The 6 Critical Components of Population Health
I was recently asked to define the term population health from a market perspective. My quick answer was that there are many definitions, but is this really true? I pulled recent reports on population health from four well-respected analysts representing Chilmark, HIMSS, International Data Corporation (IDC), and KLAS. What I found is that they all had a similar definition, which is something like this:
“To manage the health of a specific population using a network of financially incented providers and community partners.”
The next thing I looked at were the IT capabilities identified by each analyst organization. Basically, they all identified the same six categories:
- Data Aggregation
- Patient Stratification
- Care Coordination
- Patient Engagement
- Performance Reporting
Apparently, I was wrong about the variety of definitions. There is, at least among industry analysts, a common definition and understanding of the six components of population health.
- Data Aggregation – Every analyst report seems to start with this capability, and I agree. Population health is a concept that extends beyond the borders of the integrated delivery system. Even Kaiser Permanente, arguably the largest population health organization in the U.S., has some patient care delivered to its members by non-Kaiser providers. A data warehouse is required to collect member information from the many data sources (claims, clinical, behavioral, mental, administrative, patient reported, etc.) and organizations (integrated delivery networks, specialty hospitals, multi-specialty physicians, small clinics, long-term care and skilled nursing facilities, home health, etc.) that must come together to address the health of a population.
- Patient Stratification – This is sometimes called patient risk stratification or patient identification. Many people I talk to are focused on risk stratification, most likely because, traditionally, this is how health plans identified and stratified their members. What we are finding out is that not only are there multiple risk models (some are open source, some are proprietary to a vendor), some models are better than others for identifying particular events or diseases. Clearly, one model does not fit all needs. We are also seeing that the purpose of this stratification is really to identify the health needs of individuals in the population, such as complex care (multiple comorbid conditions), chronic diseases at multiple levels of acuity, people with a rising risk of disease or other complication that would increase utilization, and keeping the majority of people healthy and well.
- Care Coordination – Also called care management, this is most often focused on the top five percent of people who account for nearly 50 percent of health care costs.
Figure 1: National Institute for Health Care Management, Data Brief, July 2012
The Agency for Healthcare Research and Quality (AHRQ) in their Care Coordination Atlas 2014, goes beyond the Organizational Design Model, which focuses on specific outlier patients and coordination activities, to include the Donabedian Model for quality improvement (structure, process, and outcomes), which affects every patient receiving care (the inliers).
Figure 2: AHRQ Care Coordination Alas 2014, Chapter 3, Care Coordination Measurement Framework
Many population health experts either assume that health providers are already engaged in quality improvement programs, thus this aspect does not need to be addressed, or, like traditional health plans, they are focused on the shorter financial impact. As a Health Catalyst team member, I could easily rant about this short sightedness for many hours. The bottom line is that successful population health requires improvement for both outliers and inliers.
- Patient Engagement – This requires providers and community partners to actively create personal relationships with patients. Health IT alone is not the answer, as proven by the lack of patient portal success in the CMS Meaningful Use program. The proposed Merit-Based Incentive Payment System, appears to move toward outcomes versus functionality, which opens up new avenues to explore, such as old-school telephonic broadcasts, 24/7 hotlines, extended office hours, and mobile device applications supporting multi-user threaded messaging systems that connect care teams, patients, and caregivers. Nielsen data on smartphone usage by people age 65 and above indicates a steady increase of about 10-15 percent per year. This data, combined with increased recognition and encouragement by family or friend caregivers (who are often younger), is a strong argument for promoting mobile apps to support patient engagement.
- Performance Reporting – The old saying that “you can’t manage what you can’t measure” is as true in healthcare as for anything else. It could be argued that there are so many measures and metrics (PQRS, MSSP, CMS Core Measures, TCPI, commercial payer measures, etc.) that providers have a difficult time figuring out on which ones to focus. Humans, this includes most providers, are only capable of focusing on four to five things at a time, according to research from the University of Oregon. Simple dashboards of metrics have not been shown to significantly change outcomes. Instead, analytics are needed to find the four or five measures that will yield the biggest return (outcome) on investment (resource effort). Dashboards are not enough. One of the healthcare systems we work with recently identified the need to not only monitor these measures at the individual provider and patient level, but also the need for the analytics = to figure out why something is not improving, whether it’s the protocol, the people, or the training. You can’t do this analysis with a simple metric-based dashboard.
- Administrative/Business – Managing and administering the multitude of new value-based business models is something totally new for providers and, to some degree, health plans, as well. Maybe I should have started with this one because this seems to be what the market has focused on first. How do I know if I am making or losing money for a given contract, provider, disease, etc.? Per member per month (PMPM) is now a common part of our language. The new business of population health must be enabled by the data aggregation system to ingest and correlate claims from multiple health plans (every health plan has a slightly different way for supplying their claims to providers) with a rich set of clinical and patient reported data. Balancing patient satisfaction against provider quality and cost requires new systems that can be deployed in weeks, rather than months or years.
Figure 3: Effective population health management must be powered by population-sized multitasking.
I am a picture/graphics person, so the old expression of “a picture is worth a thousand words” works for me. When I think about population health, this is the graphic I draw, with data aggregation and analytics (a data warehouse) at the center, surrounded by overlapping capabilities that must all be addressed in order to be successful. Identifying which capabilities to address first and then drafting the roadmap to build out these capabilities will enable successful organizations.
Would you like to use or share these concepts? Download this presentation highlighting the key main points.