# A Clinical Surveillance Software Platform for Early Identification of Severe Asynchrony in Mechanically Ventilated Patients in the Intensive Care Unit

> **NIH NIH R43** · AUTONOMOUS HEALTHCARE, INC. · 2020 · $299,854

## Abstract

Severe patient-ventilator asynchrony affects 12-30% of ventilated patients in the intensive care unit (ICU) or
approximately 200,000-500,000 patients annually in the US. Many patients will become asynchronous with the
ventilator and will be attempting to exhale when the machine (ventilator) is attempting to move air into the
lungs, and vice versa. Severe asynchrony, where asynchrony index (quantifying the fraction of asynchronous
breaths) exceeds 10%, is associated with a 5x increase in ICU mortality and was associated with 6 extra days
of mechanical ventilation. In other words, in the US, patients with severe asynchrony incur an extra $5-12
billion of critical care costs. Currently, no commercially available software exists to detect asynchrony in real-
time. The first step needed to address the problem of asynchrony is recognition, and unfortunately, due to
constraints on healthcare providers, patients may be “fighting the ventilator” well before recognition by the
providers. In addition, studies show that clinicians have a poor sensitivity in detecting asynchrony using
waveform analysis. Our overall goal is to assist respiratory therapists in identifying episodes of severe
asynchrony earlier and improving their accuracy and speed in interpreting waveforms, which are major steps
required to address asynchrony. Our specific aims are: 1. Developing and Testing a Clinical Surveillance
Dashboard for Respiratory Therapists to Monitor Asynchrony in Multiple Patients. In this specific aim,
we will extend the Syncron-ETM software to analyze data from multiple ventilators. In addition, we will perform a
simulation study with real patient data, where two respiratory therapists will assist in a proof-of-concept clinical
utility testing. In Scenario A, the respiratory therapists will monitor 10 patients where the asynchrony
information is provided. In Scenario B, another set of 10 patients (with similar asynchrony behavior) are
monitored without any information on asynchrony. Respiratory therapists are asked to identify episodes of
severe asynchrony (asynchrony index>10%) for a period of more than 5 minutes. In the end, we will compare
the number of correctly detected episodes of severe asynchrony (comparing sensitivity and specificity) and
timing of such detections. 2. Developing the Capability to Assist Respiratory Therapists in Improving
Waveform Interpretation. In this specific aim, we propose to develop a capability to assist respiratory
therapists in the interpretation of waveforms more accurately and rapidly. In order to ensure clinical adoption,
we intend to avoid a “black box” approach. Specifically, we intend to provide adequate information to the user
and allow the user to make the ultimate decision regarding asynchrony by “auditing” the system. First, we will
add a capability to visually annotate waveforms and highlight detected “landmarks”. Next, we will perform a
simulation study based on previously collected patient data, where two respirato...

## Key facts

- **NIH application ID:** 10079676
- **Project number:** 1R43HL154833-01
- **Recipient organization:** AUTONOMOUS HEALTHCARE, INC.
- **Principal Investigator:** Behnood Gholami
- **Activity code:** R43 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2020
- **Award amount:** $299,854
- **Award type:** 1
- **Project period:** 2020-08-01 → 2023-01-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10079676

## Citation

> US National Institutes of Health, RePORTER application 10079676, A Clinical Surveillance Software Platform for Early Identification of Severe Asynchrony in Mechanically Ventilated Patients in the Intensive Care Unit (1R43HL154833-01). Retrieved via AI Analytics 2026-05-26 from https://api.ai-analytics.org/grant/nih/10079676. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*