AHI System

The AHI System utilizes pattern analysis of lead-II ECG waveform data, specifically heart rate variability and morphology, to provide real-time assessment of hemodynamic stability and predict the likelihood of future hemodynamic instability in adult hospitalized patients. Its purpose is to act as an adjunctive clinical decision support tool, enabling earlier recognition of patient deterioration by stratifying risk based on continuous physiological monitoring.

The AHI System is a multiparameter software device designed to aid healthcare professionals in managing adult, hospitalized patients undergoing continuous ECG monitoring. It features two analytics: the Analytic for Hemodynamic Instability (AHI) and the Analytic for Hemodynamic Instability Predictive Indicator (AHI-PI). AHI analyzes heart rate variability (HRV) and ECG morphology from a single lead-II ECG to provide a frequently updated binary output (Stable/Unstable) indicating the patient's current hemodynamic status. Hemodynamic instability (HI) is defined as hypotension (systolic blood pressure <90 mmHg or MAP <70 mmHg) combined with tachycardia (heart rate ≥ 100 bpm). AHI-PI utilizes up to 30 minutes of AHI outputs to predict the likelihood of a future HI episode, defined as 10 continuous minutes or more of HI. The system presents this risk as a color-coded indicator (red, yellow, green). Clinical data demonstrated that AHI-PI can distinguish risk levels, with a high-risk (red) indication being 51 times more likely and a moderate-risk (yellow) indication 9 times more likely to experience HI in the next hour compared to a low-risk (green) indication. For patients who experienced HI, the AHI-PI High Risk indicator predicted deterioration in 89% of cases with a median lead time of 48 minutes. The AHI System is intended for adjunctive use and not to independently direct therapy, aiming to enhance vigilance and early identification of patients at risk.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. How does the AHI-PI algorithm's performance vary across different patient subgroups not extensively covered in the study, such as those with specific cardiac comorbidities or arrhythmias?; What is the clinical impact on patient outcomes (e.g., ICU length of stay, mortality, intervention rates) when clinicians actively use the AHI System compared to standard monitoring alone?; How does the AHI System account for potential confounding factors that might affect ECG morphology or heart rate variability, such as medications or patient movement artifacts, and what is the false alarm rate?; What is the specificity and negative predictive value of the AHI-PI indicators, particularly for the 'Low Risk (Green)' category, in a broader, more diverse hospital population?; Given the AHI-PI's reliance on up to 30 minutes of prior AHI outputs, how robust is its predictive capability during rapidly evolving hemodynamic changes or immediately post-intervention?