Privacy-preserving screen capture: towards closing the loop for health IT usability

Area of Science:

• Health Informatics
• Computer Science
• Human-Computer Interaction

Background:

• Information technology is increasingly integrated into healthcare, necessitating effective documentation of system usage.
• Existing screen capture methods face challenges with privacy-sensitive data and the complexity of integrated health IT systems.
• Health IT systems often involve multiple vendors and closed-source components, limiting instrumentation possibilities.

Purpose of the Study:

• To present a novel tool for capturing and analyzing tricky or troublesome usage scenarios in health IT systems.
• To address the privacy constraints associated with sensitive data in healthcare settings.
• To develop a system that is independent of specific applications and operating systems.

Main Methods:

• The study introduces a system combining keyboard video mouse (KVM) capture with automatic text redaction and interactive unredaction.
• KVM capture ensures application and OS independence by bypassing software interface dependencies.
• The system was evaluated using a corpus of Electronic Medical Record (EMR) screenshots.

Main Results:

• Empirical measurements demonstrated the effectiveness of the automatic text redaction technique.
• Processing latency was measured to assess system performance.
• The system successfully produced precise technical content from EMR screenshots.

Conclusions:

• The developed tool effectively captures and redacts sensitive information from health IT systems.
• KVM-based capture combined with redaction offers a versatile solution for documenting complex usage scenarios.
• These techniques can enhance stakeholder communication and improve end-user influence on health IT system evolution.