low-quality evidence
For adult inpatients with severe CAP, we suggest systemic corticosteroids	Conditional Low-quality evidence	Short time interval between symptom onset and presentation.	Longer time between symptom onset and presentation

表 4
		Ability to administer corticosteroids early after meeting criteria for severe CAP (within 24 hours)ICU-admittedRespiratory failure (P:F<300)Elevated inflammatory markers (ie, CRP, IL-6)	Longer time since onset of severe CAP (e.g. >72 hours)Lack of respiratory failureNormal or low inflammatory markersContraindications to corticosteroids (ie, influenza, Aspergillus, uncontrolled diabetes, recent gastrointestinal bleeding)Pregnancy**
Exclusion criteria from key studies*Recommendation and factors listed are for patients without another established indication of corticosteroids.

表 5
Validated definition includes either one major criterion or three or more minor criteria.
Major criteria
Septic shock with need for vasopressors
Respiratory failure requiring mechanical ventilation
Minor criteria
Respiratory rate ≥ 30 breaths/min
$Pa_{02}$/$Fi_{02}$ ratio ≤ 250
Multilobar infiltrates
Confusion/disorientation
Uremia (blood urea nitrogen level ≥ 20 mg/dl)
Leukopenia (white blood cell count < 4,000 cells/µl)
Thrombocytopenia (platelet count < 100,000/µl)
Hypothermia (core temperature < 36°C)
Hypotension requiring aggressive fluid resuscitation

表 6
Table 3. Strengths of Recommendation
	Strong Recommendation ("We recommend . . .")	Conditional Recommendation ("We suggest . . .")
For patients	The overwhelming majority of individuals in this situation would want the recommended course of action and only a small minority would not.	The majority individuals in this situation would want the suggested course of action, but a sizeable minority would not.
For clinicians	The overwhelming majority of individuals should receive the recommended course of action. Adherence to this recommendation according to the guideline could be used as a quality criterion or performance indicator. Formal decision aids are not likely to be needed to help individuals make decisions consistent with patient values and preferences.	Different choices will be appropriate for different patients, and you must help each patient arrive at a management decision consistent with her or his values and preferences. Decision aids may be useful to help individuals make decisions consistent with their values and preferences. Clinicians should expect to spend more time with patients when working towards a decision.
For policy makers	The recommendation can be adapted as policy in most situations including for the use as performance indicators if supported by high- or moderate-quality evidence.	Policymaking will require substantial debates and involvement of many stakeholders. Policies are also more likely to vary between regions. Performance indicators would have to focus on the fact that adequate deliberation about the management options has taken place.

表 7
Table 4. Key Criteria for Establishing Expertise in Lung Ultrasound Exams
Factor	Requirements
Ultrasound Equipment	Either a cart-based or handheld ultrasound device with a low-frequency ultrasound probe that provides adequate penetration, typically 14-16cm in adults, is needed to assess for pneumonia.
Training	Requisite training in LUS must provide background knowledge; practice in image acquisition, optimization, and interpretation; and knowledge in clinical integration. Mastery of LUS knowledge and skills through formal assessments should be demonstrated prior to use in clinical practice as recommended by specialty guidelines. [140-143]
Imaging Protocol	A standardized protocol evaluating the superior and inferior portions of the anterior, lateral, and posterior chest wall should be utilized.[32]
Image Archival	Dynamic ultrasound images, typically 2-4 second video loops, should be recorded, labelled per local convention, and saved in a retrievable image archival.
Documentation	Documentation of the operator, indications, exam performed, and ultrasound findings of the pleura and lung parenchyma - including location of abnormalities using standard terminology - should be included as a report within the patient's medical record.
	Findings from different imaging modalities shall be compared and periodic quality assurance checks of clinicians using LUS should be performed at the same level of radiology images. Discrepancies of imaging findings associated with negative outcomes should be reviewed for quality improvement.
Patient	Patient factors that limit LUS imaging including obesity, drains, scars, wounds, and uncooperativeness, should be considered when choosing imaging modality.

表 8
Table 5. Comorbidities that may warrant antibiotic therapy for outpatients with CAP who test positive for a respiratory virus. Conditions are ranked by % of committee members who would prescribe antibiotics for patients with each condition, in descending order.
Comorbidity (see footnotes for further definitions and examples)	% committee members who voted this condition that may warrant antibiotics
Greater than 50% agreement
Chronic pulmonary disease other than asthma	82%
End-stage liver disease	71%
End-stage renal disease	65%
Cardiovascular disease	53%
Alcoholism	53%
Neoplastic disease	53%
Less than 50% agreement
Neurological disease	47%
Chronic liver disease	35%
Malnutrition	35%
Current smoker	35%
Corticosteroid therapy* (<20mg daily or <4 weeks)	30%
Diabetes mellitus	29%
Chronic kidney disease	24%
Human immunodeficiency virus* (CD4>200)	24%
Asthma	21%
Rheumatological diseases* (not on immunosuppressants)	18%
Obesity (BMI>30)	12%
*Patients with solid organ transplant on anti-rejection medications, corticosteroid therapy more than 20mg/day for 4wks, human insufficiency virus with CD4 count <200, or rheumatological diseases on immunocompromising medication should be considered immunocompromised hosts, to whom the CAP guidelines do not apply. Refer to References 6 and 7 for guidance on diagnosis and management of these patients.
Chronic pulmonary diseases other than asthma= chronic obstructive pulmonary disease [COPD], bronchiectasis or interstitial lung disease. End-stage liver disease = ascites, variceal hemorrhage, hepatic encephalopathy, or renal impairment. End-stage renal disease =GFR <15 mL/min lasting >3 months. Solid organ transplant recipient=not taking immunosuppressive anti-rejection medication. Cardiovascular disease =congestive heart failure, coronary artery disease, or poorly controlled hypertension. Alcoholism = recurrent or ongoing alcohol use despite inability to fulfill obligations, or despite social or interpersonal problems exacerbated by alcohol use. Neoplastic disease = not taking immunosuppressive chemotherapy. Neurological disease such as Parkinson's Disease, dementia, myasthenia gravis or amyotrophic lateral sclerosis. Chronic liver disease defined as abnormal liver function tests, coagulopathy, or other evidence of chronic liver damage lasting >3 months. Malnutrition = weight loss, BMI <18.5, reduced muscle mass, or reduced food intake or assimilation. Current smoker, including cigarettes and marijuana. Corticosteroid therapy not at immunosuppressive doses such as a cumulative dose >600 mg of prednisone. Chronic kidney disease defined as GFR 15-60 mL/min, albuminuria >30 mg / 24 hours, or other markers of kidney damage lasting >3 months. Human immunodeficiency virus (HIV) with CD4 >200 and no AIDS defining illness. Rheumatological diseases such as rheumatoid arthritis or systemic lupus erythematosus, not taking immunosuppressive medication. Obesity defined as BMI >30.

