Taking Accountability for Population Health: The Role of Data

Ray Herschman, President, xG Health Solutions

For providers to accept accountability for population health and payments based on episodes of care, they need control. This control comes, in large part, from information.

Many providers now have access to rich and timely clinical data through their electronic medical records (EMRs). However, it is not enough. Providers taking responsibility for not only health processes and outcomes, but also the total cost of care for a population, must have visibility to the place of service, relative costs, and overall quality of care for their patients beyond their four walls. This information resides in administrative data held by payers.

Accountable care organizations (ACOs) and clinically integrated networks (CINs) are emerging as contracting entities through which the full continuum of providers — hospitals, physicians, ancillary service providers, and skilled nursing facilities — share accountability for the health of a defined population. These organizations will increasingly be paid via value-based reimbursement models that align economic incentives among payers and providers.

Implicit within value-based payment arrangements must be access to this data so providers have the total picture they need to effectively manage population health. Complete data and related analytic and decision support capabilities are foundational to provider success in value-based care delivery. As providers start down the road of valuebased contracting, they face several choices related to data and analytics. These choices have long-term implications for provider organizations’ budgets and focus.

How practices access payer information
Providers and payers who agree that sharing information is key to measurably improving quality and costs are in the midst of sorting out their roles in this new world. Payers have long held data close, but new accountabilities are driving at least three different ways for providers to access administrative data.

Borrow. Some providers rely on quarterly paper reports or payerspecific Web portals to access information about their patients. While any information is welcome for providers who have made do without it for so long, this path has several challenges.

First, there is significant variation among payers in what reports they provide, how they stratify risk, and how they categorize and group the data. Second, each payer has only a percentage of a provider’s total patient panel, resulting in a lack of credibility and stability in performance measures. Third, the information is often aggregated and latent, making it difficult to use for patient outreach and provider-level improvement.

Finally, because payers are challenged with having to provide reports to their entire provider network, reporting is relatively static and standardized. Far more important than standardized reports is data interpretation which, at the in-depth level necessary to be a high-performing ACO/CIN, has historically not been a core competency of either carriers or providers. From a provider perspective, using multiple sets of metrics and payer portals to manage a patient population is not practical or sustainable over time.

Build. Recognizing these limitations, some providers are investing millions of dollars in technical infrastructure, labor, and consultants to develop their own data intake, transformation, data warehousing, and analytic solutions.
This can be all-consuming even with one payer, not to mention multiple payers, each with their own data formats and idiosyncrasies.

This path also requires corresponding investments in recruiting, hiring, and training clinical and medical economics analysts. Committing to tools, upgrades, and staff can be costly, overwhelming even large practices and drawing time and financial resources away from running the business, delivering high-quality care, and developing critical performance-improvement initiatives.

Partner. An emerging alternative is for providers to work with a neutral, third-party intermediary with experience managing the complexities of aggregating multi-payer data on behalf of providers. This approach leverages technical infrastructure and staff across a broad pool of providers and patients to drive down the costs for all users.

Providers gain a “virtual single-payer view” of the patient population they are accountable for managing. This “utility” model is also beneficial to payers, as they are relieved of the ongoing cost burden of dealing with hundreds of point-to-point data transfers with each provider entity. Instead there is a trusted, neutral third party that acts as a single hub through which their data is transmitted to providers in a secure and confidential manner.

Leveraging payer data
Once information is secured, whose job is it to interpret the information, size and prioritize interventions, and recommend interventions? The answer for most providers is that this is no one’s job. Some organizations find themselves with a host of new tools and reports, few of which provide meaningful insights into how and where to prioritize energy and resources to measurably improve patient care and meet financial targets. Reports and tools are necessary, but insufficient for providers to succeed as population health managers.

There are several options delivery systems have for moving from reports and information to action.

Build. Well-trained, highly skilled clinical and medical economics analysts are essential for sustainable and successful population health management organizations. Provider organizations can hire their own analytics teams, but most have historically not needed to develop and hire associates with these cognitive skill sets. Acquiring and developing these resources is extremely difficult in today’s competitive marketplace. Furthermore, delivery systems will need to plan and budget for scaling their analytic capacity across their entire enterprise to ensure the reliability, consistency, and sophistication necessary to drive sustained performance improvement as valuebased payment evolves over time.

Partner. Alternatively, many provider organizations are now seriously considering the value of contracting with emerging service organizations focused on providing data analytic interpretation as a service. This “service as a service” model is akin to “software as a service” in that both centralized and distributed resources can be made available on demand, at a far lower cost than if those resources were owned or hired by each individual provider organization. This new model, however, goes several steps further to include regular access to highly specialized, dedicated analysts that are accountable for and measured on their success in helping providers improve clinical outcomes and financial performance.

While some software and IT vendors are quickly recognizing the necessary shift toward analytics, not all organizations have analytic expertise tried and tested in a clinical environment. Key elements of an effective analytic approach in either the build or partner model include the following:

  • Dedicated, highly trained analytic staff who have experience supporting clinical care teams with interpretation of results and drill-down analyses based on provider questions and hypotheses;
  • Direct access to experienced clinical, informatics, statistical, and other subject matter expertise to provide advanced analytic insights; and
  • Ongoing innovation and learning, in order to continuously bring new, market-leading insights, learning, and methods to providers.


To improve value for patients, consistently improve the clinical care quality, and drive affordability overall, providers need an unprecedented level of collaboration and partnership from payers. This collaboration should include these key elements:

  • Ready access to timely and complete administrative and claims data
  • Eligibility and benefits data for those patients attributed to them;
  • New reimbursement models to align incentives with value
  • Investments in both care management services within provider organizations and actionable information at the patient, population, provider, and integrated delivery system levels


This type of collaboration and this level of investment were not priorities or easily justifiable under traditional feefor-service reimbursement models. With so many forces aligned around moving from volume- to value-based care, providers have a platform to negotiate the necessary control (through data) in exchange for taking accountability for improving population health and entering into reimbursement models that align incentives intended to achieve this end.

Ray Herschman is the President and Chief Operating Officer of xG Health Solutions, the primary provider of Geisinger’s Health Care Performance Improvement Intellectual Property (IP). xG Health provides experienced professionals to partner in developing and implementing strategies focused on improving quality and reducing the longterm cost of care.