SBIR Phase II: A Multi-Modality Sensing Approach for the Objective Assessment of Agitation and Sedation

The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase II project involves advancements in the state-of-the-art in biomedical technology, improvement of the quality of care for intensive care unit (ICU) patients, and a reduction in healthcare costs. The proposed agitation and sedation system will assist in realizing an improved distribution of ICU workload by allowing the medical staff to focus on patients with increased agitation and in need of immediate attention. It is expected that this technology will also improve the quality of care in military ICUs. The proposed project focuses on the development of a multi-modality agitation and sedation assessment system for patients in ICUs. The primary consequence of oversedation is the prolongation of mechanical ventilation and ICU stay. Prolonged ventilation is expensive and is associated with known risks, such as inadvertent extubation, laryngo-tracheal trauma, and, most significantly, ventilator-associated pneumonia. Alternatively, agitated patients can do physical harm to themselves by dislodging vital life support and monitoring devices with excessive musculoskeletal activity. Undersedation and oversedation result in excessive economic burden on the healthcare system. The annual cost of oversedation in the US is $400 million, whereas reducing mechanical ventilation in oversedated patients by just one day can result in an annual savings of $4 billion in the US alone. Current clinical practice in patient critical care requires the nursing staff to assess the patient?s agitation and sedation state and provide sedatives to ameliorate the patient?s agitation. This process relies on subjective assessments and can be influenced by personal bias. This project proposes to use machine learning algorithms to integrate data from multiple sensing modalities to objectively identify cases of undersedation and oversedation in ICU patients.