Review question
We reviewed the evidence for antibiotics to prevent bacterial lower respiratory tract infections (LRTI) in children aged 12 years and under who are at increased chance of contracting such infections.
Background
Annually, about 1.4 million children die from respiratory tract infections. One in five of these deaths occur in children under five years old. Using antibiotics to prevent LRTI in high-risk children aged 12 years and under may be beneficial. However, no review currently describes the evidence for using antibiotics to prevent LRTI in this population.
Study characteristics
In February 2015, we searched for studies examining the effect of antibiotics to prevent LRTI in children aged 12 years and under with an increased risk of contracting such infections.
We included 10 studies; three examined the effectiveness of antibiotics in 1345 children with HIV, four in 429 children with cystic fibrosis, one in 219 children with sickle cell disease, one in 160 children undergoing treatment for cancer, and one in 40 children who were underweight at birth with underlying breathing problems.
The study duration was between 18.9 and 24.7 months for children with HIV, between six and 36 months in children with cystic fibrosis, 15 months in one study of children with sickle cell disease, 13 to 24 months in one study of children with cancer, and seven days in low birth weight newborns with underlying respiratory problems.
Study funding sources
Of three studies that included children with HIV, one was funded by an International Aid Agency, another by government and the third by a charity.
Of four studies that included children with cystic fibrosis, three were funded by charities and one by government.
The one study that included children with sickle cell disease was funded by government and the study that included children with cancer was funded by government, industry and charity.
The funding source for the study that included newborns with underlying respiratory disorders was not specified.
Key results
In children with HIV, antibiotics did not reduce the likelihood of contracting tuberculosis. Overall there was no improvement in death rates, but there was a marked reduction in hospital admission with antibiotic use in one study.
In children with cystic fibrosis, antibiotics did not reduce the likelihood of infection with Pseudomonas bacteria, but there was a reduction in flare-ups, where bacterial infection makes it harder for the child to breathe.
In one study of children with sickle cell disease, antibiotics reduced the likelihood of contracting blood infections.
In one study of children with cancers, antibiotics reduced the likelihood of contracting Pneumocystis infection in the lungs.
In low birth weight newborns with respiratory disorders, antibiotics did not reduce the likelihood of lung infections.
Quality of the evidence
In children with HIV, we rated the overall quality of the evidence for two studies that examined the effect of antibiotics for preventing tuberculosis as moderate because of differences in how the antibiotics were administered. We rated the quality of the evidence for antibiotic use in reducing deaths as moderate because of differences in the types of antibiotics used. Due to lack of data, we could not rate the quality of the evidence for hospital admissions.
In children with cystic fibrosis, we rated the overall quality of the evidence for two studies that examined the effect of antibiotics in reducing the likelihood of Pseudomonas infection as moderate because of differences in the types of antibiotics used. We rated the quality of the evidence for the effect of antibiotics in reducing 'flare-ups' as high.
In children with cancer, we rated the quality of the evidence for the effect of antibiotics in reducing the likelihood of Pneumocystis infection as moderate because of the likelihood of indirectness of the study results.
Due to lack of data, we could not rate the quality of the evidence for the effect of antibiotics in reducing the likelihood of blood infections in children with sickle cell disease, or in reducing the likelihood of lung infections in low birth weight newborns with respiratory disorders; the two studies had low probability and high probability of bias respectively.
There is inconclusive evidence that antibiotic prophylaxis in certain groups of high-risk children can reduce pneumonia, exacerbations, hospital admission and mortality in certain conditions. However, limitations in the evidence base mean more clinical trials assessing the effectiveness of antibiotics for preventing LRTIs in children at high risk should be conducted. Specifically, clinical trials assessing the effectiveness of antibiotics for preventing LRTIs in congenital heart disease, metabolic disease, endocrine and renal disorders, neurological disease or prematurity should be a priority.
Lower respiratory tract infections (LRTIs) in young children account for 1.4 million deaths annually worldwide. Antibiotics could be beneficial in preventing LRTIs in high-risk children, and may also help prevent school absenteeism and work days missed by children and/or carers. While it is well documented that the efficacy of antibiotic prophylaxis for RTIs decreases over time, there are no reviews that describe the use of antibiotic prophylaxis to prevent LRTIs in high-risk children aged 12 years and under.
To assess the effectiveness and safety of antibiotic prophylaxis in the prevention of bacterial LRTIs in high-risk children aged 12 years and under.
