Health plan principles and management strategies are the key to reducing avoidable utilization and improving outcomes for defined populations.
Value-based payment models are forcing hospitals and health systems to assume increasing levels of financial risk, prompting many to adopt population health management as a means for controlling costs and improving patient outcomes. Many of these organizations are just beginning to explore the world of population health management, whereas others are increasing their investments in existing population health programs.
In both groups, hospital and health system leaders confront the same challenge in knowing how to plan and launch a population health initiative or grow it strategically. Most leaders are contending with unanswered questions about where to begin, which tactics are most effective, what resources are needed, and how to measure success.
The best way to answer these questions is to learn from the organizations that have the most experience in managing risk at the population level: health insurance plans. Health plan leaders are skilled at using data to understand their biggest population health opportunities. They also have developed many effective strategies for improving the quality of population-level outcomes and cost performance.
Health system leaders can move successfully into population health by adopting a health plan mindset. The first step is to master some basic principles and practices.
Lay the Groundwork
Health plan management teams have a distinctive way of conceptualizing the healthcare market and framing their challenges. Hospital and health system leaders can adopt this framework by making three changes in their outlook and practice.
Think like a health plan. The first step is to reframe the organization’s relationship with the people it serves. For hospitals, these people are traditionally referred to as patients. By contrast, health plans think of their customers as members.
The difference is subtle but important. By definition, a patient is someone who comes to a provider and receives treatment. Providers have direct control over what happens to patients. In contrast, a member is someone who is sometimes within the provider’s direct control (as a recipient of care) and sometimes outside the provider’s control (the rest of the time).
The cornerstone of effective population health management is the realization that the provider organization’s responsibility extends beyond the four walls of the hospital. This shift in focus is critical to moving beyond mere quality improvement and achieving better health for the populations the organization serves. Health systems will continue to be challenged to find innovative ways to move health care into other settings and promote new concepts such as “hospital at home,” and before this can become widely adopted, a robust population health strategy must be deployed.
Measure like a health plan. The next step is to change the metrics the organization tracks. The key is to choose measures that track member health and total cost of care, not just patient and procedure volumes in the health system’s own institutions. Utilization metrics are key to better controlling and managing overall cost and quality and are used by health plans to measure the performance of a specific population. For example, hospital leaders traditionally follow metrics like admissions, patient days, and total discharges—measures that gauge hospital performance under a fee-for-service business model, but that provide little insight into the overall health of the hospital’s patient members.
To measure like a health plan, hospitals and health systems undertaking a population health strategy should begin by tracking total hospital and skilled nursing facility (SNF) admissions per 1,000 population. This measure gauges both overall health and total costs while simultaneously improving quality outcomes, and it can help the organization begin to identify population health issues that require attention.
The provider organization also should begin tracking risk scores and enrollment data by service line monthly. By detecting a rise in risk scores or a large influx of Medicare Advantage patients, for example, the organization can know when to begin putting appropriate clinical resources in place to ensure the organization can continue to optimize health for high-acuity patients.
These and other member-focused measures should be tracked on performance scorecards tailored for different levels within your organization. The board scorecard should provide a high-level view of the population under management. Operational leaders should have access to a more detailed set of metrics at a frequency that lets them make timely decisions.
Create incentives like a health plan. Health plans are skilled at effecting change by implementing and fine-tuning financial incentives. Hospital system leaders can leverage the same principles to manage the health of populations. For example, payer contracts with physicians typically include incentives to keep members within the network if medically appropriate and prescribe generic drugs where possible.
Hospital systems can build these same incentives into compensation plans for both employed and contracted or “voluntary” physicians. Successful health systems, independent practice associations (IPAs), and accountable care organizations (ACOs) provide their primary care physicians with:
- Education on population health management and the flow of funds under managed care contracts
- Timely, actionable data on their panel’s total cost of care, out-of-network utilization, and gap closure
- Meaningful financial incentives for successfully reducing the total cost of care and improving quality for member patients
Combined, such activities can elicit primary care physicians’ interest and engagement in becoming accountable for meeting the health system’s organizational and community health goals.
