Generating the data to support sophisticated analytics
Privia, a national physician organization that helps providers move toward value-based care, relies on enhanced reporting and analytics as the engine of its operation. “Data is the most critical building block for that,” Kathuria said. “We pride ourselves on being a data-rich organization, but that doesn’t mean we were data-rich from day one.”
Over the years, Privia has incorporated various tools to generate the requisite data. These include:
Electronic health records. All Privia-affiliated providers are on the same EHR.
Payer claims. Privia established partnerships with payers, including CMS, to receive direct feeds.
Hospital admission, discharge and transfer feeds. As an example, Privia gets a virtually real-time flow of data from Maryland’s state health information exchange.
Ancillary data. These sources include imaging, lab and pharmacy.
The various sources feed into a centralized data warehouse that serves as the foundation for Privia’s reporting and analytics initiatives, Kathuria said.
Enhanced reporting processes to support advanced care
Privia guides its client-practices along a five-stage journey to develop the type of data-reporting processes needed to support advanced clinical models, Kathuria said (some practices may already be in a later stage when they join Privia).
“The pace at which an organization moves through these five phases is a function of how bought-in the senior leadership is,” he added.
“Gut” reporting. The organization has little to no data-reporting infrastructure and makes business decisions based on gut reactions and past experience rather than data.
Tactical reporting. Reporting functions are siloed in departments, but the organization is starting to build capabilities such as Excel-based reporting and is incorporating data analysts.
Enterprise reporting. Reporting becomes centralized, perhaps utilizing a data warehouse. Common definitions of metrics are established, business intelligence engineers are brought aboard, and static reporting tools are converted to visually oriented and interactive dashboards.
Strategic reporting. The organization can conduct “what-if” reporting, Kathuria said, and it looks to leverage artificial intelligence to implement analytical forecasting. The role of data scientist is established.
Immersive reporting. This stage is distinguished by the embedding of relevant information and decision-making tools into a user’s day-to-day workflow. “That’s what makes this 10 times more powerful,” Kathuria said. “This is where you can finally claim to be a more data-driven organization.”
Converting data into advanced analytics
A practice’s analytics maturity cycle follows four stages:
- Descriptive/retrospective analytics (what happened)
- Diagnostic analytics (why it happened)
- Predictive analytics (what will happen)
- Prescriptive analytics (how to make it happen)
As an example of how an increasingly sophisticated analytics infrastructure can be applied, Kathuria described an organizational initiative to enhance cash collection efficacy.
Descriptive. Privia has developed a metric called net realization rate (NRR). In this analytics phase, a practice can use this metric to gauge how effectively it is collecting cash.
Diagnostic. With these capabilities, a practice can examine why its cash collection may be suboptimal. Privia provides its practices with a tool that puts collections data through various filters, such as by payer or by number of times a claim has been “touched.”
In one instance, Wood said, Privia saw that a practice’s NRR was being hampered by a low self-pay yield rate. It turned out the front-desk staff was not collecting outstanding balances or efficiently sending delinquent accounts to collections. Follow-up training helped improve the staff’s collection methods and, in turn, the practice’s NRR.
Predictive. This phase allows a practice to consider its NRR, factor in certain variables and forecast the metric over a given period. “This gives us an opportunity to put in additional resources and get that practice, [or] get that physician, back on track,” Kathuria said.
Prescriptive. To address situations when NRR is low because of high denial rates, Privia is working on a tool that allows a practice to predict whether a claim will be denied based on various factors.
If confidence that a claim will be accepted is below a certain threshold, as determined by the tool, the claim could go to a revenue cycle team member for additional review instead of going directly out to the payer.
The tool will be embedded in the electronic health record, with relevant claims redirected to the appropriate team member’s inbox, making it illustrative of both prescriptive analytics and immersive reporting. “Both of them overlay in terms of our approach to solving this problem,” Kathuria said.