Development of data-driven clinical pathways: the big data clinical evidence-based pathways.

Area of Science:

• Healthcare Management
• Health Informatics
• Clinical Pathway Development

Background:

• Traditional clinical pathways often rely on expert experience rather than objective data.
• There is a need for standardized treatment protocols and integrated payment standards.
• Existing systems may lack mechanisms for detecting upcoding and overtreatment.

Purpose of the Study:

• To develop clinical evidence-based pathways (CEBPWs) for standardizing treatment protocols.
• To align diagnosis-reimbursement criteria and detect healthcare fraud like upcoding.
• To enable early warnings for potential overtreatment and enhance healthcare quality.

Main Methods:

• CEBPWs were developed using patient-level data from 166 public hospitals in Shanghai (2022-2024).
• The study analyzed over 5.3 million cases, encompassing diagnosis-therapy groups, hospitalisation, and outpatient charges.
• The GROWTH algorithm identified frequently charged items within diagnosis-therapy combinations.

Main Results:

• CEBPWs integrate objective evidence, classification, value/frequency weighting, and temporal sequencing.
• Application to 224 diseases identified upcoding, overtreatment, and fragmented care, achieving high coverage for diagnostics, therapy, and consumables.
• Pilot evaluation revealed violations with significant frequency and cost deviations across diagnosis, therapy, drugs, and consumables.

Conclusions:

• CEBPWs offer an objective, data-driven alternative to experience-based clinical pathways.
• These pathways are critical tools for hospital management and regulation, addressing drawbacks of traditional methods.
• CEBPWs enhance healthcare quality by providing standards for diagnosis, therapy, and payment.