Long wait times, increased patient volumes and old infrastructure built for a different time, to name just a few challenges, have pushed emergency departments and their staff to the breaking point.
Already a Subscriber? Sign in
The latest data on wait times in Ontario, from January, shows that emergency patients who are admitted to hospital spent an average of 23.3 hours waiting for an in-patient bed. Only about one in four of those patients got a bed within the eight-hour provincial target time.
But there are rays of light.
That hospitals are facing intense pressure exacerbated by COVID-19 has been widely reported, but what has received less attention is the fact that the pandemic also provided opportunities to offer patients more flexibility and, in many cases, faster and more convenient care.
This week, the Star spoke with leaders from four Toronto hospitals to learn how they’re using technology to improve the patient experience, whether that’s cutting emergency department (ED) wait times, using existing capacity more efficiently or providing emergency care outside the hospital.
Here’s what we learned.
Virtual EDs are cutting wait times
In the last few months of 2020, as COVID cases across the province began rising to become the pandemic’s second wave, worries about contracting the virus while sitting in a waiting room were top of mind for many.
This prompted doctors and administrators at University Health Network (UHN) and Sunnybrook to create a place online that would let patients suffering from urgent but not imminently life-threatening conditions see a physician or nurse practitioner without having to set foot in a hospital.
In December 2020, the virtual ED program was born, and it’s still going strong at both hospitals. In a nutshell, it allows patients to book a time slot to see a health-care provider online that day. Before they can register, they must answer online questions to help them decide whether this sort of consultation or an in-person visit to the ED is a better choice.
“We don’t want to see patients virtually who are having heart attacks or strokes or experiencing significant trauma,” says Dr. Sameer Masood, a UHN emergency physician and the ED virtual care lead.
Patients can choose from between 15 and 20 time slots per day and visits last between 10 and 20 minutes, on average. Providers can order blood work, ultrasounds and X-rays, if necessary (in these cases, patients would have to physically visit a clinic), and relay any prescriptions to the patient’s pharmacy.
Uptake has been steady, with UHN seeing between 200 and 250 patients virtually per month, and Sunnybrook reporting about 300.
“Historically, the right place has been within the four walls of the hospital ED, but we do know that for a significant proportion of patients, the right place could be their home or another location more convenient for them,” says Masood, also noting that having “ a specific time slot means you can plan for that. You’re not waiting around for three, four or five hours … with this uncertainty of when you’ll be seen.”
AI is changing the game
When Devin Singh was a physician-in-training, he saw firsthand how many patients were waiting an unreasonably long time for care in EDs. It lit a fire in his belly not only to decrease wait times but also to help ED staff identify which patients needed care most urgently.
So, at age 32, he went back to school and got a master’s degree in computer science, learned to code and opened his eyes to the possibilities of the power of artificial intelligence.
Now, the emergency physician and clinical co-lead of Artificial Intelligence in Medicine at SickKids is in the early phase of the launch of a new app he designed. Beacon is the product of Hero AI, a startup he founded with the hospital in 2020. The app sends alerts to physicians when ED patients with urgent conditions start to wait too long for care.
How does it work? When patients see the triage nurse upon their arrival at the SickKids ED, a whole bunch of patient information — age, weight, heart rate, breathing rate, oxygen level, blood pressure, temperature, the reason for their visit — is collected. Triage nurses also assign patients a level of seriousness based on the Canadian Triage and Acuity Scale (CTAS). Level 1 (Resuscitation), Level 2 (Emergent) and Level 3 (Urgent) are given to patients who cannot wait long for medical care, while Level 4 (Less Urgent) and Level 5 (Non Urgent) are for patients who are able to wait.
“Let’s say I have 10 patients at CTAS Level 2 who all need urgent care. When you have that many patients at that level, are there a few of them that absolutely need to be seen right away, such as transplant patients or those who are immunocompromised? And are there some that can wait a bit longer?” Singh explains. The patients’ information is sent to the Hero AI software, which uses machine learning to identify which patients might need to be seen the soonest. The app will then send an alert to staff to flag higher risk, vulnerable patients.
