Applying AI in Healthcare: Netsmart AI Data Lab

Over the past decade, billions of federal dollars have been invested in reshaping how hospitals and clinics manage their health records. But in human services, post-acute and community-based services — the segments of care dedicated to mental health, substance use, home care, hospice and the environments people live in — change has been slower to come.

Healthcare organizations across the nation are working to address clinician burnout and are more constrained than ever. That’s why Netsmart, an industry leader in electronic health records (EHRs) for human services and post-acute care, and Amazon Web Services (AWS) joined forces to advance artificial intelligence (AI) for community-based care providers, through the development of an AI Data Lab. By combining the scale and power of AWS with the Netsmart CareFabric® platform, Netsmart is driving intelligent automation, generative AI, predictive analytics, natural language processing (NLP) and risk models to improve care delivery and outcomes for community-based care providers.

The first-of-a-kind AI Data Lab built on AWS speeds up innovation, allowing Netsmart to rapidly deploy new technology and capabilities designed for clinician satisfaction, streamlined operations and enhanced outcomes. Early initiatives include a fully AI-enabled digital assistant utilizing cloud services that can summarize, identify gaps in treatment and screening and suggest recommendations and actions in real-time at the point of care. Netsmart and AWS are building new capabilities within the AI Data Lab to scale predictive modeling to help clinicians anticipate and intervene before a person is in crisis or need of hospitalization and leverage NLP to lift data from unstructured notes and voice dictation, making that data actionable and insightful for providers and clinicians.

In addition to integrating new technologies, such as Amazon Aurora and Amazon Translate, Netsmart is embarking on ambitious AI/ML projects using Amazon Bedrock, Amazon Redshift, Amazon Chime SDK, AWS Healthscribe, and Amazon SageMaker to develop a digital assistant that can summarize patient history and offer suggestions when prompted, detect risk of hospitalization and suicide risk, and offer real-time translation and parse free text into usable data.

“Netsmart is leading the way in applying AI in healthcare delivery, and their success in leveraging AWS native services for a decade is rapidly accelerating our ability to deploy and add value for behavioral health and post-acute providers,” Tehsin Syed, general manager of health AI, AWS.

We’ve started with a digital assistant powered by generative AI that automates patient history reviews, a routine and time-consuming process that’s a leading cause of provider burnout. By leveraging services like Amazon Bedrock and Amazon SageMaker, Netsmart can create short-form summaries of relevant treatment milestones, freeing up clinicians’ time to focus on providing care.

“We believe equipping providers with AI technology is the future of actionable, predictive analytics within the healthcare ecosystem,” said Netsmart chief executive office (CEO), Mike Valentine. “MBHC is at the forefront of the industry, and we’re thrilled to be partnering with them on effectively taking their predictive analytics and wiring it into their population health and provision of care functions, empowering staff to make real-time decisions based on the broad data set available. This is only the beginning of our impactful innovation efforts with AWS, and we look forward to what’s on the roadmap, including becoming a qualified health information network (QHIN), advancing interoperability and care coordination for community-based providers nationwide.”

The AI Data Lab is also deploying predictive models that can help prevent hospital admissions. Using data amassed from EHR and health plan claims, the algorithm will identify a patient’s risk, as a fast follow on, they’ll address one of the most difficult challenges in behavioral health: how to accurately identify the risk of suicide. Alerts from these new tools can be harnessed across the US healthcare ecosystem in emergency rooms, primary care clinics, and other settings.

The AI Data Lab is also taking on a project for NLP that starts with real-time translation for non-English speaking case managers and will also handle automating paper records into structured data that can help identify patient needs.

“The work we’re doing with Netsmart and AWS has been instrumental in implementing cutting-edge analytics and technology, propelling the success of the Certified Community Behavioral Healthcare Clinics (CCBHCs) model across the state of Missouri and nationwide,” said Missouri Behavioral Health Council CEO, Brent McGinty. “Through these collaborations, we are not only achieving remarkable outcomes but also setting new benchmarks for funding opportunities in CCBHCs and statewide programs, revolutionizing the way we deliver mental health services.”

MBHC leverages the Netsmart CareManager™ population health management platform, enabling the organization to access real-time data from multiple sources, such as EHRs, claims data and more, powered by AWS for scalability and real-time analytics, including machine learning. MBHC takes a holistic approach to individual care, efficiently coordinating services across various providers and stakeholders, including law enforcement and crisis response teams. This approach leverages predictive modeling to prioritize both reactive and proactive actions, enabling the organization to target at-risk individuals based on social determinants of health data and tailor interventions to their needs. The CareManager platform is instrumental in this effort, providing real-time alerts driven by predictive analytics, covering everything from potential hospitalizations to Medicaid costs. Since implementing CareManager, MBHC has shared they have achieved significant results, including 14% lower hospital costs, a 16% decrease in emergency department visits, and a 35% overall increase in access to care.

Furthermore, MBHC’s proactive stance extends to early intervention and prevention, exemplified by its recent launch of a health-risk profile using metabolic syndrome screening data to identify individuals at higher risk of developing various diseases. This comprehensive approach not only enhances overall health but also lowers long-term healthcare expenses. In a collaboration with Netsmart and AWS, MBHC is working to replicate these successful algorithms to identify suicide risk. By harnessing alerts from emergency rooms, primary care clinics, and other healthcare settings, MBHC aims to intervene and prevent suicides, ultimately saving lives.

More early adopter organizations, including those who provide post-acute care, will be announced soon. Netsmart and AWS have committed to two other rapid-cycle pilots, including telehealth transcription and predictive patient scheduling. To learn more about Netsmart, visit www.ntst.com.

Learn about how AWS is addressing patient and clinician experience, explore use cases, and other related resources.