On the basis of all the findings and conclusions, this Delphi study defined the content and layout of a CDSS for delivering exercise programs to cancer survivors, addressing disparities in access to supportive care for cancer survivors. To our knowledge, this study represents the first development of a CDSS for exercise oncology worldwide. This work directly responds to calls from the broader exercise oncology community, most notably the Exercise Oncology Knowledge Mobilization Initiative [17], which identified the accessibility of cancer exercise programs as a top priority among international stakeholders, by translating those priorities into a structured, operational decision support framework.

The core components and structure of the system, which was initially designed to enhance exercise prescription in Switzerland, are readily adaptable for implementation in other countries with minimal modifications, owing to the expertise of international specialists on the steering committee and Delphi panel. The modular and transparent development process enables extensions and adaptations, ensuring the transferability of the system across diverse global contexts and its alignment with emerging evidence. This work bridges the gap between research and real-world practice, equipping health professionals with practical tools to deliver safe, effective, and personalized exercise interventions for cancer survivors in diverse clinical settings.

Professional medical advice (e.g., from physicians or nurses) significantly increases patient engagement with and adherence to exercise regimens during cancer therapy [18]. The list of referral indications defined in this study enhances the clinical workflow by structuring and supporting the referral process with evidence.

An indication can be defined as the reasonable professional judgment that a medical procedure is suitable and useful for achieving a specific therapeutic goal with a certain probability [19]. Unlike our major reference works [1, 2], in this study, in alignment with our preparatory scoping review, we chose to group outcomes under distinct treatment indications, which we validated through the consensus process. In contrast to the “outcome” concept, the “indication” concept always supports treatment intentionality. For example, the “Physical Function” outcome does not necessarily imply the need for intervention, whereas the “Impaired Physical Function” indication implies a need for intervention and justifies the costs covered by insurance. The concept of “indication” is central to the CDSS architecture. It allows for the grouping of outcomes from the literature and is well suited for the development of decision trees or taxonomies, which are essential for process digitalization; this can be observed in the development of a CDSS, since it is classified in the Systematized Nomenclature of Medicine—Clinical Terms (SNOMED) [20] with the concept ID 432678004, thus contributing to the simplification and standardization of terminology [21].

Exercise across different phases of the patient journey has been an important issue for clinicians and researchers for two main reasons: to prevent possible harm from exercise and to detect indications for exercise arising from the physiological effects of medical or surgical treatments [22,23,24,25]. With respect to exercise prescription, both the type and timing of medical treatment in relation to the exercise oncology program are relevant factors. In our study, we decided to combine these two factors into the concept of “treatment status,” which encompasses both. This concept is particularly appropriate for organizing findings from published evidence according to treatment indications, as randomized controlled trials (RCTs) in the field of exercise oncology frequently reference treatment status in their titles or inclusion criteria (e.g., during chemotherapy, during radiotherapy, after mastectomy, before lung resection, and during endocrine treatment).

In the context of health care, the principle of profitability refers to health-related outcomes achieved relative to costs rather than financial profit alone. This principle reflects the extent to which patients benefit from interventions, making value the core concept [26]. In terms of value determination, intervention costs serve as the denominator, whereas intervention effects constitute the numerator. These effects should be measured in a patient-centered manner, ideally using patient-reported outcome measures (PROMs) [27]. However, to accurately evaluate the effects of clinical interventions, the selected measurement instruments must possess appropriate measurement properties for evaluative purposes [28]. In particular, they must be capable of detecting clinically important changes and demonstrating validity. In this Delphi study, we responded to the participants’ suggestions to primarily use PROMs to evaluate the efficacy of exercise interventions. We recommend a set of instruments from the EORTC Quality of Life Group [29], as these instruments demonstrate adequate clinical interpretability and validity for the German-speaking population [30,31,32,33].

The development of exercise prescriptions through consensus roundtables has established the importance of structured and personalized programs for cancer survivors [34]. Advances in exercise oncology have led to updates in the field [1, 2], which discuss the level of evidence regarding the efficacy of structured exercise prescriptions in improving specific outcomes. In this study, we described structured prescriptions not only on the basis of FITT principles but also on the incorporation of two additional major exercise principles: progression and overload. Although these principles are acknowledged in the literature [35], to our knowledge, this study is the first to integrate them into a comprehensive exercise oncology CDSS, thus improving replicability by describing prescriptions in a more dynamic manner and leveraging the opportunities offered by digitalization, which enables the management and monitoring of more complex prescriptions to enhance customization and personalization.

Safety considerations for exercise in patients with specific cancer-related conditions are available [1, 3] and can be readily integrated into clinical pathways. However, patients with advanced cancer or those undergoing cancer treatments may present with a more dynamic clinical picture [24, 36]. This necessitates a focus on additional considerations and the development of further skills to prevent adverse events or to enable their early recognition. In this study, available guidelines, relevant literature, and the perspectives of clinical experts were considered in the definition of safety considerations for the CDSS. The resulting content is based on three pillars: the early recognition of oncological emergencies, identification of exercise contraindications, and tailoring of exercise programs for patients with osseous fragility.

This Delphi study, while among the first to develop a CDSS for exercise oncology, faced several notable limitations. A significant challenge was the moderate acceptance rate of exercise prescription recommendations, with nearly half of the 112 proposed recommendations being declined. These recommendations were selected on the basis of a predefined evidence-based criterion, requiring recommended interventions to demonstrate statistically significant within-group improvements for the outcome of interest. The relatively high proportion of “declined” recommendations likely reflects two factors. First, feasibility considerations played a central role, particularly regarding intervention frequency, duration, and intensity, which participants frequently evaluated against real-world clinical constraints such as patient burden and reimbursement policies. Second, the complexity of cancer populations, including their comorbidities and treatment-specific considerations, may have reduced the perceived generalizability of some recommendations. Together, these challenges highlight the persistent gap between evidence generation and implementation in exercise oncology. The inability to revise and resubmit declined recommendations for further Delphi rounds, owing to adherence to the predefined criterion, restricted our capacity to refine proposals while maintaining methodological consistency. Despite this, we identified and described emerging trends in participant feedback, offering valuable insights for future iterations. Additionally, many interventions considered under the Delphi-defined criterion have been inadequately reported, particularly those lacking details on the FITT principles. Incomplete reporting, such as unspecified exercise types or frequencies, contributed to a higher disapproval rate among the participants. This limitation underscores the need for more comprehensive documentation in future studies to enhance the robustness and acceptance of proposed exercise prescriptions.

Future research should facilitate the transition from knowledge-based CDSSs, such as the one developed in this study, to AI-based CDSSs for exercise oncology informed by continuous real-world data. This perspective extends beyond the direct findings of the Delphi process and reflects a broader outlook on the future development of CDSSs in exercise oncology. A critical step in this direction is establishing an internationally accepted ground truth, which requires robust consensus processes among global stakeholders [37]. This process will challenge exercise oncology stakeholders, as it demands active engagement in defining standardized labeling criteria (e.g., categorizing treatment status or prescription indications) and annotating data (e.g., linking chemotherapy phases to treatment status or fatigue to specific exercise prescriptions). Despite these challenges, the growing utility of large language models underscores their potential to support evidence-informed, complex decision-making in exercise oncology. Moreover, leveraging real-world data to train AI-based CDSSs could address the evidence gaps identified in this study, such as the incomplete reporting of FITT principles, and potentially reshape our understanding of the efficacy of exercise prescriptions for diverse clinical indications.