Transforming How Patients and Clinicians Prepare for Appointments

Before their appointment in the clinic, a patient logs into HURTS and begins a conversation with its AI agent. The agent guides patients through intake questions. But unlike a traditional questionnaire, it asks follow-up questions, adapts to the patient's responses and gathers information relevant to their specific appointment. It then pulls additional information from the patient's medical record to create a summary that helps the clinician prepare to meet their patient for the first time.

Importantly, humans remain central to the process: Patients review and approve the summary before it is used, and clinicians validate the information during the visit.

"There are two very key humans in the loop," Johnston said. "There's a patient in the loop and there's a clinician in the loop. This is not agentic AI. This is AI used to the maximum utility without a lot of the risk."

In addition, the technology is hosted on Mass General Brigham servers, helping ensure data security while enabling these advanced AI capabilities.

It's all about human care

HURTS is not intended to replace clinicians. Instead, it is intended to be used as a tool to amplify clinical expertise and allow clinicians to practice at the highest level.

"We're trying to leverage this technology in a way that empowers humans, not replaces them," Huang said. "We know this technology is coming. There are right ways to do it, and there are wrong ways to do it."

According to Johnston, the system may ultimately help incorporate not only clinical facts but also patient values and preferences into care decisions.

"So much of what we do once we've established the facts of the case is to talk about what our patients value," Johnston said. "We make treatment decisions with our patients honoring those values. This tool helps us get to the core of what is most important to our patients."

Early results show promise

Preliminary evaluations have produced encouraging findings. In a retrospective review involving 424 spine patients, HURTS condensed approximately 950,000 characters of chart information into concise summaries that physicians could review in an average of 13.5 seconds. The evaluation also suggested the tool could improve surgical triage and referral efficiency.

The platform, first used in the Department of Neurosurgery, has since expanded to the Department of Anesthesiology, Perioperative and Pain Medicine and received Institutional Review Board approval for a prospective, first-in-human randomized clinical trial.

With funding from a Baker Family Initiative in Patient Experience and Outcomes award, the team hopes their project - dubbed "The Flipped Clinic: AI-Assisted Patient Intake to Transform the Specialty Care Experience" - will be adopted in many more clinics across Mass General Brigham.

"Every single specialty has a number of things that they would love to do with their clinic, if only they could have just a few pieces of key information beforehand," Huang said. "The vision is that HURTS expands and grows not only within Mass General Brigham, but hopefully one day outside of Mass General Brigham and becomes a universal tool for clinicians."

Collaboration is key to success

While the idea for HURTS stemmed from a conversation between two colleagues, it quickly became a collaborative endeavor.

Leaders such as Merit Cudkowicz, MD, MSc, executive director of the Neuroscience Institute, where she holds the Carol and James Herscot Endowed Neuroscience Institute Chair; Meg Kotarski, MBA, chief administrator of the Neuroscience Institute; and E. Antonio "Nino" Chiocca, MD, PhD, chair of the Department of Neurosurgery, gave their full support.

"I would be remiss if I didn't also mention our Neurosurgery and Anesthesiology, Perioperative and Pain Medicine colleagues Hasan Zaidi, MD; Jeff Webb; Katie Dunn; Robert Jason Yong, MD, MBA; Olivia Sutton, MD; Brian Sou, MD; and Kathryn Satko; as well as Layne Conley and his colleagues from Digital," Johnston said. "They all played a role in helping champion, test and expand the HURTS technology across departments."