A new tool developed by orthopedic trauma surgeons at NYU Langone Health is helping to accurately predict mortality risk, functional outcomes, and costs for geriatric and middle-aged patients who have orthopedic fractures. Based on a statistically validated risk algorithm, PersonaCARE supports informed care decisions for a growing patient population.

“We created this software to fill a need to better treat orthopedic patients age 55 and older who have fractures,” says Sanjit R. Konda, MD, assistant professor of orthopedic surgery. “This population has been growing rapidly, and studies show that it is stressing the financial capability of hospitals to provide quality care.”

The heart of PersonaCARE is the Score for Trauma Triage in the Geriatric and Middle Aged (STTGMA), a peer-reviewed algorithm developed by Dr. Konda and colleagues. The software also integrates with hospital accounting systems, enabling it to generate cost predictions.

“Our goal is to help more orthopedic surgeons and other trauma specialists deliver value-based care that decreases hospital costs while improving patient outcomes.”—Sanjit R. Konda, MD, Assistant Professor of Orthopedic Surgery

Orthopedic trauma surgeons at NYU Langone now use PersonaCARE to accurately predict quality measures such as length of stay, need for blood transfusion, risk of major and minor complications, and mortality risk up to one year after a hospital stay for all older hip fracture patients. The system can also predict inpatient and outpatient costs up to 90 days post-discharge.

Tool Enables Early Decision-Making

PersonaCARE uses data collected in the emergency department during a standard history and physical examination. Based on physiological variables including age, medical comorbidities, vital signs, and anatomic injuries, the software calculates a risk score and a cost score, allowing surgeons to promptly initiate appropriate clinical pathways.

“For example, the system might identify that a certain patient has a high cardiac risk, so we can put that patient immediately into a cardiology pathway with an echocardiogram and a cardiology consultation,” Dr. Konda says. “Getting these consultations early makes the entire process more efficient—in fact, it often helps us get patients with hip fractures to the operating room within 24 hours.”

PersonaCARE Aligns with Value-Based Care

By accurately predicting emergency room costs, procedure costs, and inpatient costs for patients with hip fractures, PersonaCARE enables physicians to make decisions based on care value.

“A risk profile might show us that a certain patient does not need to spend any time in the intensive care unit (ICU) and can be safely monitored in either a step-down unit or even a regular room,” Dr. Konda says. “So rather than admitting the patient to a unit where costs will accumulate rapidly, we can offer appropriate care in a less care- and cost-intensive setting.”

Cost profiles generated by PersonaCARE can predict hospital accounting and loss for an entire 90-day episode of care. According to Dr. Konda, this could help clinicians manage patients under bundled payment programs such as Medicare’s Bundled Payments for Care Improvement (BPCI) Initiative.

Improving the Algorithms for Better Patient Data

Dr. Konda and his colleagues are using machine learning techniques to refine PersonaCARE. They recently studied whether the addition of frailty markers—such as existing disability, need of an assistive device, and malnutrition—could improve the software’s underlying algorithms, though they were found to be insignificant in predicting inpatient mortality. “We are going to continually test new markers to see if there is anything we could do to improve on the accuracy we already have,” Dr. Konda notes.

NYU Langone plans to roll out a commercial version of the PersonaCARE platform in 2019. “Our goal is to help more orthopedic surgeons and other trauma specialists deliver value-based care that decreases hospital costs while improving patient outcomes,” Dr. Konda says.