The results of this study build upon the primary analysis previously published by our group, demonstrating that more than half of oncology trainees reported discussing medical cannabis with > 5 patients in the past year (Patell et al. 2022). In sum, while more than half of trainees participating in our survey have discussed medical cannabis with a substantial number of their patients, fewer than half have translated these discussions into actual recommendations for cannabis use. In the primary analysis, prior training in medical cannabis was associated with a 50% increased likelihood of ‘discussing’ medical cannabis; the secondary analysis presented here further amplifies this and shows a 160% increased likelihood of recommending medical cannabis in the setting of prior training. This affirms the significant impact of curricula in medical cannabis use during medical training with regard to how often trainees feel prepared to discuss and/or recommend its use with patients. Given the limited high-quality evidence on medical cannabis use in oncology patients, these findings highlight the importance of formulating curricula using best available data on the subject to ensure well-informed patient-provider decision making while appreciating limitations in evidence (Braun et al. 2024a, b), in concordance with current national guidelines (Braun et al. 2024a, b). As trainees indicate that fellowship training is the most common time for this to occur (if at all), these findings point to an area of unmet educational need and an important opportunity to develop evidence-based curricula in this domain.

We found that the majority of participants in the study identified oral ingestion as their clinically preferred mode of medical cannabis use, followed by vaporization. The preference for oral ingestion may reflect both provider familiarity and the perceived safety and controllability of this route. Pharmacokinetic considerations such as bioavailability, timing of onset and duration of effect are tied to route of administration (Braun et al. 2024a, b). Current literature does not clearly endorse one route as superior across all clinical contexts, but in palliative and oncology care, the optimal route may vary depending on the symptom being targeted (Braun et al. 2024a, b; Steele et al. 2019). Interestingly, 15.5% of participants indicated they did not have knowledge on a preferred mode of use, and 16.4% had no preference. The heterogeneity in these responses may reflect that some providers lack awareness on the increasingly numerous available routes of administration, lack of robust safety data, and the absence of clinical guidelines to direct the safest and most effective use in cancer populations at the time of conducting the survey (Corroon et al. 2019). When comparing our results to those available in the published literature to date, practicing oncologists were more likely to indicate no preference or a lack of knowledge on a preferred mode of medical cannabis use (Sannes et al. 2021). Further, 15.4% of practicing oncologists reported not supporting medical cannabis use in their practice compared to just 2.7% of trainees in this study—again, recapitulating the notion that discussions surrounding cannabis use are increasing amongst younger generations of our work force (Jacobs et al. 2022; Weisman and Rodríguez 2021).

In this analysis, we found that prior training in medical cannabis was associated with having an articulated opinion on the comparative efficacy, risks, modes of use, and populations most likely to benefit from medical cannabis. However, in most cases training did not significantly influence whether those opinions were favorable or unfavorable. Interestingly, simply having an articulated view—particularly regarding efficacy and mode of use—was associated with higher rates of recommending medical cannabis. A similar, though nonsignificant, trend was observed for safety-related opinions, suggesting it may also be a critical domain to target in education. For instance, participants with an articulated opinion on cannabis efficacy recommended it more frequently, regardless of whether they viewed it positively or negatively. This underscores the importance of targeted education that promotes informed, nuanced understanding of cannabis in oncology care.

It is important to note that the evidence base for medical cannabis use in oncology varies by indication, with relatively stronger support for chemotherapy-induced nausea and vomiting, more limited data for chronic pain and cachexia, and little to no evidence for symptoms such as diarrhea (Braun et al. 2024a, b; Worster et al. 2022). Given the growing evidence on cannabis-related side effects and drug interactions, training should also equip fellows to assess safety risks and counsel patients appropriately, including those using cannabis recreationally (Braun et al. 2024a, b). This includes specific domains such as drug interactions, potential interactions with cancer directed therapy such as immunotherapy, hepatotoxicity and side effects (Nachnani et al. 2024; Woerdenbag et al. 2023). There is interest in the implementation of evidence based medical cannabis curriculum (Crowley et al. 2024). Attempts to do so during oncology fellowship training suggest this is both feasible and well received (Patell et al. 2025). Our results provide valuable insights regarding specific knowledge domains that may be most likely to impact recommendations for medical cannabis use and thus prioritized during the curriculum development process: comparative efficacy and modes of administration. It is critical for curricula designed for trainees in oncology thus highlight areas of certainty/uncertainty so as to facilitate truly informed patient-provider counseling and decision-making.