Multi-faceted intervention to improve management of antibiotics for children presenting to primary care with acute cough and respiratory tract infection (CHICO): efficient cluster randomised controlled trial

Peter S Blair, Grace Young, Clare Clement, P Dixon, Penny Seume, Jenny Ingram, Jodi Taylor, Christie Cabral, Patricia J Lucas, Elizabeth Beech, Jeremy Horwood, Martin Gulliford, Nick A Francis, Sam Creavin, Janet A Lane, Scott Bevan, Alistair D Hay

Research output: Contribution to journalArticlepeer-review

2 Citations (Scopus)


Objective: To assess whether an easy-to-use multifaceted intervention for children presenting to primary care with respiratory tract infections would reduce antibiotic dispensing, without increasing hospital admissions for respiratory tract infection. Design: Two arm randomised controlled trial clustered by general practice, using routine outcome data, with qualitative and economic evaluations. Setting: English primary care practices using the EMIS electronic medical record system. Participants: Children aged 0-9 years presenting with respiratory tract infection at 294 general practices, before and during the covid-19 pandemic. Intervention: Elicitation of parental concerns during consultation; a clinician focused prognostic algorithm to identify children at very low, normal, or elevated 30 day risk of hospital admission accompanied by antibiotic prescribing guidance; and a leaflet for carers including safety netting advice. Main outcome measures: Rate of dispensed amoxicillin and macrolide antibiotics (superiority comparison) and hospital admissions for respiratory tract infection (non-inferiority comparison) for children aged 0-9 years over 12 months (same age practice list size as denominator). Results: Of 310 practices needed, 294 (95%) were randomised (144 intervention and 150 controls) representing 5% of all registered 0-9 year olds in England. Of these, 12 (4%) subsequently withdrew (six owing to the pandemic). Median intervention use per practice was 70 (by a median of 9 clinicians). No evidence was found that antibiotic dispensing differed between intervention practices (155 (95% confidence interval 138 to 174) items/year/1000 children) and control practices (157 (140 to 176) items/year/1000 children) (rate ratio 1.011, 95% confidence interval 0.992 to 1.029; P=0.25). Pre-specified subgroup analyses suggested reduced dispensing in intervention practices with fewer prescribing nurses, in single site (compared with multisite) practices, and in practices located in areas of lower socioeconomic deprivation, which may warrant future investigation. Pre-specified sensitivity analysis suggested reduced dispensing among older children in the intervention arm (P=0.03). A post hoc sensitivity analysis suggested less dispensing in intervention practices before the pandemic (rate ratio 0.967, 0.946 to 0.989; P=0.003). The rate of hospital admission for respiratory tract infections in the intervention practices (13 (95% confidence interval 10 to 18) admissions/1000 children) was non-inferior compared with control practices (15 (12 to 20) admissions/1000 children) (rate ratio 0.952, 0.905 to 1.003). Conclusions: This multifaceted antibiotic stewardship intervention for children with respiratory tract infections did not reduce overall antibiotic dispensing or increase respiratory tract infection related hospital admissions. Evidence suggested that in some subgroups and situations (for example, under non-pandemic conditions) the intervention slightly reduced prescribing rates but not in a clinically relevant way.

Original languageEnglish
Article numbere072488
Publication statusPublished - 26 Apr 2023


  • Humans
  • Child
  • Adolescent
  • Anti-Bacterial Agents/therapeutic use
  • Cough/drug therapy
  • Pandemics
  • COVID-19
  • Respiratory Tract Infections/drug therapy
  • Primary Health Care


Dive into the research topics of 'Multi-faceted intervention to improve management of antibiotics for children presenting to primary care with acute cough and respiratory tract infection (CHICO): efficient cluster randomised controlled trial'. Together they form a unique fingerprint.

Cite this