表 9
Table 6. Clinical Stability Definitions* The duration of antibiotics should be determined based upon daily assessment of clinical responses.
Temperature	≤37.8° C
Heart rate	<100 beats per minute*
Respiratory rate	<24 breaths per minute*
Arterial oxygen saturation or partial pressure	SpO2 ≥90% or PaO2 ≥60 mmHg on room air* or baseline oxygen requirement*
Systolic blood pressure	≥90 mmHg
Mental status	Normal
All criteria were needed to be met to be considered “stable” in the Dinh study [89]. Prior 2007 guidelines and Uranga study [88] required afebrile plus no more than one sign of instability and used heart rate of ≤100 beats per minute, and respiratory rate of ≤24 breaths per minute. For el Moussaoui et al study, eligibility for 3-day duration was determined by improvement of 2 or more points on a respiratory symptom scale, temperature < 38°C and ability to take oral intake. *Neither Dinh nor Moussaoi studies included patients with chronic respiratory insufficiency. Thus this factor weakens this recommendation (See Table 1)

表 10
Study	Type of Study	Location	Number of Subjects	Population	Intervention	Outcomes	Risk of Bias
Amatya 2018	Observational	Nepal	62	Patients with suspected CAP and CT scan in $ED^{a}$	Lung Ultrasound, CXR	Sensitivity, Specificity **	None
Bourcier 2014	Observational	France	144	Patients with suspected CAP and CT scan in $ED^{b}$	Lung Ultrasound, CXR	Sensitivity, Specificity**	None
Cortallero 2012	Observational	Italy	120	Patients with suspected CAP and CT scan in $ED^{c}$	Lung Ultrasound, CXR	Sensitivity, Specificity **	None
Dhawan 2022	Observational	India	85	Patients with suspected CAP and CT scan in tertiary care hospital $ICUs^{d}$	Lung Ultrasound, CXR	Sensitivity, Specificity **	None
Gibbons 2021	Observational	USA	110	Patients with suspected CAP and CT scan in $ED^{e}$	Lung Ultrasound, CXR	Sensitivity, Specificity **	None
Liu 2014	Observational	China	179	Patients with suspected CAP and CT scan in $ED^{f}$	Lung Ultrasound, CXR	Sensitivity, Specificity **	None

表 11
Corradi 2015	Observational	Italy	54	Patients with suspected CAP and CT scan in $ED^{g}$	Lung Ultrasound, CXR	Sensitivity, Specificity **	None
Fares 2015	Observational	Egypt	38	Patients with suspected CAP and CT scan in a hospital $ICU^{h}$	Lung Ultrasound, CXR	Sensitivity, Specificity **	None
Karimi 2019	Observational	Iran	280	Patients with suspected CAP and CT scan in $ED^{i}$	Lung Ultrasound, CXR	Sensitivity, Specificity **	None
Taghizadieh 2015	Observational	Iran	30	Patients with suspected CAP and CT scan in $ED^{j}$	Lung Ultrasound, CXR	Sensitivity, Specificity **	None
Bitar 2018	Observational	Kuwait	82	Patients with suspected CAP and CT scan in a hospital medical-surgical $ICU^{k}$	Lung Ultrasound, CXR	Sensitivity, Specificity **	None

表 12
Study	TP	FP	FN	TN	Sensitivity (95% CI)	Specificity (95% CI)
Amatya 2018	40	11	4	7	0.91 [0.78, 0.97]	0.39 [0.17, 0.64]
Bitar 2019	31	1	1	5	0.97 [0.84, 1.00]	0.83 [0.36, 1.00]
Bourcier 2014	23	0	0	0	1.00 [0.85, 1.00]	Not estimable
Corradi 2015	30	1	14	19	0.68 [0.52, 0.81]	0.95 [0.75, 1.00]
Cortallero 2012	25	1	1	3	0.96 [0.80, 1.00]	0.75 [0.19, 0.99]
Dhawan 2022	74	0	10	1	0.88 [0.79, 0.94]	1.00 [0.03, 1.00]
Fares 2015	28	2	6	2	0.82 [0.65, 0.93]	0.50 [0.07, 0.93]
Gibbons 2021	79	9	2	20	0.98 [0.91, 1.00]	0.69 [0.49, 0.85]
Karimi 2019	232	0	48	0	0.83 [0.78, 0.87]	Not estimable
Liu 2014	106	1	6	66	0.95 [0.89, 0.98]	0.99 [0.92, 1.00]
Taghizadieh 2015	29	1	0	0	1.00 [0.88, 1.00]	0.00 [0.00, 0.97]

表 13
Study	TP	FP	FN	TN	Sensitivity (95% CI)	Specificity (95% CI)
Amatya 2018	32	9	12	9	0.73 [0.57, 0.85]	0.50 [0.26, 0.74]
Bitar 2019	5	5	27	1	0.16 [0.05, 0.33]	0.17 [0.00, 0.64]
Bourcier 2014	12	0	11	0	0.52 [0.31, 0.73]	Not estimable
Corradi 2015	28	4	16	16	0.64 [0.48, 0.78]	0.80 [0.56, 0.94]
Cortallero 2012	18	1	8	3	0.69 [0.48, 0.86]	0.75 [0.19, 0.99]
Dhawan 2022	58	1	26	0	0.69 [0.58, 0.79]	0.00 [0.00, 0.97]
Fares 2015	24	3	6	5	0.80 [0.61, 0.92]	0.63 [0.24, 0.91]
Gibbons 2021	57	13	24	16	0.70 [0.59, 0.80]	0.55 [0.36, 0.74]
Karimi 2019	263	0	17	0	0.94 [0.90, 0.96]	Not estimable
Liu 2014	87	4	25	63	0.78 [0.69, 0.85]	0.94 [0.85, 0.98]
Taghizadieh 2015	27	1	2	0	0.93 [0.77, 0.99]	0.00 [0.00, 0.97]

表 14
Quality assessment	Quality assessment	Quality assessment	Quality assessment	Quality assessment	Quality assessment	Summary of Findings	Summary of Findings	Quality	Importance
No of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other	# patients	Effect (range)	Importance
Sensitivity and specificity: Ultrasound (using chest CT scan as the reference standard)
$11^{1,2}$	Accuracy	Not Serious	Serious3	Not $serious^{4,5}$	Serious6	None	939	Sensitivity= median 95% (range 68-100%) Specificity= median 75%$^{6}$ (range 0-100%)	⊕⊕○○ LOW
Sensitivity and specificity: Chest X-Ray (using chest CT scan as the reference standard)
$11^{1,2}$	Accuracy	Not Serious	Serious5	Not $serious^{4,5}$	Serious6	None	939	Sensitivity= median 70% (range 16-94%) Specificity= median 55%$^{6}$ (range 0-94%)	⊕⊕○○ LOW
$11^{1,2}$	Accuracy	Not Serious	Serious5	Not $serious^{4,5}$	Serious6	None	939	Sensitivity= median 70% (range 16-94%) Specificity= median 55%$^{6}$ (range 0-94%)	CRITICAL LOW

表 15
POPULATION:	Patients with suspected community-acquired pneumonia
INTERVENTION:	Lung ultrasound
COMPARATOR:	Chest x-ray
SETTING:	Inpatients and outpatients

