Antibiotics for bronchiolitis in children under two years of age

Question

We reviewed the evidence on the effect of antibiotics on clinical outcomes in children with bronchiolitis.

Background

Bronchiolitis is a serious respiratory illness that affects babies. It is most commonly caused by respiratory syncytial virus (RSV) and is the most common reason for hospitalisation in babies younger than six months. Babies usually present with runny nose, cough, shortness of breath and signs of difficulty in breathing, which can become life-threatening. Despite its viral cause, antibiotics are often prescribed. Prescribers may be expecting benefits from anti-inflammatory effects attributed to some antibiotics or be concerned about secondary bacterial infection, particularly in children who are very unwell and require intensive care. We wanted to discover if antibiotics improved or worsened clinical outcomes in children with bronchiolitis.

Study characteristics

This evidence is current to June 2014. We identified seven trials (824 participants) comparing antibiotics with placebo or no antibiotics in children with bronchiolitis. Two of these studies also compared intravenous and oral antibiotics.

Key results

Our primary outcome was duration of symptoms/signs (duration of supplementary oxygen requirement, oxygen saturation, wheeze, crepitations (crackles), fever). Secondary outcomes included duration of admissions/time to discharge from hospital, readmissions, complications/adverse events (including death) and radiological (X-ray) findings.

We included seven studies with a total of 824 participants. Four studies reported on duration of supplementary oxygen requirement and did not demonstrate a significant difference in the duration of oxygen use comparing antibiotics to placebo. We combined three studies comparing azithromycin versus placebo and again did not demonstrate a significant difference between antibiotics and placebo in the duration of oxygen requirement. Most of the included studies did not report on the primary outcomes of wheeze, crepitations and fever. One study with a high risk of bias found mixed results for the effects of antibiotics on wheeze but no difference for other symptom measures. One study found no difference in duration of fever and one study found no difference in presence of fever on day two.

In regards to secondary outcomes, six included studies did not find any difference between antibiotics and placebo for the outcomes of length of illness or length of hospital stay. For length of hospital stay, we combined data from three studies comparing the use of azithromycin versus placebo as a subtotal as part of the overall analysis of the effect of antibiotics on hospital stay. These combined results similarly showed no difference between antibiotics (azithromycin) and placebo. One small study with a high risk of bias found that three weeks of clarithromycin significantly reduced hospital readmission compared to placebo. However, this reduction in hospital readmissions was not replicated in a more recent study that randomised 97 children to receive either a single large dose of azithromycin or placebo. There were no deaths reported in any arms of any of the seven included trials and none of the studies specifically reported on adverse effects of antibiotics. Only two studies made general comments that no adverse effects were found with antibiotic use. Radiological findings were not reported as an outcome in any of the included studies.

Quality of the evidence

This 2014 updated review is stronger, owing to the inclusion of two new randomised controlled trials (RCTs). These two studies combined involved a further 138 participants in the antibiotic arm and 143 participants in the placebo arm. Prior to this only three small RCTs had examined antibiotics versus placebo, with only 72 participants in the antibiotic arms and 72 participants in the placebo arms. Consequently, this review makes a substantial contribution, especially with regards to the role of macrolides, such as azithromycin, in bronchiolitis. No new unpublished data have been included. However, the review authors have no reason to suspect that the search strategy has biased the review results. Raw data could not be obtained from one study conducted 40 years ago, nor from three other trials, which is a weakness of this review. Three trial authors did provide raw data for this review.

Conclusion

This review did not find sufficient evidence to support the use of antibiotics for bronchiolitis. Research may be justified to identify a subgroup of patients who may benefit from antibiotics.

Authors' conclusions: 

This review did not find sufficient evidence to support the use of antibiotics for bronchiolitis, although research may be justified to identify a subgroup of patients who may benefit from antibiotics. Further research may be better focused on determining the reasons that clinicians use antibiotics so readily for bronchiolitis, how to reduce their use and how to reduce clinician anxiety about not using antibiotics.

Read the full abstract...
Background: 

Bronchiolitis is a serious, potentially life-threatening respiratory illness commonly affecting babies. It is often caused by respiratory syncytial virus (RSV). Antibiotics are not recommended for bronchiolitis unless there is concern about complications such as secondary bacterial pneumonia or respiratory failure. Nevertheless, they are often used.

Objectives: 

To evaluate the effectiveness of antibiotics for bronchiolitis in children under two years of age compared to placebo or other interventions.

Search strategy: 

We searched the Cochrane Central Register of Controlled Trials (CENTRAL 2014, Issue 6), which includes the Cochrane Acute Respiratory Infection Group's Specialised Register, and the Database of Abstracts of Reviews of Effects, MEDLINE (1966 to June 2014), EMBASE (1990 to June 2014) and Current Contents (2001 to June 2014).

Selection criteria: 

Randomised controlled trials (RCTs) comparing antibiotics to placebo in children under two years diagnosed with bronchiolitis, using clinical criteria (including respiratory distress preceded by coryzal symptoms with or without fever). Primary clinical outcomes included time to resolution of signs or symptoms (pulmonary markers included respiratory distress, wheeze, crepitations, oxygen saturation and fever). Secondary outcomes included hospital admissions, length of hospital stay, readmissions, complications or adverse events and radiological findings.

Data collection and analysis: 

Two review authors independently analysed the search results.

Main results: 

We included seven studies with a total of 824 participants. The results of these seven included studies were often heterogeneous, which generally precluded meta-analysis, except for deaths, length of supplemental oxygen use and length of hospital admission.

In this update, we included two new studies (281 participants), both comparing azithromycin with placebo. They found no significant difference for length of hospital stay, duration of oxygen requirement and readmission. These results were similar to an older study (52 participants) that demonstrated no significant difference comparing ampicillin and placebo for length of illness.

One small study (21 participants) with higher risk of bias randomised children with proven RSV infection to clarithromycin or placebo and found a trend towards a reduction in hospital readmission with clarithromycin.

The three studies providing adequate data for days of supplementary oxygen showed no difference between antibiotics and placebo (pooled mean difference (MD) (days) -0.20; 95% confidence interval (CI) -0.72 to 0.33). The three studies providing adequate data for length of hospital stay, similarly showed no difference between antibiotics (azithromycin) and placebo (pooled MD (days) -0.58; 95% CI -1.18 to 0.02).

Two studies randomised children to intravenous ampicillin, oral erythromycin and control and found no difference for most symptom measures.

There were no deaths reported in any of the arms of the seven included studies. No other adverse effects were reported.