Aaron Conway, an assistant professor at the Lawrence Bloomberg Faculty of Nursing is the 2021 recipient of the AMS Healthcare Compassion & AI Grant, a funding opportunity that will support his work integrating AI and machine learning. Conway’s goal is to improve the patient experience and provide enhanced care for cardiac patients undergoing procedural sedation.
Current guidelines from the Canadian Anesthesiologist’s Society require patients to stop eating and drinking non-clear fluids from 6 hours and 2 hours respectively, prior to a procedural sedation. However, current practice usually has patients receiving standardized instructions to fast from midnight the night before, a fail-safe approach that many hospitals turn to as a result of unpredictable scheduling changes for procedures. As a result, patients are often fasting for unnecessarily long periods of time. Recent research has found that on average, patients have fasted for more than 12 hours.
“Fasting before sedation or anesthesia is important to prevent a patient from vomiting or aspirating or having other adverse outcomes,” says Conway, who is also the RBC Chair in Cardiovascular Nursing Research at the Peter Munk Cardiac Centre. “However, the current practice can lead to these excessive fasting times, which can be quite difficult for patients to tolerate.”
With the support of the AMS Healthcare grant, Conway is working with Professor Chris Beck from the Faculty of Engineering, Julie Vizza a patient partner from University Health Network (UHN) and Amanda Matthews from the cardiac triage team at the Peter Munk Cardiac Centre (UHN), to create a real-time, automated instruction system that will predict procedure start times reducing the number of hours patients are required to fast from food and non-clear fluids.
The automated fasting instruction system will aim to overcome the challenges that schedule changes and delays present for clinicians and their patients. Following preliminary discussions with patients, caregivers and clinicians, the goal is to develop a simple SMS notification system that will provide patients with accurate and updated start times for their procedures, with patients receiving two alerts, one to let them know 6 hours prior to stop eating, and another 2 hours prior to abstain from non-clear fluids.
“By utilizing machine learning in this context, I am hoping to bring compassionate care to these patients by reducing this unneeded burden and distress around fasting from food and drink during what is usually an already stressful period of time,” says Conway.
In addition to the development of the digital technology required, the grant will also be supporting the co-design process with patient partners. Conway and his team will conduct a series of qualitative interviews with patients, caregivers and physicians to inform the design of the tool and to ensure that the patient and health team requirements for optimal use of the tool are included in the implementation of this technological solution.
Conway acknowledges that previous attempts to adhere to the 6and 2-hour fasting recommendations have been unsuccessful because they have relied on an increased workload for physicians and ineffective methods of communication between hospital departments.
Using state of the art machine learning, Conway is hopeful that his automated instruction system will offer a more effective solution and reduce clinician workload while providing a better care experience for patients.