Health Data Governance and Value Creation

The FHT “health data microcosm” complements the work of research modules 1 - 3 on diverse health problems and applications of mobile digital health technologies. The well-defined and controlled “small-world system” of data and users is sufficiently holistic and interdisciplinary to mimic the complexity of real-world health systems.

This allows researchers to take a systemic approach in developing new mobile digital health technologies and to unify lessons learned from test cases in modules 1 - 3 in the form of best practice protocols, guidelines and an IT infrastructure which provides a flexible platform for testing new technologies.

The FHT health data microcosm will be implemented on this infrastructure, which will be populated with digital health avatars (i.e. model with determined parameters) that can be tracked over time to quantify either the effect of interventions or deterioration of health. The empirical model allows researchers to study the systemic effects of new health technologies by building in-silico models for health forecasting, treatment response predictions and cost estimations.

The team will define a “trustworthy data governance” concept applicable to the legal and regulatory framework of the Singaporean ecosystem. It aims to define soft law that protects privacy, data ownership and accountability, while enabling modern analytics across a variety of health data that populates the FHT health data microcosm.  

Modified Policy Delphi: Development of an ethical code on the collection, use and transfer of data in digital health technology

In 2022, the team conducted a modified Policy Delphi mixed-methods study to engage an expert panel of stakeholders in developing an ethical code that is both values-based and empirically informed on how researchers in Singapore should collect, use and transfer whatever constitutes potentially sensitive health data. The panel’s deliberations occurred over three stages starting with semi-structured interviews to generate ideas and policy options, which were then prioritised with an online survey and refined at a stakeholder workshop. Five stakeholder groups participated in the Policy Delphi: data contributors and end-users; data generators; data resources; data facilitators; and professional data users.

TIER 1: HIGHLY SENSITIVE DATA

Tier 1 has the highest degree of sensitivity and includes personal data about stigmatising conditions such as, HIV status, history of sexually transmitted diseases (STDs), mental health disorders, and fertility treatments. Some of these information are found on the Ministry of Health’s list of reportable medical information under the Personal Data Protection Act.

Examples:

• Medical history of common diseases (e.g. diabetes, stroke, heart attack etc)*
• Medical images or scans of body parts*
• A person’s direction-finding ability collected from phone apps*
  
  

TIER 2: POTENTIALLY SENSITIVE DATA

Tier 2 consists of de-identified Tier 1 health data and other information on the MOH list, such as drug and alcohol abuse, history of organ transplantation and contraceptive treatments. The panel added medical history of cancer and other common (non-stigmatising) diseases, medical images and self-reported mental health status.

Examples:

• Data in Tier 1 that is de-identified
• History of substance abuse and drug addiction (including drug abuse and alcoholism)
• History of contraceptive procedures (e.g. hysterectomy and vasectomy)
• Details about the identity of organ donation recipients, transplant procedures
• History of cancer*
• Self-reported mental health data collected from phone apps*
• Voice recording or speech data collected from phone apps*
• Sexual orientation*
  

TIER 3: NON-SENSITIVE DATA

Tier 3 contains other non-sensitive data such as medical history of common diseases (e.g. diabetes, stroke, heart attack etc), medical images or scans of body parts, a person’s direction-finding ability collected from phone apps.

Examples:

• Medical history of common diseases (e.g. diabetes, stroke, heart attack etc)*
• Medical images or scans of body parts*
• A person’s direction-finding ability collected from phone apps*
  

*Data points that are not on the MOH list of reportable medical information