Specialists also must be engaged, in collaboration with primary care physicians, and given incentives to do the following:
- Keep care in network
- Reduce total cost of care by practicing evidence-based medicine and co-managing patient members’ care with primary care physicians
- Enhance timely access to care, including consultation with primary care physicians and members to avoid emergency department (ED) and out-of-network utilization
For many hospitals and health systems, the main obstacle to getting started in managing population health is “analysis paralysis.” Leaders begin by capturing huge amounts of clinical and financial data, but they do not know how to sort the information for actionable insights. This problem is only growing as data collection systems get better and databases get bigger. For health systems just beginning to integrate population health management principles, starting with self-insured claims data for health system employees is a good Petri dish. Another important starting point is to approach health plans on their willingness to partner and share claims data.
In contrast, health plans take a methodical approach to population health management. Plan leaders do not try to solve every problem at once. Instead, they analyze their population, identify their opportunities, decide which issues to work on first, and then target them with limited interventions and pilot programs. Before providers implement large-scale programs, they can benefit from first demonstrating proof of concept through pilots, with iterations and improvements to ensure the programs are scalable and replicable.
To develop a similar analytical approach, provider organizations should begin by obtaining monthly claims data for their attributed members from their contracted health plans. Organizations with ACOs in the Medicare Shared Savings Program (MSSP) or have an MSSP or Next Generation ACO program can leverage the claims data they receive from the Centers for Medicare & Medicaid Services.
For hospitals and health systems that have focused mainly on encounter and clinical data, it will take some work to become adept at taking in and normalizing claims data to make it usable. Some organizations already have dedicated staff who can assist with analyzing claims data for opportunities to manage utilization better. Organizations that lack this advantage should begin developing these resources or consider outsourcing data analysis functions. Either way, to be successful in managing the health of a population, a provider organization must start hardwiring claims data population health analytics into its organization’s DNA.
Analyze referrals. As a starting point for population health analysis, provider organizations should use claims data to identify care delivered outside the health system, and they should benchmark primary and specialty care physicians for total cost of care. Primary and specialty care physicians with a strong pattern of out-of-network utilization and/or high risk-adjusted total cost of care should be flagged, and an in-depth analysis should be performed to identify specific causes.
In some instances, an organization will uncover clear patterns of unnecessary referrals—for example, a family practice physician who refers almost all headaches to a neurologist. In other instances, the issue may be that the referrer is unaware of in-network specialist resource. In many cases, patients seek out-of-network care without a referral, and the primary care physicians may be unaware it is occurring. In each of these scenarios, targeted physician education can help reduce unnecessary specialty and out-of-network utilization.
This exercise could also identify opportunities to enhance the services available within your system. If a health system sees a strong flow of outside referrals for gastrointestinal services, for example, the organization might better manage those services by recruiting a gastroenterologist. Alternatively, arranging for e-consultations with outside specialists could help lower overall costs.
Analyze high-cost diagnoses. Another approach is to aggregate clinical and claims data to identify the most expensive DRGs in the provider organization’s population. For example, a health system could find that patients with rheumatoid arthritis are among its most expensive members. Analyzing this population further, the hospital may find that IV immunoglobulin drugs are the main cost driver in this group. These prescription costs cannot be easily changed, but the same may not be true of the site of service. Typically, hospital care is not medically necessary for a significant percentage of these patients. Many members may be able to receive their infusions at an outpatient infusion clinic or a physician office, or via a home health service. Moving these services off campus will not only reduce costs, but also improve patient satisfaction.
Analyze length of stay (LOS) across various settings. LOS analysis can uncover opportunities to provide more appropriate care to specific populations. For instance, a hospital might find that its average LOS for hospice care is five days, which could indicate that patients at end of life are staying in the acute care facility too long. Discharging these patients to the hospice sooner would enable them to receive more appropriate palliative care and create a better patient and family experience. At the same time, it would help control total system costs.
Scale the Program
Once a population health initiative or program has proven itself, the next step is to apply learnings to larger target groups and new healthcare needs. Again, health plans have developed a disciplined approach to scaling population health initiatives. The starting point is careful attention to operating metrics. Regular and frequent iterations of initiatives and programs should be conducted to ensure goals regarding quality outcomes, patient experience, and total cost of care are met.
