BC’s Hospital at Home programs accentuate research and learning

British Columbia’s Hospital at Home (HaH) programs, which are among the first in Canada, are now on the verge of adopting AI to further improve workflow and assist doctors, nurses, and other clinicians.

In particular, AI is expected to be extraordinarily helpful when it comes to patient selection – an area where clinicians spend a lot of time.

“The big advantage is that AI enables health professionals to focus on patient care instead of patient selection,” said Dr. Kendall Ho, professor of emergency medicine at UBC’s Faculty of Medicine. “Hospital professionals are spending hours on administration just to identify eligible patients, and that’s before talking to them.”

Dr. Ho added that it’s not only a patient’s medical diagnosis that’s taken into consideration for HaH eligibility. There are several other factors, such as how far away they live from the hospital, which usually determines whether nurses and support staff can reach them.

There’s also the issue of family supports at home, and the types of therapies that need to be administered by the hospital.

Artificial intelligence could be used to automate and speed up this time-consuming selection process.

“If we can use AI to start screening so clinicians don’t need to spend hours looking through the records, that’s a huge advantage,” said Dr Ho.

To use AI for this purpose, British Columbia’s HaH programs will need to build a set of digitized inclusion/exclusion criteria that are integrated into a machine learning platform.

Importantly, Dr. Ho noted that AI can learn and adjust as it goes along and can refine its results over time to keep improving the screening process.

Of course, AI’s human overseers will be learning, too.

“We’ll be looking for the juncture points that lead to patient improvement [at home],” said Dr. Ho. “We’ll use AI to ask how we can improve the efficiency of the management of the patient, and where can we add value as a health system to continue to improve.”

As Dr. Ho noted, healthcare professionals in B.C. will deploy AI in HaH to answer important questions. “We want to find out, are there patterns of management that we didn’t know about? Are there treatments that in fact could be different from what we are using? Those are all advantages of using AI.”

B.C.’s Hospital at Home system started at Island Health, on Vancouver Island. It now consists of eight programs, with at least one in each of BC’s health authorities. There are a total of 117 beds at last count, a major increase from the program’s initial nine beds, and participants are treated as inpatients rather than discharged patients.

Moreover, HaH cares for a wide range of patients across many departments. “It spans the patient population from hospitalist to internal medicine, surgery, family medicine and community, and psychiatry,” said Dr. Ho. “So, it’s very exciting that as a province, we have this type of expertise and different programs, and that we can learn from each other.”

There’s a spirit of learning and cooperation on the AI front, as well. The different programs and health authorities are working together to create a common framework for deploying and using artificial intelligence in the Hospital at Home.

“If we all paddle the same way, it will go a lot faster,” said Dr. Ho. “And I think there’s a willingness in this province to move forward together.”

So, while some HaH programs have been tinkering with AI in a limited way, the goal is now to work on an architecture for programs province-wide.

Right from the start, B.C.’s Hospital at Home program emphasized the importance of measuring the effectiveness of what was a relatively new phenomenon in Canada.

“Anytime you start something new, I think you must assess it objectively for two things: is it efficacious, and is it safe? Those are two really important pieces to start with,” said Sean Spina, PharmD, regional manager, clinical pharmacy services at Island Health.

“You also want to have a learning system, and from the start, the Island Health leadership and the ministry [of Health] recognized that we had to have early learnings from the program,” he added.

At the beginning, said Dr. Spina, the program would call every patient that went through the program and ask them questions to identify what was good or bad about it. The program leadership would then discuss the findings.

“If we had to make changes, we would,” he said. “It was education in real-time.”

So what has research shown about the Hospital at Home programs? What has it learned?

Dr. Spina and his colleagues have conducted some interesting studies and have made important discoveries.

In a survey of 990 HaH patients from June 2023 to September 2024, the average patient age was 71, and patients ranged from 17 to 105 years old.

Genders were evenly represented (48 percent men, 52 percent women). Approximately one quarter of the patients (26 percent) lived alone. The average length of stay in the program was 8.1 days.

Of the 990 cases reviewed, only 96 (9.7 percent) resulted in escalation, meaning the patient had to move back to hospital. Four resulted in an unexpected death at home (0.4 percent).

The study also found that certain medical conditions increased the chance of having a crisis at home that required a move back to hospital. These include advanced cirrhosis, severe heart disease, metastatic cancer, and frailty.

In particular, a very high risk of escalation was associated with advanced cirrhosis; a medically stable patient with advanced cirrhosis and no other risk factors had a 49.8 percent risk of escalation. As a result of this finding, the Victoria HaH Program has added advanced cirrhosis to its exclusion criteria.

Subsequently, an escalation risk prediction tool was developed for the Victoria HaH Program.

