What is population health? There are lots of nuanced definitions for population health (or “pop health”) in the literature, but I like to think of pop health in terms of how it’s currently practiced. In my mind, this translates to how data is used to inform care delivery for hundreds or thousands of patients. By consequence, this means pop health is usually practiced by administrative staff in health systems, and not by policymakers or academic professors.
The best way to understand pop health is through an example. Consider the case of a clinic that provided care to 5,000 patients in 2019. The clinic has a standard of care, which is supported by clinical workflows, IT infrastructure, equipment, and of course, medical expertise. The clinic tracks the proportion of diabetics in their patient pool, which stood at 1,000 patients in January 2019. Fast forward to January 2020 – the clinic now has a standing size of 1,250 diabetics. During 2019, some of their existing patients were diagnosed with diabetes, some new diabetic patients began visiting the clinic that year for their first time, and some diabetics passed away. With these data, the clinic needs to determine whether something is wrong with their standard of care, or if this increase in diabetic patients is normal. This is just one example of what pop health managers do.
This example highlights an important concern for health systems, and that is healthcare spending. In the past, having more diabetic patients, who require more care and services, was simply good for business. But that equation is changing quickly with accountable care. Under accountable care, many health systems are assuming some degree of financial responsibility for the health outcomes of their patient populations, and if healthcare spending goes beyond a defined amount, the health system will have to pay money back to one or more insurers. This payment structure is known as a shared risk model. And that means health systems are looking for ways to save money!
It is also important to note that pop health is focused on a broad set of metrics applied to a population. So following the example above, we care about the proportion of diabetics across time, or the average hemoglobin A1C in a diabetic population, but not the course of disease in a single patient. Occasionally, pop health managers might zoom into the health outcomes of one patient, as a case study, but individual health is not the practical focus; improving individual health is the clinician’s job.
Why should you care about population health? Again, because health systems are looking for ways to save money. Often times, saving money on healthcare spending is aligned with the quality of care. Here’s a simple example – if a hospital improves its surgical practices and reduces the chance a patient catches an infection, then that patient is less likely to have a readmission. Since hospital readmissions are expensive, the health system is motivated to improve the quality of care. This is a clear win-win situation for the health system and the patient. But improving quality and lowering healthcare spending are not always aligned – this is a very important point to digest. There are plenty of tactics a health system can use to reduce spending with no benefit to patients or the community. Here is a short list:
Post-episode spending. Many health systems get paid through bundling care into episodes. These episodes often begin with a particular surgery, and end 90 days later. The hospital gets paid for all the care that typically occurs within that 90-day period – this is called the bundle price. If the patient’s medical spending exceeds the bundle price, the hospital must pay that portion back to the insurer. And if that patient’s medical spending falls below the bundle price, the hospital can pocket the difference.
Can you guess what we observe? Some hospitals delay certain types of care for patients until the 90-day mark is passed! If a patient has a doctor visit, physical therapy, or even a readmission after the 90-day mark, the hospital still gets paid the bundle price, without the penalty.
Denial of access. Health systems benefit from managing a large number of relatively healthy patients in a couple ways. One way is through advertising high quality care through official healthcare rankings such as US News and World Report, and many others. But if their patients are relatively healthy to begin with, then the health system is not actually responsible for those favorable outcomes (or not completely responsible).
And here is the kicker – health systems can pursue healthier patient population by simply avoiding certain zip codes. For example, zip codes in poor neighborhoods have a high number of dual-eligible patients (dual-eligible patients qualify for both Medicare and Medicaid). These patients are more likely to have poorly managed chronic disease, and adverse social conditions that include unstable housing, food insecurity, and lack of transportation, to name a few.
When you look closely at some health systems over time, you can observe declines in the number of poor patients served by that system. This is not positive proof of bad tactics by hospital leadership, but it certainly looks like it! Here is a brief article I coauthored that describes this awful trend.
To circle back to the question of why you should learn about population health and how it’s practiced by health systems -> because the decisions made by health systems directly affect your health outcomes. It’s a bit tricky to see this, but through accountable care and shared risk models, a health system takes up a similar role as a health insurer. And we all know what happens when insurers exert too much influence on clinical decision making – patients are more likely to experience limited diagnostic and treatment options, and sometimes outright denial of care.
What can you possibly do about this? To say the least, it’s quite difficult for an individual patient to keep a close eye on how a local health system is impacting care delivery. A first step is to ask your doctor about your own care pathway – what does the next 6 months look like for you, in terms of testing, regular doctor visits, and procedures? Armed with this information, you can then pursue a second opinion from a clinician who is outside your doctor’s network. Essentially, you will do a comparison of care pathways to better understand the standard of care for your health system, and whether your missing out on beneficial diagnostic or treatment options.
Unfortunately, this approach may not be free of charge, and depending on your insurance coverage, it may cost you the full price for additional primary care or specialist visits. Additionally, depending on your current state of health, you may not be in the right state of mind to do all this groundwork!
I actually favor a different, community-level approach for monitoring care pathways, that is inspired by the Medicare rules for governing an accountable care organization (ACO). The ACO is a type of health network / system that is paid under a shared risk model (mentioned above). Medicare ACOs require a representative from the patient community (a Medicare beneficiary) to hold a seat on the ACO’s governing board, or an advisory board. The community representative would not be employed by the health system, and have no conflicts of interest.
Conceivably, this person would be well positioned to address issues such as stinting care, the location of new sites of care, health equity, health disparities, and social determinants of health. In practice, we expect other board members would share similar concerns, however, a community representative on the board would not be influenced by the financial concerns of the health system.
All health systems, not only Medicare ACOs, should have community representatives on their board of directors. These representatives should be particularly engaged in the health systems’ shared risk contracts, and fully aware of gaming practices for delaying or denying care to patients. I admit the path to making this a reality is unclear, but it probably begins with having discussions with patient advocacy organizations, community based organizations, and non-profit fundraisers. And may also include setting up a learning network to share insights / examples on hospital gaming practices.
I hope a few readers out there are inspired to engage in these efforts for the benefit of their communities. Feel free to reach out if you have any questions on this topic.
~ James
