Picking up the threads: Simplifying clinical administration of multiple value-based contracts

The right data, presented in the right format, to the right people at the right time — that’s the holy grail for many medical practice administrators.

But when more providers in medical groups, hospitals and health systems enter into value-based and alternative payment models, the simple calculation of rate multiplied by volume must be updated. When quality performance is the determining factor for an increasing share of reimbursement, it’s important to consider the costs and administrative requirements to get from where you are today to the metric-driven goal.

Corey Shank, MS, associate managing director for managed care, Texas Tech University Health Sciences Center, Lubbock, Texas, is something of an expert in doing just that: taking multiple value-based contracts and threading the disparate data from payers into systems that teams at Texas Tech can actually use.

“That’s the number one factor of being successful in information: to consolidate information,” Shank said at MGMA19 | The Annual Conference. “Each payer gives us information differently, and workflows are really hard to manage in those scenarios,” as it results in different staff members managing various payers’ patient case loads. “To me, that’s kind of a pain, because then you can’t maximize someone’s productivity or subject-matter expertise” across the panel, Shank noted.

Payer portal data pains

Getting to a simplified clinical administration of multiple value-based contracts requires a focus on data management that goes beyond simply checking portals for payer data, building dashboards and having a single look into your analytics, Shank said.

“Everyone’s got a freaking portal, right?” Shank noted. “I’ve got two payers that have the same portal, but they don’t give me the same data from the portal,” underscoring the difficulty of data that is accessible yet not actionable.

“Portals are a pain,” but Shank said that the challenges faced by practice leaders can be mitigated by working on problems to be solved by care teams rather than by payers. “We want to become payer-agnostic and patient-centric” in managing data, he added.

Many payer quality measures are based on HEDIS through NCQA, which are standard and commonly based on community health needs assessments (CHNAs) in a provider’s region. Thus, consolidating disparate payer data into a unified dataset can help reduce administrative burden associated with different formats in which payers provide patient data, including:

• Microsoft Access database files
• PDF files
• HTML tables.

The trick, as Shank sees it, is logging into a portal and drilling down to what will help your clinical team make measurable improvement in quality scores, such as finding the patients whose A1c scores are above the threshold for bonus reimbursement, or diabetic patients who have not had their retinal eye exam.

Threading the data

To get to the point in which your clinical team secures the data that can be acted upon for value-based contracts, the practice first must thread the data sets together to make it easier to operationalize the clinical workflows that can improve outcomes.

Yet in many practices, the answer to that data challenge is some combination of third-party applications that can become expensive depending on practice size. Shank contended that a great deal of that work can be done in Microsoft Excel, which gives the organization more control. “I think we need to start … preparing ourselves to take ownership of our own situation — not relying on a vendor, not relying on a payer, but relying on ourselves to manage this better,” Shank said.

Shank proposes finding a means to normalize the data into a format that can be distributed to your team in preferred formats. The way to link them is using a shared key: a common field between two data systems that you have. “When we went to our payers, we just use their subscriber ID,” Shank said, noting that the subscriber ID is ideal since you may not always have another common identifier, such as a Social Security number.

Once the shared key is linking the data sets, the information then can tie into a scheduling system, for example, to better manage patients as they come in with the knowledge that they have certain conditions and need specific actions taken for proactive management of quality scores. (See Table 1 for an example of the value-based situation identifier).

“Your objective is to know when a patient is on the schedule, if they’re in a value-based situation and if their condition needs to be managed,” Shank said. “Who can better engage the patient [than] when they’re in the clinic?”

Score-improving steps

Knowing the patients whose outcomes are driving value-based contract performance is only an awareness piece for practice administrators. The next step is to build workflows to help identify patients with out-of-control clinical values (such as an A1c above 8) and make meaningful changes, whether your practice is fully managing the full patient population or targeting specific metrics.

Shank noted it’s difficult to manage 100% of your population, hence why many practices are “not on a 100% mandate for [clinical] success.” If it’s not feasible to bring all diabetic patients’ A1c scores down to an in-control range, the data work provides a starting point for identifying the patients whose quality scores can be addressed with more-immediate interventions.

“We on the financial or administration side are going to have to say [to providers], ‘you’re doing your best, but you need some help,’” Shank said. “Until we make enough money to hire many case managers or care coordinators, we’re going to have to make these difficult decisions.”

Shank recommends dividing this work into two areas for out-of-control quality scores:

1. Binary out-of-controls: Factors boiled down to a yes or no, such as whether an eye exam or vaccination was administered, or if a certain test was performed.
2. Nonbinary out-of-controls: Factors that need to be controlled over time and can fluctuate, such as a patient’s blood sugar.

In either case, the administrative team should find the workflow that best supports addressing these factors by nonphysician members of the care team. “I want the last person to have to interact with managing this information to be the physician,” Shank said. “My mindset is how can we do it without the physician so they can practice medicine?”