表 16
Test accuracy How accurate are the tests?
JUDGEMENT	RESEARCH EVIDENCE
○ CXR is a lot more accurate ○ CXR is slightly more accurate ○ Lung US is a lot more accurate ○ Lung US is slightly more accurate ○ Chest x-ray and US are comparably accurate	TEST CHARACTERISTICS The guideline committee judged the sensitivities and specificities of lung ultrasound and chest x-ray as comparable. The estimated medians might seem quite different but, when one considers that the lung ultrasound studies were likely performed by experienced operators, the committee concluded that the accuracy of lung ultrasound is likely overestimated in the studies compared with routine clinical practice. When one accounts for this likelihood, the committee concluded that the accuracy of lung ultrasound and chest x-ray are likely comparable. Lung ultrasound Sensitivity = median 95% (range 68-100%) Specificity = median 75% (range 0-100%) Chest x-ray Sensitivity = median 70% (range 16-94%) Specificity = median 55% (range 0-94%)
Desirable Effects How substantial are the desirable effects of making a diagnosis? “Substantial” refers to both the importance and magnitude of the desirable effects. As an example, a small improvement in a critical outcome might be considered more substantial than a large improvement in an unimportant outcome.
JUDGEMENT	RESEARCH EVIDENCE
○ CXR is a lot more likely to lead to desirable effects ○ CXR is slightly more likely to lead to desirable effects ○ Lung US is a lot more likely to lead to desirable effects ○ Lung US is slightly more likely to lead to desirable effects ○ Chest x-ray and US will lead to similar desirable effects	DESIRABLE PATIENT-IMPORTANT OUTCOMES Desirable effects derive from true positive and true negative results. They include the initiation or continuation of appropriate antibiotic therapy in those you have pneumonia, the elimination of the burdens and costs of seeking alternative diagnoses in those you have pneumonia, avoiding unnecessary antibiotic therapy in those who do not have pneumonia, and promoting ongoing pursuit of the correct diagnosis in those who do not have pneumonia. For both lung ultrasound and chest x-ray, a positive test result will result in the same intervention, antibiotics, and therefore eventually the same outcomes. Thus, the committee concluded that if test accuracy is comparable (see above), downstream desirable outcomes must also be comparable. Sensitivity = true positive rate:

表 17
	Lung ultrasound = 95% Chest x-ray = 70% Specificity = true negative rate: Lung ultrasound = 75% Chest x-ray = 55%
Undesirable Effects	Undesirable Effects
How substantial are the undesirable effects of making a diagnosis? "Substantial" refers to both the importance and magnitude of the undesirable effects. As an example, a small but important complication of diagnostic testing might be considered more substantial than a large but unimportant complication.
JUDGEMENT	RESEARCH EVIDENCE
○ CXR is a lot more likely to lead to undesirable effects ○ CXR is slightly more likely to lead to undesirable effects ○ Lung US is a lot more likely to lead to undesirable effects ○ Lung US is slightly more likely to lead to undesirable effects ○ Chest x-ray and US will lead to similar undesirable effects
Balance of desirable and undesirable effects Does the balance between desirable and undesirable effects favor the intervention or the comparison?
JUDGEMENT	RESEARCH EVIDENCE The guideline committee noted that chest x-ray and lung ultrasound probably lead to similar desirable effects (foregoing or discontinuing antibiotics, additional diagnostic testing to either confirm the negative result or seek an alternative diagnosis). Thus, the committee concluded that if test accuracy is comparable (see above), downstream undesirable outcomes must also be comparable. 1 - sensitivity = false negative rate: Lung ultrasound = 5% Chest x-ray = 30% 1 - specificity = false positive rate: Lung ultrasound = 25% Chest x-ray = 45%
Balance of desirable and undesirable effects Does the balance between desirable and undesirable effects favor the intervention or the comparison?
JUDGEMENT	RESEARCH EVIDENCE
○ Favors chest x-ray ○ Probably favors chest x-ray ● Does not favor either chest x-ray or lung ultrasound ○ Favors lung ultrasound ○ Probably favors lung ultrasound ○ Varies ○ Don't know	The guideline committee noted that chest x-ray and lung ultrasound probably lead to similar desirable and undesirable effects (see above). Therefore, they concluded that the balance of effects does not favor either chest x-ray or lung ultrasound.
Quality of evidence of test accuracy What is the committee's confidence in the above listed estimates of test accuracy (i.e., what is the quality of evidence)?
JUDGEMENT	RESEARCH EVIDENCE

表 18
○ Very low • Low • Moderate ○ High ○ No included studies	QUALITY OF EVIDENCE The quality of evidence for both lung ultrasound and chest x-rays was low because there are accuracy studies that were downgraded due to inconsistency (there were a wide range of estimates across studies) and imprecision (the confidence intervals were wide for most studies).
Quality of evidence of test result/management What is the committee's confidence that the test results will lead to certain clinical actions?
JUDGEMENT	RESEARCH EVIDENCE
○ Very low ○ Low • Moderate ○ High • No included studies	The committee did not evaluate published evidence regarding clinical actions that follow lung ultrasound and chest x-ray results. The committee concluded that the frequency of clinical actions following lung ultrasound and chest x-ray must be comparable if accuracy of studies is comparable (see above) since the tests lead to the same clinical actions. Clinical actions include the initiation/continuation antibiotic therapy or foregoing/discontinuing antibiotic therapy.
Quality of evidence of management/clinical outcomes What is the committee's confidence that the clinical actions prompted by the test results will lead to certain outcomes?
JUDGEMENT	RESEARCH EVIDENCE
○ Very low ○ Low • Moderate ○ High • No included studies	The committee did not evaluate published evidence regarding the clinical outcomes of antibiotic therapy in patients with CAP. However, the committee was confident that antibiotic therapy improves clinical outcomes in patients with CAP. The committee therefore concluded that clinical outcomes following lung ultrasound and chest x-ray must be comparable if the accuracy of studies is comparable (see above), since the tests lead to the same clinical actions which create those outcomes.
Quality of evidence What is the overall certainty of the effects of the tests? This is defined as the lowest quality of evidence among the qualities of evidence of test accuracy, result/management, and management/clinical outcomes
JUDGEMENT	RESEARCH EVIDENCE
● Very low ○ Low • Moderate ○ High ○ No included studies	The overall quality of evidence is very low because, even if there exists good evidence that test results effect clinical actions that improve outcomes, there is very low quality of evidence for test accuracy for both lung ultrasound and chest x-ray. ○ No included studies
Acceptability Is the intervention acceptable to key stakeholders?
JUDGEMENT	RESEARCH EVIDENCE
○ No ○ Probably no ○ Probably yes ● Yes ○ Varies ○ Don't know	Lung ultrasound and chest x-ray are non-invasive, painless, and not burdensome. Therefore, both are acceptable to most patients. This conclusion is based on the committee's non-systematic clinical observations.
Feasibility Is the intervention feasible to implement?

表 19
JUDGEMENT	RESEARCH EVIDENCE
o No	Both chest x-ray and lung ultrasound are available in most clinical settings. The primary limiting factor is the availability of experience operators and interpreters of lung ultrasound. This conclusion is based on the committee's non-systematic clinical observations.
o Probably no	primary limiting factor is the availability of experience operators and interpreters of lung ultrasound. This conclusion is based on the committee's non-systematic clinical observations.
o Yes	observations.
o Varies
o Don't know

表 20
	JUDGEMENT	JUDGEMENT	JUDGEMENT	JUDGEMENT	JUDGEMENT	JUDGEMENT	JUDGEMENT
TEST ACCURACY	CXR a lot more	CXR slightly more	CXR and lung US are comparable	Lung US slightly more	Lung US a lot more	-	-
DESIRABLE EFFECTS	CXR a lot more	CXR slightly more	CXR and lung US are comparable	Lung US slightly more	Lung US a lot more	-	-
UNDESIRABLE EFFECTS	CXR a lot more	CXR slightly more	CXR and lung US are comparable	Lung US slightly more	Lung US a lot more	-	-
BALANCE OF EFFECTS	Favors CXR	Probably favors CXR	Does not favor either CXR or lung US	Probably favors lung US	Favors lung US	Varies	Don't know
QUALITY OF EVIDENCE OF TEST ACCURACY	Very low	Low	Moderate	High	No included studies	-	-
QUALITY OF EVIDENCE OF TEST RESULT/MANAGEMENT	Very low	Low	Moderate	High	No included studies	-	-
QUALITY OF EVIDENCE OF MANAGEMENT/OUTCOMES	Very low	Low	Moderate	High	No included studies	-	-
QUALITY OF EVIDENCE	Very low	Low	Moderate	High	No included studies	-	-
ACCEPTABILITY	No	Probably no	Probably yes	Yes	Varies	Don't know	-
FEASIBILITY	No	Probably no	Probably yes	Yes	Varies	Don't know	-