In the case of a child going through an acute mental health crisis, for example, Singh says Beacon will alert a psychiatry team elsewhere in the hospital who can see the patient immediately, instead of having the patient wait in the ED to see an emergency physician. The wait for children, “ from arrival to when they’re seen by psychiatry, is now less than 90 minutes. The length of stay for those kids sitting in the ED is now reduced by almost two hours, which is a huge improvement from before,” he says.
There is also a patient-facing version of Beacon which tells them how busy the ED is, where they are in the queue and estimated wait times. Singh says this information has helped to cut down on some of the incivility that can erupt between parents and nurses during frustrating and seemingly endless waits.
The app, again using machine learning, will also send patients information about their condition based on their triage data, so that they can build up their health literacy before seeing a physician.
Since its launch in September, more than 2,000 patients have used the app and Singh has ambitions to expand its use to hospitals across the country and beyond.
“We’re just getting started,” says Singh.
You can now book an ED appointment
Long wait times for less serious emergent conditions are the scourge of EDs, and it’s worse when many patients show up at the same time.
What if EDs could help patients who don’t need immediate medical attention avoid these surge periods by offering consultations during historically less busy times?
That’s exactly what Michael Garron Hospital is doing with its ED booked appointment program, which also launched during the pandemic.
“These are unscheduled scheduled visits,” explains Dr. Kyle Vojdani, chief and medical director of Michael Garron’s ED. “It’s a program where something happens and you need emergency care, but there is a little bit of flexibility of when you can come in. For example, if you knew that the ED was busier at 11 o’clock in the morning and then right after work, would you come at those times if you could come when the volumes were lower?”
Using the hospital’s website, patients are presented with a list of criteria they must meet to ensure their condition is one that can in fact wait a short amount of time for care. They can then choose one of 10 to 12 appointments offered every day at a time that tends to be less busy for the ED.
Upon arrival, patients with booked appointments still have to go through the regular triage process, see a nurse, be assessed by a physician and potentially be sent for imaging or wound dressing, depending on their condition. But the aim, says Vojdani, is to get these patients in and out within two hours.
He stresses that the program is not clearing a path for patients who don’t need emergency care; all patients who book their appointments at Michael Garron’s ED need to be seen urgently.
To date, almost 4,500 patients have used the program and there’s capacity in the system for more daily appointments, Vojdani says. He notes the booked appointments have not raised wait times for other ED patients. In fact, he says, Michael Garron has among the lowest wait times in the city, despite having one of its highest volumes of patients.
Algorithms give accurate wait time estimates
We’ve all had that sinking feeling in our hearts when we arrive at the ED waiting room only to find it packed with other people.
How long we end up waiting is a crapshoot dependent on not only patient volumes, but also on the seriousness of others’ illnesses, staffing levels and physical space in the ED.
Sunnybrook Health Sciences Centre’s online ED wait time tool, available for mobile devices and desktop, launched last November. Using an AI algorithm that considers patient volumes and the illnesses being presented, the tool provides wait times, the number of patients waiting to be seen by a physician for the first time and the total number of patients in the ED, so that visitors can get a good sense of what to expect. The information is also on display at the ED for patients who don’t have mobile devices.
“It empowers patients to consider their reason for ED visit in the context of the wait time, and for some, they will decide they may have another choice such as attending a walk-in clinic, using a virtual ED clinic option, waiting for a next-day appointment with a family provider, or coming to the ED at a less busy time,” says Dr. Justin Hall, chief of Sunnybrook’s ED.
Initially, the tool was updated once an hour, but Hall says positive patient feedback prompted the hospital to increase that to once every 15 minutes to give a more accurate sense of what’s going on in the ED.
“Some patients have commented that it provides them not only with an expected wait but a greater appreciation for how busy the department is,” Hall adds, explaining that the ED also uses the tool and its data points to boost staff at more demanding times. And there may be more to come.
“We would like to make it more refined to get it to the individual patient level. It’s not there yet so what we have now is an average of what patients can expect.”