‘When will I die?’ Stanford’s AI guesses when hospital patients will die to help prioritise end-of-life care
Giving hospitals and hospices a better idea of when terminally-ill patients might die means that their final days in care can be improved
SCIENTISTS are using advanced artificial intelligence to guess when hospital patients might die.
A Stanford University research team applied machine learning technology to health records, in order to help hospitals and hospices give better end-of-life care to the terminally ill.
Researchers examined Electronic Health Record (EHR) data from Stanford Hospital and Lucile Packard Children's hospital.
The data, which covered health history for around two million child and adult patients, was used to train a "neural network" that is now able to predict the mortality of people with serious or terminal illnesses.
The idea is that by telling hospitals and hospices when patients are likely to die, end-of-life care can be prioritised in a more intelligent way.
"We demonstrate that routinely collected EHR [electronic health record] data can be used to create a system that prioritises patients for follow up for palliative care," the Stanford researchers explain.
The study concluded: "We find that it is possible to create a model for all-cause mortality prediction, and use that outcome as a proxy for the need of a palliative care consultation."
The researchers also added that the resulting model is "currently being piloted" for daily outreach to newly-admitted patients.
According to the experts, around 80 per cent of Americans want to spend their last days at home, but around 60 per cent die in hospital.
Having a way to predict deaths could help more people pass on in their preferred environment.
Speaking to The Sun Online, Dr Adrian Tookman, Medical Director for terminal illness charity Marie Curie, said that predicting prognosis is "notoriously difficult".
"Our own research shows that doctors, regardless of their experience, struggle to make accurate predictions."
But he warns that while estimating a patient's date of death is useful, it shouldn't be the only focus.
"What really matters is that clinicians provide the best possible palliative care based on an individual's needs – regardless of how long they expect someone to live."
"We know that palliative care increases quality of life, reduces pain, and can help some people live longer than exploring invasive medial interventions."
He also told us that he's interested to follow the progress of Stanford's AI tool, and hopes it can enable conversations about palliative care to happen "as early as possible".
Kenneth Jung, a research scientist at Stanford, admits that while the AI technology is helpful, it should be used in conjunction with medial professionals.
"We think that keeping a doctor in the loop and thinking of this as 'machine learning plus the doctor' is the way to go, as opposed to blindly doing medical interventions based on algorithms," Jung told IEEE.
"That puts us on firmer ground both ethically and safety-wise."