Practices enhance the first and only two-sided SaaS actuarial platform for VBC
At Syntax, the first questions we get from potential clients usually involve data:
How do you manage payor and provider data?
How can I add market data to my value-based care modeling?
Can I use data from multiple sources? And even;
What’s your data strategy as you grow?
I love these questions, not only because we anticipated these scenarios when we built Syntax, but also because questions like these give me the chance to talk about our Data Governance program.
In my last blog, I discussed why software engineering is the foundation of Syntax Health. Today, I want to share another critical building block: Data Governance.
The timing couldn’t be better, as Syntax just announced a new partnership with CareJourney—a data integration that will offer our clients significant supplemental data to model their VBC contracts. This partnership is possible partly because of our strong governance practices—having planned for integrations like this in our data strategy.
Many data projects fail due to integration and quality problems. What’s more, data is a representation of a world that is constantly changing. This constant change makes building anything atop data difficult—like building a house on sand. Data Governance provides a way to manage data throughout its life cycle—building a house on rock instead of sand.
Syntax’s philosophy aligns with this brief from The American Health Information Management Association regarding the key aspects of data governance:
Ownership: The entire enterprise produces and benefits from data, data challenges are enterprise-wide, and collaborative ownership of data is required.
Strategy: Data must be handled deliberately and strategically, not in a haphazard, ad-hoc manner, and investments must be made to execute data strategy.
Quality: Standard processes, tools, quality benchmarks, and metadata must be defined at an enterprise level in accordance with the strategy.
At Syntax, there are several ways we use data governance to achieve results for our clients. Here are a few real examples:
Collaborative Ownership
At Syntax, data governance is jointly owned by engineering, analytics, and business. When Syntax recently onboarded a provider-sponsored health system, before doing anything with their data, I met with our Chief Analytics Officer, Emily Walker, to discuss the semantic meaning of the data from a business standpoint. We leverage the deep domain expertise of our Syntax analytics and business partners to ensure we understand the use cases supported by the data, the value of the data, and the meaning of the data. Then, we leverage engineering’s expertise in data modeling, architecture, and pipelining to manage and leverage the data. The key thing here is we—engineering—do not unilaterally own the data or the data process. It is something collaboratively owned and determined by multiple stakeholders.
Strategic Execution
Prior to the integration of any data at Syntax, we collectively assess its strategic value, and then the data is carefully and efficiently modeled and understood, avoiding ad-hoc technical debt. For example, when Syntax began to integrate CareJourney data into our platform, instead of focusing first on how to get their data into our environment, we focused on the semantic meaning of the data and on its value. We took time to anticipate how our user experience would change when adding CareJourney data to their modeling work. Then, we moved forward with integrating the data in a way that fit nicely with our existing data and showcased its semantic meaning well.
Tech-Enabled Quality
There’s a human component to data quality, but human effort alone isn’t enough. Syntax’s data platform was architected to minimize the toil involved in data quality. We selected our entire data platform’s technology stack with data quality in mind–knowing that data is messy. One key component of our stack is DBT, which makes it trivial to implement automated baseline data quality checks, such as checking for null/empty/missing or duplicate values. DBT also makes it easy to add bespoke, custom data quality checks–so as we find more errors, we can add further data quality checks to detect errors and prevent downstream damage. DBT also facilitates the use of packages, opening up a rich ecosystem of open-source tools we can leverage to avoid reinventing the wheel.
Conclusion
Without good data governance, nothing else is possible in a repeatable, scalable, or long-term way. If you’d like to improve your VBC modeling and contracting—using our platform built with strong data governance—reach out to us at Syntax; we’d love to help.
Will Hudgins is a dynamic technology leader who is passionate about transforming healthcare for the better. He brings an engineering leadership that balances technical excellence with attention to the unique nuances of healthcare.