I, Doctor: Patient Preference for Medical Diagnostic Artificial Intelligence

Abstract

Background: Artificial intelligence and automation have the ability to positively alter the practice of medicine through streamlined diagnostic timelines, increased diagnostic accuracy, and reducing employee workload. However, patients and providers alike may feel wary of implementing these technologies into their care. This study aims to evaluate four factors that may influence an individual’s preference for the use of these technologies: Accuracy, Efficiency, Invasiveness, and Risk. Methodology: We implemented a survey which presented hypothetical medical scenarios followed by questions relating to preference for an automated medical intervention against a traditional, non-automated human intervention among 60 psychology undergraduate students. Results: The study found that the accuracy and efficiency of the intervention greatly influenced participant preference for it, with higher accuracy or efficiency of the automation relating to a higher preference for the automation. It was also found that invasiveness did not significantly influence preference for an automated method, with participants failing to significantly choose the automated intervention even when it presented a less physically invasive option compared to the traditional method. Finally, it was found that participants significantly preferred the human over the automated intervention in higher-risk medical scenarios. Conclusion: By discussing the benefits of accuracy and efficiency in using automated healthcare tools, such as their ability to reduce wait times and diagnostic timelines, and implementing these technologies starting in low-risk scenarios, patients and providers alike may be more likely and willing to see the benefits these tools have to offer.