Prealize Blog

3 ways AI is making proactive healthcare a reality

Written by Sri Gopalsamy | Aug 27, 2020 11:39:57 PM

For centuries, reactive healthcare – waiting for an illness or issue to occur before addressing it – has served as the industry status quo. This has exacerbated risks for patients, insurers, and providers, even before the global coronavirus pandemic. In just a few months, the worldwide healthcare crisis has revealed how acting after the fact leads to poor outcomes, confusion, and a lack of clear answers.

 

Based on global events as well as recent conversations with key players in the healthcare industry, the need for proactive healthcare is more evident and imperative than ever. Where reactive healthcare leads to expensive and avoidable illnesses and treatments, not to mention a strained-to-the-verge-of-breaking system, a proactive model aims to identify and tackle patient issues and needs before they become critical.

 

Ultimately, the current healthcare crisis and climate are highlighting a stark need for change, and next-generation health analytics powered by artificial intelligence now enable accurate targeting of individual patients as well as proactive measures that can preserve health, improve engagement, and enhance outcomes.

 

The stark difference between reactive and proactive

 

When it comes to reactive versus proactive, the numbers add up quickly: a whopping $205 billion is wasted each year on reactive, inefficient, and uncoordinated healthcare. Much of this waste is spent reactively managing patients with chronic conditions –– those who are most likely to go to the ER, visit multiple doctors, and require hospitalization, for example. Fully $32 billion of healthcare spending comes from unnecessary ER visits alone.

 

Further, many individuals have recently postponed care for issues considered non-urgent due to fear of COVID-19 exposure, other medical phobias, or an inability to afford care. Sooner or later, all of these issues are likely to require attention and will be enormously expensive to manage if people wait until they become urgent before seeking treatment.

 

In contrast, proactive healthcare does not wait for illness; rather, it focuses on early outreach and engagement. Thanks to the use of next-generation predictive analytics, organizations can access patient risk and offer mitigation before a health issue becomes a crisis. To that end, more insurers are making proactive healthcare a priority to improve health outcomes.

 

Unlike traditional population health tools, which rely on historical indicators such as past ER utilization or a static view of patient co-morbidities to make generalized predictions about the future, proactive healthcare solutions dive deeper. As part of this deep dive, they train on specific member population data to uncover underlying clinical drivers and individual patient insights. 

 

What took so long? Despite the need for change, first-generation analytics lacked the precise and actionable insights required to power effective proactive healthcare – which is where artificial intelligence comes in. Where many population-health tools capture broad strokes, AI moves the dial forward with specificity in terms of demographics and behaviors. AI-based technology and analytics make it easier for organizations across the industry to target individuals and address their particular issues with effective measures.

 

While many forms of change are challenging for organizations and employees, the shift to proactive, AI-based technologies can be life-changing for companies and the people they care for, making it a must-adopt transition.

 

Enter AI: Taking proactive healthcare from wish to reality

 

The gap between reactive and proactive can be effectively and affordably bridged with artificial intelligence. The following 3 proactive steps can make a tremendous difference in national healthcare and real-world outcomes:

 

  1. Making sense of it all

The first phase of AI in healthcare involves discovery – processing millions of data points that are readily available to insurers in the form of medical claims, which average 1 to 4 million new claims per month for most companies, as well as prescription medication usage, lab results, and patient engagement with the system overall. To put this into practice, our engineering/operations team members can process these data into a standard format for analysis, and then our data science experts can use that information to build AI/ML models and continually train and improve those models.

 

Such models can lead to significant annual predictions, helping our customers identify patients whose next-year costs will exceed $30,000, those who will have an increase of $50,000 or more compared to their average as well as patients who may move from the bottom 10 percent of annual cost to the top 10 percent. Data can also indicate: a) emergency room utilization, b) inpatient admissions and readmissions, c) clinical reasons for increase in costs and d) patient likelihood of responding to outreach.

 

All of this matters because it makes it much easier to identify patients with critical needs and to develop programs and outreach to stop some of these issues in their tracks or mitigate them. This first step uses AI to make a profound difference in people’s health and lives, something that is meaningful before, during, and after a pandemic. In other words, discovery is all about being proactive.

 

2. Making an impact

Information is powerful, but information alone is not powerful enough.

 

Armed with data, our clinical and data science team members then created algorithms to determine clinical impactability, a score representing the likelihood of a successful intervention based on a patient’s issues. This incorporates such features as medical condition and healthcare utilization patterns. Importantly, these algorithms will allow insurers to identify the members with whom they can make the largest potential impact. Then, it is time to plan for that impact.

 

Prealize works in a consultative manner with customers to apply appropriate ranking based on the customer’s resource capability and priorities. An understanding clinical impactability, organizations can optimize workflows and targeting to create real action and change as efficiently as possible.

 

3. Making and taking action

Some organizations have the capabilities to collect and even analyze data, but not all of them excel at leveraging that information to drive better outcomes. 

 

AI tools can lead to highly impactful action plans that enable health plans and provider organizations to deploy limited resources in the most effective way possible. Determining which patients will most benefit from such action and connecting them with resources –– such as care management, digital health programs, chronic disease management, wellness programs, behavioral health support and digital outreach –– gives companies the best chance at reducing costs while also improving member health, creating the ultimate win-win.

 

Taking action is the critical third step in deploying AI in the transition from reactive to proactive healthcare.

 

Taking the next step

At Prealize, we’re in the fortunate position of spending a lot of time talking with innovative payers. Again and again, they share that reactive care continues to drive most healthcare spending. As a result, payers are increasingly turning to AI tools and insights to further the shift to proactive healthcare. AI-powered next-generation analytics help insurers lead the charge to better health, better budgeting, and better outcomes for payers, providers and patients alike.

 

About the author:

Sri Gopalsamy is CTO of Prealize Health. Prealize uses machine learning to transform healthcare from reactive to proactive, so more people can live healthier lives.