We need to consider several things when we ask a question like “Who Should we focus on to reduce healthcare costs?” In healthcare a modest percentage of patients define a disproportionate amount of expense, i.e., 20-30% of our unhealthy account for 70-80% of healthcare costs. Therefore, it is logical to focus on these individuals if we want to make an impact on future costs. However, this approach requires one to assume that those same individuals will continue to remain in that high expenditure category. If we look closer, it turns out, this scenario is not entirely accurate.
Recently, Figueroa et al. reviewed spending patterns of persistently high-cost Medicare patients and discovered that of those in the top 10 percent of spending in a given year, only 28 percent of them remained in the high-cost category in the subsequent two years. So, if our models focus on only those 10 percent high-cost patients, we will effectively be missing the vast majority.
There are various reasons for this occurrence. For one, high-cost medical encounters often occur from unique circumstances and transient events such as a hip fracture. Furthermore, an illness that only lasts a finite period will return to a baseline. In addition, certain healthcare events like pneumonia requiring hospitalization may be a high expenditure but are limited in duration. All that to say, we need to think about ways to impact more a little differently.
We need to pivot and address the question “Who will be high-cost, or rising risk, who is high-cost and will remain so, and who is high-cost temporarily and is that cost appropriate or inappropriate?” This paradigm, accordingly, will require different analytics and predictive capabilities and an adjustment of the utilization of our resources.
Numerous factors are influential in our attempts to adjust our models. Healthcare providers have an inherent bias to focus on those that are in a “high-cost” moment in a manner aimed at preventing further high-cost. Undoubtedly, there are opportunities to lower our expenditures during that event, but we also must decide whether there is a requirement for additional interventions if we want to reduce our spending. Simultaneously, shifting our resources to those that do not appear high-cost today will require us to understand better the drivers that influence movement into such a category. It is essential to identify and stratify both physiologic and psychosocial factors. Engagement and activation also play a crucial role. Unfortunately, we are nascent in our ability today, and predictive analytic innovations in this realm are young in their development.
Providers also must examine the drivers of differences in the likelihood of being persistently high-cost and those that are in a growing risk category. Are there cultural differences? To what degree do disparities of treatment play a role?
As we define high-cost, high-need patients, it is imperative to have a greater understanding of the potential variables. This is an exciting opportunity, and a more focused approach will allow us to affect those that need our attention and resources the most. Therefore, the advantages of targeting high-cost patients without a nuanced approach will handicap our ability to have the effect we ultimately seek.