The New York State Health Home program is designed for the neediest Medicaid patients and aims to reduce overall healthcare costs by decreasing inpatient costs (and utilization) by addressing social determinants of health such as housing, transportation and food.
THE PROBLEM
The COVID-19 pandemic has led to a major SDOH crisis in New York, resulting in an increased demand for Health Home care-coordination services – not only for current patients but for a new pool of patients who need these services, as well.
"Our program members required promptly timed outreach to identify their needs and align resources to meet their needs," said Dr. Sumir Sahgal, lead physician and founder at Essen Health Care in Bronx, New York.
"Furthermore, program regulatory requirements had to be met, which included the conducting of timely assessments and creating care plans that are patient-centric.
"With the overwhelming demand for our Health Home services, we soon discovered we weren't able to hire care coordinators fast enough," he continued. "In a nutshell, the team was swimming in work that needed to be done. We needed to scale up – quickly."
PROPOSAL
MyndYou is a vendor of an AI-powered virtual-care assistant called MyEleanor. According to Sahgal, the company offered an effectiveautomated solution that would let Essen Health Care respond to patients in need much more efficiently.
"MyndYou proposed a member outreach program using MyEleanor that would help us conduct hundreds of calls in a span of hours and thousands of phone calls in a span of just a few days," he explained. "We knew we needed to talk to and engage our members, but didn't have enough time, staff or budget to do so as quickly and efficiently as MyEleanor can."
"This allowed for the care coordinator to spend more time in completing the service for the patient and means they are calling the patient with a focused reason and to set the stage for a longer assessment call."
Dr. Sumir Sahgal, Essen Health Care
MyEleanor contacts Essen Health Care patients by phone and listens not only to what they say, but also to how they say it. AI, natural language processing and proprietary algorithms work together.
"Patients open up to Eleanor, often telling her things they won't even tell their doctor – but she can also detect worsening medical conditions from what they're not saying, too, including falls risk, behavioral health issues and more," Sahgal noted.
Almost all of Essen's patients are underprivileged and many only speak Spanish. Eleanor conducts calls in both English and Spanish and provides care coordinators with a HIPAA-compliant transcript after each call. This alleviates the need for a translator.
The vendor also offers customized health questionnaires. Patient answers provide care coordinators with information about their caseload and program-requirement needs for assessment, care plans and documentation.
More importantly, they offer immediate feedback to coordinators on patient concerns, unique needs and information about the support patients might need.
"We get customized alerts based on our preset parameters and the transcript of each call," Sahgal said. "We liked that the program would allow us to focus on addressing high-risk patients with ongoing conversations for more intensive, meaningful and high-touch engagement.
"Members who recently were discharged from the hospital or who are part of a special program could participate in our Transition of Care program with more frequent calls to identify acute needs for care coordinators and prevent rehospitalization."
MyndYou's team proposed Essen Health Care use MyEleanor for outreach to patients with chronic conditions in order to enable caregivers to maintain a strong relationship with them in a short amount of time. The vendor's team conducted multiple cohort outreaches to gauge patient engagement and involvement, then set up customized outreach to help maintain the condition and information.
This tool offered an ability for Essen's clinical staff to conduct calls that could identify the signs of a worsening condition or other issues.
"By getting patients to engage in preventative care, we helped reduce hospitalizations and ED overutilization, and helped supply the patient with more personalized care and treatment," Sahgal said.
"This ongoing check of patient calls allowed clinical staff to see firsthand the development of conditions and provide direct and timely intervention."
MEETING THE CHALLENGE
In August 2021, Essen Health Care started going through its entire Health Home membership and selected 4,000 members who needed to be reached that month. Staff members worked with the vendor's team of technical and clinical experts to set up the campaign, and they helped create a few pointed questions Essen staff would ask the members, including:
- Their preferred language (English or Spanish).
- The best time of day to call them (morning, afternoon or either).
- Any alternative phone numbers.
- Whether they are satisfied with care coordination services and, if not, what Essen can do better.
- An open-ended question about what they need at the moment.
In addition to the outreach, Essen flagged 500 patients who would be a fit for the program because they are considered high-risk. The outreach also included an option for patients to enroll in an ongoing weekly call with Eleanor.
