Automated Robots Help to Triage Patients, Protect Care Providers
In March of this year, Mass General Brigham launched a telephone hotline for its patients, employees and community as part of its coronavirus response efforts. The goal: to answer any questions and quell concerns that people might have about the virus, its symptoms and their care options.
Ultimately, the Boston-based healthcare system, formerly known as Partners HealthCare, aimed to facilitate triage and limit exposure of its care team by directing higher-risk patients to the proper care location up front, whether that meant a testing site, a respiratory illness clinic or, if deemed necessary, the emergency department.
But the hotline quickly became overwhelmed, with peak wait times averaging around 30 minutes and callers hanging up before an operator could respond. The health system swiftly pivoted its approach and began exploring automated solutions to address the problem.
After seeing the success achieved by Providence St. Joseph Health through its own online screening and triage tool, Mass General Brigham rolled out a similar AI-based chatbot. The tool, known as the Partners HealthCare COVID-19 Screener, presents patients with questions based on information provided by the Centers for Disease Control and Prevention and Mass General Brigham experts to rapidly differentiate between possible COVID-19 cases and less threatening ailments.
Although the chatbot is new, project team members write in the Harvard Business Review that they expect it to “alleviate high volumes of patient traffic to the hotline, and extend and stratify the system’s care in ways that would have been unimaginable until recently.”
The team shared that two Mass General Brigham network hospitals — Massachusetts General Hospital and Brigham and Women’s Hospital — are considering the use of intelligent robotic technologies within their inpatient clinics and COVID-19 surge wards to reduce viral transmission between patients and providers.
The solutions, developed at the Massachusetts Institute of Technology and Boston Dynamics, would enable hospital caretakers to perform routine tasks, such as medication delivery or obtaining a patient’s vital signs, remotely instead of needing to be physically present in the patient’s room.
Hospitals Embrace AI to Streamline Clinical Decision-Making
To further drive efficiencies in hospitals during the pandemic, some healthcare systems are looking to AI algorithms to aid their clinical decision-making process. Take UC San Diego Health, for example.
The UCSD Health team partnered with a cloud service provider to develop and apply a new AI algorithm to more than 2,000 unique lung X-ray images, helping clinical staff to more quickly identify early signs of pneumonia in COVID-19 patients — critical to determining who may require admission or extended hospitalization.
The algorithm, which was trained by the team with over 22,000 notations by human radiologists, overlays color-coded maps on X-ray images to demonstrate pneumonia probability.
“As we prepare for a potential surge in patients with COVID-19, it’s not just patient rooms and supplies that may become limited, but also physician and staff capacity,” Dr. Christopher Longhurst, CIO and associate chief medical officer for UC San Diego Health, said in an organization press release. “So, it’s tremendously helpful to have tools that allow physicians who are not as experienced as radiologists in reading X-rays to get a quick idea of what they’re looking at, especially frontline emergency and hospital-based physicians.”
The tool is already paying off: An emergency department patient who was not exhibiting signs of the coronavirus had chest X-rays performed for unrelated reasons; however, the AI analysis of his images identified signs of early pneumonia. After learning this, the USCD Health clinical staff tested the patient for COVID-19, and he was found to be positive.
“We would not have had reason to treat that patient as a suspected COVID-19 case or test for it, if it weren’t for the AI,” said Longhurst. “While still investigational, the system is already affecting clinical management of patients.”