Calibrating Readmission Risk Prediction Models for Determining Post-discharge Follow-up Timing

Area of Science:

• Health Services Research
• Healthcare Management
• Predictive Analytics in Healthcare

Background:

• Soaring hospital readmission rates significantly strain US healthcare financial resources.
• Timely outpatient follow-up post-discharge is a proven, cost-effective intervention to reduce readmission risk.
• Physician shortages in primary and specialty care create a dilemma for transitional care planning.

Purpose of the Study:

• To explore the utility of institution-specific readmission risk prediction algorithms.
• To stratify patient populations into high- and low-risk strata for readmission.
• To assign risk-concordant timing for outpatient follow-up post-discharge.

Main Methods:

• Developing and applying institution-specific algorithms to assess diverse risk factors (administrative, clinical, socioeconomic).
• Classifying hospital patient populations into distinct high- and low-readmission risk strata.
• Assigning follow-up appointment timing based on identified risk strata.

Main Results:

• Algorithms can effectively stratify patients based on readmission risk.
• Risk stratification enables tailored follow-up scheduling.
• This approach aids in equitable human resource allocation.

Conclusions:

• Institution-specific readmission risk prediction algorithms are valuable tools for transitional care.
• Risk-stratified follow-up optimizes outpatient physician utilization.
• This strategy can effectively reduce hospital readmission rates and improve resource allocation.