Building awareness, auditing and reporting

In Shank’s organization, there were three care coordinators for 500 physicians, which meant they were very busy. Understanding that email alerts may not be the best way to regularly draw attention to the binary and nonbinary factors, the organization printed big, blue sheets that the care coordinators brought to the clinics and nurses. Despite being a low-tech option, “that was one of the most successful things that we did in managing these binaries, because it was so obvious” and hard to miss, Shank noted.

While the care coordinators preferred to call patients and manage patient care through their patient navigator training, Shank pointed out that prioritizing patients daily in this way allowed them to do those simple things that provided the foundation for addressing the nonbinary factors once tests or annual wellness visits were completed.

From there, the focus shifted to auditing for whether binary factors were addressed and flagging the providers who were struggling with completing them. Shank admits that although physicians and nurses were not happy about that, it offered insight to the entire team as to how their success would be measured (see Table 2).

Those efforts also had to be timely. In the past, the organization only knew when it was doing well or poorly on the day it assembled the annual report of metrics. Shank said that practices, once the data threading work is done, should shift into the mindset of knowing how well they are doing today.

“If you don’t know that from day to day, then you may or may not be successful, because I will promise you this: I’ve been in ACOs … where they just count on this physician dashboard to be this miracle thing,” Shank said. “And that is not going to be successful.”

Tracking patients in value-based measures allowed Shank to assign an actual value to each patient’s out-of-control quality scores, based on the clinic’s specific contract goal and the valuation of the quality measure reimbursement.

These dollar values also help establish the value proposition of setting staffing levels of care coordinators. While realizing that 15 more patients would need to meet quality scores for earning a quality reimbursement isn’t especially compelling, Shank said, emphasizing that each new patient brought into control helps unlock an incentive built into the contract.

“At the end of the day … the care coordinators became one of the most valuable sets of people, because they were the ones reporting this every day,” Shank said. “When we did report success, they were the ones responsible for it.”

To get even more value out of the care coordinators, Shank said his department considered having them become medical assistants so they could do more clinical work. “You want to maximize the productivity of a person to their credential,” Shank added.

Customizing reports to roles

To wit, a performance dashboard should be “tailored to the person that you’re asking to act on” it, Shank said. When sharing dashboard data, there should not be too much emphasis on what’s not working, but rather the areas for improvement.

“If you’re responsible for presenting information to the practice, do your best to not make it dire, because you’re not going to engage someone to take action,” Shank said. “If my care coordinators are working on something, I want to drill it down to them and say, ‘these are the five patients that you need to impact because these are the closest to being in control.’ ... Your dashboard is these five patients, not the 150,000 that you can’t manage.”

Starting with those operational clinical team members is the first step to cascading the reporting, Shank added. Weekly reviews of patient data in value-based situations should identify where some patients might fall through the cracks, and those results should filter into monthly management reports to ensure team members who are at risk for incentives can see successes or lack thereof. Those monthly results then should be reviewed quarterly at a practice or department leadership level.

“Every person, whether it’s daily, weekly, monthly or quarterly, they should know their role with this value-based [performance],” Shank said. This organization-wide awareness helps underscore the importance of the data work and ensuring both better care outcomes and bottom-line results.

“At some point, the risk of not doing something is greater than the risk of doing something and not getting reimbursed with it on your claim,” Shank noted.

Action steps

Corey Shank recommends three action steps for dealing with payers’ information:

1. Normalize the data: Normalize data files from payers depending on the columns, rows, tabs and other methods in the file(s) you receive.
2. Use shared keys: Find a field in the payers’ data sets that link to your data system, also known as a “shared key.” Good examples of shared keys are billing IDs used, such as a subscriber ID.
3. Tie your data systems together: Once data is normalized and the shared keys are defined, your systems’ data can be linked to scheduling systems, health records and other systems used within your practice.

Reporting and communication tips

Shank recommends these tips for building out reporting of performance data:

• Make it consumable: Provide a dashboard or report meaning for the user and keep it as simple as possible. Instead of emphasizing measures in which providers are lagging, present information on the areas that can be improved.
• Make it timely: Don’t wait until a monthly or quarterly meeting to share data if there are opportunities to improve. Instead, build a cascading delivery of the data. Begin with a weekly operations update that leads into a monthly management review and ultimately a quarterly report for leadership.
• Make it targeted: Tailor the data for each audience. An operations audience needs to know their missed opportunities on a frequent basis along with the status of case-managed patients, whereas managers need date-specific performance for measuring team goals, and practice leadership need to assign critical values to performance (e.g., bonuses).

Learn more

For a video demonstrating how to thread payer data in Microsoft Excel, visit bit.ly/VBC-Threads.