表 21
Strong recommendation for chest x-ray	Conditional recommendation for chest x-ray	Conditional recommendation for chest x-ray or lung ultrasound	Conditional recommendation for lung ultrasound	Strong recommendation for lung ultrasound	Out
○	○	○	○	○	○	○

表 22
Study	Type of study	Location	Number of subjects (I/C)	Population	Outcomes
$^{1}El$ Moussaoui - 2006	RCT	Netherlands	119 (56/63)	Inpatients	Clinical cure
$^{2}D^{I}Ignazio$ - 2004	RCT	Worldwide (Canada, Chile, India, Lithuania, Mexico, Peru, Russia, and United States)	363 (174/189)	Outpatients	Clinical cure
$^{3}Dinh$ - 2021	RCT	France	310 (157/153)	Inpatients	Mortality, clinical cure, length of stay
$^{4}Drehobl$ - 2005	RCT	Worldwide (United States, Canada, Argentina, Russia, India, Estonia, and Lithuania)	411 (202/209)	Outpatients	Clinical cure

表 23
Study	Antibiotics (I VS C)	Intervention days	Control days
Inpatient	Inpatient	Inpatient	Inpatient
$^{1}El$ Moussaoui - 2006	Patients who improved after 3 days of IV amoxicillin were randomized to placebo or 750 mg of oral amoxicillin TID	3	8
$^{2}Dinh$ - 2021	After 72 hours of beta-lactam treatment, patients were randomized to receive placebo or 500 mg amoxicillin plus 62.5 mg of clavulanate TID	3	8
Outpatient	Outpatient	Outpatient	Outpatient
$^{3}D'Ignazio$ - 2004	A single 2 gm dose azithromycin microspheres vs 500 mg oral levofloxacin	1	7

$^{4}Drehobl$ - 2005	A single 2 gm dose azithromycin microspheres vs clarithromycin	1	7

表 24
Study or Subgroup	Less than 5 days	Less than 5 days	5 or more days	5 or more days		Risk Ratio	Risk Ratio	Risk Ratio
Study or Subgroup	Events	Total	Events	Total	Weight	M-H, Random, 95% CI	M-H, Random, 95% CI	Risk Ratio
D'ignazio - 2005	164	211	182	212	49.3%	0.91 [0.83, 0.99]
Drehobl - 2005	175	202	172	209	50.7%	1.05 [0.97, 1.14]
Total (95% CI)		413		421	100.0%	0.98 [0.84, 1.13]
Total events	339		354
Heterogeneity: Tau* = 0.01; Chi* = 5.83, df = 1 (P = 0.02); P = 83%
Test for overall effect: Z = 0.30 (P = 0.76)
Study or Subgroup	Events	Total	Events	Total	Weight	M-H, Random, 95% CI	M-H, Random, 95% CI	Risk Ratio
D'ignazio - 2005	164	211	182	212	49.3%	0.91 [0.83, 0.99]
Drehobl - 2005	175	202	172	209	50.7%	1.05 [0.97, 1.14]
Total (95% CI)		413		421	100.0%	0.98 [0.84, 1.13]
Total events	339		354
Heterogeneity: Tau* = 0.01; Chi* = 5.83, df = 1 (P = 0.02); P* = 83%
Test for overall effect: Z = 0.30 (P = 0.76)

表 26
Study or Subgroup	Less than 5 days	Less than 5 days	5 or more days	5 or more days	Risk Ratio	Risk Ratio	Risk Ratio
Study or Subgroup	Events	Total	Events	Total	Weight	M-H, Random, 95% CI	Risk Ratio
Dinh - 2021	105	141	107	141	49.5%	0.98 [0.86, 1.12]
Moussaoui - 2006	47	52	49	56	50.5%	1.03 [0.90, 1.18]
Total (95% CI)		193		197	100.0%	1.01 [0.92, 1.11]
Total events	152		156
Heterogeneity: Tau* = 0.00; Chi* = 0.33, df = 1 (P = 0.57); P* = 0%
Test for overall effect: Z = 0.15 (P = 0.88)

表 27
Study or Subgroup	Less than 5 days	Less than 5 days	5 or more days	5 or more days	Mean Difference	Mean Difference
Study or Subgroup	Mean	SD	Total	Mean	SD	Mean Difference
Dinh - 2021	6	3.7	152	6.3	151	-0.30 [-1.13, 0.53]
Total (95% CI)		152			151	100.0%
Heterogeneity: Not applicable						-100 -50
Test for overall effect: Z = 0.71 (P = 0.48)						Favors [less than 5 days]

表 28
Certainty assessment	Certainty assessment	Certainty assessment	Certainty assessment	Certainty assessment	Certainty assessment	Certainty assessment	№ of patients	№ of patients	Effect	Effect	Certainty	Importance
№ of studies	Study design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	less than 5 days	5 or more days	Relative (95% CI)	Absolute (95% CI)	Certainty	Importance
Clinical cure rate at 1-2 weeks
$4^{1,2,3,4}$	RCT	Not $serious^{5}$	$Serious^{6}$	Not serious	$Serious^{7}$	none	524/612 (85.6%)	543/620 (87.6%)	RR 0.98 (0.91 to 1.05)	18 fewer per 1,000 (from 79 fewer to 44 more)	⊕⊕○○ Low	CRITICAL
Clinical cure rate at 3-4 weeks
$4^{1,2,3,4}$	RCT	Not $serious^{5}$	$Serious^{6}$	Not serious	$Serious^{7}$	none	491/606 (81.0%)	510/618 (82.5%)	RR 0.99 (0.92 to 1.07)	8 fewer per 1,000 (from 66 fewer to 58 more)	⊕⊕○○ Low	CRITICAL

表 29
Certainty assessment	Certainty assessment	Certainty assessment	Certainty assessment	Certainty assessment	Certainty assessment	Certainty assessment	Certainty assessment	Effect	Effect	Effect	Certainty	Importance
№ of studies	Study design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	less than 5 days	5 or more days	Relative (95% CI)	Absolute (95% CI)	Certainty	Importance
Clinical cure rate at 1-2 weeks
$2^{1,2}$	RCT	Not $serious^{3}$	Not serious	Not serious	Not serious	none	361/413 (87.4%)	387/414 (93.5%)	RR-0.94 (0.90-to 0.98)RR 0.96 (0.91 to 1.01)	56 fewer per-1,000 (from 93 fewer-to 19 fewer)45 fewer per 1,000 (from 87 fewer to 10 more)	⊕○○○ Moderate	CRITICAL
Clinical cure rate at 3-4 weeks
$2^{1,2}$	RCT	Not $serious^{3}$	$Serious^{4}$	Not serious	$Serious^{5}$	none	339/413 (82.1%)	354/421 (84.1%)	RR 0.98 (0.84 to 1.13)	17 fewer per 1,000 (from 135 fewer to 109 more)	⊕⊕○○ Low	CRITICAL

