Return visits to the emergency department: An analysis using group based curve models.

Area of Science:

• Health Services Research
• Computational Health Informatics
• Public Health

Background:

• Repeated emergency department (ED) visits represent a significant cost and utilization challenge in healthcare systems.
• Patient risk stratification is crucial for identifying individuals who may benefit from targeted interventions.
• Group-based trajectory modeling (GBTM) offers a method to identify distinct patient pathways over time.

Purpose of the Study:

• To apply GBTM to a large cohort of emergency department (ED) patients to identify distinct risk trajectories.
• To understand the characteristics associated with higher-risk ED revisit patterns.
• To inform healthcare organizations on stratifying patients for targeted interventions.

Main Methods:

• Utilized group-based trajectory modeling (GBTM) on a dataset of 37,416 patients with up to 10 ED visits between 2006 and 2016.
• Integrated electronic health record (EHR) data, state health information exchange data, and area-level social determinants of health.
• Identified three distinct patient trajectory groups based on ED visit patterns.

Main Results:

• Three distinct patient trajectory groups were identified.
• Higher-risk trajectories were associated with behavioral diagnoses, injuries, alcohol and substance abuse, stroke, and diabetes.
• These findings highlight specific patient characteristics linked to frequent ED utilization.

Conclusions:

• GBTM is an effective computational technique for understanding patient risk in large populations.
• Identifying high-risk trajectories allows for patient stratification and the development of targeted, early interventions.
• This approach can optimize resource allocation and improve outcomes for frequent ED users.