According to a new review from the University of Birmingham, patient-reported outcomes are slowly being integrated into healthcare Artificial Intelligence (AI).
Researchers from the University of Birmingham and University Hospitals Birmingham examined over 600 interventional studies on healthcare AI technologies. The team found that only 24% of studies included a patient-reported outcome element in their research. However, in recent years the number of studies including patient outcomes has risen by almost two-thirds.
The study, titled ‘The role of patient-reported outcome measures in trials of artificial intelligence health technologies: a systematic evaluation of ClinicalTrials.gov records (1997–2022)‘, has been published in The Lancet.
Healthcare AI can facilitate automated problem-solving of issues traditionally carried out by doctors. It can also make quicker and more reliable decisions and apply problem-solving techniques that humans alone would not be capable of.
Examples of innovative healthcare AI include BrainWear, a system that asses the progress of brain tumours, which is being developed at Imperial College London and HeartFlow, a tomography system that creates a digital model of the arteries in the heart, identifying where blockages may be restricting blood flow.
Healthcare AI can revolutionise treatment
“The opportunities for AI to revolutionise healthcare are only going to make patients’ lives better if those models consider how patients actually feel and respond to healthcare interventions. Our review shows that patient-reported outcomes, such as measures of symptom burden and quality of life, are increasingly being incorporated into AI studies which are very encouraging,” said Dr Samantha Cruz Rivera from the Centre for Patient-Reported Outcomes Research at the University of Birmingham.
According to the researchers, healthcare AI may be capable of tech analysing and raising an alert when a patient’s health is in decline in the future. For this to happen, large-scale patient-reported outcome datasets are needed so that AI can support care in specific conditions and incorporate the patient experience.
“Integrating patient-reported outcomes within AI can support the humanisation of AI for health and ensure that the patient’s voice is not lost in a rush to digitise and automate health care,” added Dr Cruz Rivera.
Chronic health conditions lead the way
According to the review, patient-reported outcomes for chronic health conditions, such as mental health and arthritis, have been integrated into AI studies more frequently than other conditions.
Research on patient-reported outcomes has been a priority for National Institute for Health and Care Research (NIHR) and the Birmingham Biomedical Research Centre. The researchers have said that the adoption of patient-reported outcomes for testing AI healthcare technologies in chronic conditions demonstrates the importance of patient feedback for long-term health management.
“Managing long-term health conditions places a huge burden on patients and their families, but also the NHS and social care system. AI systems can help support patients and healthcare systems to aid decision-making, improve workflow and lead to more efficient care with improved outcomes,” said Melanie Calvert, professor of outcomes methodology at the University of Birmingham.
The researchers have been encouraged by increased research into AI tech solutions for chronic conditions and how patient-reported outcomes have been incorporated into these new studies.
“It’s clear that having technology that can analyse and predict patient outcomes to help prioritise care is going to be a part of healthcare’s future. However, we must ensure that the patient-reported outcome data used to train the AI systems are applicable to the population they are intended to serve. If we don’t do this, the gaps between advantaged and disadvantaged populations will only worsen, ” Calvert concluded.
