An Electronic Health Record Tool Designed to Improve Pediatric Hospital Discharge has Low Predictive Utility for...

Area of Science:

• Pediatric Hospital Medicine
• Health Informatics
• Patient Safety

Background:

• Improving pediatric hospital discharge is crucial for reducing readmissions.
• Electronic health record (EHR) tools can potentially streamline discharge processes.
• Identifying high-risk pediatric patients early is essential for targeted interventions.

Purpose of the Study:

• To develop and evaluate an EHR tool designed to predict 30-day unplanned readmissions in pediatric patients.
• To assess the tool's utility in flagging children with potential discharge complications.

Main Methods:

• Development of an EHR tool incorporating three discharge risk factors: home health, polypharmacy (≥6 medications), and non-English speaking caregiver.
• Implementation of the tool across 29,542 discharged pediatric patients.
• Evaluation of the tool's predictive accuracy for 30-day unplanned readmissions using logistic regression and C-statistic.

Main Results:

• 28% of patients had a composite score of 1, and 8% had a score ≥2.
• A composite score of 1 (OR: 1.7) and ≥2 (OR: 4.2) were significantly associated with increased odds of 30-day readmission.
• The model achieved a C-statistic of 0.6259, indicating low discriminatory performance.

Conclusions:

• The developed EHR tool identifies pediatric patients with higher odds of readmission based on specific discharge risk factors.
• Despite a positive association, the tool's low discriminatory performance suggests further research is needed to enhance its practical benefit.
• Additional studies are required to validate and improve the tool's effectiveness in improving pediatric hospital discharge quality.