Aidoc’s AI Solution for PE Management Evaluated in Three Single-Center Studies

September 22, 2023—Aidoc, a provider of health care artificial intelligence (AI) solutions, announced the presentation of studies in AI-driven pulmonary embolism (PE) care.

According to the company, the investigations were conducted at three institutions utilizing Aidoc’s PE AI solution and demonstrated its clinical value in reducing mean hospital length of stay, improving patient access, and correctly alerting care teams of suspected PEs with potential for advanced interventions. The centers were the University of Texas Medical Branch (UTMB) in Galveston, Texas; Jamaica Hospital Medical Center in Jamaica (Queens), New York; and The University of Chicago Department of Medicine in Chicago, Illinois.

The findings were presented during the 9th annual Pulmonary Embolism Symposium held September 21-23 in Austin, Texas.

Aidoc stated that the three studies highlighted the care advantages of utilizing Aidoc in the management of acute PEs, including numerous patient outcome benefits. Also, they illustrated the ongoing evolution of Aidoc’s AI-powered PE solution from an algorithm to enterprise-wide platform used by Pulmonary Embolism Response Teams (PERTs) across entire health systems.

Aidoc’s “always-on” AI runs in the background examining both dedicated CT exams and nondedicated CT exams for both expected and unexpected PE, and brings suspected findings forward to care teams, faster, stated the company.

Aidoc’s solution for PE management and care serves as an end-to-end workflow used by multidisciplinary PERTs and diagnostic radiologists to manage acute PE. The solution automates the notification and suggests prioritization of CT imaging to aid radiologists in the triage and coordination of the disease’s treatment.

With the technology, PE cases are directed to a mobile workflow, where PERTs can view imaging, access the electronic medical record data, and communicate to expedite treatment decisions and improve patient care. PERTs can combine their expertise with the operational efficiency of AI to address acute PE directly.

As outlined by the company, key findings include:

1. From UTMB, 37% hospital length-of-stay reduction with AI-triggered PERT activation and initiation of advanced therapies.
2. From Jamaica Hospital, 68% increased access of catheter-directed interventional therapy for intermediate-high to high acuity PE patients.
3. From University of Chicago, a highly sensitive (95%) early alerting system that successfully identified critical PE patients.

First, the study at UTMB showed how Aidoc’s AI platform, the aiOS, helped operationalize PE workflows and communications. UTMB assessed the clinical and patient outcomes after implementation of Aidoc’s AI-powered PERT solution during a 1-year period.

Each AI finding had a chart review either retrospectively for PERT consultation eligibility (pre-AI) or prospectively to determine if a PERT consultation had occurred (post-AI).

The mean length of stay in the intensive care unit decreased by 33.5% (pre-AI: 2.30 days; post-AI: 1.53 days), while the mean length of stay in the hospital decreased by 36.7% (pre-AI: 6.70 days, post-AI: 4.24 days). These findings highlight the ability to enhance patient management and outcomes in the PE patient population, noted the company.

Second, the Jamaica Hospital study found Aidoc increased PE interventions—the company noted that even in settings where a PERT is available, a substantial percentage of PE patients in the traditional workflow are treated without the involvement of the PERT.

The investigators examined the change in the volume of PE patients referred by the PERT for catheter-directed therapies after implementing of Aidoc in the PERT workflow. They found that the annual volume of patients referred for interventional therapies rose by 68% compared to the period before the implementation of Aidoc PE AI.

Third, the investigators from The University of Chicago published an abstract involving Aidoc’s PE AI, assessing the retrospective predictive power of the AI to identify patients that will require escalated care. Aidoc PE AI achieved a sensitivity of 95% in identifying patients that ultimately required advanced therapy in the pre-AI standard of care PERT.