Ten Core Concepts for Ensuring Data Equity in Public Health

Area of Science:

• Public Health Data Science
• Health Informatics
• Computational Public Health

Background:

• Public health decisions increasingly depend on large-scale data and technologies like artificial intelligence (AI) and mobile health.
• Underrepresentation of certain populations (rural, disabled, homeless, low/middle-income) in health datasets leads to biased findings and suboptimal outcomes.
• Addressing data inequities is critical for equitable health outcomes from technological advancements.

Purpose of the Study:

• To propose a framework of 10 core concepts for improving data equity in public health research and practice.
• To integrate computer science principles (fairness, transparency, privacy) with public health data science best practices.
• To provide a structured approach for evaluating data practices throughout the data lifecycle.

Main Methods:

• Developed a framework integrating computer science principles (fairness, transparency, privacy) with public health data science best practices.
• Applied concepts across the data life cycle: study design, data collection, analysis, interpretation, and policy translation.
• Focused on mitigating information and selection biases, learning causality, and ensuring generalizability.

Main Results:

• Proposed 10 core concepts to operationalize data equity in public health.
• The framework offers a structured approach to evaluate data practices for representation and service to all populations.
• Data equity is foundational for trustworthy inference and actionable evidence.

Conclusions:

• Data equity is essential for ensuring technological and digital health advances improve outcomes for everyone, not widen existing gaps.
• The 10 core concepts provide a practical method to operationalize data equity in public health.
• Advancing data equity requires parallel efforts in information theory and structural changes for informed decision-making.