Nurse Generated EHR Data Supports Post-Acute Care Referral Decision Making: Development and Validation of a Two-step...

Area of Science:

• Health Informatics
• Clinical Decision Support Systems
• Healthcare Management

Background:

• Discharge planning (DP) involves complex decisions regarding post-acute care (PAC) referrals.
• Current methods for PAC referral decisions can be subjective and variable.
• Accurate and timely PAC referrals are crucial for patient recovery and reducing hospital readmissions.

Purpose of the Study:

• To develop and validate a clinical decision support (CDS) algorithm for optimizing PAC referral decisions.
• To determine the appropriate site of care for patients requiring PAC.
• To reduce subjectivity and variation in discharge planning.

Main Methods:

• Utilized electronic health record (EHR) data to create case studies.
• Engaged 171 interdisciplinary experts to judge case studies and inform model development.
• Generated prediction models to support referral decisions.

Main Results:

• A two-step algorithm was developed and validated.
• Achieved an area under the curve (AUC) of 91.5% for predicting the need for referral (yes/no).
• Achieved an AUC of 89.7% for determining the appropriate site of PAC.

Conclusions:

• The validated CDS algorithm effectively supports discharge referral decision-making using EHR data.
• Automating assessment through CDS can identify high-need patients earlier.
• This approach has the potential to decrease hospital readmissions and adverse events.