bluetree blog

see all posts

What data do I need before experimenting with at-risk contracts?

Value-based healthcare has been a burning topic for many organizations this year. In this post, I’ll dissect the role of analytics in shared-risk/full-risk (capitation) contracts. I’ll also show you what it takes to get the data you need to make smart business decisions while fully utilizing your Epic platform.

There are various levels of risk that are currently being tested with value-based contracts. Here's a chart from The Brookings Institution on the various approaches being tested to move our healthcare system away from fee-for-service.

              * via

Performance-based models that include a quality incentive payment also heavily need metrics and analytics to maximize their profitability.

Today, we want to focus on full at-risk contracts with a comprehensive capitation payment, such as an employee-owned plan or Medicare Advantage.

In a full-risk, full-capitation contract, the HCO assumes financial responsibility for all patient care. As in any contract, it all boils down to profitability. To measure and improve profitability, we need to be able to take a holistic view of all the contract’s components. In basic terms, it consists of the total cost of providing care and total revenue. Cost is defined by all care provided, all the way from the ED to primary care. The revenue is made up of total members x agreed per-member, per-month (PMPM) fee.

Let's dig into the revenue number. At its most basic level, Revenue = # of active members x PMPM. As enrollment changes month-to-month, so does your monthly revenue. So it's important to be able to accurately identify enrolled patients on a monthly basis.

So now we know the metrics needed to track to measure revenue. But practically speaking, what does it take to capture this data and begin to analyze it for meaningful insight? There are two main pieces to this puzzle:   

  1. The contract details that delineate monthly payments and any risk adjustment that comes into play
  2. Monthly patient enrollment data

This data is extractable via Clarity if you're using Epic/Tapestry for your contract management. But your contracting team likely has its own system to track this information. We recommend working with your contracting team to identify how this data is stored and the best way to extract it on an ongoing basis. Once those data points are identified--whether they’re stored in Clarity or a separate system-- we recommend building out a data structure. This data structure can be an extension of Caboodle or a standalone data mart that is integrated with your data warehouse.

Having your data structure connected to the rest of your data warehouse should be your end goal. This will give you the ability to easily report on PMPM revenue and connect that revenue number to the rest of the metrics in the profitability equation.

download our upgrade 1-pager

The second piece of the revenue puzzle is accurate enrollment reporting. Typically, this information is sent over by the ACO/insurer as a CSV file and contains patient information, enrollment data, and their PCP. The biggest challenge in getting this data into your system revolves around patient matching. One approach is to import the data into your data warehouse utilizing Epic's patient matching algorithm. Depending on your system build, a data-link action can be utilized to set a smart-data element or a patient record field to indicate the patient is part of a specific plan. You could also potentially utilize the coverage/registration module in Epic to store this information.

Building out these two foundational components will ultimately deliver the revenue metrics you want to track and measure.

As previously mentioned, the other main component is cost. Getting external claims data helps you determine the real cost associated with providing patient care. Epic has built out data structures to support importing claims data into the data warehouse. We recommend that as a good first step so you can capture cost for a specific patient population. It will also let you analyze any charges that are generated within your system, allowing you to get a near real-time view of the cost information. Reconciling the charges you generate with the claims those charges link up to is a common challenge we see here at Bluetree. Ultimately, you should strive to avoid any double counting.

                 * via

At the end of the day, my main takeaway seems simple: all you need is revenue and cost data to optimize the value of an at-risk contract. But the effort required for getting all the data from disparate sources, integrating it, and preparing it for reporting can be a massive undertaking. Dealing with multiple payor formats and different contract structures increases the complexity pretty quickly.

download our upgrade 1-pager

In my next few posts, I’ll dig further into how you can analyze cost, utilization, and revenue data, as well as the best ways to identify and address leakage (patients getting care outside of your system). Later on, I’ll walk through some of the more technical detail involved with getting this data into usable shape. Finally, I’ll also address shared-risk and performance-based contracts that rely on quality metrics and the analytics that can be helpful in those circumstances.

Thanks for reading. Feel free to reach out with any thoughts or questions at

like this post?