Payer-Provider Partnership, Data Management Promote Population Health

Payer-Provider Partnership

Both Nowak and Davis articulated the importance of open communication between payers and providers.

"It's so easy from the payer perspective to be like, 'Oh, I think you want this. You need this.' Instead of going out and asking what providers are trying to achieve and prioritize,” said Davis. “It has to be a partnership.”

Nowak noted the importance of educating providers about their value-based contracts. At Mercy, the frontline providers have a three-call rule – they attempt to reach a patient three times. If they are unsuccessful, the patient is removed from their patient list. But if this patient shows up in the emergency department three days later, a provider could be penalized in a value-based contract.

“The frontline provider doesn’t know because that patient was removed from her schedule,” said Nowak. “The connection between all these metrics and the actual performance of the contract is so important to understand.”

Nowak also emphasized that showing providers how they were performing relative to their peers can be a motivating factor.

“Providers have to know how they’re performing. They need to know what their ED-per 1,000 utilization rate is, their inpatient-per-1,000 utilization rate, and how that rate affects the reimbursement that we would be getting as an organization,” she said. “That’s the connection that isn’t always made for providers.”

“When providers are able to see where they fall in relation to the metrics that are going to affect the performance of the contract, then their compensation is directly affected. There has to be some level of accountability there,” Nowak argued.

On the payer side, Davis noted that they report these measures regularly to their provider partners.  

“We’re working right now to partner with one of our community health providers to build a provider profiling tool. It’s a packet to help providers go out and have those physician-to-physician conversations,” she explained. “One of the aspects on there is ED visit rate. The next one is the average cost of those ED visits, so this is the financial impact of a provider’s rate being higher than others in his specialty.”

Davis stressed that having this information can be empowering to the physician leaders and grab their attention. It can prompt conversations about referral patterns, patient care, and care quality.

When trying to implement new initiatives for population health management, Davis noted that these provider relationships can pay off.

“Work with your provider partners and have them sit at the table with you. Having a provider champion out in the community as you’re trying to roll out new initiatives is much more successful when you do it with them versus to them,” she stated.

Data Management

Effective data management is a critical component of this partnership as both Nowak and Davis noted provider-level utilization metrics can drive better care.

The most common data source used by both Mercy and Neighborhood Health Plan of Rhode Island was claims data.

Being a payer, this information is readily available to Neighborhood Health Plan of Rhode Island, but Davis pointed out how this information can lag. The reports built from this data should be used to identify patterns rather than individual members to contact.

“We give aggregate, member-level data to our provider partners but always with the nuance to not use this for outreach. It’s more about looking at patterns,” Davis explained. “It’s the wrong type of data for specific member-level outreach. Rather it’s for looking at patterns to say these kinds of people are more likely to visit the ER. Let’s go approach the pattern versus the particular individual.”

Other data sources notably used included the health information exchange and social determinants of health.

With multiple data sources, it can be challenging to manage and integrate all the sources.

“The combining of all of these pieces of data is really important,” said Nowak.

To manage data from multiple sources, Nowak said their organization uses a centralized data warehouse. Additionally, Mercy has a governance structure integrated into its framework, a change management system, and platforms to display the data.

“With a centralized data warehouse, you have a lot of capabilities. If you’re going to bring in data from all various sources, you have to have that capability on your side or be ready to outsource it,” Nowak stressed. “In this data and analytics space, a healthcare organization needs to determine whether this is going to be one of their pieces of expertise or if they contract out to an analytics company? It's important to have that conversation internally and decide what you are capable of bringing in-house."

Davis insisted on the same, noting that they have the resources in house to integrate information on state program eligibility they obtain because of Rhode Island’s state Medicaid accountable care organization.

Integrating this data has allowed Neighborhood Health Plan of Rhode Island to prioritize their work and advised others to do the same.

“From an analytic perspective, get to know your data. Figure out what your top three things are you want to work on,” Davis asserted. “Start small and build those key reports that look at outcomes that can enable the performance of low-hanging fruit.

Davis also noted the importance of ensuring the data tells a story. Reporting aggregate numbers may not help plans or providers understand the underlying causes or motivations for patients’ utilization rates and costs.

“If you look at individual data sources, you may just get pieces of understanding. You’re not telling enough of the story to really make any type of impact,” she argued.

Nowak furthered this point by articulating how real-time data incorporation could help give a better understanding of the whole patient. Mercy uses text messaging to communicate with some of their higher-risk patients.

“It’s easy to look at claims data and see that someone has visited the ED four times in the past six months, but it’s more real-time when you’re able to communicate with that patient in order to predict if this patient is going to the ED in the next five days,” Nowak explained. “That real-time outreach that we are using for our patients is another piece of data that we can add to this space.”

Not only does this real-time data allow Mercy providers to reach out to patients sooner, but it is also reflective of trends in the industry and patient expectations.

“Especially in this generation where everyone is becoming more and more reliant on their mobile devices and virtual care, it is an opportune time to be able to acquire data from your members in a much more real-time fashion and then be able to make a difference on the back end,” Nowak continued.