Eleanor makes as few attempts as needed to reach the patient between Monday and Friday from 9am to 5pm, in addition to weekends and evening hours.
"After the one-week setup, the campaign started with 4,000 patients, and the initial results were very promising: Our early contact rate was 35%," Sahgal reported. "That did dip as we made subsequent calls to patients who hadn't answered on the earliest calls, but MyEleanor can also clean up our contact lists.
"She's able to determine if numbers are outdated or out of service," he continued. "That saves us the time of doing that work ourselves. We also had Essen appear as the caller ID to encourage patients to answer the calls."
Patients' responses included giving MyEleanor up-to-date connect information, outreach preferences, additional phone numbers and self-reported information about their specific needs.
"Reports on the calls come to us daily in an Excel spreadsheet after each round of calls," Sahgal explained. "We distributed them to our supervisors via internal communication, and they in turn shared the reports with their care coordinators.
"Our care coordinators used the report to update the contact information for patients, the call time preference, as well as address the identified need of the patient prior to calling them for a completed core service," he continued.
"This allowed for the care coordinator to spend more time in completing the service for the patient and means they are calling the patient with a focused reason and to set the stage for a longer assessment call, which patients usually avoid."
The ability to start fast and work with the daily Excel reports allowed Essen to tweak and improve the workflow prior to integration, he added.
MyndYou also provided a HIPAA-compliant recording of each call on their portal for quality assurance purposes.
In addition, 130 out of the 500 patients identified as high-risk signed up for the ongoing call with Eleanor. The campaign is still under a pilot model led by one of Essen's RN supervisors. The RN supervisor will receive the transcript report and distribute it to the care coordinator for action.
MyEleanor set up the ongoing call to also include "flagged issues" or "urgent concerns." These are forwarded on the same day to the RN supervisor, who will work with the patient, care coordinator and care team to address the needsthe patient reported on the call This feature is important in prompting immediate action to rising-risk patients and getting timely and reliable information for the care coordinator to act on, and hopefully prevent hospitalization, Sahgal said.
"Now that we have optimized, efficient, validated processes and workflows in place, we'll focus on further incorporating MyEleanor into our day-to-day workflow," he noted. "We think of her as a virtual assistant to our care coordinators.She helps them, and they work together to manage overall patient care coordination.
"Our goal is to integrate our campaign even more seamlessly, such as by importing transcript reports directly into our eCARES platform, so care coordinators can identify the actionable tasks and also document what they have done."
RESULTS
Following is an overview of the Health Home program results:
- 48% of the reachable target population members (defined as initially unengaged) participated in a phone call with MyEleanor to collect their updated outreach preferences and determine their clinical need.
- The Essen care team responded, and offered core services to 78% of the patients that Eleanor identified as needing additional help.
- In the Ongoing Care Support pilot with high-risk patients, 94% of patients who initially engaged with MyEleanor remained active in the second month of the program.
In initial outreach efforts during the first few months of the Chronic Care Management Program, 12% of MyEleanor patient check-up calls resulted in a clinical action. For example:
- 20% led to physician appointments.
- 5% led to post-hospital-program referrals.
- 12% led to coordinated delivery of patient medical equipment and supplies, such as glucose monitors and blood pressure machines.
"By addressing the needs of patients, our team was able to increase access to medical equipment that helped empower the patient in managing their own care," Sahgal said.
ADVICE FOR OTHERS
"Our patient population is getting older and living longer: There will be an additional 20 million Baby Boomers added by 2030," Sahgal said. "As the population ages, demand for healthcare services will increase tremendously. All of this is before the COVID-19 pandemic, which has further increased demand for care.
"For affordable care, we must decrease cost while providing accessible and quality care," he continued. "This will be led by innovation in healthcare technology. We see our daily lives changing with technology– smartwatches, smartphones and even smart cars. All of this will be applied to healthcare."
For example, telemedicine has evolved from nascent services to a standard care model in just a year, he added.
"Healthcare organizations have traditionally been slow in adopting technology, cost and disruption of care being some of the reasons for hesitation," he observed. "However, soaring patient demand, slowing reimbursement and market positioning [are]forcing the change.
"Advancement in software, artificial intelligence and natural language processing has allowed people to command machines into desired actions, bringing a new level of engagement by ease of use," he concluded.
