Applying CHAID algorithm to investigate critical attributes of secured interoperable health data exchange

Area of Science:

• Health Informatics
• Information Security
• Healthcare Systems

Background:

• The increasing exchange of electronic health records (EHRs) in interoperable healthcare environments has led to significant concerns regarding data security and confidentiality.
• Interoperable healthcare systems are complex, comprising numerous independent components, necessitating a robust security framework.

Purpose of the Study:

• To identify critical security attributes and their dependencies within interoperable healthcare systems.
• To determine the most influential factors for reducing security threats during the sharing of electronic health records.

Main Methods:

• Utilized a real-time study of patient-doctor relationships in hospitals with functional EHR systems.
• Employed the chi-squared automatic interaction detection (CHAID) classification technique to identify data dependencies.
• Analyzed and verified the resulting decision tree using regression analysis.

Main Results:

• Identified salient features that significantly control security threats in EHR sharing.
• Established dependencies among security attributes crucial for a secure interoperable healthcare framework.

Conclusions:

• Understanding and addressing key security attribute dependencies is vital for mitigating risks in EHR exchange.
• The findings provide a foundation for designing more secure interoperable healthcare systems to protect patient confidentiality.