Electronically delivered interventions to reduce antibiotic prescribing for respiratory infections in primary care: cluster RCT using electronic health records and cohort study

Martin C Gulliford, Dorota Juszczyk, Toby Prevost, Jamie Soames, Lisa McDermott, Kirin Sultana, Mark Wright, Robin Fox, Alastair Hay, Paul Little, Michael Moore, Lucy Yardley, Mark Ashworth, Judith Charlton

Research output: Contribution to journalArticle (Academic Journal)peer-review

5 Citations (Scopus)
179 Downloads (Pure)


Background: Unnecessary prescribing of antibiotics in primary care is contributing to the emergence of antimicrobial drug resistance.

Objectives: To develop and evaluate a multi-component intervention for antimicrobial stewardship in primary care; to evaluate the safety of reducing antibiotic (AB) prescribing for self-limiting respiratory infections (RTI).
Health technology being assessed: A complex intervention, developed as part of this study, including a webinar, monthly reports of general practice-specific data for AB prescribing; decision support tools to inform appropriate AB prescribing.

Methods: A cluster randomised controlled trial was conducted in 79 general practices in the UK Clinical Practice Research Datalink (CPRD). The primary outcome was the rate of antibiotic prescriptions for self-limiting RTI over the 12-month intervention period. A separate population-based cohort study was conducted in 610 CPRD general practices that were not exposed to the trial interventions. Data were analysed to evaluate safety outcomes for registered patients with 45.5 million person-years of follow-up from 2005 to 2014.

Results: There were 41 intervention trial arm practices (323,155 patient-years) and 38 control trial arm practices (259,520 patient-years). There were 98.7 AB prescriptions for RTI per 1,000 patient-years in the intervention trial arm (31,907 AB prescriptions) and 107.6 per 1,000 in the control arm (27,923 AB prescriptions); adjusted AB prescribing rate ratio (RR) 0.88 (95% confidence interval 0.78 to 0.99, P=0.040). There was no evidence of effect in children less than 15 years (RR 0.96, 0.82 to 1.12, P=0.632) or adults aged 85 years and older (RR 0.97, 0.79 to 1.18, P=0.742), AB prescribing was reduced in adults 15-84 years (RR 0.84, 0.75 to 0.95, P=0.006). One antibiotic prescription was avoided for every 62 (95% confidence interval 40 to 200) patients aged 15-84 years per year. Analysis of trial data for 12 safety outcomes, including pneumonia and peritonsillar abscess, showed no evidence that these might be increased as a result of intervention. Analysis of data from non-trial practices showed that if a general practice with the average list size of 7,000 patients reduces the proportion of RTI consultations with antibiotics prescribed by 10%, then it may observe 1.1 (0.6 to 1.5) more cases of pneumonia per year and 0.9 (0.5 to 1.3) more cases of peritonsillar abscess per decade. There was no evidence of that mastoiditis, empyema, meningitis, intracranial abscess or Lemierre’s syndrome were more frequent at low prescribing practices.

Limitations: The research was based on electronic health records that may not always provide complete data. The number of practices included in the trial was smaller than initially intended.

Conclusions: This study found evidence that overall general practice AB prescribing for RTI was reduced by this electronically-delivered intervention. AB prescribing was reduced for adults aged 15 to 84 years, but not for children or senior elderly.

Future work: Strategies for antimicrobial stewardship should employ stratified interventions that are tailored to specific age-groups. Further research into the safety of reduced AB prescribing is also needed.

Funding: NIHR Health Technology Assessment programme (13/88/10).
Original languageEnglish
Pages (from-to)1-70
Number of pages70
JournalHealth Technology Assessment
Issue number11
Publication statusPublished - 1 Mar 2019

Structured keywords

  • Digital Health
  • Physical and Mental Health


Dive into the research topics of 'Electronically delivered interventions to reduce antibiotic prescribing for respiratory infections in primary care: cluster RCT using electronic health records and cohort study'. Together they form a unique fingerprint.

Cite this