表 30
Certainty assessment	Certainty assessment	Certainty assessment	Certainty assessment	Certainty assessment	Certainty assessment	Certainty assessment	№ of patients	№ of patients	Effect	Effect	Certainty	Importance
№ of studies	Study design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	less than 5 days	5 or more days	Relative (95% CI)	Absolute (95% CI)	Certainty	Importance
Overall mortality	Overall mortality	Overall mortality	Overall mortality	Overall mortality	Overall mortality	Overall mortality	Overall mortality	Overall mortality	Overall mortality	Overall mortality	Overall mortality	Overall mortality
$1^{1}$	RCT	Not $serious^{2}$	Not serious	Not serious	$Serious^{3}$	none	17/535 (3.2%)	18/371 (4.9%)	RR 0.63 (0.27 to 1.49)	18 fewer per 1,000 (from 35 fewer to 24 more)	⊕○○○ Moderate	CRITICAL
Clinical cure rate at 1-2 weeks
$2^{1,4}$	RCT	Not $serious^{2}$	$Serious^{5}$	Not serious	$Serious^{3}$	none	163/199 (81.9%)	156/206 (75.7%)	RR 1.06 (0.90 to 1.24)	45 more per 1,000 (from 76 fewer to 182 more)	⊕⊕○○ Low	CRITICAL
Clinical cure rate at 3-4 weeks
$2^{1,4}$	RCT	Not $serious^{2}$	Not serious	Not serious	$Serious^{3}$	none	152/193 (78.8%)	156/197 (79.2%)	RR 1.01 (0.92 to 1.11)	8 more per 1,000 (from 63 fewer to 87 more)	⊕○○○ Moderate	CRITICAL
Hospital length of stay
$1^{1}$	RCT	Not serious	Not serious	Not serious	$Serious^{3}$	none	172	171	-	MD 0.35 lower (1.17 lower to 0.47 higher)	⊕○○○ Moderate	IMPORTANT

表 31
Study	Type of Study	Location	Number of Subjects	Population	Intervention	Outcome s	Risk of Bias
Blum 2015	RCT	Switzerland	785	Hospital admission with $CAP^{a}$	Prednisone 50 mg daily for 7 days	Mortality Clinical stability Adverse drug events Length of stay Antibiotic duration	None
Confalonie ri 2005	RCT	Italy	46	ICU admission with severe $CAP^{b,c}$	Hydrocortisone 200 mg IV bolus followed by 10 mg/hour for 7 days	Mortality Adverse drug events Length of stay	Serious s
Dequin 2023	RCT	France	800	ICU admission with severe $CAP^{d}$	Hydrocortisone 200 mg per day for 4 days with predefined criteria to administer for a total of 8 or 14 days with gradual taper	Mortality Adverse drug events	Serious s
Fernandez 2011	RCT	Spain	45	Hospital admission with $CAP^{e}$	Methylprednisole 200 mg IV followed by 20 mg/6 h for 3 days, then 20 mg/12 h for 3 days, then 20 mg/day for 3 days	Mortality Length of stay	Serious s
Fitzgerald 2022	RCT	Multination al	79	Hospital admission with CAP and new pleural effusion	Dexamethasone 4 mg IV every 12 hours for 48 hours	Clinical stability Adverse drug events	Very Serious s

表 32
						Length of stay Antibiotic duration
Marik 1993	RCT	South Africa	30	ICU admission with CAPf	Hydrocortisone 10 mg/kg	Mortality Length of stay	Very Serious
Meduri 2022	RCT	USA	584	ICU admission with severe CAPg	Methylprednisone 40 mg IV, then 40 mg/day days 1-7, 20 mg/day days 8-14, 12 mg/day days 15-17, and 4 mg/day days 18-20	Mortality Adverse drug events Length of stay	Serious s
Meijvis 2011	RCT	Netherlands	304	Hospital admission with CAPh	Dexamethasone 5 mg IV followed by 5 mg daily for 3 days	Mortality Length of stay Antibiotic duration	None
Mikami 2007	RCT	Japan	31	Hospital admission with CAPi	Prednisolone 40 mg IV daily for 3 days	Length of stay Antibiotic duration	Very Serious s
Nafae 2013	RCT	Egypt	80	Hospital admission with CAPh	Hydrocortisone 200 mg IV bolus followed by 10 mg/hour for 7 days	Mortality Length of stay Antibiotic duration	Serious s
Sabry 2011	RCT	Egypt	80	ICU admission with CAPg	Hydrocortisone 200 mg IV bolus followed by 12.5 mg/hour for 7 days	Mortality Length of stay Antibiotic duration	Serious s
Snijders 2010	RCT	Netherlands	213	Hospital admission with CAPi	Prednisone 40 mg daily for 7 days	Mortality Treatment failure Clinical stability	Serious s

表 33
						Adverse drug events Length of stay
Torres 2015	RCT	Spain	120	Hospital admission with severe CAPc,h	Methylprednisolo ne 0.5 mg/kg IV every 12 hours for 5 days	Mortality Treatment failure Clinical stability Length of stay	Serious s
Wagner 1956	RCT	USA	113	Hospital admission with confirmed pneumococc al pneumonia	Hydrocortisone PO 80 mg once, then 60 mg every 6 hours for 3 doses, then 40 mg every 6 hours for 4 doses, then 20 mg every 6 hours for 4 doses, then 10 mg every 6 hours for 4 doses, then 10 mg every 12 hours for 2 doses	Mortality	Very Serious s
Witterman s 2021	RCT	Netherlands	412	Hospital admission with CAPk	Dexamethasone 6 mg PO daily for 4 days	Mortality Length of stay	Serious s

表 34
Study or Subgroup	Corticosteroids	Corticosteroids	Corticosteroids	Control	Control	Control	Mean Difference IV, Random, 95% CI
Study or Subgroup	Mean	SD	Total	Mean	SD	Total	Weight	Mean Difference IV, Random, 95% CI	Mean Difference IV, Random, 95% CI
1.11.2 RCTs Only
Blum 2015	6.33	0.74	392	7.33	0.74	393	18.7%	-1.00 [-1.10, -0.90]
Confalonieri 2005	22.25	11.15	23	29.25	17.89	23	0.5%	-7.00 [-15.62, 1.62]
Fernandez-Serrano S2011	10.6	3.16	23	13	7.13	22	3.0%	-2.40 [-5.65, 0.85]
Fitzgerald 2022	7	3.81	51	6.17	2.34	28	9.8%	0.83 [-0.53, 2.19]
Marik 1993	4.3	3.8	14	4.6	5.9	16	2.6%	-0.30 [-3.81, 3.21]
Meduri 2022	7.67	5.96	297	9	8.2	287	11.2%	-1.33 [-2.50, -0.16]
Meijysis 2011	6.83	2.99	151	8.07	4.71	153	13.5%	-1.24 [-2.13, -0.35]
Mikami 2007	11.3	5.5	15	15.5	10.7	16	1.0%	-4.20 [-10.14, 1.74]
Nafae 2013	9.27	2.4	60	16.5	2.24	20	11.3%	-7.23 [-8.38, -6.08]
Snijders 2010	10	12	104	10.6	12.8	109	2.9%	-0.60 [-3.93, 2.73]
Torres 2015	10.83	4.94	61	11.17	5.32	59	7.0%	-0.34 [-2.18, 1.50]
Wittermans 2021	4.5	0.75	203	5	0.6	198	18.6%	-0.50 [-0.63, -0.37]
Subtotal (95% CI)	10.83	4.94	61	11.17	5.32	1324	100.0%	-1.53 [-2.14, -0.91]
Heterogeneity: Tau* = 0.52; Chi* = 163.82, df = 11 (P < 0.00001); P* = 93%
Total (95% CI)	1394						1324	100.0%	-1.53 [-2.14, -0.91]
Heterogeneity: Tau* = 0.52; Chi* = 163.82, df = 11 (P < 0.00001); P* = 93%
Test for overall effect: Z = 4.89 (P < 0.00001)
Test for overall effect: Z = 4.89 (P < 0.00001)

