Industry Voices—They key to successful health-focused AI tools? Quality data

Artificial intelligence is revolutionizing the healthcare industry through advancements in predictive modeling and analytics that help with delivering data-driven insights and assist in decision-making. 

However, AI tools can only be effective if served with the right kind of data. To bolster individualized care and enhance overall healthcare outcomes, the quality of the data that feeds AI algorithms should be the top priority as the industry begins implementing these new digital technologies into everyday practice. 

To capture “clean” or more complete data, AI adopters in the health sector must take three courses of action: record accurate data, develop robust data governance protocols and establish responsible AI principles. 


Document data accuracy
 

An AI tool is only as good as its data. 

Accurate records—clinical, claims data, PDFs, etc.—provide the groundwork for health-focused GenAI models. They are also crucial to ensuring the correct and most effective healthcare outcomes. For example, accurate, centralized data formed the basis of the Sydney Health app, resulting in precise, personalized healthcare recommendations and services. 

We also know that there can be inaccuracies in data for a number of possible reasons—errors in entering data into the system, duplicate data, misinterpretation of data, etc. Technologists need processes in place in order to identify and correct these data inaccuracies before analytics and algorithms are performed on these data. 


Centralized and well-governed data
 

Quality, in the context of data, is not merely about the information being correct. It's also about how efficiently it is managed and readily available when needed. 

Healthcare organizations can prioritize data quality through a centralized data repository and robust data governance protocols. Centralizing data ensures that all insights and models are based on the most comprehensive information available. Data governance protocols provide the structure, policies and procedures essential to managing data consistently and accurately, ensuring AI and GenAI models are trustworthy and reliable. 

At Elevance Health, we have the Enterprise Data Governance Executive (EDGE) program, a governance structure designed to safeguard data and maximize its potential. There are five strategically aligned committees within our structure to address data use, data quality and management, privacy and security, regulation of offshore engagements, and responsible AI. 

Safeguarding data is crucial as the industry develops new AI solutions, and a robust governance structure is the key to achieving quality and accurate data across the board. 


Responsible AI 
 

When developing AI solutions, healthcare organizations must establish a strong framework and a set of guiding principles to ensure data are being handled ethically and responsibly. This includes considering the power and impact AI can have before, during and after solutions are implemented.  

Our teams at Elevance Health have established the Office of Responsible AI (ORAI), a program responsible for developing, implementing and maintaining AI governance. The ORAI aims to enable AI innovation while also mitigating and minimizing disparate impacts. 

To ground AI development in our values and help ensure adherence to regulations, the ORAI adopted five guiding AI principles: fair and inclusive, robust, explainable and transparent, accountable, and private and secure. 

When developing new technology, people should always be the priority. These guiding principles put consumers, providers and communities at the center by protecting their data as we create personalized digital experiences.  

To conclude, quality, clean data are pivotal in an era where digitization, AI and GenAI are fundamental drivers of healthcare innovation and excellence. The journey toward an improved healthcare experience is best traveled together. Healthcare organizations using AI should work closely with their partners, providers and other stakeholders to ensure the foundation of data-driven insights—the data themselves—are accurate so that we can move forward together, appropriately.

Ashok Chennuru is the chief data and digital transformation officer at Elevance Health.