Perform a cost-benefit analysis. Every population health management initiative or pilot program should include a plan for measuring success. The best approach is to track the cost of running the initiative and measure the benefits. This analysis should consider the staff, resources, and interventions that were applied to the target problem and the impact of each of these elements on clinical outcomes and costs.
Through the cost-benefit analysis, the organization can measure the ROI for each intervention. Interventions that demonstrate a positive ROI may be replicated on a larger scale. If an initiative breaks even or even yields a slightly negative return, the organization should consider making adjustments and retesting. For example, it can be challenging for a hospital to view reductions in emergency department volume as a benefit to the bottom line, but in risk-based arrangements, the savings in the total cost of care can benefit the hospital as that volume is shifted to more appropriate and lower-cost care settings. A hospital should conduct a “waterfall” analysis of each of the costs associated with the utilization shift and how the shift can improve overall quality performance.
Define the model. The cost-benefit analysis will yield a model that details both staffing and resource inputs and expected clinical outcomes and cost benefits. For example, the analysis may determine that a successful population health initiative aimed at diabetes requires a basic staffing model that includes 1.5 home health nurses per 1,000 members.
Lessons learned in one area also can be applied to other disease populations. To return to the example of reducing hospital-based infusion services, leaders could adapt the model they developed for high-cost arthritis drugs to other high-cost populations, such as people with repeated hepatitis C diagnoses.
Review regularly. The best health plans review their population data on a regular basis. It is important to keep in mind that populations are not stagnant. A member population is like an amoeba—it moves and changes constantly. Significant shifts in demographics, socioeconomics, diagnoses, and health status can occur within a short timeframe.
Moreover, any merger, acquisition, or new service line can quickly alter a provider organization’s member population and network adequacy in dramatic ways. After any such undertaking, it is imperative that the organization assess its readiness and gaps to manage attributed risk populations in the new configuration. Whether or not the network is undergoing changes, regular reviews of attributed population characteristics such as status, geographic utilization, and home and work zip codes should be conducted to verify the effectiveness of population health initiatives and make course corrections where needed.
Develop Advanced Capabilities
Health plans use a number of strategies to build success into their population health initiatives. The following strategies are advanced, but hospitals and health systems should focus on developing them early.
Create a unified care management strategy. Care management is an essential part of a proactive approach to managing population health. However, very few hospital systems invest in a care management infrastructure, and those that do often fail to achieve strong results. The main problem is poor coordination. In many instances, patients receive care management from several different coordinators, department navigators, social workers, and clinical liaisons. These efforts are not well synchronized, which creates confusion for members and leads to waste.
In contrast, the best health plans approach care management strategically by investing in staff and resources around clearly defined goals. The entire infrastructure is coordinated to ensure members receive timely services to optimize outcomes and minimize costs. Although outcomes can be debated, the way health plans approach care management can provide key learnings for health systems.
To create a unified care management strategy, a provider organization should begin by inventorying all points in the care continuum where individuals interact with any kind of care manager, including both hospital-based navigators and social workers and care management staff employed by the organization’s contracted payers. Once all these points of entry have been identified, organizational leaders will begin to see where they need to streamline, orchestrate, and standardize care management services.
This entire exercise should be based on a systemwide vision of the care management experience. This vision should encompass the organization’s determinations regarding key care management considerations, such as:
- Which patients should access care management and when
- How services should be coordinated
- Whether primary care physicians will serve as gatekeepers to the entire process
Having such a clearly defined vision can help a provider establish consistent care management processes that are standardized across the organization.
Care management strategy often involves a “build versus buy” decision. For instance, a hospital may have piloted several population health initiatives using an outsourced care management vendor. As the programs develop, however, it may make sense to reduce costs by bringing care management functions in house. Conversely, analysis may show that the key to scaling an organization’s programs is to outsource care management to an efficient vendor.
When developing a care management strategy, health system leaders should leverage their strengths. Although health plans have good infrastructure and principles, health systems are in a better position to deliver effective care management because they can interact with the patient member more often. Health systems also have closer connections with direct healthcare providers. The bottom line: Health systems should learn from health plans on using data and infrastructure to improve care management, but they should take advantage of their unique relationship with the member population to ensure the best care management results.
