Listen to this article

The basics:

• Saint Peter’s reduced ER visits by 7% using Lightbeam AI
• Tech identifies social risk factors such as food and transport issues
• Patients screened and matched to support programs in real time
• Model boosts care team efficiency and supports health equity

Saint Peter’s University Hospital is using AI to help improve clinical outcomes, reduce cost of care and increase communication among its team. Since adopting technology from Lightbeam Health Solutions, the partners say the New Brunswick-based facility has recorded a decrease of more than 7% in emergency department visits among high-risk patients by identifying social determinants of health on an individual basis.

The technology is also helping Saint Peter’s use existing support programs and match them to community members who need them.

Social vulnerabilities account for 80% of health outcomes, according to figures cited by Saint Peter’s and Lightbeam.

Saint Peter’s told NJBIZ it implemented the Lightbeam Population Health Module in August 2022. The health care provider declined to disclose the cost of the programming.

However, it was described as “cost effective” for improving outcomes for high and rising-risk populations within a value-based care model. Additionally, the model integrates easily with existing systems, Saint Peter’s noted.

Founded in 2012, Lightbeam offers an artificial intelligence platform to assist and promote the transition to a value-based care model. Its team is informed by experience in ambulatory, inpatient and payer expertise.

In a case study produced by the company, it noted New Brunswick-based Saint Peter’s University Hospital – a 478-bed, Catholic, acute care teaching hospital and safety-net provider – came to the partnership with an understanding about the importance of addressing social determinants of health “before they negatively impact physical outcomes and drive avoidable utilization and cost.”

In 2023, Saint Peter’s began screening patients for social determinants, identifying transportation and food insecurity as major concerns. Speaking with the American Medical Association at the time, Saint Peter’s Healthcare System Director, Transformational Population Health & Outcomes Ishani Ved explained, “We do a quick questionnaire, and we hook them up with resources that they would need based on whatever needs are identified, or if there are potential government subsidies.”

With Lightbeam, Saint Peter’s is able to efficiently scour data to also offer proactive care to produce better outcomes.

Making connections

Saint Peter’s worked with community-based organization Healthier Middlesex to conduct community needs assessments to understand the broad essentials for its base. According to that group’s 2022 report, residents in every focus group and interview highlighted social determinants, including housing, transportation, economic security and food access.

“While Saint Peter’s had a screening process in place, it still needed a way to scale its ability to ID risk factors and SDOH. And, a way to deliver at the point of care,” Lightbeam noted in its report.

The partners established three goals:

• Identify high-needs patients and their specific social determinants
• Increase staff productivity and impact through efficient referrals to existing social service programs
• Measure and report the clinical and financial impacts of their health equity and charity care work to meet charity care reimbursement objectives and ensure financial sustainability; including reaching out to and serving the most vulnerable patients

“Through partnerships and grant funding, we had existing programs to support food accessibility and transportation but not a way to efficiently identify which patients needed them,” Ved commented in July when the figures were revealed. “Lightbeam AI identified patients with food and transportation needs, empowering our clinical team to make efficient referrals to existing programs, ultimately enabling us to more effectively serve a growing and predominantly low-resource population.”

Lightbeam AI’s model, SDOH Individual, provided Saint Peter’s the ability to predict which patients were heading toward high utilization in the next 90 days using only “social vulnerability” data. And by integrating the tech with patient scheduling systems, team members developed the ability to take proactive steps with the information.

The model identifies individual social determinant factors, which then allows clinical teams to offer support that directly aligns with specific and personal challenges, such as food insecurity or transportation — thus addressing immediate needs.

“Saint Peter’s care team reviewed the risk factors and recommendations in order to mitigate high utilization,” Lightbeam’s report noted.

Lightbeam explained that SDOH Individual lists a patient’s top 10 risk factors, along with the five best recommendations to help reduce that person’s risk. Patients are classified as either high, medium or low risk. And Lightbeam says this categorization is done with minimal patient data inputs.

Output risk factors and recommendations include:

• Poor indoor air quality – Educate patient on integrated pest management to reduce risk from pests
• Low transportation availability – Explore transportation barriers to care
• Lack of other adults in household – Assess for social support system

Additional benefits to come

Despite the AI at work, communication is still key to success; as is teamwork. Clinical teams also collaborated with other providers, social workers and the Managed Care team – both in the medical record and in real life – about the screening and findings from Lightbeam’s AI agent.

Studies indicate ED visits have increased broadly as issues accessing primary care and addressing underlying conditions also rise.

Nonetheless, “emergency care presents a high-yield opportunity for SDOH screening,” suggests a 2024 report from American College of Emergency Physician’s Clinical Emergency Data Registry. “Combined with downstream referral processes and high-impact interventions, the development of such screening programs could potentially transform long-term health trajectories in vulnerable populations, reduce costs and contribute to physician well-being with the support of reimbursement mechanisms.”

Looking ahead, the company anticipates additional outcomes, such as reductions in appointment no-show rates, missed appointment costs and increased capture of Z codes for comprehensive documentation. According to Centers for Medicare & Medicaid Services, Z Codes are used to document SDOH data. They refer to factors influencing health status or reasons for reaching out to health services that aren’t classifiable elsewhere as diseases, injuries or external causes.

“This collaboration highlights the power of AI and the integration of SDOH to improve care, enhance care manager capacity, and achieve meaningful cost savings,” said Lightbeam CEO Pat Cline. “Our market-leading, purpose-built AI models are transforming population health management for clients like Saint Peter’s to reduce ED visits and avoidable admissions while enhancing patient satisfaction.”

Noting it is still early days, Lightbeam concluded its case study saying that Saint Peter’s has achieved promising results. And is confident in its long-term impact on a patient-centered, data-driven approach to access, outcomes and financial stability.