We searched the Cochrane Central Register of Controlled Trials (CENTRAL 2015, Issue 1) and the Database of Abstracts of Reviews of Effects (DARE), MEDLINE and MEDLINE In-Process (OvidSP) (1946 to 13 February 2015), EMBASE (OvidSP) (1974 to 12 February 2015), Science Citation Index Expanded (1945 to 13 February 2015) and Conference Proceedings Citation Index-Science (Web of Science Core Collection) (1990 to 13 February 2015). We searched for ongoing studies on ClinicalTrials.gov and the World Health Organization ICTRP. We handsearched the bibliographies of retrieved full texts of relevant studies.
We included randomised controlled trials (RCTs) comparing oral or intravenous antibiotics versus placebo or no treatment to prevent infections in high-risk children aged 12 years and under. We used a combination of the Centers for Disease Control and Prevention (CDC), National Health Service (NHS), American Academy of Paediatrics (AAP) and National Institute for Health and Care Excellence (NICE) guidelines to define conditions at higher risk of complications. Our primary outcome was the incidence of bacterial lower respiratory infections. Secondary outcomes included clinical function, hospital admission, mortality, growth, use of secondary antibiotics, time off school or parental work, quality of life and adverse events.
We extracted data using a customised data extraction sheet, assessed the risk of bias of included studies using the Cochrane 'Risk of bias' criteria, and used the GRADE criteria to rate the quality of the evidence. We used a random-effects model for meta-analysis. We presented the results narratively where we could not statistically combine data.
We included 10 RCTs of high-risk children using antibiotics (azithromycin, ciprofloxacin, co-trimoxazole, isoniazid, oral penicillin V or vancomycin) to prevent LRTIs. Three studies included HIV-infected children (n = 1345), four cystic fibrosis (n = 429) and one each sickle cell disease (n = 219), cancer (n = 160) and low birth weight neonates with underlying respiratory disorders (n = 40). The study duration ranged from seven days to three years. The quality of the evidence from studies including children with HIV infection, cystic fibrosis or cancer was moderate. Due to inadequate data, we were unable to rate the quality of the evidence for two studies: one in children with sickle cell disease (low risk of bias), and another in low birth weight neonates with underlying respiratory disorders (high risk of bias).
In HIV-infected children receiving continuous isoniazid prophylaxis, there was no significant difference in the incidence of pulmonary tuberculosis (risk ratio (RR) 0.64, 95% confidence interval (CI) 0.32 to 1.29, I2 statistic = 47%, P value = 0.21). There was no significant effect on mortality with co-trimoxazole or isoniazid prophylaxis (RR 0.82, 0.46 to 1.46, I2 statistic = 76%, P value = 0.58); however, analysis of one study that used co-trimoxazole showed a significant reduction in mortality (RR 0.67, 95% CI 0.53 to 0.85, P value = 0.001). There was a significant decrease in the rates of hospital admission per child-year of follow-up with co-trimoxazole prophylaxis in one study (P value = 0.01). There was no evidence of increased adverse events due to antibiotic prophylaxis (RR 1.10, 95% CI 0.75 to 1.64, I2 statistic = 22%, P value = 0.28); however, there was scant reporting of antibiotic resistance - the one study that did assess this found no increase.
In two studies of children with cystic fibrosis receiving ciprofloxacin prophylaxis, there was no significant difference in Pseudomonas infections (RR 0.76, 0.44 to 1.31, I2 statistic = 0%, P value = 0.33). In two studies assessing the benefit of azithromycin prophylaxis, there was a significant reduction in the frequency of pulmonary exacerbations (RR 0.60, 95% CI 0.48 to 0.76, I2 statistic = 0%, P value < 0.0001). The effect of antibiotic prophylaxis on growth in children with cystic fibrosis was inconsistent across the studies. There was an increased risk of emergence of pathogenic strains with either azithromycin or ciprofloxacin prophylaxis in two studies reporting this outcome. There was no significant difference in the quality of life (one study). In three studies, there was no significant increase in the frequency of adverse events with prophylaxis with azithromycin (two studies) or ciprofloxacin (one study). There was no evidence of increased antibiotic resistance in two studies.
In the one study of children with sickle cell disease, a significantly lesser proportion of children with pneumococcal septicaemia was reported with penicillin V prophylaxis (P value = 0.0025).
In the one study of children with cancer there was a significant decrease in Pneumocystis carinii pneumonia with trimethoprim-sulfamethoxazole prophylaxis (RR 0.03, 95% CI 0.00 to 0.47, P value < 0.01). There was no significant increase in the frequency of adverse events with antibiotic prophylaxis.
In low birth weight children with underlying respiratory disorders, there was no significant difference in the proportion of children with pulmonary infection with vancomycin prophylaxis (P value = 0.18).
No included studies reported time off school or carer time off work.