Skip to main navigationSkip to main content
The University of Southampton
ARTIC PC - Antibiotics for lower Respiratory Tract Infection in Children presenting in Primary Care

Public and Scientific summaries of the research

Public summary

Although chest infections are one of the most common infections managed in children seen in primary care, there have been no placebo controlled trials to show what effect antibiotics have - in contrast to adults where there is now good trial evidence to show that for most people antibiotics do not work, and also modest effects in important clinical features (these are listed below).

The trouble with prescribing for most children is that we are using antibiotics too much causing the bacteria to become resistant, which is likely to lead in the future to serious infections in children becoming untreatable from ‘superbugs’. Most children who see the doctor with a chest infection currently get antibiotics, and the groups of children that are even more likely to get antibiotics at the moment are those who have one or more particular features:

• Phlegm,

• Fever,

• Shortness of breath,

• or rattly noises heard in the chest when the doctor listens with the stethoscope.

It is a real priority to show which groups of children that GPs prescribe for currently benefit and which do not, so that antibiotics can be targeted appropriately and the effectiveness of antibiotics can be conserved for future generations.

The study will randomise about 1000 children with chest infections presenting in primary care to either get amoxicillin (an antibiotic) or not get amoxicillin for 1 week, and see whether antibiotics make any difference to symptom severity, or the duration of illness. All children will be given advice about using painkillers (and anti-fever medication) and will be followed up carefully during the next month. The study will be large enough to be able to show which, if any, groups of children that GPs currently prescribe for benefit from antibiotics (such as those with fever compared to those with no fever), and which groups do not.

Parents and children who are happy to have further tests will have any of or all of:

• a pinprick blood sample,

• a swab of the throat and

• an X ray done.

This is to see whether simple markers of inflammation and infection or the presence of bacteria, or any lung involvement seen on the X ray can predict benefit from antibiotics.

Scientific summary

RCT nested in an observational study.

Setting: Primary care.

Prior Literature: Cochrane/Medline searches identified no placebo-controlled trials of antibiotics for children with uncomplicated Lower Respiratory Tract Infection (LRTI).

Target population: Children 6 months to 12 years.

Health technologies: Amoxicillin 50mg/kg/24 hours (in divided doses for 7 days) or placebo.

Inclusion Criteria:

• Acute uncomplicated LRTI (acute cough as the most prominent symptom and lower tract symptoms/signs (sputum/’rattly chest’/coarse rhonchi; breathless; pain).


• Non-infective (e.g. reflux, PE) or croup (viral aetiology very likely). Those with clinically suspected pneumonia or very unwell and/or unwilling to be randomised will be invited to participate in an observational study collecting the same data.

Baseline measures:

• structured history/examination;

• pulse oximetry;

• X ray (for clinically undetected consolidation);

• and optional samples: throat swabs (for bacteriology) and pinprick (for CRP and Full Blood Count).

Measurement of costs/outcomes: Primary outcome: duration of moderately bad symptoms (from validated symptom diary).

Secondary outcomes: symptom severity (days 2-4); the development of new or worsening symptoms: complications.

Primary Outcome: duration of moderately bad symptoms(from validated symptom diary completed until the child is well)

Health related quality of life will be measured by proxy methods in which the EuroQoL (EQ5D) will be completed by patients or carers (on days 1,7,14,21,28).

Follow-up (at 1 month): measure lung function (if aged 6 years+).

Sample size: 938 children are required (alpha=0.01;beta=0.2;80% follow-up) to estimate effectiveness in 5 pre-specified subgroups to detect a 2 day difference in duration of symptoms rated moderately bad or worse for a subgroup representing at least 30% of the sample (our data suggests 30-65%). Subgroups: sputum seen and/or heard by parents (’rattly chest’) or clinicians; history of fever; physician rating of being unwell; short of breath; chest signs (non-focal coarse crepitations/rhonchi/wheeze).

Analysis: Multiple linear regression models will provide estimates both overall and for subgroups and estimates of interaction with intervention by subgroup. In addition we will explore to what extent results from additional tests (bacteriology, biomarkers, abnormalities on chest-X rays etc.) are potential effect modifiers, and will explore differences between the purely observational data set and the trial data and our other large observational cohorts (TARGET and 3C studies). We will conduct economic evaluation alongside the clinical trial. Costs will take an NHS and personal social service (PSS) perspective (primary analysis) and societal perspective (secondary analysis). Outcomes will be measured as costs per symptomatic day prevented and the incremental cost per QALY gained. Cost effectiveness acceptability curves will also be produced. In addition, a decision analytic model will be developed to explore the cost-effectiveness implication over the long term and investigate the potential implication of the costs associated with antibiotic resistance.


Useful Downloads

Need the software?PDF Reader

Useful Downloads

Need the software?PDF Reader
Privacy Settings