Among U.S. healthcare payers and providers, data governance has long been considered a best practice, and today, you’d be hard pressed to find an organization that doesn’t at least have the basics in place. However, as more and more tools powered by artificial intelligence (AI) come into use, data governance is evolving from a best practice to an essential priority and “having the basics in place” is far from enough.

In our experience, data governance is one of those areas where too many organizations make reactionary changes only after something has gone awry. With the rapid pace at which new AI tools are being used in the workplace, it’s clear that now is the time to institutionalize a new approach that proactively makes data governance a business imperative.

The need for data governance strategy

In healthcare, data governance is a key factor in an organization’s ability to mitigate risk and improve compliance across a myriad of global regulations primarily focused on protecting sensitive information. And it’s not just regulators who are concerned about data; today’s patients and healthcare professionals also demand more authority over their own data and how it’s being used. This means managing, optimizing, and ensuring the accuracy of data are all critical to commercial success.

Data governance generally includes several key capabilities:

• Optimizing overall data quality.
• Reducing data duplication.
• Common understanding and definition of the data.
• Repeatability in the use of data.
• Providing easy, consistent access to data.
• Democratizing data across the entire business.
• Delivering accurate, real-time responses to important business questions.

Despite sharing these common attributes, data governance is not a “one size fits all” solution. Every organization should take the time to craft a strategy best suited to their particular interactions with data.

Developing a data governance strategy starts with identifying an organization’s level of maturity and where they aspire to be. Not all organizations are ready to be at a level four or five maturity if they are currently at a one or two. Concrete steps to improve are key, allowing success to build upon success. Identify the most critical data elements and start to govern and define those first. 

It’s also important to be selective in identifying data owners. The ownership role comes with some important responsibilities that require time and attention. Don’t think of it as just another add-on responsibility that simply needs a manager’s name attached to it.

Finding the right technologies to help implement your data governance strategy comes later. Technology will not fix a poorly managed data governance office, nor satisfy unreasonable, poorly planned expectations.

Benefits of data governance

One important part of getting buy-in for developing a data governance strategy — as with any business initiative — is to show real, measurable value. Since the use of trusted, governed data impacts nearly every part of the modern healthcare organization (for payers and providers alike), value can be found in such capabilities as:

• Improving regulatory compliance.
• Generating faster, more accurate business analytics/insights.
• Extending access and utility across a wider portion of the organization.
• Accelerating and streamlining new initiatives.
• Reducing waste, inefficiency, and costs associated with multiple parties using different datasets while attempting to solve the same problem or answer the same question.
• Reducing errors stemming from and work required to fix poor-quality data.

In terms of AI-specific value, achieving any of these benefits requires the right input data (clean, uniform, structured, tagged with metadata, etc.) so the AI engine can deliver optimal results. AI makes the old adage “garbage in, garbage out” more relevant than ever — but the flipside is equally true: “Value in, value out.”

Strategy first

Achieving the benefits and value promised by AI requires a comprehensive and well-planned data governance strategy. Some of the critical pieces of your new strategic framework that need to be clearly defined include:

• Roles and structures (who has ownership, decision rights, data access, etc.).
• Policies and standards (e.g., treatment of internal vs. external data, when new data needs to be defined and how).
• Well-established guardrails (e.g., access to PHI/PII).
• Architecture and integration (cloud storage, access technologies, analytics tools, etc.).

Technology, of course, plays a role in data governance, but a lot of work first goes into analysis and planning to ensure that technical solutions are suited to your actual needs and can evolve with you in the future.

Now is the time to start

IQVIA’s data governance customers may come to us after realizing they should streamline initiatives and invest in protecting their data, or having anticipated the need for data governance in an increasingly complex environment. Whatever the circumstances, committing to data governance starts at the top. As with any critical business initiative (and yes, data is no longer simply an IT initiative — it’s very much a business initiative), this one must be prioritized for resource allocation and must have executive leadership.

The good news is you don’t have to do it all at once. Start with just a single step: Make an assessment of where you are now and, most importantly, where you want the data journey to take you. It’s also important to realize that this truly is a journey; don’t think you can define one single endpoint and be done with it. Data and regulations are always changing, and now AI is accelerating that change on a magnitude seldom seen. In the face of all this change, a well-planned strategy will ensure your data governance is flexible and able to deliver real value.