The aim of this study is to test if BPS (Bacterial Pneumonia Score) guided antibiotic use in children with non severe community acquired pneumonia (CAP) will reduce antibiotic use as compared to standard care practice (current guidelines for CAP).

Background: Pneumonia is a leading cause of mortality in children. Despite more than 50% of pneumonias are due to viruses, because it is difficult to rule out bacterial etiology, initial management of pneumonia in children usually includes antibiotics, often unnecessary. In 2006 was designed and validated a clinical prediction rule (BPS: Bacterial Pneumonia Score) which accurately identifies hospitalized children's risk of bacterial pneumonia. However, BPS efficacy on guiding therapeutic decision in children with community acquired pneumonia (CAP) has not been yet assessed.

Aim: The aim of this study is to test if BPS guided antibiotic use in children with non severe community acquired pneumonia will reduce antibiotic use as compared to standard care practice (current guidelines for CAP) Design: This is a randomized, controlled, blinded trial, to assess antibiotics use regarding two methods for initial management of children aged 3-60 months with non severe community acquired pneumonia. Children will be randomly allocate to be managed according to BPS or currently enforced guidelines. Use of antibiotics (%) and clinical outcome of both groups will be compared.

Setting: Tertiary children hospital in Buenos Aires, Argentina. Patients: Consecutive children aged 3-60 months assisted for non severe community acquired pneumonia as outpatients. Patients with wheezing, severe pneumonia, pulmonary or cardiovascular chronic disease, or antibiotic use or hospitalization in the previous two weeks will be excluded.

Endpoints:

Primary: Use of antibiotics in each group (proportion) Secondary: Treatment failure (proportion) in each group Endpoints will be assessed at baseline and after 1, 2, 5, 7 and 10 days by a blinded investigator.

Intervention: Patients with CAP will be randomized (1:1) to BPS versus enforced guidelines. In the BPS group antibiotics will be indicated in patients with a BPS ≥ 4 points, while in the control group antibiotics will be indicated according to current guidelines.

Variables and measurement: Antibiotic use will be defined as initial use of any antibiotic, immediately after diagnosis. Treatment failure will be defined as persistence of fever after 2 days, or tachypnea or diminishing in respiratory rate less than 5 bpm. after 2 days, or signs of severe pneumonia or requiring or changing antibiotics at any time.

Study hypothesis: BPS antibiotic use guidance will reduce at least 20% antibiotic use, as compared to standard care practice.

Analyses: These will be done based on an intention-to-treat and a per-protocol principle. With an assumed 20% less use of antibiotics in the intervention group, a maximum of 5% losses to follow-up, a confidence of 5% and power of 90%, the total sample size is 60. This will allow detecting a difference in clinical outcome of 28%. Proportion will be compared by Chi square test.

Interim monitoring: Regular review of serious adverse events, quality and integrity of the study by an independent data safety and monitoring board. Safety interim analysis after 50% of the patients recruited.

Significance: Due to the high prevalence of CAP in children, this study will offer the potential for a substantial reduction in health costs and antibiotics resistance.

Inclusion Criteria:

• Children aged 3-60 months assisted as outpatients for non severe community acquired pneumonia.

Exclusion Criteria:

• Wheezing
• Severe pneumonia
• Pulmonary or cardiovascular chronic disease
• Antibiotic use in the previous two weeks
• Hospitalization for any reason in the previous two weeks