The program uses the risk prediction tool as part of the consent process, informing potential patients of their likelihood of returning to the hospital unexpectedly. The tool is also used to maintain the target escalation rate of 6-8 percent.

Moreover, the program may decide not to admit a patient with a high risk of escalation.

(See “A Tool for Predicting the Risk of Escalation in Hospital at Home,” March 2026, in The Journal of Advanced Home Medicine.)

Overall, preliminary survey results have shown outstanding acceptance of B.C.’s HaH by patients.

In fact, according to a 2023 study led by Dr. Spina, 100 percent of 75 patients surveyed and 95 percent of 57 family caregivers (FCGs) had an overall positive experience with the program.

(The study appeared in the Patient Experience Journal and is titled “Patients’ and family caregivers’ experiences with a newly implemented hospital at home program in British Columbia, Canada.”)

Moreover,100 percent of these patients and 96 percent of the FCGs would recommend the program to their friends and family, while 97 percent of the patients and 96 percent of the FCGs would choose the program again if faced with the same situation.

Those are impressive stats, even if they involved relatively small numbers of patients and family members.

Now that the program is more mature, changes have been made to integrate the findings from the research completed to date.

What remains to be done, said Dr. Spina, is research on the technology that’s being used.

“We talk a lot about technologies for the healthcare team and what they need. But with Hospital at Home, it’s the patients who are facing the technology. So, I think that’s an area of innovation, as well, where we need to engage with patients and companies,” said Dr. Spina.

He believes the technology could still be tweaked to improve efficiencies and ease of use for patients.

Both Dr. Spina and Dr. Ho are high-profile leaders of the Hospital at Home sector – not only in Canada, but also worldwide. They regularly speak at conferences and offer their insights to colleagues here and in other countries.

Recently, they were presenters at the World Hospital at Home Congress in Vienna, Austria, and spoke in Australia, as well as at a national conference held earlier this year in Montreal.

They’re also learning from Canadian and international colleagues. Dr. Spina asserted that Australia is where HaH began, some 30 to 35 years ago. It has a lot to teach others, especially on the topic of patient-centred care in the home.

Asked which other leaders he follows most closely, he noted the following:

• Dr. Michelle Grinman in Alberta. She is also the lead for next year’s Canadian Hospital at Home Conference (CHAHC).
• Dr. Bruce Leff at Johns Hopkins University School of Medicine.
• Dr. David Levine, an associate professor of medicine at Harvard Medical School and a leader in Mass General Brigham’s home-hospital program.
• Dr. Michael Maniaci, physician lead for Mayo Clinic’s Advanced Care at Home program in Florida.
• Dr. Michael Montalto, at Epworth HealthCare’s Hospital in the Home program in Melbourne, Australia, a global leader in this field.

When it comes to technology in the home, Dr. Ho breaks things down into four areas. The first comprises communications between patient and caregivers. He notes that communication in this area is aimed at patient safety and relationships with healthcare providers.

“It’s technology like video conferencing, where health professionals can be in touch with the patients, talk with them and to family members,” he said.

It also includes texting, emails, and the traditional telephone, as well as electronic call bells for emergencies.

Secondly, there are technologies for monitoring the patients. These items focus on physiological monitoring and include devices to measure blood pressure and oxygen levels, weight, heart rate, and temperature.

A third area involves the electronic health record. “It’s very important that patients, health professionals, nurses, physicians, pharmacists can enter data in different locations,” said Dr. Ho. “We can then unify the information and act on it together.”

nd he added, “Finally, the latest technology we’re now looking at is how do we use AI? Once you digitize the patient’s journey, understand the diagnosis and their background, their medications, and the physiology, AI could lead to novel ways of monitoring.”

Something else that could be improved, despite years of work on this front, is interoperability across regions. For HaH programs to work collaboratively across a province – and even within an individual health authority – they need better connectivity between the electronic health record systems used.

“In British Columbia, we have two primary electronic health record vendors,” said Dr. Ho. “And even with the two vendors, there are different instances of how data is being formatted and used. So that creates a kind of barrier for interoperability.”

At the same time, Dr. Ho believes it’s just a matter of time before the kinks in this type of interoperability challenge are worked out. It’s not permanent.

There’s also a challenge of what he calls ‘semantic interoperability’, or the way that different clinicians use and understand medical terms. “So, if I say flu and somebody else says flu, do we mean the same thing?”

This is another area where generative AI and natural language processing could help, said Dr. Ho. It could prevent misunderstanding between clinicians in different parts of the healthcare continuum. “Across the province, it will be very important for us to harmonize semantic interoperability,” he asserted.

And finally, secure systems to protect data must be developed.

“All three of these areas need to click for us to use the data,” said Dr. Ho.

The approach in B.C., to solving these challenges, as well as others, is to work together – not only with colleagues across the province, but with peers in other provinces and around the world.