Primary objective : to estimate impact of CT-scan on diagnostic for emergency department (ED) patients with suspected Community-acquired Pneumonia (CAP).

Secondary objective: to estimate impact of CT-scan on treatment (antimicrobial therapy) and site of care for ED patients with suspected CAP.

Rational: Community-acquired pneumonia (CAP) is a frequent infectious disorder in patients visiting the ED. CAP is responsible for high morbidity and associated-mortality is increasing in Western countries. CAP corresponds to invasion of the lung by pathogens. Diagnosis depends on clinical and X-ray assessment. However, these signs and symptoms are poorly specific and are often lacking. As prognosis depends on precocious and fitted antimicrobial treatment, making CAP diagnosis in a short time span (4-8 hours) is mandatory. Preliminary studies suggest that chest CT-scan could over-performed X-ray for diagnosis of CAP. Consensus conferences suggest the use of CT-scan in patients with uncertain diagnosis and unusual presentation and outcome. Because CT-scan is currently easily available, its use in a first intent is questionable for ED patients with suspected CAP.

Primary objective: to estimate impact of CT-scan on diagnostic for ED patients with suspected CAP.

Secondary objective: to estimate impact of CT-scan on treatment (antimicrobial therapy) and site of care for ED patients with suspected CAP.

Prospective multicenter study to measure chest CT-scan impact. 350 patients visiting the ED of 4 inner tertiary teaching hospitals in Paris, France, with suspected CAP.

Management: Patients will be managed according to current guidelines, including conventional chest X-ray.

Evaluation criteria. Attending ED physicians will implement pre- and post-test proforma for diagnosis (CAP) level of certainty, treatment (antimicrobial agents), site of care, before and after chest CT-scan. Comparison of ED physician's answers before/after CT-scan. Patients will be followed until day 28. An adjudication committee (1 pneumologist, 1 infectiologist, 1 radiologist)will review patients' data for gold standard diagnosis.

Statistical considerations: The investigators hypothesize that chest CT-scan wil modify diagnosis certainty in 20%. This implies that 300 participants should be enrolled to allow assessment of changes in 15 % et 25 %. Undue changes will be calculated a posteriori when diagnosis gold standard will be established by adjudication committee.

Anticipated results: Chest CT-scan should improve diagnosis certainty, treatment and site-of-care in patients visiting the ED with suspected CAP. If this is observed in at least 20%, the investigators will measure impact of chest CT-scan in a prospective randomized interventional study.

Inclusion Criteria:

• Patient above 18 years of age
• Patient with a presumptive diagnosis of CAP according to attending ED physician
• Patient experiencing at least one systemic sign (T°>38°C or < 36°C, HR>90/min, RR>20/min)
• Patient experiencing one respiratory sign (cough, lateral chest pain, localized crackles, dyspnea) that recently appeared
• Patient with a prior medical examination, the results have been or will be communicated
• Patient gave written informed consent or in cases of emergency parent/ support person who gave written informed consent if he/she is present on the day of inclusion

Exclusion Criteria:

• Pregnancy
• Patient with shock
• Patient with respiratory distress and immune suppression
• Patient with other criteria for immediate ICU referral to ICU
• Patient with living conditions making it impossible to follow 28 days
• Patient not affiliated with a social security system