Determining preventability of pediatric readmissions using fault tree analysis.

Area of Science:

• Healthcare Management
• Pediatric Medicine
• Quality Improvement

Background:

• Distinguishing preventable from non-preventable hospital readmissions presents efficiency and consistency challenges.
• Previous methods for readmission review lacked standardization and were time-consuming.

Purpose of the Study:

• To evaluate the efficiency and reliability of a web-based fault tree tool for physician-led chart reviews to determine readmission preventability.
• To identify root causes of general pediatric readmissions and quantify the proportion deemed preventable.

Main Methods:

• General pediatricians utilized a web-based fault tree tool to classify the root causes of all 15-day general pediatric readmissions in 2014.
• The tool guided reviewers through a structured question process to assign one of 18 root causes.
• Inter-rater reliability was assessed by cross-checking 20% of cases.

Main Results:

• The overall 15-day readmission rate was 3.4% (248 of 7252 discharges).
• Potentially preventable readmissions accounted for 6.0% of all readmissions (15 cases, 0.2% of total discharges), with premature discharge being the most frequent cause.
• High inter-rater reliability (κ = 0.79) was achieved, with 86% agreement on root cause classification.
• Each review using the tool took approximately 20 minutes.

Conclusions:

• The web-based fault tree tool enables efficient and consistent identification and classification of hospital readmission root causes.
• The study confirmed that a small percentage of general pediatric readmissions are potentially preventable.
• This tool supports quality improvement initiatives by standardizing the readmission review process.