11358
Blog
Deepak Srikant

Case in Point: AI’s Effectiveness in Pulmonary Embolism Care

Incidental Pulmonary Embolism, Length of Stay, Pulmonary Embolism, Time to Treatment, Workflow
Table of Contents

Insights from a Mount Sinai Health System study Presented at the 10th Annual PERT Symposium

AI continues to reshape clinical workflows, particularly finding an edge in how it collects, analyzes and applies data to patient care. At the 10th annual PERT Symposium, a study conducted at Mount Sinai Health System was presented, showcasing the powerful potential of AI in managing pulmonary embolism (PE), a life-threatening condition that requires precise and timely treatment decisions.

Using Aidoc’s AI-driven platform, this study demonstrated how automation can significantly streamline data collection and ultimately improve patient outcomes. We spoke with Farnaz Dadrass, MD at Mount Sinai, who was the lead author on the study and shared her insights on the implications of these findings for both clinicians and patients.

1. Inside the Study

The study, conducted between July and December 2023 at Mount Sinai, utilized Aidoc’s AI to track over 1,000 patients diagnosed with PE. The goal was to automatically collect and analyze crucial clinical data from EMRs, including factors such as heart function, oxygen saturation and blood pressure, among other biomarkers. This data allowed researchers to categorize patients into four distinct risk groups: low, intermediate, intermediate-high and high risk based on PE severity. 

The AI streamlined data collection, which traditionally requires manual input from healthcare professionals, by pulling key variables like right ventricular to left ventricular (RV-LV) ratio and biomarkers such as troponin and D-dimer levels. These parameters play a critical role in determining the severity of PE and informing treatment pathways.

According to the study’s abstract, of the 1,024 patients analyzed, almost half (48.4%) were categorized as low-risk, while 39.7% were intermediate risk, 9.7% intermediate-high risk and 2.2% high risk. These stratifications allowed for more precise treatment strategies based on the risk category, improving clinical decision-making.

2. What Surprised Dr. Dadrass

Dr. Dadrass noted several surprising aspects of the study, particularly how much the AI-driven system could expedite data analysis. “The speed at which this data was pulled and categorized was remarkable,” she explained. “In traditional settings, gathering this much data manually could take days or even weeks. The fact that the AI did this instantly allows us to act faster, which can be life-saving in acute cases.”

She was also struck by the large volume of incidental PE findings. “Nearly 27% of the PEs were found incidentally, which is a significant proportion. These patients might not have been diagnosed as quickly without this level of automated triage.” 

Dr. Dadrass also commented on the value the AI provided in categorizing patients by risk level. “Having the AI collect and compare so many data points helps us risk-stratify patients more accurately. For instance, seeing that 21% of patients with elevated troponin and right heart strain had systolic blood pressures below 90 mmHg underscores how important it is to identify and manage high-risk patients quickly.”

3. Patient-Centered Benefits: How AI Enhances Outcomes

From a patient perspective, the AI-driven system can offer substantial improvements in both experience and outcomes. By reducing the time it takes to diagnose and stratify PE cases, patients can receive the right level of care faster, minimizing the risk of complications and, in the worst cases, death.

For patients in the low-risk category, for example, the AI can allow for quicker identification of those who may not need aggressive treatment, reducing unnecessary hospital admissions and interventions. Conversely, high-risk patients can be identified sooner, ensuring they receive appropriate life-saving treatments without delay. 

“Speed is everything when managing conditions like PE,” Dr. Dadrass emphasized. “For a patient, the difference between waiting hours or even days for a diagnosis and receiving immediate care can be the difference between life and death. AI accelerates this process significantly.” 

Additionally, as the AI continues to aggregate large data sets, it will contribute to personalized care by uncovering trends that might not be immediately obvious in smaller patient populations. As the study’s abstract highlights, the large-scale data collection enabled by AI “opens the door for identifying disparate populations, ensuring that care becomes even more tailored to the needs of individual patients.

Learn more about AI’s role in advancing acute PE care in this whitepaper.

Explore the Latest AI Insights, Trends and Research

Follow a manual added link

Webinars

Aidoc experts, customers and industry leaders share the latest in AI benefits and adoption.

Follow a manual added link

eBooks

Explore how clinical AI can transform your health system with insights rooted in real-world experiences.

Follow a manual added link

AI Strategy

Learn how to go beyond the algorithm to develop a scalable AI strategy and implementation plan.

Follow a manual added link

Solutions

Explore how Aidoc can help increase hospital efficiency, improve outcomes and demonstrate ROI.

Deepak Srikant

© Copyright 2024 Aidoc