How to use Ottawa Hospital to capture AI Ambient Voice to reduce your doctor’s exhaustion by 70 %, and achieve 97 % of the patient’s satisfaction

Doctors and patients (or not) is an important problem in health care – because they can struggle to reach care and doctors because they simply have a lot to do it.

Ottawa Hospital (TOH) began facing this challenge with merge Microsoft’s Dax Copilot last year. It had already a significant impact: the early results of seven minutes appear provided for each meeting, and a 70 % reduction in fatigue and exhaustion that the doctors reported, and 93 % of patients are about the experience of better care or reward.

“It is possible that access to care will be one of the biggest issues faced by patients,” said Glen Kirins, EVP and CIO from TOH, for Venturebeat. “If we are able to improve productivity, and even two patients for each doctor in each seizure, you double this through 10 doctors in preparing care, then strike this by 365 – this is not an illogical increase in access to care.”

AI AI as an active assistant

TOH was the first Canadian Hospital to try the Dax Copilot from Microsoft, which was directly integrated with the widespread electronic health record platform epic (EHR). In March, Microsoft collected Dax Copilot with Dragon Medical One (DMO) at Microsoft Dragon Copilot assistant, which says a technology giant, used by more than 600,000 doctors.

Dax Copilot, which comes out of the box, picks up Doctor and patient conversations Through the mobile phone application and creates a draft clinical notes in real time, Ken Harper, head of the Microsoft project for Dragon, explained to Venturebeat.

He said: “Doctors can start a recording from a mobile phone, put their phone down, examine the patient, and speak to the patient, and is not different from what they do anyway.”

The system then extracts details based on the context of the visit (symptoms, diagnosis, treatment plans, and follow -up), which are immediately available in EHR; All that the doctor must do is review and finish quickly.

“Instead of having to write something from the zero point and remember all the fine details or write in front of the patient, this works automatically,” Harper explained, noting that they are receiving a first “very accurate” draft once the visit is over.

To adjust Dragon Copilot, Microsoft uses “a huge warehouse of clinical data organized over the years,” Harper explained. The engineers are constantly, large language models (LLMS) with that data so that the system can understand and reliable medical information.

To increase accuracy, models are improved through specialization – whether it is an emergency doctor, dermatologist, cardiologist, or any other medical specialist.

In the feedback rings, the teams analyze the first draft prepared by the machine and compare that with what the doctor changed after that.

“He makes sure that as an average over time, as data continues and we learn from this data, editing is decreased,” said Harper.

How to improve visits and reduce normal fatigue

According to the Canadian Medical Association, doctors spend all over 10 hours a week In administrative tasks, such as updating the plans after the patient’s dates.

Kearns explained that TOH has developed a strong evaluation plan for Dax Copilot that includes monthly updates through the Microsoft Power BI. This includes reactions from doctors, patient surveys and data from an epic.

“This framework continuously helps us monitor improvements and direct improvements,” he said.

He explained that at the end of the day or transformation, doctors must return to the documents from the patient’s visits. But the tool has been reduced after working hours, planning work and documentation for “all categories of categories doctors“This not only saves them time, but also helps reduce fatigue because they have less arduous work to do.

The tool has also improved the cognitive load of doctors during the visits: instead of focusing on the details of the patient’s entry and movement in documents and models, it is able to “engage differently and better”, Kerns said. In addition, “we have seen an increasing level of productivity, and more patients for each seizure, for each doctor.”

Kearns confirmed that all patients are required to approve before registering appointments and that they can reach notes at their patient’s Mychart portal. They are also provided politely about the program and confirm that their health records are always secret and safe.

The patient also accepted “very positive”, as it was 97 % that their experience with the artificial intelligence tool was good like a typical or better date. “We attribute this to the doctor’s opportunity to interact with them differently and more deliberately during the visit, which is often pressed in time,” said Kirins.

He said that the tool can be used in scenarios such as the detection of vital signs and social determinants of health (non -medical issues that affect a person’s health, such as lack of nutrition or lack of transportation. Moreover, pre -decline and referral orders can be.

“There is a lot of path in the area of ​​capture documents,” said Kirins.

Fellow team colleagues to overcome employment issues

Another field where TOH integrates artificial intelligence is “Digital teammates”. Last summer, the attachment developed use cases with DELOTTE and launched Sophie, which speaks several languages.

Interestingly, it is able to explain the patient’s feelings and behavioral response.

“I hate to say this, but patients lie to the doctors,” Kerns admitted. “You know,” what is your pain scale? Oh, I am fine, it’s five out of 10. “But then, Sophie will get the ability to take a look at your face and go,” Well, it does not seem to be five. “

She may then ask about the patient’s definition of five, and can be burned based on objective data.

TOH will make another avatar in the spring to help patients navigate and reach the health care system and engage in prior examination.

“I don’t think anyone does not realize the human resource crisis in the world,” Kerns. “We really want to try to support and serve patients more powerful than we can today.”

For example, the facility makes follow -up calls to patients after certain procedures. However, due to resource restrictions, they can only follow up with higher patients. Kearns’s goal follows with each patient, as it provides a Avatar explaining any questions for the patient and confirming if they understand their discharge orders, or they are able to reach the pharmacy, or they follow the orders of doctors. The avatar can escalate to a nurse or the clinical patient team if necessary.

“One of the things that health care is proud of is a human touch,” said Kirins. “This is a way to ensure that we increase and improve human touch resources, but also make sure patient support well through their health care journey.”

However, he indicated that he was still early in the game. One of the important future steps is to enable the digital team teammate to interact with the information and the epic environment.

“We have a lot of work to do there, we are still focusing on the adoption side,” said Kirins. “We are still a healthy care system that interacts with the health condition of patients. We would like to get to the place where we are active.”