How behavioral analytics can fine-tune patient engagement
To succeed in value-based care, providers need to get consumers Involved in their care. Analytics can help them determine which approaches are most effective at inducing participation.
There is probably no better illustration of the difficulties of predicting human behavior than the current Presidential election cycle.
Polling organizations spend millions of dollars building models and surveying voters to make their predictions of the outcomes of primaries and caucuses, only to end up surprised (and sometimes embarrassed) a couple of hours after the polls close. Candidates also hire experts and spend millions more dollars building messages they hope will resonate with voters, then spend their post-election remarks explaining why they missed the mark.
It all comes down to the old adage “actions speak louder than words.” No matter what people say they want, or well-meaning experts believe they want, there is no better indicator of what will work in any given situation than the actual behaviors those people display when faced with a decision.
This is an important lesson for hospitals and healthcare organizations to learn as they work on the patient engagement strategies that are a critical component of the transition to value-based care. No matter how brilliant or logical a patient engagement strategy may seem in the abstract, if it doesn’t translate into the desired behaviors from patients, it’s not going to have the desired effect of improving patient (and population) health, reducing readmissions, increasing patient satisfaction and lowering the cost of care.
Encouraging patients to follow a specified plan of care is nothing new. What’s different is that it has become crucial in affecting provider compensation.
Under the old fee-for-service model, if patients failed to follow care instructions after release (such as making a follow-up appointment with their primary care physician or filling and taking prescribed medications) it would negatively impact their health, potentially leading to readmissions. From a provider perspective, readmissions meant more revenue.
Under the value-based care model, however, readmissions or deteriorations in health are costly to the provider as well as the patient. With incentives now becoming aligned with performance, it is in everyone’s best interests for patients to be actively engaged in their own care. The remaining question, then, is how best to make it happen.
When healthcare organizations build patient engagement strategies, they typically start with a few assumptions. For example, if they are developing a program to reduce pregnancy or sexually transmitted diseases in teens, the logical assumption would be to build an app for smartphones.
According to Pew Research Center, 73 percent of teens age 13 to 17 have access to smartphones, and 91 percent use the Internet on a mobile device, with 92 percent of them using it at least once a day. That’s some pretty strong evidence on the surface.
Yet the reality may turn out to be entirely different. It could be that the teens in the target audience would prefer to receive this information in another way, such as in-person consultations, an email newsletter or a hard copy pamphlet they can pick up and read later.
Before making significant investment in building the app and rolling it out to the entire population, a better plan would be to create a pilot and test it against other, more traditional methods of disseminating information. As part of the test the organization can try including different incentives for taking actions. For example, if participants review a set of information and then take and pass an online test, they can earn a reward. The rate of redemption will show how important rewards are to this population.
The organization can refine this concept further by offering different types of rewards (such as direct cash payment, gift cards or points toward prizes) to the test groups to determine which are more effective at driving the desired behavior. The more understanding there is around the target population and what motivates it, the more effective the organization can be in designing programs that lead to the desired health outcomes—and lead to downstream improvements in financial performance.
Behavioral analytics can also be used to improve patient engagement (and guide resources) in an existing program, such as diabetes management. The organization can gather information regarding the types of programs it is offering, such as care management, automated care gap identification and notifications, group counseling sessions and self-study information on diet and exercise. It can compare the usage rates for each of its program elements, along with the effect it has on individual and population health, and subsequently on in-patient stays and emergency department visits. Armed with this behavioral data, the healthcare organization can determine which interventions are the most effective—and therefore worth greater investment—as well as those that are no longer worthy of continued investment, freeing up budget to develop new options or enhance existing ones.
By coupling this information with analytics showing the impactability (likelihood that a particular intervention will have a significant impact on patient or population outcomes) and intervenability (likelihood that the patient or population will become actively engaged in their own care), the healthcare organization can further refine its programs and maximize its chances of delivering the right care to the right patient at the right time.
Clearly, measuring behaviors when determining how to engage patients more effectively for value-based care can have a significant impact on success, and avoid costly missteps. Why aren’t more organizations doing it?
The reason is that, up until now, they haven’t had the analytics to do it. The first generation of analytics were focused primarily on clinical outcomes, showing what happened in the past.
Next-generation analytics provide the opportunity not only to dig deeper to determine which programs, offerings and strategies drove those outcomes, but also to start asking “what if we did things differently?” These predictive analytics are helping hospitals and healthcare organizations take advantage of the data they have about patient behaviors to improve their engagement strategies. That can help them create an even greater impact on both patient and population health.
