We conducted a cross-sectional study using an anonymous web-based survey to investigate the professional use of social media apps among medical residents and to identify the factors influencing their adoption in health care. The study was designed and reported in accordance with the CHERRIES (Checklist for Reporting Results of Internet E-Surveys) [40] (Checklist 1).

Eligible participants were medical residents currently enrolled in postgraduate training programs in France. Participants were recruited between June 25 and November 27, 2024. Recruitment was conducted using a nonprobability convenience sampling strategy. The survey link was disseminated via institutional mailing lists, university networks, and social media through medical residents’ groups. Participants were also encouraged to share the survey with peers, enabling snowball sampling. The recruitment message included a brief description of the study objectives, eligibility criteria, estimated completion time, and a link to the survey. Participation was voluntary, and no incentives were provided.

The questionnaire was developed to assess the professional use of social media applications among medical residents and the factors influencing their adoption. It included both descriptive and theory-based components, with the latter based on the UTAUT2 model [39], adapted to the health care context. It was administered online using LimeSurvey and was designed to be completed in approximately 15 minutes. Before dissemination, it was pilot-tested on 3 medical residents to ensure clarity, relevance, and usability.

The survey was structured into 4 sections covering participant characteristics, personal and professional use of social media apps, frequency and contexts of professional use, and factors influencing adoption and use. Participants accessed the survey through a web link and could complete it using a computer or mobile device. They were able to review and modify their responses before submission. No randomization of items was implemented. Adaptive questioning was used based on the apps selected by participants to reduce the number and complexity of questions. To minimize duplicate entries, survey settings were configured to restrict multiple submissions from the same participant.

Only fully completed questionnaires were included in the analysis. Incomplete responses were excluded to ensure data integrity. All statistical analyses were conducted using RStudio (version 4.3.3; Posit Software, PBC). Descriptive statistics were used to summarize participant characteristics, frequency and context of app use, and UTAUT2 construct scores. Categorical variables were reported as counts and percentages, and continuous variables as means and SDs. To investigate factors influencing technology adoption and use, PLS-SEM was conducted for WhatsApp as described earlier. A 2-sided significance level of P<.05 was considered statistically significant.

This study was approved by the CERAP-HP Center Ethics Committee (IRB: IORG0010044). Participants were informed about the study objectives, eligibility criteria, and procedures on the survey’s introductory page. Informed consent was obtained electronically, as completion of the questionnaire implied consent to use anonymized responses for research purposes. No personally identifiable information was collected. Data were anonymous, stored on secure servers, and scheduled for deletion after publication. No financial or material compensation was provided.

A total of 137 medical residents were included in the study (4 were excluded from the analysis due to inconsistent responses). Most of them were women (n=87, 63.5%) and were aged between 26 and 40 years (n=102, 74.5%). A total of 40 distinct specialties were represented, with pediatrics (n=18, 13.1%), general medicine (n=12, 8.8%), and radiology and medical imaging (n=10, 7.3%) being the most represented (Multimedia Appendix 3).

In private life, the most used apps among the 137 medical residents were WhatsApp (n=133, 97.1%), Messenger (n=132, 96.4%), Facebook (n=120, 87.6%), and Instagram (n=105, 76.6%). LinkedIn (n=10, 7.3%) and TikTok (n=12, 8.8%) were less common.

This study examined the professional use of 6 popular social media platforms among 137 French medical residents across 40 different specialties. Regarding the context of professional use, over 90% of residents reported using WhatsApp professionally, particularly for written transmissions, patient case discussions, and seeking specialist advice. Messenger was used less often for similar purposes. Surprisingly, WhatsApp was also the top app for professional networking, far ahead of LinkedIn, which was used by less than 15% of residents. About 25% of residents used Facebook, mostly as the top app for medical training. Conversely, Instagram was rarely used, mainly for education, and TikTok was never used. Regarding the factors influencing adoption and use, WhatsApp ranked highest overall, leading in performance expectancy, behavioral intention, and use behavior, and showing high effort expectancy, indicating it was perceived as highly effective, easy to use, and likely to remain in use. LinkedIn ranked highest for social influence (peer effect), while Instagram ranked highest for hedonic motivation (entertainment value). Technology trust was low across apps, particularly for Facebook, Instagram, and Messenger, reflecting common security and privacy concerns. Notably, Habit was the only construct significantly associated with both the intention to use and actual use of WhatsApp, suggesting that routine use may outweigh security concerns. These results suggest that informal, noncompliant apps are widely used in clinical practice, mainly because of habit and convenience, even when trust in their security is low.

Regarding the context of professional use, our study found WhatsApp to be the most used app among residents (127/137, 93%), a result that is similar to O’Sullivan et al [7], who reported 100% use among Irish hospital residents. In contrast, Van Ravenswaay et al [8] found Facebook to be the most popular platform among health care professionals (71%) in a survey conducted by the Clinical Education Alliance, based in the United States. In our study, WhatsApp was mainly used for patient care, including transmissions and specialist advice (86/137, 63%) and case discussions (55%), which is similar to the findings in Australia, where 85% of physicians use it for clinical communication (eg, patient management details, imaging reports) [6]. Similarly, Cherrez-Ojeda et al [4] reported a 93% WhatsApp usage rate for physician communication in Ecuador, without detailing clinical contexts. Surprisingly, our study found WhatsApp to be the top app for professional networking, surpassing LinkedIn, which is specifically designed for this purpose. Youssef et al [11] also found that German surgeons preferred messenger apps (34%) over employment platforms (27%). Lastly, our results found limited and mostly passive use of social media for education, consistent with previous studies [8,11,12].

