Patient and public perspectives on artificial intelligence in healthcare and health research: comfort, perceived utility, and concern.
TL;DR
This study found that AI is broadly acceptable to the public across a variety of different healthcare and health research scenarios and should be built on to interrogate the acceptability of AI across different contexts and over time.
Abstract
Background
Public perceptions of artificial intelligence (AI) in healthcare have been found to be generally favorable if specific areas of concern are addressed. This exploratory survey aimed to elicit levels of comfort, perceived utility, and concern among people with and without chronic health conditions across a range of healthcare, health training, and health research scenarios. RESEARCH
Design
AND
Methods
A cross-sectional survey was conducted with adults based in the United Kingdom. Quantitative data were presented using descriptive statistics, and non-parametric tests were used to detect differences between scenarios and the influence of sociodemographic characteristics. Free-text responses were analyzed using thematic analysis.
Results
Study participants had a mean age of 49.04 years (range 19-92) and were predominantly male (57.14%). The sample contained people with (49.07%) and without (50.93%) chronic health conditions. Most participants reported they were comfortable in each scenario (56.5-71.4%), found AI to be useful (60.2-73.3%), and indicated they were only slightly or not at all concerned (59.6-70.2%).
Conclusions
This study found that AI is broadly acceptable to the public across a variety of different healthcare and health research scenarios. Future research should build on these exploratory findings to interrogate the acceptability of AI across different contexts and over time.