Polling organizations spend millions of dollars building models and surveying voters to make their predictions of the outcomes of primaries and caucuses, only to end up surprised (and sometimes embarrassed) a couple of hours after the polls close. Candidates also hire experts and spend millions more dollars building messages they hope will resonate with voters, then spend their post-election remarks explaining why they missed the mark.
It all comes down to the old adage “actions speak louder than words.” No matter what people say they want, or well-meaning experts believe they want, there is no better indicator of what will work in any given situation than the actual behaviors those people display when faced with a decision.
This is an important lesson for hospitals and healthcare organizations to learn as they work on the patient engagement strategies that are a critical component of the transition to value-based care. No matter how brilliant or logical a patient engagement strategy may seem in the abstract, if it doesn’t translate into the desired behaviors from patients, it’s not going to have the desired effect of improving patient (and population) health, reducing readmissions, increasing patient satisfaction and lowering the cost of care.Encouraging patients to follow a specified plan of care is nothing new. What’s different is that it has become crucial in affecting provider compensation.
Under the old fee-for-service model, if patients failed to follow care instructions after release (such as making a follow-up appointment with their primary care physician or filling and taking prescribed medications) it would negatively impact their health, potentially leading to readmissions. From a provider perspective, readmissions meant more revenue.
Under the value-based care model, however, readmissions or deteriorations in health are costly to the provider as well as the patient. With incentives now becoming aligned with performance, it is in everyone’s best interests for patients to be actively engaged in their own care. The remaining question, then, is how best to make it happen.
When healthcare organizations build patient engagement strategies, they typically start with a few assumptions. For example, if they are developing a program to reduce pregnancy or sexually transmitted diseases in teens, the logical assumption would be to build an app for smartphones.
According to Pew Research Center, 73 percent of teens age 13 to 17 have access to smartphones, and 91 percent use the Internet on a mobile device, with 92 percent of them using it at least once a day. That’s some pretty strong evidence on the surface.
Yet the reality may turn out to be entirely different. It could be that the teens in the target audience would prefer to receive this information in another way, such as in-person consultations, an email newsletter or a hard copy pamphlet they can pick up and read later.
Before making significant investment in building the app and rolling it out to the entire population, a better plan would be to create a pilot and test it against other, more traditional methods of disseminating information. As part of the test the organization can try including different incentives for taking actions. For example, if participants review a set of information and then take and pass an online test, they can earn a reward. The rate of redemption will show how important rewards are to this population.
The organization can refine this concept further by offering different types of rewards (such as direct cash payment, gift cards or points toward prizes) to the test groups to determine which are more effective at driving the desired behavior. The more understanding there is around the target population and what motivates it, the more effective the organization can be in designing programs that lead to the desired health outcomes—and lead to downstream improvements in financial performance.
Behavioral analytics can also be used to improve patient engagement (and guide resources) in an existing program, such as diabetes management. The organization can gather information regarding the types of programs it is offering, such as care management, automated care gap identification and notifications, group counseling sessions and self-study information on diet and exercise. It can compare the usage rates for each of its program elements, along with the effect it has on individual and population health, and subsequently on in-patient stays and emergency department visits. Armed with this behavioral data, the healthcare organization can determine which interventions are the most effective—and therefore worth greater investment—as well as those that are no longer worthy of continued investment, freeing up budget to develop new options or enhance existing ones.
By coupling this information with analytics showing the impactability (likelihood that a particular intervention will have a significant impact on patient or population outcomes) and intervenability (likelihood that the patient or population will become actively engaged in their own care), the healthcare organization can further refine its programs and maximize its chances of delivering the right care to the right patient at the right time.
Clearly, measuring behaviors when determining how to engage patients more effectively for value-based care can have a significant impact on success, and avoid costly missteps. Why aren’t more organizations doing it?
The reason is that, up until now, they haven’t had the analytics to do it. The first generation of analytics were focused primarily on clinical outcomes, showing what happened in the past.
Next-generation analytics provide the opportunity not only to dig deeper to determine which programs, offerings and strategies drove those outcomes, but also to start asking “what if we did things differently?” These predictive analytics are helping hospitals and healthcare organizations take advantage of the data they have about patient behaviors to improve their engagement strategies. That can help them create an even greater impact on both patient and population health.
More for you
Loading data for hdm_tax_topic #better-outcomes...