Regarding the factors influencing adoption and use, our study showed that Habit significantly influenced both the intention to use and the actual use of WhatsApp among medical residents, a finding consistent with another study using elements of the technology acceptance model and UTAUT2 [35]. Additionally, other factors such as perceived usefulness [24,35] and ease of use [24] have also been identified as predictors of social media use. However, comparing with other studies is challenging because they differ in terms of studied population (eg, physicians, pharmacists, allied employees [35]); theoretical models used (eg, technology acceptance model [24,25,35], UTAUT2 [35]); or factors examined (eg, habit is not systematically studied [24,25]). The absence of significant effects of other factors in our study may reflect our population: medical residents, a digitally fluent group likely in a late-stage adoption phase, where WhatsApp use is already routine. In such contexts, technology use often becomes automatic, and habit can outweigh the influence of early adoption factors such as perceived usefulness or ease of use, a pattern well documented in postadoption research [56-58].

Our study showed the widespread use of social media in health care, despite security and regulation concerns. Restrictive policies cannot be a solution to prevent security breaches and may even be counterproductive. Instead, various strategies should be adopted at different levels:

IT level: Secure health care apps should be developed with the desirable features of current social media apps while incorporating strong security safeguards. As shown in a prior study [59], required features should include usability (eg, easy and fast), interoperability with electronic health records, safety (eg, secure authentication and secure certified hosting), and functionalities, such as “instant messaging,” “group discussion,” “notification,” or “data storage in electronic health records” [59]. Additionally, the Medical Informatics community has recognized the importance of human factors in the safe and effective adoption of new technologies [60]. As suggested by Kushniruk and Kaufman [61] in 2024, this requires a multilevel perspective, from the individual user to the complex social and organizational context. In parallel, the American Medical Association stressed the need to integrate physicians’ perspectives throughout development so that digital solutions enhance care, fit into workflows, and address concerns such as liability, transparency, and trust [62].

Educational level: Training in digital health, including security, patient confidentiality [63-65], and regulations [66,67], should be implemented in medical curricula [59]. Digital professionalism (ie, responsible use of social media and professional integrity) should be taught as well [68]. Evidence shows that targeted digital professionalism programs improve knowledge of appropriate online behavior, professional conduct, and confidence in privacy and security practices [69]. Innovative approaches, such as interactive workshops reviewing social media content, can lead to meaningful behavior changes [70]. Likewise, promoting critical thinking in digital health [71] and interprofessional team-based training [72] can further strengthen competencies, enhance safety culture, and prepare physicians to practice effectively in a digital health care environment.

Institutional level: Hospitals and health institutions should establish clear, enforceable policies for digital communication, including acceptable tools [59]. They should also provide professional smartphones to separate personal and work use, thereby reducing security risks [3]. Additionally, they should implement and promote the adoption of secure apps, focusing on a bottom-up approach (eg, direct engagement with frontline clinicians and designation of an experienced colleague to lead adoption), which has been proven to increase adoption [73,74]. Finally, institutions should allocate resources to secure communication as part of broader efforts to improve patient safety, efficiency, and regulatory compliance. Financial investment is essential to enable practice change and technology adoption. For example, Sevick et al [75] demonstrated that implementing a digital tool not only improved provider and patient satisfaction but also reduced per-patient costs, illustrating that such investments can be both impactful and cost-effective.

This work has several limitations. First, it focused on French medical residents, which may limit the generalizability to other professional groups, health care systems, or cultural contexts. However, residents are a relevant group as they are heavily involved in clinical work and represent the next generation of physicians likely to use social media in clinical practice. The sample size (N=137) is relatively small, but it spanned 40 specialties, enabling analysis across various disciplines. Due to the sample size, structural modeling (PLS-SEM) could not be performed for all apps [55]. However, we still conducted descriptive factor analysis for each app and used free-text responses to ensure a broad interpretative scope despite the analytical constraints.

Second, reliance on self-reported data may introduce “selection bias,” possibly overrepresenting residents more engaged with social media [76], which may have been further influenced by the use of social media as one of the recruitment channels, potentially favoring more active users. Responses may also be affected by “social desirability bias,” as residents may have described behaviors that do not fully reflect their actual practices due to a desire to appear more responsible or avoid judgment [77]. To limit this bias, the survey was anonymous, which encouraged participants to answer honestly. Likewise, “recall bias” might occur [78], as participants may not remember exactly how they used each app. We mitigated this by designing a structured, detailed questionnaire, breaking down distinct dimensions of use (frequency, professional tasks, and adoption factors), encouraging residents to reflect more precisely on their practices.

Finally, we limited the analysis to 6 adoption factors (performance expectancy, effort expectancy, social influence, facilitating conditions, hedonic motivation, habit), as defined by the validated and widely used UTAUT2 model for technology acceptance and use behavior. We also added a technology trust factor to address health care–specific concerns regarding data privacy and security. However, other factors, such as institutional factors (eg, hospital management policies [25]), could also be considered in further studies.

This study found that medical residents overwhelmingly used WhatsApp for professional communication, particularly in patient care and networking, with Messenger serving similar functions at a lower frequency. Facebook, Instagram, and LinkedIn were used far less often and mainly for limited purposes, such as education and career development. The analysis of factors influencing adoption and use showed that habit was the only significant factor influencing WhatsApp’s use, highlighting the dominance of routine over security concerns. Secure alternatives with similar features to popular social media apps, along with proactive institutional policies and digital professionalism training, are needed to help physicians better consider security when using social media apps.