From Wiki Journal Club
Jump to: navigation, search
Yadav K, et al. "A Multifaceted Intervention Improves Prescribing for Acute Respiratory Infection for Adults and Children in Emergency Department and Urgent Care Settings". Academic Emergency Medicine. 2019. 26(7):719-731.

Clinical Question

In pediatric and adult patients that present to the emergency department with an acute respiratory illness that does not respond to antibiotics, does a more aggressive educational and stewardship program provided to patients and providers decrease the inappropriate prescribing of antibiotics, as compared to a less-involved program.

Bottom Line

Implementing an antimicrobial stewardship program in the emergency department, including patient and provider education, physician champions, audit and feedback, and peer-comparison, will decrease inappropriate antibiotic prescribing.

Major Points

As we are rapidly approaching the post-antibiotic era, stewardship of this likely limited and precious resource is becoming paramount. Inappropriate prescribing are likely contributing to the development and spread of multi drug resistant organisms. Antibiotics are commonly prescribed in the emergency department to both out-patients and those being admitted. The MITIGATE trial was a multi-centred, cluster randomized effectiveness trial indented to test the effects of two antimicrobial stewardship programs, one “adapted" from the CDC’s Core Elements for Outpatient Antibiotic Stewardship and one “enhanced” that added a additional visual aids and clinician peer-comparisons.

Including 291 providers spread across 9 academic-associated centres (and two states) included 44 820 visits to the emergency department for acute respiratory illness. Enrolling cases from July 2017 to February 2018, the primary outcome of this trial was to see if the enhanced intervention decreased the rate of inappropriate antibiotic prescribing. This trial did not show that more intensive behavioural nudging decreased inappropriate, however, did demonstrate an overall decrease of inappropriate prescribing, decreasing from 6.2% to 2.4%. With the adjustment model the difference was calculated to be 2.2% vs 1.5% post-intervention (OR 0.67).

Unfortunately, without a control arm it is unclear if the difference in prescribing is due to the intervention (likely) or due to some other change, such as a change in practice or a less severe “flu season.” There was a loss of randomization between sites as there was up to an 80% cross-over so that may have contributed to no difference being found. Safety end-points would have been an important addition to this trial but the additional follow-up required may not have been pragmatic. Finally, case identification using ICD-10 codes may have introduced bias or missed key cases. Overall this trial suggests not only that running an antimicrobial stewardship is possible within the emergency department, but things such as education for providers as well as patients and the use of visual aids are effective methods for such a program.


As of September 2019, no guidelines have been published that reflect the results of this trial.


  • cluster-randomized effectiveness, multi-centred, controlled trial
  • N=291 providers
    • Enhanced Intervention: 5 sites (n=196 providers)
    • Adapted Intervention: 4 sites (n=95 providers)
  • Setting: 4 Emergency Departments (ED) and 5 Urgent Care Centres (UCC) across three academic centres (Ca and Co) in America
  • Enrollment: July 2017 to February 2018
  • Analysis: interrupted time-series random and fixed effects analysis
  • Primary Outcome: provider-level systemic antibiotic prescribing rate for ARI diagnosis


Inclusion Criteria

  • All clinicians working in either ED or UCC
    • Adult and Pediatric Emergency physicians, internists, or pediatricians
  • ICD-10-CM codes consistent with conditions usually not responsive to antibiotic therapy
    • nonsuppurative otitis media, H65*; acute nasopharyngitis, J00*; laryngitis, J041*; supraglotti- tis, J043*; croup, J050*; influenza, J09*/ J10*/ J11*; viral pneumonia, J12*; viral bronchitis, J203*/J204*/ J205*/J206*/J207*/J208*; unspecified bronchitis, J209*; bronchiolitis J21*; lower respiratory tract infection unspecified, J22*; vasomotor and allergic rhinitis, J30*; chronic nasopharyngitis, J31*; bronchitis not otherwise specified, J42*; and asthma, J45*
  • Patient baseline visit

Exclusion Criteria

  • Resident physicians or fellows not working as attending physicians
  • non-acute respiratory infection (ARI) bacterial infections
  • antibiotic-appropriate ARI
  • non-systemic antibiotic treatment

Baseline Characteristics

Enhanced Group displayed

  • Baseline Antibiotic Prescription Rate for antibiotic-nonresponsive ARI: 4.3
  • Baseline Antibiotic - nonresponsive ARI Visits per Provider per Month: 9


  • Adapted Intervention: provider education (presentations, electronic reminders), patient education (discharge handouts), Physician champion, Departmental Feedback (monthly aggregate prescribing data)
  • Enhanced Intervention: Adapted Intervention plus: Provider Commitment-Enhanced Provider commitment (personalized and signed posters/education), Provider peer comparison (personalized monthly summary of prescribing practices and designating “top performer” or “not a top performer")


Comparisons are pre-intervention vs. post-intervention.

Primary Outcomes

Antibiotic prescription rate for antibiotic - nonresponsive ARIs
4.3% vs. 1.8% (effect size 1.9%, 95% CI –0.7% to 4.6%) P = 0.06

Secondary Outcomes

Monthly visits antibiotic - non responsive ARIs per provider
11 vs. 13
Antibiotic prescribing for ARI visits
6.2% vs. 2.4%
Antibiotic prescription rate for antibiotic - nonresponsive ARIs, adjusted
2.2% vs. 1.5% (OR 0.67, 95% CI 0.54-0.82)

Subgroup Analysis

Comparison between the two interventions was reported as non-significant


  • Trial was intended to be a comparison between interventions, change that was found may have been associated with change in practice over time
  • no safety endpoints were included
  • generalizability limited to academic centres
  • ICD-10 codes used to identify cases may have introduced bias
  • cross-over of providers between interventions may have introduced bias
  • low baseline inappropriate prescribing of antibiotics may have led to the trial being underpowered


  • not stated

Further Reading