# Data Fit for Health Equity: Learning Health Systems, AI, and the STANDING Together Recommendations

## TL;DR

This paper discusses how AI in healthcare can worsen health inequities and proposes the STANDING Together recommendations to address bias in data and AI systems.

## Contribution

The novel contribution is the STANDING Together framework for identifying and reporting bias in AI healthcare data curation and development.

## Key findings

- Learning Health Systems can use the STANDING Together recommendations to promote health equity through transparent data reporting.
- The recommendations help identify and mitigate biases in AI healthcare technologies.

## Abstract

Artificial Intelligence (AI) tools may deliver significant improvements in healthcare and Learning Health Systems are well positioned to benefit. However, during the adoption of AI, Learning Health Systems should consider the potential for AI to exacerbate health inequity and perpetuate biases that exist in healthcare and its associated data.

The STANDING Together recommendations provide a method to identify and report potential bias during the curation of datasets for AI and the development of AI from those datasets. The recommendations could form a key learning cycle within a Learning Health System ensuring transparent reporting of healthcare data use and the implementation of AI healthcare technologies that promote health equity.

Learning Health Systems are well placed to adopt the STANDING Together best practice recommendations for using healthcare data as they are likely to have both the capabilities to implement the recommendations and the strategic goals that will realize the value of health data and AI that promotes health equity. The STANDING Together recommendations are available from www.datadiversity.org.