Many in enterprise Data Management know the challenges that rapid business growth can present. Whether through acquisition or organic growth, the amount of enterprise data coming into the organization can feel exponential as the business hires more people, opens new locations, and serves new customers. The enterprise IT team, in particular, plays a significant role where they are tasked with consolidating new and existing data sources and enforcing quality and governance rules to ensure that the data is consistent, accurate, and available to the stakeholders who need it. But it’s easy for growing organizations to fall into the trap of “siloed” data. For instance, project teams may care only about the data and applications in their direct purview without consideration for company-wide access, much less interoperability. As data is scaled across the organization, the disparities in formatting, taxonomy, and structure can make enterprise-wide BI and analytics nearly impossible.

Rapid Business Growth and the IT Sprawl

We recently experienced this rapid growth at my own company, a physician-led and -owned partnership that helps health care organizations raise the standard of patient care and improve their performance metrics. We have grown rapidly over the past few years, now encompassing over 450 practices that care for eight million patients per year. Dealing with this growing footprint also meant coping with a substantial amount of clinical, medical billing, and other data flowing into our enterprise systems from multiple data sources that all had to be properly validated and merged. We quickly learned that these siloed data sources resulted in poor visibility, operational errors, inefficiencies, and unnecessary redundant effort on the part of our team – all of which could hinder our growth and ultimately affect our operational efficiency and patient experience. Not to say that our source system data was bad – only that it had been created in multiple systems, each with different rules and requirements. As a result, we realized that to drive our digital transformation efforts, we needed to impose some Data Governance and Data Quality rules so that we could ultimately have a single, integrated source of truth for reporting, decision-making, and operational efficiency.

Data Issues Across Source Systems and Data Types

As we started to analyze the scale of our enterprise data and the various source systems we would be working with, we realized that the problems we were trying to solve involved multiple data types. For example, when we looked at potential inefficiencies in our billing process, it involved harmonizing data on our providers, including their credentialing, as well as our locations, where simple spelling errors like abbreviating St. Luke’s when the managed care organization that processes our invoices writes it out as S-A-I-N-T.

But beyond billing and scheduling, we needed to arm our clinical partners with the information they needed to care for our patients. Cleaning and consolidating this data – and making it available to all our locations – was critical for tracking patients throughout their entire continuum of care. We needed to have clean, reliable data to provide clinical and operational insights to our clinical partners to ultimately improve processes and patient outcomes.

Although these issues encompassed several source systems, we decided to work with one domain, or data type, at a time starting with our most critical data. For us, that was our provider data, which includes the doctors and advanced providers (nurse practitioners and physician assistants) maintained in our systems. We first tried deploying a custom-coding solution to merge all this data into a custom-built data warehouse, but this required a lot of effort, was difficult to maintain, and provided no visibility into matching or survivorship rules. Ultimately, we decided to move forward with a dedicated Master Data Management platform to accommodate multiple key master data domains and the data interactions around them. But more than just implementing a technical solution, we found out that we needed to develop improved processes around Data Governance and quickly realized that the MDM program could help us on that journey as well.

Once we proved successful in our ability to collect and collate provider data from multiple sources and consolidate it based on matching and Data Quality rules to create a single source of truth, we quickly moved on to other domains such as facility/location. These included important identifiers for each contracted location, such as site identifiers, contract names, and patients, including all patient data for billing, such as insurance, addresses, and guarantor information.

Re-Framing the Data Governance Conversation by Demonstrating Value

We also framed our MDM and Data Governance program in terms of our core business objectives of providing compassionate care to the patients who need us. Rather than talking about the academic aspects of data warehouses or Data Stewardship, our project charter was to arm our clinical partners with the information they needed to care for our patients. 

Our initial goal was to improve the consistency, accessibility, and usability of our enterprise data. We approached the initiative by thinking about our issues holistically across data types and source systems. As a result, we were able to include all our critical domains in our governance and MDM programs and execute a phased delivery plan, knowing we could always add additional use cases as we matured and got additional buy-in from leadership.

Some of the immediate benefits we realized fell under two categories: 

• Operational efficiency: We initially sought out improved operational efficiency, including synchronized, trusted data leading to reduced data entry, errors, and duplicative work – which impacts all areas of the business, from accurate credentialling to patient billing. And this well-managed data also allows to identify future Data Quality or Data Governance issues and then implement necessary changes quickly and efficiently.
• Improved reporting, metrics and insights: We also sought out improved reporting, metrics, and insights we could use to guide our business. Without consistent trusted data, we couldn’t create meaningful reports or get insight into how our business was running … from utilization to profitability. Now that we knew our data had been matched, de-duplicated, and enriched, we could trust the critical insights around effective health care delivery and clinical-quality metrics like provider-level throughput and patient satisfaction. 

Building a Strong Data Culture

Beyond the tangible business and operational outcomes, one of the more remarkable benefits has been our ability to build a stronger data culture throughout the company. Once we started this journey, other business owners started coming to us and saying, “We know you have a ready technology solution for this, so help us with this problem.” Going forward, we plan on adding payor contracts to standardize the names and contract details with the private insurance and public payors who reimburse us for the care we provide followed by reference data. This last domain will involve mastering and managing the data used across our data types and include CPT codes, ICD-9/10 codes, CCS Diagnostics codes, and other codes that we use to organize our clinical and billing data to standardize reporting and reduce the time our medical billing and coding teams spend entering and cleaning data.

By taking the time to develop a grassroots, educational posture, we have been able to expand our use cases of Data Governance because others saw the benefits as opposed to it being a top-down directive. By not taking on too many data types at first, this grassroots approach allowed us to quickly demonstrate value and improve stakeholder buy-in for our team.