Population Health Management (PHM) and Analytics

December 21, 2013 Austin Wiggins

Every industry is defined by some type of equation of balance. As the factors influencing that industry change, that equation moves out of balance. The winners in the industry are able to continually change to stay as close to that appropriate balance as possible.

In the health care industry, that balance is being driven by a concept known as population health management (PHM). And, PHM is moving from a dream to a possibility through the use of (primarily data) analytics.

Population management process chart InfographicPHM is a continuum. To understand it, we need to understand how the economics of health care have changed and continue to change. There wasn’t always health insurance in the US. In the 19th century, if a sick or injured person needed to see a doctor, that person paid (or bartered) a charge for medical services. Eventually, even though it wasn’t the way that it happened, individuals and then employers made a move toward pooling of risk through insurance plans. An individual or company would buy a policy from an insurer who paid doctors and hospitals for their services. Generally speaking, the amounts paid were whatever was charged.

Through the development of networks, insurers began to negotiate reduced payments to health care providers. The model was fairly simple. An insurer would contract with a number of providers to be in-network. The insurer provided incentives for its insureds to seek care from in-network providers. The insurer negotiated discounts with those in-network providers. The financial model worked well for the insurers. In the short term, it served to curb inflation in the health industry. But, with competition, that reduction in inflation was only temporary and other techniques were needed. Negotiating prices would only get us so far.

Somewhere along the line, we began to realize that the total dollars paid out for health care were a combination of the price for services and the frequency of services. We also worked out that with higher quality services, the necessary frequency of services would decline.

Quality is generally difficult to measure. I think it is reasonable to say that every medical provider would tell you that they provide high quality care. Each could find patients of theirs who would tell you that as well.

The landscape has changed. Physicians are no longer compensated as much (inflation-adjusted) for each patient that they see. But, in addition to per patient fees, they are also paid for meeting a number of targets. Like much of the rest of the professional working world, they have been moved to a compensation structure consisting roughly of base pay (regular fees) and incentive compensation for meeting pre-defined targets. Here, the insurance companies and the government are now defining these targets and motivating physicians to adopt their practices to meet or exceed those targets. Those who do that well will succeed financially.

This is the PHM model. In order to implement it, payers including insurers and the government must have masses of data and be able to analyze that data. But, these are large organizations that have or can build those capabilities. On the other side, there are the providers, largely physicians, physicians practice groups, and hospitals. Between them, none have data management as a core competency. Generally, none have considered it to be of value.

Welcome to the new world of health care delivery. Remember what I said earlier. One component of physician income is now incentive compensation. Physicians have to meet targets to get their incentive payouts. There are two ways to do this. First, they can just see patients and provide the best care that they know how to and hope that the numbers work out. Second, they can use data analytics to understand the incentives and align their practices to the goals that they now must meet.

That is all well and good, but it would historically be an unusual provider or provider organization with analytics capabilities. In fact, before this decade, I think that most physicians had never heard of data analytics. Now, among those that have, most are not prepared to use them as a part of PHM. What do they need to do?

I think we can divide this provider need into two categories:

  • What do they need to do (functional capabilities)?
  • What organizational adjustments do they need to make (infrastructure changes)?

Functional Capabilities

  • Segment the patient population. Provider organizations will need to break their total patient populations into well-chosen subpopulations. Dividing the population into groups or suites based on chronic conditions will allow providers to target care opportunities.
  • Risk assessment. Sophisticated models are being built that use a combination of health care claims costs for a patient, diagnostic codes (ICD-9, ICD-10), and national drug codes (NDC) to predict clinical risk. Physicians are trained to diagnose a medical problem and to treat it (hopefully cure it). Understanding which existing factors produce the highest levels of risk will allow them to do that better.
  • Support the clinical decisions. In a perfect world, physicians would know every last detail about every patient and know exactly why and when each patient needs a care intervention. Historically, that information needed to be stored in the collective brains of the health practice. That’s just not possible. Data analytics will allow for performance management and systems of automated alerts.
  • Track results. As the equation of balance shifts from volume to value, provider organizations will need to track that value as value will be represented by performance. Performance will be judged by comparing provider results to those of risk-adjusted national and regional benchmarks. Provider organizations will need to understand their strengths and areas for improvement and develop plans to improve patient value.
  • Integrate data. No single physician can know every detail about each patient’s medical history. For his own record of care, he can look to see that piece of a patient’s history, but that may not be sufficient. The ability to share data with all relevant stakeholders and to coordinate care will improve the value proposition for the providers and the patients alike.
  • Engage with patients. The historical method of treating patients has been to tell them what to do. It’s been an entirely one-way street. By sharing the decision-making process with patients, physicians will bring their patients into the solution. This engagement is expected to raise health awareness and consciousness among the patient populations.

Infrastructure Changes

  • Leadership buy-in. Leaders must be at the head of transformation not lagging it. They must set goals, communicate them regularly, and communicate progress toward those goals just as regularly. Leaders must move from being physician administrators to business leaders whose business happens to be health care delivery and management.
  • Common goals. Some would call this a shared vision. Whatever we call it, health care business leaders and providers need to be united in it. They must see a common definition of patient value and together transform the delivery model to achieve that necessary value.
  • Structure to support those goals. In the business world, we typically refer to this as good governance. Once agreeing on those common goals, in order to guarantee acceptance of accountability, provider organizations must ensure that a structure exists that supports success rather than setting up for failure. Such a governance structure will need to include, at the very least, coordination of service, data analytics, and analytics reporting and understanding to support it.
  • Live the value model. When an organization changes its business model, to succeed, it must live that new model. For most, this is a big change. Managing this change will require a culture that supports it. Strong leadership, good governance, and state-of-the-art health care information technology combined with an effective change management strategy will build the trust necessary to move to the new value-based model.

Population health management is an entirely new concept for the health care community. It’s new because for the first time, it’s not a one-time change in the way things are done, but it’s a continual process. Organizations that enhance their functional capabilities and invest in enhancement of their infrastructures to support those functional capabilities will have the highest likelihood of succeeding in a value-based model. Ultimately, it’s going to be about a complete integration of care and data.

Previous Article
Requirements and Recommendations for Monitoring Access to Care
Requirements and Recommendations for Monitoring Access to Care

  Healthcare reform means millions of people will gain access to health insurance, prompting some concern a...

No More Articles


Company Name
First Name
Last Name
Thank you!
Error - something went wrong!