artificial intelligence could reduce emergency room mental health visits at cheo
3 minute read
cheo has a higher rate of emergency visits for mental health needs than other comparable pediatric institutions across canada, according to hospital officials.
errol mcgihon
/
postmedia
researchers at cheo have enlisted artificial intelligence to help with a difficult problem: the high number of children and youth who return to the hospital’s emergency department within six months of seeking urgent mental health care. ai could help reduce some of those emergency room visits.
revisits, as they are called, are worrisome because children and youth mental health patients who make more than one trip to emergency within six months are at higher risk for adverse health outcomes, including higher rates of suicide, self-harm and overall mortality.
they also add to the swelling demand for care in already burdened emergency departments, especially in recent years. child and youth mental health visits to emergency departments grew between two and five times from the mid-2000s until the covid-19 pandemic. rates began to drop between 2020 and 2022, but mental health visits did not drop as much as other emergency visits, according to research led by dr. kathleen pajer, director of the cheo research institute’s precision child and youth mental health collaboratory, which integrates biological, social, environmental and clinical factors to create a more complete picture of each young person’s mental health needs.
cheo has a higher rate of emergency visits for mental health needs than other comparable pediatric institutions across canada, according to hospital officials. pajer said as many as 20 per cent to 30 per cent of those children and youth would come back, sometimes repeatedly, to emergency for urgent help.
she led research, recently published in bmc medical informatics and decision making, on developing a predictive algorithm using machine learning to help identify those patients at highest risk for revisits in order to make sure they had more support.
advertisement
the research found that the machine learning, a type of artificial intelligence, did a better job of predicting which patients were most likely to return to the emergency department compared to the work of doctors and others. it did so by selecting certain characteristics and using them to generate a prediction. the computer predicted with just over 70 per cent accuracy who would come back to emergency, pajer said.
dr. kathleen pajer is director of the cheo research institute’s precision child and youth mental health collaboratory, which integrates biological, social, environmental and clinical factors to create a more complete picture of each young person’s mental health needs.
cheo
/
handout
there is still work to be done, she said, to refine the results using more variables and incorporating clinical input. once the work is finalized, clinicians at the hospital will help develop a protocol to use when a young patient is deemed high risk. that might include more support to get connected with an out-patient program or more contact to assess whether the patient is getting the help they need.
most patients who attend emergency with mental health needs leave with a referral for follow-up, but those can often take a long time, pajer said. “maybe we need to be in closer touch,” with those deemed to be at highest risk.
the goal, she said, is to identify those children and youth and help ensure they are connected with a program that meets their needs and offer additional support until they do. she noted that some young people are returning to emergency because the waits for outpatient care are so long.
tammy degiovanni is senior vice-president of clinical services and chief nursing executive at cheo.
cheo
/
handout
tammy degiovanni, who is senior vice-president of clinical services and chief nursing executive at cheo, said the hospital’s clinical team already worked hard to identify who they should be most worried about when it came to patients seeking emergency care for mental health issues.
artificial intelligence enables technology to mine through electronic records to support that work by identifying patients in need of additional support. that work, degiovanni said, reinforces hunches of medical staff, supports newer clinicians with less experience and liberates their brain space to do more.
advertisement
“it enhances clinical care rather than replacing it.”
she said cheo had a track record of using artificial intelligence to look at issues and to put those findings into the hands of clinicians.
she said it would be better for individual patients to be followed closely and connected with the best partners for them in the community. it should also help the emergency department by reducing repeat visits.
our website is your destination for up-to-the-minute news, so make sure to bookmark our homepage and sign up for our newsletters so we can keep you informed.
share story
