JMIR Publications

ABSTRACT

Background:

Direct-to-Consumer Medical Artificial Intelligence Applications (DTC Medical AI Apps) are flourishing. However, the academic community lacks a systematic comprehension of the research overview, existing barriers, and future design directions for such applications.

Objective:

The aim of this study was to provide an overview of the study characteristics, existing barriers, and design directions of DTC Medical AI Apps in the existing literature, as well as to serve as a reference for future research, design, and implementation of such applications.

Methods:

A scoping review was conducted using the Arksey and O'Malley's 5-stage framework. Articles of DTC Medical AI Apps from inception until March 27, 2023, were searched across five databases: Web of Science, Scopus, ACM Digital Library, IEEE Xplore, and PubMed, along with a snowball sampling of related articles’ reference lists for more articles. In addition, Google Scholar was used to find gray literature. Using the methodology and principles of thematic analysis, the evidence from the included articles was analyzed.

Results:

Of the 2898 articles retrieved, 34 (1.17%) were included for review. We mapped the research characteristics of the existing literature; without time constraints, but all of the relevant articles included were recently published (2018–2023). Most (24/34, 71%) were from developed countries. The medical field of applications was mostly general practice (8/34, 23%). The functions of applications included self-diagnosis (11/34, 32%), self-management (8/34, 23%), diagnosis (6/34, 18%), pre-diagnosis (1/34, 3%), and general health care inquiries (4/34, 12%). Not only for the single consumer group (19/34, 56%), but also for other users while focusing on consumers, including physicians (11/34, 32%), health departments (5/34, 15%), nursing staffs (1/34, 3%), industry professionals (1/34, 3%), local governments (1/34, 3%), and patients' family members (1/34, 3%). Nine barriers and their specific contents of DTC Medical AI Apps mentioned in the literature were also identified: Explainability, Empathy, Usability, Privacy, Accountability & Regulation, Trust, Specialization, and Physician-Patient Relationship; and seven design directions and their specific contents: Explainability, Empathy, Usability, Privacy, Accountability & Regulation, and Diversity.

Conclusions:

This is the first academic study to systematically summarize the profile of DTC Medical AI Apps, as far as the authors are aware. Future research needs to take into account the existing barriers identified in this study and refer to the refined design directions for the research, design, and implementation of such applications. This would enhance the services offered by DTC Medical AI Apps.