While claims data is most frequently used by payers and providers alike, integrating multiple data sources can help paint a more holistic view of the patient. Eliminating data siloes and promoting data integration can help providers and payers better understand their patients.

“Any data we can get our hands on is good data,” Nowak noted.

Population Health Management

Collecting data from a variety of sources allows payers and providers to more effectively manage the health of their population and, in turn, improve their value-based care outcomes.

Mercy focuses on its risk-based populations, which include Medicare Advantage, Medicare accountable care organizations, co-workers, and direct-to-employer contracts.

“Within those populations, we employ all of our data strategies to really understand that population in order to be able to identify those high-cost patients and then segment the population to go beyond our high-cost patients to truly manage that population,” Nowak explained.

Identifying high-cost patients begins with looking at claims data. After identifying these members, analysis can examine common trends in these high-cost patients to help providers understand what aspects of care they should leverage.

Reporting metrics such as emergency department utilization that Nowak previously discussed are examples of the initial stage of identification in population health management.

Davis noted a similar strategy at Neighborhood Health, where initial analysis focuses on identifying the high-cost, high-utilization population, with further investigation looking for commonalities among these members.

Because the state of Rhode Island’s Medicaid ACO is relatively young, their work is also focusing on understanding the baseline population health by provider.   

“A lot of the initial framework is looking at things aggregated at a practice and provider level, showing cost, use, and quality,” Davis said. “For many of our provider partners, it was the first time they were seeing all of this aggregate data on their patient population.”

Echoing a similar sentiment to Nowak’s comments on giving providers information to drive change, providers with access to population-health level data on their members are empowered to understand their role in value-based care contracts.

While often the high cost, high utilizing patients are emphasized the most by analytic reports, Nowak insisted that organizations should not focus solely on these members.

"We have to balance focusing on our high-cost, high-spend patients, but then focus on our patients who are still healthy and keeping them healthy. We have to juggle in that space," she stressed.

Value-Based Contracting

The standard outcomes used in most value-based reimbursement models focus on utilization and cost.

“In this value-based care space, it’s all about cost and utilization. Understanding where our patients are utilizing the system is first,” noted Nowak.

“These outcomes are the standard value-based outcomes,” Davis reiterated. “That’s what we’re looking at.”

One best practice Nowak discussed was having conversations with providers about closing gaps in care early in the year.

“A key with data and all of these value-based contracts is being able to connect with your members early in the year,” Nowak reported. “We use analytics to drive that connection. In order to be able to get patients in throughout the course of the year so that we can document appropriately, have appropriate risk scores reflected in the patient chart, and close gaps in care. We start very early in the year. That outreach is prioritized based on the patient-level of acuity.”

Davis also pointed out the importance of understanding the root cause behind these gaps in care. In their member population, they are examining how to address turnover, gender, and language affect access to care.

“For our male patients, when they had male doctors in community health centers, they had better access to care than when they had a different gender of a physician,” Davis said.

Clinicians, though, must have the resources to act on this information to avoid it becoming wasted data.

“There is an advantage to having data constantly generated. But you have to then be able to have the actions to go along with it. It is a great tool to start looking at instead of having a human always have to come up with the questions. You can start identifying some of the hot-spot issues and focus on actions,” Davis explained.

Nowak emphasized this point when discussing how Mercy is using its social determinants of health data. She provided an example of how communication with providers allowed them to understand the challenges to population health management and adjust their strategy.

Mercy deployed a social determinants of health flow sheet that was used to document these measures in the electronic medical record. After provider criticism, they stopped implementation.

“Providers felt like they needed to have a solution for their patient. So, if they didn’t have a roof over their head, what were they going to do? We quickly found that we didn’t have the back end. We didn’t have the resources connected within the electronic health record or within the community in order to be able to help those patients when they asked questions,” Nowak explained.

Mercy was then able to make the appropriate adjustments to begin finding resources for providers to refer patients.

“When you’re screening for social determinants of health, being able to have the link right there for referral is key,” Davis echoed.

Merely reporting the cost and utilization measure to providers does not drive or motivate change in care. Providing clinicians with actionable and holistic insight into the drivers behind these key metrics can help change care delivery and ultimately improve the outcomes for value-based contracts.

However, Nowak and Davis advised organizations to start simple and prioritize objectives; otherwise, things can quickly become overwhelming and unmanageable.

“Don’t make this difficult,” Nowak cautioned. “Make sure that you get the low-hanging fruit first and then dive into the more difficult pieces of analytics. That’s where you’re going to win over your providers, win over your leadership, win over the board, and whoever you’re trying to influence.”

Leveraging manageable tools is also important. The duo noted how the industry’s current growth of artificial intelligence and machine learning could help alleviate some administrative burden but warned organizational infrastructure must be able to maintain the technology.

“Don’t try to go too big,” Davis cautioned. “Don’t jump to machine learning, because it’s way down the road.”

Generating small wins early value-based contacts can promote effective communication between providers and payers and set up future work for success.

“Make sure you get those quick wins early, but you have to do that by understanding and knowing your data and then allowing the data to work for you as you plan your outreach,” Nowak concluded.