The early use of Antibiotics for at Risk CHildren with InfluEnza (ARCHIE): a double-blind randomised placebo-controlled trial

Talk Code: 
Kay Wang
Malcolm G Semple, Michael Moore, Alastair D Hay, Tricia Carver, Sharon Tonner, Ushma Galal, Rafael Perera, Ly-Mee Yu, Jill Mollison, Paul Little, Andrew Farmer, Christopher Butler, Anthony Harnden
Author institutions: 
University of Oxford, University of Liverpool, University of Southampton, University of Bristol


‘At risk’ children with underlying medical conditions are more likely to develop complications from influenza/influenza-like illness (ILI) than otherwise healthy children. The UK government stockpiles co-amoxiclav to treat bacterial complications during influenza epidemics and pandemics. However, it is unclear whether early antibiotics prevent influenza/ILI-related clinical deterioration. We aimed to determine whether early co-amoxiclav treatment in ‘at risk’ children with influenza/ILI reduces risk of re-consultation due to clinical deterioration.


We recruited ‘at risk’ children aged 6 months-12 years who presented within five days of ILI onset. We opened 151 general practices, 44 hospitals and one walk-in centre for recruitment. Participants were randomly assigned to oral co-amoxiclav 400/57 twice daily or placebo for five days. Randomisation used a non-deterministic minimisation algorithm to balance age and vaccination status. Participants, caregivers and investigators were blinded to treatment allocation. Our intention-to-treat analysis included all randomised participants with available primary outcome data. The primary outcome was re-consultation due to clinical deterioration within 28 days of randomisation. Proportions of children re-consulting were compared using a log-binomial regression model with adjustment for region, age and seasonal influenza vaccination status.Our safety analysis included all randomised participants. We conducted an exploratory subgroup analysis in participants with laboratory-confirmed influenza. Influenza was detected by real-time PCR analysis of nasal swabs. Trial registration: ISRCTN 70714783. EudraCT 2013-002822-21.


Between February 11, 2015 and April 20, 2018, we recruited 271 children of whom 37 had laboratory-confirmed influenza (13.7%). Primary outcome data were available for 265 children. No significant difference was observed between groups in re-consultation rates (co-amoxiclav 33/133, placebo 28/132, unadjusted risk ratio [RR] 1.17, 95% confidence interval [CI] 0.75-1.82 risk difference[RD] 3.6%, 95% CI -6.5%-13.7%). This finding did not change after adjustment for baseline covariates (adjusted RR 1.16, 95% CI 0.75-1.80). There was no significant difference between groups in proportions of children with serious adverse events, none of which were considered related to study medication (co-amoxiclav 8/136, 5.9%; placebo 7/135, 5.2%; RD 0.7%, 95% CI -5% to 6%).The proportion of children with laboratory-confirmed influenza who re-consulted was lower among those who received co-amoxiclav (5/21, 23.8%) versus placebo (6/16, 37.5%). However, the difference between groups was not statistically significant (unadjusted RR 0.63, 95% CI 0.24-1.71; adjusted RR 0.55, 95% CI 0.20-1.55). We therefore did not find evidence that early co-amoxiclav treatment in ‘at risk’ children with influenza/ILI reduces re-consultations due to clinical deterioration. However, there may be a benefit in children with laboratory-confirmed influenza.


Immediate antibiotics should not be given to ‘at risk’ children presenting early with influenza/ILI without evidence of bacterial infection. Antibiotics may be associated with clinical benefit during periods of high influenza activity. However, further research is needed to more accurately identify children who may benefit from early intervention.

Submitted by: 
Kay Wang
Funding acknowledgement: 
This article presents independent research funded by the National Institute for Health Research (NIHR) under its Programme Grants for Applied Research Programme (Grant Reference Number RP-PG-1210-12012). The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.