"This will bring wide adoption, especially in our aging population, and serve as a portal to more automation and access to needed services. All healthcare organizations will need to adopt such technology."
Twitter:@SiwickiHealthIT
Email the writer:bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.
FAQs
How is AI currently being used in healthcare? ›
By automating mundane tasks, such as data entry, claims processing and appointment scheduling, using artificial intelligence in healthcare can free up time for providers and healthcare organizations to focus on patient care and revenue cycle management.
What is an example of AI in healthcare? ›AI in healthcare examples that are meant for improving communication involves, for instance, platforms for automotive appointment systems, real-time health status monitoring (handy for chronic diseases such as diabetes), or developing patient engagement solutions.
What are the benefits of AI in healthcare? ›AI can track patient data more efficiently than humans can. Through AI and machine learning, health organizations can connect disparate information that previously might not have been gathered and analyzed, allowing a more unified look at patients' health. One example is diabetes.
How is AI ML used in healthcare? ›For example, machine learning can predict which patients are at risk of developing diabetes and provide personalized recommendations for diet and exercise to prevent the disease. AI can also monitor patients' health remotely, alerting doctors to changes in patients' conditions before they become serious.
How is AI used in healthcare 2023? ›AI algorithms can catalyse the rapid analysis of health data, leading to precise diagnoses and timely interventions. Predictive models powered by AI can detect patterns and trends, aiding disease prevention and personalized treatment plans. However, prioritizing privacy and security are essential.
What are 4 AI examples? ›- Virtual assistants like Siri and Alexa.
- Recommendation systems used in e-commerce platforms.
- Fraud detection in financial institutions.
- Autonomous vehicles.
- NLP for chatbots and customer service.
- Image and facial recognition in security systems.
- Medical diagnosis and healthcare systems.
Examples of AI in Medicine and Healthcare
AI can improve healthcare by streamlining diagnoses and improve clinical outcomes. A critical part AI's power in the healthcare industry is its ability to analyze a vast amount of data sets. Digital health startup Thymia is a prime example.
Already, AI- and machine learning-enabled technologies are used in medicine, transportation, robotics, science, education, the military, surveillance, finance and its regulation, agriculture, entertainment, retail, customer service, and manufacturing.
What is the conclusion of AI in healthcare? ›Conclusion: With the introduction of more innovative and new generation AI tools, healthcare is more advanced in the sense of more awareness, efficiency in delivering care, identification of developing complications, accurate diagnosis of diseases ahead of time, and most recent approaches for interventions.
How is AI used in healthcare essay? ›Artificial intelligence (AI) has the potential to significantly improve operational efficiency and cost-effectiveness in healthcare in several ways. One way is automating tasks such as data entry, analysis, and decision-making, reducing the need for human labor and increasing accuracy (Lee, (2021).
How do AI and machine learning improve patient outcomes? ›
For example, AI and ML can help doctors diagnose diseases, recommend treatments, monitor patients, predict outcomes, detect anomalies and prevent errors. AI and ML can also empower patients to manage their own health, access information, communicate with providers and track their progress.
What is an example of machine learning in healthcare? ›For example, suppose a doctor prescribes a specific medication for a patient. In that case, machine learning can validate this treatment plan by finding a patient with a similar medical history who benefitted from the same treatment.
What is the future of AI in healthcare? ›As a result, we expect to see limited use of AI in clinical practice within 5 years and more extensive use within 10. It also seems increasingly clear that AI systems will not replace human clinicians on a large scale, but rather will augment their efforts to care for patients.
What is an example of AI in nursing? ›Many different types of AI are used, including robotic devices, predictive analytics using machine learning, and virtual health assistants. Recently, robots have been developed to take over particular nursing tasks and to accompany patients.
What are some examples of AI therapy? ›AI Therapists
One example of a therapeutic chatbot like this is Woebot, a chatbot that learns to adapt to its users' personalities and is capable of talking them through a number of therapies and talking exercises commonly used to help patients learn to cope with a variety of conditions.
Social media platforms are using AI to give a customized experience to users. One of the best examples is the friend suggestions we receive on Facebook, Instagram, Twitter, and other platforms. These apps show a list of people we might know or those who are on our contact list.