Using a Clinical “Insights in the Moment” Approach to Data Governance

This is the third of three articles, powered by CHIME Digital Health Insights and sponsored by Philips Healthcare, exploring how AI can help improve patient outcomes and operational efficiencies by addressing the challenges of data overload and fragmented systems in healthcare provider organizations.

In an era of artificial intelligence and liquid data, traditional approaches to data governance need a refresh. The volume of data that AI requires, the opaque nature of AI models, and the privacy and regulatory concerns of analyzing large patient datasets all require discussion.

An effective approach is to consider how AI is deployed at the point of care. When organizations view governance through the lens of clinical insights in the moment, governance can evolve to meet modern data needs.

Empower Clinicians with Insights in the Moment

There are three principles of clinical insights in the moment.

• Provide a longitudinal view of a patient’s medical history and wellbeing.
• Integrate AI-powered analytics and actionable insights into clinical workflows to enable informed decision-making.
• Balance the need for timely insights with data privacy and security considerations.

It’s critical to integrate the functions that enable these insights with existing applications. Expecting clinicians to exit the electronic health record (EHR) and log into another system causes interruptions, frustrating end users already challenged by fragmented workflows.

Simply replicating existing interfaces isn’t enough. User-friendly dashboards are necessary to transform the output of highly technical AI models into insights clinicians can interpret at a glance. Augmented reality is one option for creating 3-D visualizations that are easier to process than a series of flattened charts or tables. The less time clinicians spend staring at the data, the more time they have to use their expertise to help patients in need.

Clinical insights in the moment can have tremendous impact. More than a decade can elapse between symptom onset to diagnosis for mental illness or Type 2 diabetes – conditions that affect hundreds of millions of Americans. Real-time access to decision support can help clinicians spot a possible diagnosis in minutes instead of years. That can contribute to lower care costs, better outcomes, and improved quality of life.

Implement an “Insights in the Moment” Data Governance Approach

To realize this potential, organizations need to modernize their approach to data management and governance. A previous post outlined the key management component – a that enables open access to standardized data and frees data from siloed environments. Meanwhile, governance must balance traditional pillars such as data quality, security, access, and lifecycle management with additional considerations.

One is usage permissions. It’s no longer enough to define permissions based on roles. Access must account for clinical context as well. Nurses floating among units of the hospital should have access to different data sets and decision support tools depending on where they’re working that day.

Context matters for researchers, too. Data anonymization practices vary based on the type of research being conducted and who will see the results. Dynamic anonymization and de-identification capabilities apply the right privacy policies in the right context – and can create synthetic data sets if necessary.

Finally, modern governance benefits from modern tools. Automated network and data monitoring, coupled with anomaly detection and response, enhances security by further ensuring only authorized entities access the right data at the right time. Along with peace of mind, this provides an audit trail to improve compliance and satisfy cybersecurity insurers.

Address Ethical and Legal Considerations

Of course, data governance today must consider the ethics of AI utilization. Clinical teams should trust that AI models are transparent, trained on data sets that minimize the impact of bias in AI, and deployed to complement their expertise without replacing their decision-making capabilities.

The World Health Organization (WHO) has provided guidance on AI governance. This covers how AI models should be developed and used in the context of ethical standards and human rights for those using the models (clinicians) and for those whose data powers the models (patients). WHO notes that governance and enforcement, including audits of AI models, may require governments to establish new legal frameworks or regulatory bodies.

The Future of Clinical Data Governance: A Continuous Journey

The most effective governance policies are living documents. That’s true more than ever as AI capabilities and regulations evolve. Continuously re-adapting governance can seem challenging, especially at a time when organizations face competing priorities and pressures.

However, policies that evolve with the times help to foster a culture of continuous learning and improvement while encouraging collaboration among governance, AI development, and clinical practices. That collaboration is essential to encourage AI adoption and realize its benefits at the moment.

Read the first two articles in this series:

1. Breaking Down the Walls: Liquidating Data Silos for Enhanced Insight Extraction.
2. Evolving to a Utility Infrastructure Model for Enhanced Data Management

About Philips

Royal Philips is a leading global health technology company focused on improving people’s health and well-being through meaningful innovation, employing about 74,000 employees in over 100 countries. Our mission is to provide or partner with others for meaningful innovation across all care settings for precision diagnosis, treatment, and recovery, supported by seamless data flow and with one consistent belief: there’s always a way to make life better. For more information, please visit https://www.philips.com/global.