Chronic and Behavioral Health Webinar: January 19 2021 Join for a Discussion on Payers’ Uses of Remote Patient Monitoring [Watch the Webinar] How Mental Health Center of Denver Engaged and Supported Behavioral Health Patients with Deviceless RPM

News

CareSignal Develops AI to Improve Patient Engagement

December 4, 2020 at 3:57 PM

Improving patient engagement is on the wish list of providers and payers alike. Both patient health outcomes and provider financial outcomes depend on it. Organizations that can keep patients enrolled in programs in which they’ll benefit — and confidently scale those programs — will improve clinical outcomes and, ultimately, lower costs

Yet disengagement is a reality, and providers don’t have the time or data to efficiently catch patients who’ve fallen off. We developed CareSignal AI to predict disengagement and proactively reengage patients. To understand the value, let’s first consider the clinical consequences of low engagement using a hypothetical example.

Without CareSignal AI, Life Happens and Patients Become Disengaged

An elderly Medicare Advantage patient with diabetes needs to check his blood sugar three times a day and log that information in a journal, which he brings to his doctor every three months. Initially, the patient takes an active role in managing his condition: checking his glucose levels each morning, afternoon, and night and using those insights to regulate his diet and activity levels.

However, he gets distracted and forgetful, and he gradually begins to check his blood sugar less frequently. Over time, his condition deteriorates until it becomes unmanageable and requires hospitalization. Paradoxically, patients who are doing well enough that they don’t require frequent clinical intervention might actually disengage more rapidly.

In the above scenario, the patient isn’t the only one who suffers. Not only does poor engagement lead to excess costs from a previously avoidable ED visit, but it also impacts overall patient satisfaction, which may cause the individual to seek care elsewhere.

[FINAL] 20Q4 CareSignal.ai One-Pager 2a

With CareSignal AI, Disengaged Patients Are Identified and Reengaged

Let’s say another patient with diabetes is using CareSignal’s Deviceless Remote Patient Monitoring™ platform informed by CareSignal AI. He would receive daily SMS text messages or IVR phone calls — on his existing phone — reminding him to submit his glucose readings and providing insight into his glucose trends.

If submissions become less frequent, CareSignal AI would identify the pattern, predict the likelihood of the patient disengaging with treatment altogether, and generate an actionable alert indicating the patient would benefit from an educational call to reengage him and get him back on track. CareSignal Engagement Specialists will then reach out in a matter of days or even hours to help course correct.

Preventive Reengagement Is Cost-Effective

In addition to having a clinical impact, high engagement has financial benefits. Typically, the longer a patient engages with the platform to improve health outcomes, the greater the impact in financial costs from lowered utilization and ED visits. However, we also look at the hidden financial savings of activating, engaging, and retaining a patient. We know that it is over 50% more cost-effective for our client-partners to reengage and retain patients on the platform than to enroll new patients.

How Does CareSignal AI Work?

The machine learning algorithm powering CareSignal AI is informed by 20,000 patient-years’ worth of healthcare data (and counting). In addition to patient engagement data, CareSignal collects metadata on patient disease profiles, demographics, geographic locations, and other attributes.

Our technology uses best-in-class data analysis techniques, including:

  • Ensemble-based random survival forests
  • Semi-parametric Cox regressions
  • Survival analysis methods

With these techniques, we’ve built seven machine learning models with more than 80 features that ultimately make complex questions easy to answer and even easier to act on. To date, our algorithm has proactively identified patients at risk of becoming disengaged at an over 80% accuracy rate, and it is continuously improving. Now, you can answer the question: “Which of my patients will become disengaged and when?”

AI-Informed Preventive Outreach Delivers Dramatically Better Engagement

We go beyond predicting disengagement. We’ve developed patient reengagement strategies that have led to as many as 57% more patients remaining engaged than would have without a reengagement touchpoint. Our team of Engagement Specialists handle the responsibility of reengaging patients through proactive identification and outreach, relying on a proven behavioral science-based framework to convey the importance of our remote patient monitoring platform and addressing any concerns. Consequently, patients report higher rates of satisfaction with the level of communication and quality of care from the provider.

The Engagement Specialist’s AI-informed workflow is much more efficient than traditional care manager outreach, allowing the Specialist to reengage 11.4 patients per hour compared to one patient per hour. By relying on CareSignal AI to identify patients at risk of disengagement and allowing CareSignal’s Engagement Specialists to keep those patients engaged, providers can shift their focus from enrollment to providing optimal patient care that drives retention and improves their bottom lines. Over the course of our engagements, CareSignal has been able to consistently lower patient retention costs, reduce ED visits, and generate between 4.5 times and 10 times ROI for our partners.

The impact of AI on healthcare promises to be nothing short of transformative, and CareSignal is proud to be leading the charge. By working with us, healthcare organizations find that they’re able to solve some of the most pressing problems and be better prepared to face the challenges that tomorrow will bring.

To learn more, visit CareSignal.AI