top of page

ARTIFICIAL INTELLIGENCE HELPS RESIDENTS THINK ABOUT HOSPITAL TRANSFERS

Dr. Daniel Kobewka wants decision support to honour what matters most

    For people living with serious illness in long-term care, getting sick can force a difficult choice: go to the hospital or stay at the long-term care home.

    Hospital services may offer a chance of treatment and recovery, but it comes with noise, needles, stretchers, and unfamiliar faces that can also bring discomfort, confusion, and distress, especially for someone living with dementia.

    The alternative is to remain in the familiarity of the long-term care home, surrounded by trusted caregivers focused on comfort and quality of life, but without the same benefits of hospital-based care.

    Sometimes the choice is clear. Often, it is not.

    “Every person facing these decisions should have the support they need to make choices that honour their values, hopes, and fears,” says Dr. Daniel Kobewka, Investigator with the Bruyère Health Research Institute. “That’s why we developed Care Guide, an AI agent available by phone that supports residents and caregivers when faced with decisions about hospital transfer.”

Designing Care Guide with artificial intelligence

    Conversations about serious illness are among the most difficult in health care, but too often, those conversations don’t happen at all. Illness can progress quickly, decisions are often made in moments of crisis, and clinicians with time and expertise to support thoughtful discussions are not always available.

    Dr. Kobewka and his team believe there is an opportunity to change that by ensuring every person has access to meaningful conversations before difficult decisions must be made.

    To create Care Guide, the research team worked through what a skilled clinician does in these types of conversations—sharing prognosis, eliciting what matters most, exploring fears and hopes, and helping people weigh whether the possible benefit of hospital care is worth the burden it can bring.

    “These are not just medical questions; they are deeply human ones.” says Dr. Kobewka. “That’s why compassion has shaped the design of Care Guide at every level, including the balance of being honest about prognosis, yet gentle in delivery. Serious news shouldn’t be softened until it’s misleading, but it can be compassionately framed it in a way that is accurate while leaving room for emotional responses and nuance.”
IAF Lisa Sheehy Photo.png

“Every person facing these decisions should have the support they need to make choices that honour their values, hopes, and fears. That’s why we developed Care Guide, an AI agent available by phone that supports residents and caregivers when faced with decisions about hospital transfer.”

Dr. Daniel Kobewka

Investigator with Bruyère Health Research Institute

Bridging clinical accuracy and compassion

    Artificial intelligence is often associated with efficiency and automation. For Dr. Kobewka, its greatest potential is in creating more time and space for reflection.

    Caregivers who have used Care Guide said they feel heard rather than hurried. They open up because there's no time pressure, no clipboard, no sense that the conversation needs to wrap up. Care Guide listens for as long as it takes and doesn’t flinch at hard questions.

    Dr. Kobewka’s team has also worked to ensure Care Guide treats staying in long-term care as an active, caring choice. Rather than framing long-term care as “doing nothing” compared to the route of “taking action” at the hospital, this AI tool is built to recognize that staying in the residence is a deliberate plan focused on comfort, calm, familiarity, and quality of life.

    “Some people worry that using artificial intelligence might make us less human, but I see a future where technology helps us lean into our humanity,” says Dr. Kobewka. “I want people to feel heard without judgement, safe sharing their hopes and fears, and empowered to make decisions that honour their deepest values.”

Help make this future a reality.

Donate to support research.

bottom of page