Medical students’ digital therapeutics adoption depends on more than enthusiasm alone
Digital therapeutics are moving from concept to clinic, but this study makes clear that acceptance among future prescribers is still uneven. Among 309 medical postgraduates surveyed across multiple universities in China between October 2024 and January 2025, the reported intention to adopt digital therapeutics reached 61.5%. That figure suggests a meaningful level of openness, yet it also leaves a large minority outside the adoption pathway at a time when these software-based interventions are trying to gain a place beside drugs, devices, and counseling in routine care. The gap matters because the students answering these questions are not casual observers; they are tomorrow’s clinicians, and their willingness to prescribe or recommend digital tools will shape whether those tools ever reach patients in a credible way.
A brief but important finding sits at the center of the analysis: behavioral intention emerged as a positive predictor of adoption, with an odds ratio of 1.753 and P < 0.05. In practical terms, the more strongly students intended to use digital therapeutics, the more likely they were to report adoption. The study also identified two negative predictors. A higher cognitive level of digital therapeutics was associated with lower adoption, with an odds ratio of 0.704, and low perceived necessity carried an odds ratio of 0.422, also with P < 0.05. One result is especially striking: greater knowledge did not automatically translate into greater uptake.
Education, necessity, and professional fit shaped uptake in opposite directions
The pattern points to a familiar but often underappreciated problem in implementation science. Education can expand awareness of a therapy’s scope, limitations, and regulatory complexity, and that broader cognitive load may dampen confidence if the clinical role of the product remains fuzzy. For digital therapeutics, where software, behavioral guidance, monitoring, and disease-specific indications often overlap, knowing more may reveal more uncertainty rather than less. Students who understood the category more deeply may have been less likely to adopt it if they were unsure where it fits in the treatment sequence or whether it belongs in their future prescribing routine.
Perceived necessity worked in the opposite direction. When a student saw a real clinical need for the intervention, adoption became more likely. That is biologically and behaviorally plausible: treatments are usually embraced when they solve a problem that conventional care does not solve well enough, whether because of access, adherence, self-management, or the need for repeated reinforcement between visits. Digital therapeutics are designed to do exactly that in many disease areas, but the study suggests the perceived gap between potential and necessity still determines whether future clinicians view them as essential or merely novel.
The investigators used a validated questionnaire grounded in the Technology Acceptance Model and the Theory of Planned Behavior, then paired logistic regression with fuzzy-set Qualitative Comparative Analysis. The two methods tell different stories. Logistic regression isolates single predictors, while fsQCA tests how conditions combine into pathways. In a field like digital therapeutics, where adoption rarely hinges on one belief alone, that combination is clinically useful. It captures the reality that a student may need usefulness, necessity, and professional compatibility at the same time before the threshold for adoption is crossed.
Five configurations produced high adoption, but education absence predicted non-adoption
The fsQCA results identified five configurations associated with high adoption. Rather than relying on a single universal formula, the analysis showed that students could arrive at adoption through different combinations of beliefs and educational exposure. High perceived usefulness and high perceived necessity formed a core configuration, while professional compatibility appeared as a peripheral condition in some pathways. In other words, digital therapeutics were more likely to be adopted when students believed the tools were both valuable and relevant to their future clinical role.
The absence of DTx-related education stood out as a near-necessary condition for non-adoption, with a consistency of 0.924. In plain clinical language, not being taught about digital therapeutics strongly tracked with not adopting them. Yet even this negative association was not absolute. A strong core of usefulness and necessity could compensate for limited educational exposure. That finding has direct relevance for curriculum design. If medical schools want future prescribers to use digital therapeutics thoughtfully rather than ignore them, teaching cannot stop at definitions or theoretical advantages. Students need to see where these tools fit in patient care, which conditions they address, and why they are worth considering alongside established options.
The study’s central message is not that education alone will solve adoption resistance. It is that the pathway to uptake is layered. Cognitive understanding, perceived necessity, usefulness, and professional fit all interact, and any one of them can weaken uptake if it is missing. Digital therapeutics are unlikely to enter routine use because students are simply told they are innovative. They enter practice when future clinicians understand the clinical problem, recognize the device’s value, and can imagine prescribing it as part of normal care.
What this means for medical training and implementation strategy
For educators, the findings argue for more than a descriptive lecture on digital health. Training programs need to address how digital therapeutics compare with medication, counseling, and follow-up care in real patients, especially in situations where adherence or access remains a barrier. The study also suggests that professional identity matters. If students do not see these interventions as compatible with their future role, adoption will lag even when they understand the technology itself.
For developers and health systems, the lesson is equally concrete. Adoption strategies should not focus only on technical performance or marketing language. They need to build clinical credibility by showing usefulness, necessity, and fit within workflows that future prescribers already recognize. The survey was cross-sectional and limited to medical postgraduates in China, so it cannot prove how beliefs will translate into actual prescribing behavior once these students become practicing physicians. Even so, the data offer a clear map of where resistance may form and where education can intervene before habits harden.
The practical question now is whether training programs can turn conceptual familiarity into clinical readiness. This study suggests they can, but only if education is paired with a convincing case for necessity and a clear explanation of how digital therapeutics belong in everyday care.