Infinitus has launched Clinical Escalations, a new AI-powered risk detection system designed to improve patient safety during healthcare conversations by identifying urgent clinical issues in real time. The technology continuously evaluates patient interactions with AI agents, allowing the system to recognize when an inquiry that initially appears administrative—such as a prescription refill request—may actually signal a medical emergency. By bringing clinical risk assessment into AI conversations from the very beginning, Infinitus aims to reduce the risk of under-triage and ensure that patients are directed to appropriate care more quickly.
Unlike traditional automated phone systems that rely on patients to recognize and report emergencies themselves, Clinical Escalations monitors every exchange throughout a conversation and dynamically adjusts its assessment as new information emerges. The platform uses the Schmidt Thompson Call Prioritization Index, a widely adopted nurse triage framework, to assign severity scores on a turn-by-turn basis. According to the company, the system has achieved a 0% under-triage rate in real-world deployments, meaning it did not miss any patient emergencies during testing. Every escalation decision is also recorded to provide transparency and allow human clinical teams to review AI recommendations.
The technology has been trained on real-world healthcare conversations involving patients with diverse communication styles, literacy levels and medical complexity, enabling it to recognize high-risk situations even when symptoms are described indirectly or inconsistently. Infinitus says the platform is intended for healthcare providers, health plans and pharmaceutical patient support programs managing large volumes of patient-facing communications. The launch builds on the company's broader AI platform, which already includes tools for creating healthcare AI agents and analyzing conversations, and reflects growing interest in AI systems that can safely augment clinical operations while maintaining human oversight.
Click here to read the original news story.
