Abstract Secure messaging (SM) has seen increasing adoption over the past decade, prompting growing interest in understanding its impact on healthcare delivery. This review examines key themes in the existing SM research, such as usage patterns, perceived benefits, and persistent challenges, to identify research gaps and inform opportunities for sociotechnical solutions in the artificial intelligence (AI) era that could enhance patient–provider communication effectiveness. Searches were conducted in PubMed, Scopus, IEEE Xplore, ACM Digital Library, Cochrane CENTRAL, CINAHL, and Web of Science. Following the PRISMA guideline, we identified 366 relevant peer-reviewed studies published from January 2009 to September 2025. Existing research has primarily investigated (1) the effect of SM use on clinical outcomes, (2) adoption patterns, and (3) user experiences. The literature shows that SM promotes patient engagement, care coordination, and patient–provider communication. However, significant challenges persist, including privacy concerns, limited access for vulnerable populations, patient misuse, and increased clinician burden. Educational initiatives and patient-centered design are essential for promoting appropriate and accessible use of SM. Emerging AI solutions also show promise in enhancing SM use, particularly for message triaging and drafting replies. The integration of such AI solutions must be guided by robust governance frameworks to ensure transparency, maintain trust, and align with evolving clinical, billing, and regulatory environments. Future research should include more diverse care settings and populations, prioritizing the development of equitable sociotechnical tools and interventions that can be seamlessly integrated into clinical workflows.
Guo et al. (Sun,) studied this question.