Behavioral analytics point way to the most effective treatments
While there have been a plethora of large and small changes to healthcare over the past few years, nothing has impacted the industry like the transition from fee-for-service to value-based care, and for good reason.
Rather than having a blank check for delivering care as they have in the past, providers are increasingly being asked to produce better outcomes while having spending limits imposed on them. They are also taking on risk surrounding reimbursement that is tied to those outcomes, something they never had to in the past.
It’s no wonder there are concerns that it could all lead to a conundrum where providers may someday be forced to choose between delivering quality care and meeting cost requirements.
It doesn’t have to be an either-or dilemma, however. In fact, a new generation of behavioral analytics is helping healthcare organizations learn how to increase quality while reducing costs by taking advantage of new sources of big data—including socioeconomic data and data gathered through population health initiatives—to deliver insights that more precisely predict both needs and outcomes of patients and populations. These insights are designed to quantify two key indicators of a value-based program’s success: impactability and intervenability.
One of the most important ways behavioral data can be used is to assess the impactability of individuals within a particular population. Understanding the impact that successful interventions will have helps set priorities for which resources to devote to which patients, as well as how much time and money to put into it. Much of this revolves around closing care gaps.
Take a diabetic population, for example. The typical analytics process will examine claims or clinical data to identify patients who have care gaps. Then, the organization will spring into action in an attempt to close those gaps to reduce both cost and risk.
The problem with this approach is that too many of the healthcare organization’s limited resources may be spent where they will deliver little value because they don’t take other factors (such as education, which is often an indicator of health literacy) into account. They also tend to place equal value on closing all types of care gaps, regardless of whether they will have a significant impact on the patient’s outcomes.
Behavioral analytics, coupled with finely tuned predictive algorithms, help organizations set priorities so they can apply their limited personnel and financial resources where they will deliver the best returns.
Understanding impactability starts with identifying the number of care gaps, along with the severity of those gaps. If one patient has six care gaps and another has one, priority will be given to addressing the needs of the patient with six care gaps, especially since those with six usually have more serious needs.
Behavioral factors such as education, income, geographic location, gender, race and others are used to further refine the predictions, based on their known influence for patients with these conditions and traits. Once all this additional data has been analyzed, a single impactability score that looks at the difference between taking the recommended actions and not taking them is created. This score helps healthcare professionals set priorities to decide which care gaps to address first, elevating the quality of care overall.
It also helps them decide how to address them, for example, through conversation with a physician, a call from a nurse, a call from a case or care manager, or via text or email. The goal is to apply the right resource to produce the desired result at the lowest possible cost.
Understanding impactability also helps healthcare organizations understand which interventions have little value, helping them eliminate unnecessary tests or other treatments that, while common, will not have a significant effect on patient outcomes.
The other area where behavioral analytics is invaluable to predicting successful outcomes is intervenability. This is essentially patients’ willingness to participate actively in their own care.
Take medication adherence, for example. The impactability score may show that a particular medication will be effective in helping a particular patient get her high blood pressure under control. But when the behavioral analytics are applied to her intervenability, the organization discovers that people at her income level who live in her neighborhood tend not to take their medication as prescribed, because of the high cost. Instead, they either don’t take it all, or they try to stretch the supply by only taking it every third day rather than every day.
Knowing this, the physician, nurse or care manager can have a conversation with that patient, explaining the importance of taking the medication every day. The healthcare professional can also work with the patient to find alternatives, such as a lower-cost generic or supplementing the prescription with free samples or rebates, to make it more affordable.
In value-based care, understanding intervenability also means not expending limited, valuable resources where patients are simply unwilling to do what needs to be done to improve their outcomes. Quality care will still be provided to those patients, but the type and degree of follow-up will likely be more limited. In the meantime, the provider can work on finding ways to get the patients more engaged in their own care.
The transition to value-based care is one of the greatest challenges facing healthcare organizations today. As more of their reimbursement is tied to outcomes, they must gain a better understanding of how and where to use all of their resources.
By providing insights into impactability and intervenability, behavioral analytics can help providers predict the outcomes of their actions and programs more precisely, helping them place their focus where it will deliver the greatest value. All of which helps improve the quality of care while driving down the cost.