Risk analysis of a patient monitoring system using Bayesian Network modeling

Area of Science:

• Health Informatics
• Information Systems Security
• Risk Management

Background:

• Complex information systems necessitate proactive failure prediction.
• Healthcare information systems require rigorous risk analysis due to their critical nature.
• Existing methods may not adequately address the interdependencies of system failures.

Purpose of the Study:

• To propose a novel risk analysis framework for health information systems.
• To integrate CCTA Risk Analysis and Management Methodology (CRAMM) with Bayesian Networks.
• To enhance the prediction and prioritization of system failures in critical healthcare applications.

Main Methods:

• Utilized CCTA Risk Analysis and Management Methodology (CRAMM) for asset, threat, and vulnerability identification.
• Employed Bayesian Networks for graphical modeling of interrelationships between undesirable events.
• Conducted "what-if" studies using the developed Bayesian Network model.

Main Results:

• Successfully identified and valued assets, threats, and vulnerabilities within the health information system.
• Developed a Bayesian Network model that visually represents the interactions of potential system failures.
• Prioritized critical events through "what-if" scenario analysis, demonstrating the framework's efficacy.
• Applied the framework to a vital signs monitoring system (VITAL-Home) in a homecare telemedicine setting.

Conclusions:

• The proposed integrated method effectively analyzes risks in health information systems.
• Bayesian Networks provide a powerful tool for visualizing and understanding complex failure interactions.
• This framework enhances the ability to anticipate and mitigate failures in safety-critical healthcare IT systems.
• The VITAL-Home system case study validates the practical applicability and effectiveness of the risk analysis approach.