FeBRILe3: Safety Evaluation of Febrile Infant Guidelines Through Prospective Bayesian Monitoring

Area of Science:

• Pediatrics
• Infectious Diseases
• Health Services Research

Background:

• Evidence supports earlier discharge for well-appearing febrile infants at low risk of serious bacterial infection (SBI).
• However, prolonged hospital admissions (≥48 hours) remain common for these infants.
• Prospective safety monitoring is crucial for broader guideline implementation.

Purpose of the Study:

• To evaluate a new hospital guideline recommending early discharge for low-risk febrile infants.
• To assess the safety of early discharge using a sequential Bayesian safety monitoring framework.
• To determine if hospital readmission rates within 7 days of discharge exceed a predefined safety threshold.

Main Methods:

• A sequential Bayesian safety monitoring framework was employed.
• Infants under 3 months admitted for fever without source were enrolled in Western Australian hospitals (August 2019-December 2021).
• Readmissions within 7 days were assessed against a priori safety thresholds (8 per 100 infants).

Main Results:

• Readmission rates remained below the safety threshold throughout the study.
• A total of 20 readmissions occurred (4 per 100 infants), with a >0.99 probability of being below the threshold.
• Early discharge (before 48 hours) occurred for 38% of infants, including 52% of low-risk infants, with only 3% of early-discharged low-risk infants readmitted for unrelated issues.

Conclusions:

• A Bayesian monitoring approach demonstrated that early discharge for low-risk febrile infants is safe.
• This framework supports guideline implementation to reduce low-value care.
• No increased risk of readmission was observed with early discharge.