The Anonymous Data Warehouse: A Hands-On Framework for Anonymizing Data From Digital Health Applications

Area of Science:

• Digital Health
• Health Informatics
• Data Security

Background:

• The digital health sector is expanding, increasing the demand for sharing anonymized health data.
• Current medical literature lacks comprehensive coverage of data anonymization techniques.
• Digital health data presents unique properties and challenges for anonymization.

Purpose of the Study:

• To offer a practical framework for data anonymization tailored to digital health applications.
• To synthesize literature trends and common anonymization methods into a usable structure.
• To address the specific opportunities and challenges inherent in digital health data.

Main Methods:

• Synthesizing literature on anonymization techniques and digital health data trends.
• Developing a framework focusing on separate data storage and localized anonymization.
• Implementing strategies such as minimizing static attributes and ensuring k-anonymity.
• Utilizing aggregation, rounding, and capping to prevent quasi-identifiers.

Main Results:

• A proposed framework for anonymizing digital health data is detailed.
• The framework emphasizes separate data storage, localized anonymization, and selective data export.
• Methods to ensure user and attribute k-anonymity and prevent quasi-identifier risks are outlined.

Conclusions:

• A pragmatic approach to data anonymization is crucial for preserving data utility while minimizing reidentification risk.
• The proposed framework provides a adaptable structure for digital health data anonymization.
• The framework requires modification based on specific dataset characteristics.