A Hybrid Markov-SPC Approach to Assess Cost Differences in Urgent Care Utilization Using Patient-Reported Data in...

Area of Science:

• Health Services Research
• Health Economics
• Data Analytics

Background:

• Estimating healthcare costs in real-time is challenging for quality improvement.
• Patient-reported outcomes (PROs) offer a valuable data source for cost analysis.
• The IBD Qorus Learning Health System (LHS) collects crucial PRO data.

Purpose of the Study:

• To develop and prototype a real-time cost-estimation method.
• To assess cost differences between usual care and improvement initiatives.
• To leverage PRO data, Markov modeling, and Statistical Process Control (SPC) analytics.

Main Methods:

• Utilized patient-reported outcomes (PROs) including emergency department visits and hospitalizations.
• Employed Markov modeling to estimate annualized per-patient cost differences.
• Applied Statistical Process Control (SPC) analytics for real-time visualization and monitoring of cost estimates.

Main Results:

• Developed a real-time Markov-informed SPC visualization prototype.
• The prototype analyzes and monitors monthly annualized per-patient cost savings.
• Validation against claims data is currently in progress.

Conclusions:

• The hybrid Markov-SPC approach provides near real-time cost difference estimates.
• This method can enhance improvement, advocacy, and strategic planning.
• Successful validation could establish a new standard for cost analysis in healthcare improvement.