表 35
Study or Subgroup	Corticosteroids Events	Control Total	Events	Total	Weight	Risk Ratio M-H, Random, 95% CI	Risk Ratio M-H, Random, 95% CI
1.6.1 RCTs Only
Blum 2015	96	392	61	393	19.1%	1.58 [1.18, 2.11]	M-H, Random, 95% CI
Confalonieri 2005	18	23	6	23	9.7%	3.00 [1.46, 6.17]
Dequin 2023	63	400	88	395	19.0%	0.71 [0.53, 0.95]
Fitzgerald 2022	25	51	10	28	12.4%	1.37 [0.78, 2.43]
Meduri 2022	167	297	162	287	22.2%	1.00 [0.86, 1.15]
Snijders 2010	41	104	37	109	17.5%	1.16 [0.82, 1.65]
Subtotal (95% CI)	1267	1267		1235	100.0%	1.21 [0.90, 1.62]
Total events	410		364
Heterogeneity: Tau² = 0.10; Chi² = 24.53, df = 5 (P = 0.0002); P = 80%
Test for overall effect: Z = 1.29 (P = 0.20)
Total events	410		364
Heterogeneity: Tau² = 0.10; Chi² = 24.53, df = 5 (P = 0.0002); P = 80%
Test for overall effect: Z = 1.29 (P = 0.20)
Test for subgroup differences: Not applicable

表 36
Study or Subgroup	Corticosteroids	Corticosteroids	Control	Control	Risk Ratio	Risk Ratio	Risk Ratio
Study or Subgroup	Events	Total	Events	Total	Weight	M-H, Random, 95% CI	Risk Ratio
Blum 2015	3	392	4	393	14.9%	0.75 [0.17, 3.34]
Confalonieri 2005	1	23	1	23	4.5%	1.00 [0.07, 15.04]
Dequin 2023	9	400	13	395	47.2%	0.68 [0.30, 1.58]
Meduri 2022	6	297	4	287	21.1%	1.45 [0.41, 5.08]
Sabry 2011	2	40	2	40	9.1%	1.00 [0.15, 6.76]
Torres 2015	0	61	1	59	3.3%	0.32 [0.01, 7.76]
Total (95% CI)		1213		1197	100.0%	0.83 [0.47, 1.48]
Total events	21		25
Heterogeneity: $Tau^{2}$ = 0.00; $Chi^{2}$ = 1.37, df = 5 (P = 0.93); $F^{2}$ = 0% Test for overall effect: Z = 0.62 (P = 0.54)
Study or Subgroup	Events	Total	Events	Total	Weight	M-H, Random, 95% CI	Risk Ratio
Blum 2015	5	392	2	393	14.4%	2.51 [0.49, 12.84]
Fitzgerald 2022	2	51	0	28	4.3%	2.79 [0.14, 56.13]
Meduri 2022	4	297	3	287	17.3%	1.29 [0.29, 5.71]
Snijders 2010	4	104	3	109	17.7%	1.40 [0.32, 6.09]
Torres 2015	1	61	0	59	3.8%	2.90 [0.12, 69.87]
Wittermans 2021	10	209	7	293	42.6%	2.00 [0.77, 5.18]
Total (95% CI)	1114	1114	1169	1169	100.0%	1.85 [1.00, 3.44]
Total events	26	26	15	15
Heterogeneity: $Tau^{2}$ = 0.00; $Chi^{2}$ = 0.68, df = 5 (P = 0.98); $P^{2}$ = 0% Test for overall effect: $Z = 1.95$ (P = 0.05)

表 38
Study or Subgroup	Corticosteroids Events	Control Events	Control Events	Weight	Risk Ratio M-H, Random, 95% CI	Risk Ratio M-H, Random, 95% CI
Blum 2015	13	392	14	393	11.3%	0.93 [0.44, 1.95]
Confalonieri 2005	0	23	4	23	0.8%	0.11 [0.01, 1.95]
Dequin 2023	39	400	44	395	31.4%	0.88 [0.58, 1.32]
Fitzgerald 2022	2	51	3	28	2.2%	0.37 [0.06, 2.06]
Meduri 2022	59	297	56	287	43.4%	1.02 [0.73, 1.41]
Meijvis 2011	7	151	5	153	5.1%	1.42 [0.46, 4.37]
Snijders 2010	10	104	4	109	5.1%	2.62 [0.85, 8.09]
Torres 2015	1	61	0	59	0.7%	2.90 [0.12, 69.87]
Total (95% CI)		1479		1447	100.0%	0.99 [0.76, 1.29]
Total events	131			130
Heterogeneity: $Tau^{2}$ = 0.01; $Chi^{2}$ = 7.61, df = 7 (P = 0.37); $F^{2}$ = 8%
Test for overall effect: Z = 0.06 (P = 0.95)

表 39
Quality assessment	Quality assessment	Quality assessment	Quality assessment	Quality assessment	Quality assessment	Quality assessment	Summary of Findings	Summary of Findings	Summary of Findings	Summary of Findings	Quality	Importance
No of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other	Steroids	Placebo	Relative Effect (95% CI)	Absolute Effect (95% CI)	Quality	Importance
Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality
131-13	RCT	Serious14	Not serious	Not serious	Not serious	None	111/1809 (6.1%)	161/1766 (9.1%)	RR 0.68 (0.53 to 0.86)	29.12 per 1000 (42.77 to 12.74)	⊕○○○ Moderate	CRITICAL
Treatment Failure	Treatment Failure	Treatment Failure	Treatment Failure	Treatment Failure	Treatment Failure	Treatment Failure	Treatment Failure	Treatment Failure	Treatment Failure	Treatment Failure	Treatment Failure	Treatment Failure
210-11	RCT	Not serious	Serious15	Not serious	Serious16	None	42/165 (25.5%)	42/168 (25.0%)	RR 0.83 (0.25 to 2.80)	52.53 per 1000 (231.75 to - 556.20)	⊕⊕○○ Low	CRITICAL
Clinical Stability	Clinical Stability	Clinical Stability	Clinical Stability	Clinical Stability	Clinical Stability	Clinical Stability	Clinical Stability	Clinical Stability	Clinical Stability	Clinical Stability	Clinical Stability	Clinical Stability
41,10-11,17	RCT	Serious14	Serious15	Not serious	Serious18	None	682	654	-	MD-0.45 (-1.77 to 0.86)	⊕○○○ Very Low	CRITICAL
Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events
61,2-3,6,10,17	RCT	Not serious	Serious15	Not serious	Serious18	None	410/1267 (32.4%)	364/1235 (29.5%)	RR 1.21 (0.90 to 1.62)	-61.95 per 1000 (29.5 to - 182.9)	⊕⊕○○ Low	CRITICAL
Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay
121,2,4-8,10,11,13-17	RCT	Serious14	Serious15	Not serious18	Not serious18	None	1394	1324	-	MD-1.53 (-2.14 to - 0.91)	⊕⊕○○ Low	IMPORTANT
Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration
51,7-8,17-19	RCT	Not serious	Serious15	Not serious18,20	Serious16	None	669	610	-	MD-2.01 (-4.46 to 0.45)	⊕○○○ Very Low	IMPORTANT
Footnotes:	Footnotes:	Footnotes:	Footnotes:	Footnotes:	Footnotes:	Footnotes:	Footnotes:	Footnotes:	Footnotes:	Footnotes:	Footnotes:	Footnotes:
No of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other	Steroids	Placebo	Relative Effect (95% CI)	Absolute Effect (95% CI)	Quality	Importance
Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality
61-6	RCT	Serious7	Not serious	Not serious8	Not serious	None	81/824 (9.8%)	122/810 (15.1%)	RR 0.62 (0.41 to 0.94)	57.38 per 1000 (89.09 to 9.06)	⊕○○O Moderate	CRITICAL
Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events	Adverse Events
31-2,5	RCT	Not serious	Serious9	Not serious8	Serious10	None	248/720 (34.4%)	256/705 (36.3%)	RR 1.12 (0.69 to 1.82)	43.56 per 1000 (112.53 to - 298.66)	⊕⊕○○Low	CRITICAL
Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay
41,3,5,6	RCT	Serious7	Not serious	Not serious8	Not serious	None	395	385	-	MD - 1.06 (-2.01 to -0.12)	⊕⊕○○Low	IMPORTANT
No of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other	Steroids	Placebo	Relative Effect (95% CI)	Absolute Effect (95% CI)	Quality	Importance
Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality	Mortality
$2^{1-2}$	RCT	Serious3	Not serious	Not serious4	Serious5	None	13/354 (3.7%)	18/351 (5.1%)	RR 0.73 (0.36 to 1.461)	13.77 per 1000 (32.64 to -23.46)	⊕⊕○ Low	CRITICAL
Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay	Length of Stay
$4^{1-2,6-7}$	RCT	Serious3	Not serious	Not serious4	Serious5	None	420	395	-	MD - 0.52 (-1.33 to 0.28)	⊕⊕○ Low	IMPORTANT
Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration	Antibiotic Duration
$3^{1,6-7}$	RCT	Not serious	Not serious	Serious4,8	Serious5	None	217	197	-	MD - 0.99 (-3.93 to 1.96)	⊕⊕○ Low	IMPORTANT

