A study led by researchers from the Nanyang Technological University Singapore has revealed that individuals are more likely to accept preventive health interventions suggested by AI with the involvement of human health experts.
It also found lesser trust in AI-powered preventive care than in interventions led by human experts.
The study has inquired into users’ perceptions of preventive health interventions, such as health screening and physical activity prompts, proposed by AI as compared to those recommended by humans. It involved about 15,000 participants in South Korea using an undisclosed mobile health application.
The first set of 9,000 participants was grouped into three: one group was given AI-recommended daily steps; another group received steps recommendations from human experts, and a third controlled group received a neutral intervention that mentioned neither AI nor a health expert.
It was found that almost one in five of those who received AI suggestions accepted the intervention while 22% of people in the second group accepted the recommendations from human experts.
Later, another set of participants was recruited with one group receiving an intervention that disclosed the use of AI in tandem with health experts and another group receiving an intervention that explained how AI came up with steps recommendations.
From this cohort, the researchers noted that individuals are more accepting of AI-suggested health interventions that are complemented by human experts than those interventions based on purely AI or humans. There is also a higher rate of trust in transparent AI-generated interventions.
WHY IT MATTERS
The study’s findings, which have been published in the journal Production and Operations Management, indicate that the human element remains important even as the health system moves to further adopt AI for screening, diagnosing, and treating patients.
“Our study shows that the affective human element, which is linked to emotions and attitudes, remains important even as health interventions are increasingly guided by AI, and that such technology works best when complementing humans rather than replacing them,” said Hyeokkoo Eric Kwon, an associate professor from the NTU Nanyang Business School who led the study.
THE LARGER TREND
Given the growing ubiquity of machine learning and AI in healthcare settings, it has become more crucial to design digital technology with the users in mind to ensure these become an integral part of care interventions.
In a HIMSS forum late last year, Jai Nahar, a pediatric cardiologist at Children’s National Hospital in the US, said that “whenever we’re trying to roll out a productive solution that incorporates AI, [the patients should be involved] right from the designing stage of the product or service”. Clinicians too must also be included in this process, he added.
Meanwhile, another mobile health study in South Korea published earlier this year found that mobile health apps could moderate the effects of social determinants on the health of South Koreans. Based on a survey of over 1,000 participants, it was revealed that frequent use of mobile health technologies could ease the effects of SDOH, such as societal economic inequality, on a person’s capacity for self-health management and on the personal view of their health.