Build coalitions. When it comes to improving health, the provider bias is toward direct action. Physicians, nurses, and other caregivers improve patient health by directly furnishing diagnostics and interventions. In contrast, health plans affect member health by working with and through partners. To successfully manage population health, hospital leaders must become skilled at building healthcare coalitions.
This effort may involve looking outside the health system’s employed or affiliated physician groups to other provider organizations that support the organization’s population health vision. For instance, a health system with a population health goal of reducing inappropriate ED utilization could partner with one or more federally qualified health centers to provide preventive care to high-utilizing populations.
Coalition-building also involves identifying community-based organizations (CBOs) that can play an important role in population health. The aim here is to address the social determinants of health, such as employment status, education, food security, and language barriers—factors that can be far more predictive of system costs than medical status alone.
Provider organizations should begin by analyzing the target population and understanding their needs, which includes understanding the social determinants that affect the population’s ability to access healthcare services and benefit from them. For example, a health system that is creating a maternal/child health program in an urban area might need to address issues of transportation and food security. In such a case, the health system could partner with a medical transport service and a social service agency that aids with Women, Infants, and Children (WIC) Program enrollment.
Alternatively, an organization’s health initiative might focus on senior citizens in rural areas. For this population, it might make sense to partner with a Meals on Wheels program. This population might also benefit from telehealth services, so the organization might want to consider partnering with a commercial telehealth vendor to fill the gap.
Grant-funded CBOs benefit from these partnerships because they drive the referrals that lead to additional revenue. In some cases, it may make sense to pay CBO partners a small per-member-per-month (PMPM) fee, with the opportunity to earn bonus incentives for meeting population health goals. For instance, a CBO that provides behavioral health services could earn incentives for helping to reduce ED visits for patients with high-risk behavioral health issues.
Take on administrative and proactive utilization review functions. Increasingly, providers are being given the opportunity to earn a larger percentage of health plan premiums by assuming traditional health plan functions. This trend supports hospital-driven population health in two ways. First, it represents an important new mechanism for funding population health initiatives by offsetting losses in traditional utilization. Second, it allows hospital systems to gain new experience in managing populations effectively. Investments in the technology and data analytic capabilities needed to adequately manage population health should be carefully planned. In many cases, advanced health systems will hire their own actuaries to help with predictive analytics and modeling—a critical success factor for health plan population health management.
For example, New York State launched the VBP Innovator Program in 2017, which provides financial incentives to ACOs, IPAs, and other provider organizations that take on traditional health plan management and administrative functions such as utilization review, claims administration, disease management, and member customer service. Eligible organizations can earn 90 to 95 percent of the global managed care premium.
One health system in New York City is now creating a management services organization (MSO) to take on several functions previously administered by its contracted health plans. For this health system, delegated tasks include credentialing and claims payment. System leaders expect these efforts will help them develop both their population health management capabilities and their funding.
Other health systems are developing proactive utilization review programs utilizing admission, discharge, and transfer (ADT) feeds from EDs to have physicians perform concurrent review and avoid admissions by redirecting the member to a more appropriate care setting.
Similar programs are springing up in other parts of the United States. For hospitals and health systems, these programs provide a new opportunity to fund population health management activities by providing patients with optimal care and keeping total costs down. For these organizations, the key to success is knowing exactly what parts of the healthcare business for which they want to remain accountable would be best managed in house versus contracted out, and what part of the business of health care they would like the health plan to continue providing.
Some functions of population health management such as population health analytics are essential for providers and simply should not be left to health plans to manage. Other important functions that providers should seek to have delegated to them by health plans, and for which providers should build local talent and infrastructure, include care management, utilization review, and if possible utilization management. Claims processing is a commodity that a provider can control virtually via the health plan as long as the provider has utilization management and control over claims adjudication.
For many hospital leaders, an important question remains: Is our organization too small to undertake population health management?
Although size and strength are useful, neither is decisive nor even essential. A statement often attributed to Charles Darwin puts the case well: “It is not the strongest of the species that survives, nor the most intelligent that survives. It is the one that is most adaptable to change.”
Organizations do not need to be large or well-resourced to thrive in population health. The key to success for any health system is to begin adapting to the new environment of value-base care. Systems that adopt the health plan mindset and start building the requisite capabilities can effectively manage the health of populations, improve patient outcomes, and control overall costs.