表 42
Should hospitalized patients with severe community-acquired pneumonia receive systemic corticosteroids rather than no systemic corticosteroids?
POPULATION:	Hospitalized patients with SEVERE community-acquired pneumonia
INTERVENTION:	Systemic corticosteroids
COMPARISON:	No systemic corticosteroids
SETTING:	Inpatients

表 43
Desirable Effects How substantial are the desirable effects? "Substantial" refers to both the importance of the outcomes and the magnitude of effect. As an example, a small improvement in a critical outcome might be more substantial than a large improvement in an unimportant outcome.
JUDGEMENT	RESEARCH EVIDENCE
○ Trivial ○ Small ● Moderate ○ Large ○ Varies ○ Don't know The magnitude of the effect is small but mortality is a very important outcome to most patients. Beneficial effects are seen in heterogeneous populations of patients with CAP. These effects appear to be driven by	DESIRABLE EFFECTS Decreased mortality: CAP patients- 13 RCTs, 3575 patients, 6.1% versus 9.1% (NNT 34), RR 0.68, 95% CI 0.53-0.86. Severe CAP- 6 RCTs, 1634 patients, 9.8% versus 15.1% (NNT 17), RR 0.62, 95% CI 0.41-0.94. Non-severe CAP- 3 RCTs, 705 patients, 3.7% versus 5.1% (NNT 72) RR 0.73, 95% CI 0.36 to 1.46. Decreased length of stay: CAP patients- 12 RCTs, 3403 patients, MD -1.11 days, 95% CI -1.66 to -0.55 days. Severe CAP- 4 RCTs, 780 patients, MD -1.06 days, 95% CI -2.01 to -0.12 days. Non-severe CAP- 5 RCTs, 1556 patients, MD -0.61 days, 95% CI -1.36 to 0.14 days. RCT= randomized controlled trial, NNT= number needed to treat, RR = risk ratio, CI = confidence interval, MD = mean difference.

表 44
patients with severe CAP.
Undesirable Effects How substantial are the undesirable effects? "Substantial" refers to both the importance of the outcomes and the magnitude of effect. As an example, a small increase in a critical adverse outcome might be more substantial than a large increase in an unimportant adverse outcome.
JUDGEMENT	RESEARCH EVIDENCE
○ Trivial ● Small ○ Moderate ○ Large ○ Varies ○ Don't know The magnitude of the effect is small and is of average importance since it is treatable and reversible upon discontinuation.	UNDESIRABLE EFFECTS Increased hyperglycemia: CAP patients- 7 RCTs, 2476 patients, 14.5% versus 9% (NNH 18), RR 1.71, 95% 1.21-2.40. Severe CAP- no research evidence. Non-severe CAP- no research evidence. RCT= randomized controlled trial, NNH = number needed to harm, RR = risk ratio, CI = confidence interval.
Certainty of evidence What is the committee's confidence in the accuracy of the above listed estimates (i.e., what is the quality of evidence)?
JUDGEMENT	RESEARCH EVIDENCE
○ Very low ● Low ○ Moderate ○ High ○ No included studies	QUALITY OF EVIDENCE Overall quality of evidence is the lowest quality of evidence among the critical outcomes. For the critical outcome of adverse effects, there is low-quality evidence because there are RCTs downgraded for inconsistency and imprecision.
Balance of effects Does the balance between desirable and undesirable effects favor the intervention or the comparison?
JUDGEMENT	RESEARCH EVIDENCE

表 45
o Favors the comparison o Probably favors the comparison o Does not favor either the intervention or the comparison o Probably favors the intervention • Favors the intervention o Varies o Don't know	The desirable effects (decreased mortality + decreased length of stay) were judged to outweigh the undesirable effects (increased hyperglycemia), thereby favoring systemic corticosteroids in patients with severe CAP.
Acceptability Is the intervention acceptable to key stakeholders?
JUDGEMENT	RESEARCH EVIDENCE
o No o Probably no o Probably yes • Yes o Varies o Don't know	The intervention is most often delivered intravenously or via the gastrointestinal tract, which is acceptable to most patients. This is based on the committee's non-systematic clinical observations.
Feasibility Is the intervention feasible to implement?
JUDGEMENT	RESEARCH EVIDENCE
o No o Probably no o Probably yes • Yes o Varies o Don't know	The intervention is most often delivered intravenously or via the gastrointestinal tract, which is feasible in all hospitals. This is based on the committee's non-systematic clinical observations.

表 46
	JUDGEMENT	JUDGEMENT	JUDGEMENT	JUDGEMENT	JUDGEMENT	JUDGEMENT	JUDGEMENT
DESIRABLE EFFECTS	Trivial	Small	Moderate	Large		Varies	Don't know
UNDESIRABLE EFFECTS	Trivial	Small	Moderate	Large		Varies	Don't know
BALANCE OF EFFECTS	Favors the compariso n	Probably favors the compariso n	Does not favor either the intervention or the comparison	Probably favors the interventio n	Favors the interventio n	Varies	Don't know
CERTAINTY OF EVIDENCE	Very low	Low	Moderate	High			No include d studies
ACCEPTABILITY Y	No	Probably no	Probably yes	Yes		Varies	Don't know
FEASIBILITY	No	Probably no	Probably yes	Yes		Varies	Don't know

表 47
Strong recommendation against the intervention	Conditional recommendation against the intervention	Conditional recommendation for either the intervention or the comparison	Conditional recommendation for the intervention	Strong recommendation for the intervention	Strong recommendation for the intervention
○	○	○	○	○	○

表 48
Should hospitalized patients with severe community-acquired pneumonia received systemic corticosteroids rather than no systemic corticosteroids?
POPULATION:	Hospitalized patients with NON-SEVERE community-acquired pneumonia
INTERVENTION:	Systemic corticosteroids
COMPARISON:	No systemic corticosteroids
SETTING:	Inpatients

