Diagnosis of pneumonia in the elderly is difficult because of the poor sensitivity and specificity of clinical signs as well as images from chest radiography (RT). New diagnostic tools such as thoracic low-dose computed tomography (CT), which exposes the patient to a weak dose of irradiation, could improve diagnosis. Moreover, low-dose CT could provide additional accuracy in the etiological clarification of pneumonia in elderly people.

As a first step, the investigators aim to perform a 1 year (12 months of inclusion + 3 months of follow-up) prospective study including the Divisions of Internal Medicine, Rehabilitation, Geriatrics and Radiology of the University Hospitals of Geneva. In this study, patients >65 years old with a clinical suspicion of low respiratory tract infection (LRTI) will be included. They will be prescribed antimicrobial therapy. Both chest radiography and low-dose thoracic CT will be performed within the first 72 hours after admission, as will blood tests and a nasopharyngeal swab.

The clinician's diagnosis, both before and after the results of the CT, will be compared at the end of the study to the adjudication committee's diagnostic opinion which will have access to all available clinical, laboratory and chest X-ray data and which will be considered the gold standard. At the end of the study, all the CT images will be blind-reviewed by two experts in radiology. The impact of CT scanning in the diagnosis of pneumonia will be assessed, both for its sensitivity and specificity in this population.

During the first 12 months of the study, all patients will undergo a systematic nasopharyngeal swab at admission and at discharge, from which eluates will be conserved. During the next 12 months, virological and bacteriological polymerase chain reactions (PCR) will be performed, using new diagnostic tools, in order to determine the etiological diagnosis in this population and to evaluate the impact of the new tools in the management of pneumonia for this population.

Analysis of these data will allow clinical, radiological and microbiological correlation.

The project's principal aim is to assess the impact of low-dose CT scanning versus chest radiography in the diagnosis of pneumonia in elderly patients.

One of the secondary objectives is to estimate sensitivity and specificity of CT scanning in pneumonia diagnosis in the elderly (gold standard: adjudication committee advice).

This prospective study should be considered a preliminary feasibility study in preparation for a future randomized clinical trial comparing patient management, antibiotic prescription and clinical outcomes between one group of patients with LRTI randomized to routine thoracic CT scan and one group randomized to usual care.

The second phase of this study will explore the etiologies of pneumonia in the elderly using new diagnostic tools. This will lead to identify radiological patterns associated with clinical presentation and microbiological results.

Primary objective: to estimate the impact of CT scanning in the diagnosis of pneumonia in elderly patients, in a monocentric study (but within two sites, Service de Médecine Interne Générale and Hôpital des Trois-Chêne).

Secondary objectives: to estimate sensitivity and specificity of CT scanning in pneumonia diagnosis (gold standard: adjudication committee advice), and to identify microbiological etiologies of pneumonia in the elderly.

Study-design: prospective interventional study.

Intervention:

Patients will be managed according to HUG guidelines. Frontal (and lateral if possible) chest radiography (RT) will be interpreted by the physician; he will assess the level of certainty for a diagnosis of pneumonia on a Likert scale, as well as its localization. Subsequently, a thoracic CT scan will be performed within the first 48-72 hours of admission, with conclusion by a radiologist (on a Likert scale). After the CT scan, the physician will get the result and assess a new level of certainty of diagnosis (on a Likert scale). If the clinician changes his therapy, this will be noted on the case report form (CRF).

Patients will undergo two nasopharyngeal swabs (on admission and discharge) for storage in a biobank, and microbiological (bacteriological and viral) PCR will be performed at a later date.

Patients will be followed until day 90. Information concerning rehospitalization and death will be obtained by review of the patient's medical chart and the civil registry.

At the end of the study, chest radiographs (RT) and CT will be blinded-analyzed by two referring radiologists (MS and XM).

An adjudication committee (one internist, one infectiologist, one pneumologist) will review patients' data in order to provide a gold standard diagnosis which will be made according to international guidelines (8, 9): clinical symptoms, recent new abnormalities on chest radiograph and outcome under antimicrobial therapy.

After the end of the clinical study, the virology group will analyze the eluates from the nasopharyngeal swabs using new PCR tests. After completing routine PCR (Adenovirus, Bocavirus, Coronavirus, Enterovirus, Influenza A and B viruses, Parainfluenza virus, Parechovirus, Picornavirus, Respiratory Syncytial Virus, Chlamydophila pneumoniae, Mycoplasma pneumoniae), two different bacterial PCR kits will be used to test each eluate (admission and discharge): the Fast Track Diagnostics (FTD) Bacterial pneumonia CAP (for Community Acquired Pneumonia) kit (Streptococcus pneumoniae, Haemophilus influenzae, Moraxella catarrhalis, Staphylococcus aureus, Mycoplasma pneumoniae, Chlamydia pneumoniae, Legionella spp) and the FTD Bacterial pneumonia HAP (for Hospital Acquired Pneumonia) kit (Klebsiella pneumoniae, Pseudomonas aeruginosa).

The expenses of additional examinations will not be charged to the patient or to his health insurance company, but will be paid for by the HUG within the framework fixed price for hospitalization.

Evaluation criteria:

• Assessments by the physician, before and after CT scanning, will be compared: number of upgraded or downgraded diagnoses; sensitivity and specificity of chest radiography RT and CT; correlation between the physician and the radiologist.
• Descriptions of the etiologies of pneumonia, with number of viral and bacterial pneumonia.
• Clinical, radiological, microbiological correlations.
• Outcome (re-hospitalization, mortality) at 30 and 90 days.

Statistical considerations:

The investigators hypothesize that chest CT will upgrade the certainty of diagnosis in 19% of cases (14). To significantly increase the level of diagnosis of pneumonia, 46 patients with suspicion of pneumonia are needed (α = 0.05, β 0.1). The investigators estimate the rate of pneumonia in this population to be 45% (15); this implies that 100 participants should be enrolled in order to significantly increase the accuracy of diagnosis.

Inclusion in this study could begin in June 2015, for 12 months, with the last follow-up for the last patient coming after a total of 15 months (12 months plus 90 days).

Anticipated results:

The investigators anticipated that low-dose CT scan will improve the diagnosis of pneumonia in the elderly. The study's final objective is to prepare a second randomized study measuring the impact of chest CT scans on the management of patients in two cohorts of patients - one which only undergoes a chest X-ray, and one which only undergoes a CT scan - for assessing diagnosis and treatment.

Inclusion Criteria:

• Patients above 65 years old
• AND at least one infectious sign: T° >38°C or <35°C; C-reactive protein (CRP) >10 mg/L; leucocytes >10,000/mL with >85% polynuclear neutrophils or left deviation,
• AND at least one respiratory sign: cough; purulent sputum; chest pain; localized crackles; recently appeared dyspnea; oxygen saturation (SpO2) <90%; respiratory frequency >20/min,
• AND who will be prescribed antimicrobial therapy for suspicion of low respiratory tract infection,
• AND who will give consent himself or through his support person.

Exclusion Criteria:

• Diagnosis of pneumonia in the previous six months,
• AND/OR more than 48h of antimicrobial treatment given before hospitalization,
• AND/OR thoracic CT scan performed before hospitalization or inclusion of the patient.
• Each patient will be included only once.