表 49
Desirable Effects How substantial are the desirable effects? "Substantial" refers to both the importance of the outcomes and the magnitude of effect. As an example, a small improvement in a critical outcome might be more substantial than a large improvement in an unimportant outcome.
JUDGEMENT	RESEARCH EVIDENCE
● Trivial (if any) o Small o Moderate o Large o Varies o Don't know Beneficial effects are seen in heterogeneous populations of patients with CAP, but the effects appear to be driven by patients with severe CAP. There is no evidence of benefit in patients with non-severe CAP.	DESIRABLE EFFECTS Decreased mortality: CAP patients- 13 RCTs, 3575 patients, 6.1% versus 9.1% (NNT 33), RR 0.68, 95% CI 0.53-0.86. Severe CAP- 6 RCTs, 1634 patients, 9.8% versus 15.1% (NNT 19), RR 0.62, 95% CI 0.41-0.94. Non-severe CAP- 3 RCTs, 705 patients, 3.7% versus 5.1% (NNT 72) RR 0.73, 95% CI 0.36 to 1.46. Decreased length of stay: CAP patients- 12 RCTs, 3403 patients, MD -1.11 days, 95% CI -1.66 to -0.55 days. Severe CAP- 4 RCTs, 780 patients, MD -1.06 days, 95% CI -2.01 to -0.12 days. Non-severe CAP - 5 RCTs, 1556 patients, MD -0.61 days, 95% CI -1.36 to 0.14 days. RCT= randomized controlled trial, NNT= number needed to treat, RR = risk ratio, CI = confidence interval, MD = mean difference.
Undesirable Effects How substantial are the undesirable effects? "Substantial" refers to both the importance of the outcomes and the magnitude of effect. As an example, a small increase in a critical adverse outcome might be more substantial than a large increase in an unimportant adverse outcome.

表 50
JUDGEMENT	RESEARCH EVIDENCE
○ Trivial ● Small ○ Moderate ○ Large ○ Varies ○ Don't know The magnitude of the effect is small and it is average importance since it is treatable and reversible upon discontinuation.	UNDESIRABLE EFFECTS Increased hyperglycemia: CAP patients- 7 RCTs, 2476 patients, 14.5% versus 9% (NNH 18), RR 1.71, 95% 1.21-2.40. Severe CAP- no research evidence. Non-severe CAP- no research evidence. RCT= randomized controlled trial, NNH = number needed to harm, RR = risk ratio, CI = confidence interval.
Certainty of evidence What is the committee's confidence in the accuracy of the above listed estimates (i.e., what is the quality of evidence)?
JUDGEMENT	RESEARCH EVIDENCE
○ Very low ● Low ○ Moderate ○ High ○ No included studies	QUALITY OF EVIDENCE The overall quality of evidence is determined by the lowest quality of evidence among critical outcomes. For the critical outcome of mortality, there is low-quality evidence because there are RCTs downgraded for risk of bias and imprecision.
Balance of effects Does the balance between desirable and undesirable effects favor the intervention or comparison?
JUDGEMENT	RESEARCH EVIDENCE
● Favors the comparison ○ Probably favors the comparison ○ Does not favor either the intervention or the comparison ○ Probably favors	The undesirable effects (increased hyperglycemia) outweigh the desirable effects (there might be no desirable effects in non-severe CAP), thereby favoring no systemic corticosteroids in patients with non-severe CAP.

表 51
the intervention o Favors the intervention o Varies o Don't know
Acceptability Is the intervention acceptable to key stakeholders?
JUDGEMENT	RESEARCH EVIDENCE
o No o Probably no o Probably yes ● Yes o Varies o Don't know	The intervention is most often delivered intravenously or via the gastrointestinal tract, which is acceptable to most patients. This is based on the committee's non-systematic clinical observations.
Feasibility Is the intervention feasible to implement?
JUDGEMENT	RESEARCH EVIDENCE
o No o Probably no o Probably yes ● Yes o Varies o Don't know	The intervention is most often delivered intravenously or via the gastrointestinal tract, which is feasible in all hospitals. This is based on the committee's non-systematic clinical observations.

表 52
	JUDGEMENT	JUDGEMENT	JUDGEMENT	JUDGEMENT	JUDGEMENT	JUDGEMENT	JUDGEMENT
DESIRABLE EFFECTS	Trivial	Small	Moderate	Large		Varies	Don't know
UNDESIRABLE EFFECTS	Trivial	Small	Moderate	Large		Varies	Don't know
BALANCE OF EFFECTS	Favors the comparison	Probably favors the comparison	Does not favor either the intervention or the comparison	Probably favors the intervention	Favors the intervention	Varies	Don't know
CERTAINTY OF EVIDENCE	Very low	Low	Moderate	High			No included studies
ACCEPTABILITY	No	Probably no	Probably yes	Yes		Varies	Don't know
FEASIBILITY	No	Probably no	Probably yes	Yes		Varies	Don't know

表 53
Strong recommendation against the intervention	Conditional recommendation against the intervention	Conditional recommendation for either the intervention or the comparison	Conditional recommendation for the intervention	Strong recommendation for the intervention
•	○	○	○	○

图示：Study or Subgroup Less than 5 days Events Total 0.1 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100

图示：Study or Subgroup Less than 5 days Events Total 164 211 182 212 29.1% 19.8% 19.9% 19.0% 19.1% 19.2% 19.3% 19.4% 19.5% 19.6% 19.7% 19.8% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9% 19.9

图示：92 records identified from literature search 0 duplicate records removed before screening 67 records removed* -38 for study type -28 for drug -15 for publication type -6 for outcome -2 for population 25 reports assessed for eligibility 10 studies excluded -4 for study type -2 for intervention -2 for pneumonia type -1 for no clinical outcomes -1 unavailable some studies removed for multiple reasons

图示：Study or Subgroup Events Corticosteroids Control Control | 1.2.1 RCTs Only | Snijders 2010 | Torres 2015 | Subtotal (95% CI) | 1.78 | 61 | 165 | 42 | 166 | 157 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42 | 42

图示：Study or Subgroup Mean SD Total Control Mean SD Total Weight Mean Difference IV, Random, 95% CI 1.2.3 Severe CAP Confalonieri 2005 Marik 1993 4.3 3.8 14 4.6 5.9 6.8 7.6 8.0 9 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2% 1.2%

图示：Study or Subgroup Mean SD Control Mean Control Control\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q for\q

图示：Study or Subgroup Events Corticosteroids Events TakY = 0.10; ChiP = 24.53, df = 5 (P = 0.0002); P = 80% Test for overall effect: Z = 1.29 (P = 0.20) - 1.40; ChiP = 14.20, df = 2 (P = 0.0002); P = 80% Total events 410 - 1.40; ChiP = 24.53, df = 5 (P = 0.0002); P = 80% Test for overall effect: Z = 0.45 (P = 0.20) - 1.40; ChiP = 24.53, df = 5 (P = 0.0002); P = 80% Test for overall effect: Z = 0.45 (P = 0.20) - 1.40; ChiP = 24.53, df = 5 (P = 0.0002); P = 80% Test for overall effect: Z = 0.45 (P = 0.20) - 1.40; ChiP = 24.53, df = 5 (P = 0.0002); P = 80% Test for subgroup differences: Not applicable

图示：Study or Subgroup Events Control Events Total Events Total Weight M-H, Random, 95% CI Cl M-H, Random, 95% CI Blumzard 2022 13.1.2 Severe CAP Definonier 2005 18.2 19.3 12.3 14.2 15.3 16.2 17.2 18.3 19.4 19.5 19.6 19.7 19.8 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.9 19.

引用 / 出处

2025 ATS临床实践指南：社区获得性肺炎的诊断和管理. 2025.