Abstract
Financial toxicity is increasingly recognized as a critical barrier to high-quality cancer care, treatment adherence, and positive survivorship outcomes. Despite the availability of financial and social support resources, patients often face challenges in accessing and navigating them effectively. To address this gap, the University of Kansas Cancer Center developed the Oncology Navigation Empowerment for Treatment (ONCO-NET) platform—an innovative hybrid intervention that integrates an AI-driven resource-matching system with human-centered financial navigation support. Through a structured intake process that includes demographic information, the Functional Assessment of Chronic Illness Therapy Comprehensive Score for Financial Toxicity (FACIT-COST), and social determinants of health data, ONCO-NET identifies and prioritizes the most relevant local resources for each patient. The platform provides real-time, personalized recommendations while offering access to trained financial navigators for additional support. By combining digital technology with human expertise, ONCO-NET enhances accessibility, scalability, and equity in financial navigation. This approach has the potential to improve treatment adherence, reduce disparities, and serve as a scalable model for addressing financial burdens in cancer care and other chronic disease contexts.
Financial toxicity, which describes the economic distress experienced by cancer patients during and after treatment, has become increasingly recognized as a critical barrier to quality care, treatment adherence, and positive survivorship outcomes. 1 Financial toxicity is an immediate and pervasive issue for cancer patients, with over half experiencing catastrophic health expenditures during their care. 2 Over 20% of cancer survivors are estimated to have delayed or missed medical care in the past year because of cost-related barriers. 3 Traditional financial navigation programs aimed to mitigate these issues, though impactful, often suffer from scalability and accessibility limitations, especially in low-resource or rural settings. To address these gaps, we developed an innovative, in-house hybrid intervention that integrates intelligent technology with human-centered navigation services to reduce financial toxicity among cancer patients and survivors proactively. Building the platform internally has enabled us to retain control over the development timeline, adapt to institutional needs in real-time, and ensure seamless alignment with patient-centered design principles. This approach empowers patients to access high-quality, timely resources while overcoming structural and informational barriers that impact their economic well-being.
At the heart of this model is the Oncology Navigation Empowerment for Treatment (ONCO-NET) Resource Matching system, which introduces a personalized, algorithm-driven framework for matching patients with high-quality, relevant local resources tailored to their individual financial and social circumstances. Using the ONCO-NET online or mobile application, patients complete a structured intake process; this includes short surveys on demographics, the Functional Assessment of Chronic Illness Therapy Comprehensive Score for Financial Toxicity (FACIT-COST),
4
and social determinants of health (SDoH), which is the University of Kansas Health System’s modified version of the Health Leads SDoH toolkit
5
(Figure 1). The survey package includes 40 items (FACIT-COST: 11, demographics
6
: 16, SDOH screener: 13), with an estimated completion time of 5-8 minutes, a burden consistent with standard patient-reported outcome measures.
7
Each response is systematically scored and categorized, with responses informing a real-time resource ranking engine. For example, if a patient indicates housing insecurity, the algorithm prioritizes housing-related resources in the final list presented to the user. While users can view all resources tied to the platform, the ONCO-Net system highlights the top 5 most relevant resources, providing a streamlined and intuitive support experience that strives to not overwhelm users. Cancer patient resources were identified from the Kansas Cancer Partnership website
8
and compiled into a local dataset. This dataset was then validated by using R script to verify the listed website URLs were functional, when a nonfunctional URL was identified it was removed from the dataset. ONCO-NET Mobile Application
Technically, the project employs a decision tree model, a non-parametric machine learning approach that is ideal for classification and matching tasks. This choice strikes a balance between model interpretability and performance, enabling researchers and navigators to understand why specific resources are recommended, which sets ONCO-NET apart from current resource database search tools. Future versions of the system will incorporate deep learning approaches—particularly transformer-based architectures—to enhance complex pattern recognition and enable more tailored personalization. This forward-looking strategy lays the groundwork for a continuously learning system that adapts to the evolving needs of patients and the resource landscape.
What distinguishes this initiative the most is not just its technical sophistication but its thoughtful integration of human support and digital infrastructure. The model introduces 2 synchronized layers of assistance: (1) an automated, AI-powered resource matching algorithm that offers real-time recommendations based on patient input, and (2) access to trained financial navigators employed by the University of Kansas Cancer Center. Through the app, patients and caregivers can schedule an appointment with a navigator, who then conducts a secure Zoom video or phone consultation to review resources and provide financial optimization strategies tailored to the patient’s needs. The in-app chat function is designed primarily for scheduling and quick logistical support, while the more substantive, empathetic interactions occur during these consultations. In addition, navigators can proactively reach out to patients flagged as high-risk—such as those with elevated COST scores—through a dashboard dedicatedly built for Navigators to track patient journey, ensuring timely intervention even if the patient has not initiated contact. This hybrid approach ensures that patients receive data-driven recommendations while also preserving the empathy and contextual nuance that only human navigators can provide. There is also a separate dashboard interface designed for financial navigators and healthcare professionals to view patient survey results, therefore enabling efficient follow-up with patients identified as experiencing high financial distress (Figure 2). The prospective design of the navigator-facing interface involves flagging high COST scores to quickly identify and assist patients experiencing a high financial burden. While the current version of the app is available in English, we are actively working with our Hispanic community advisory group to develop a Spanish-language version. Spanish is the second most commonly spoken language among oncology patients at The University of Kansas Cancer Center, and this future iteration will ensure accessibility and cultural relevance for Spanish-speaking patients and caregivers. ONCO-NET Financial Navigator Dashboard (Email Addresses Have Been Censored)
Equity and accessibility are core design principles of this project. Recognizing that some patients may lack consistent access to smartphones or internet services, we have integrated alternative means of surveying and providing resources to ensure that financial navigation services remain accessible. These include phone-based support and clinic-based financial navigation. The ONCO-NET app complements existing clinic-based financial navigation by enabling social workers to assess patient financial risk scores and proactively connect high-risk patients to navigators for timely, personalized support via face-to-face or virtual consultations. This level of inclusiveness is crucial for ensuring that the intervention does not exacerbate existing disparities in access to cancer care.
While the current paper highlights the quantitative and clinical feasibility aspects of the application, patient and caregiver input is being used in parallel to guide refinements and maintain usability. Another innovative aspect of the intervention is its focus on novel financial outcomes that extend beyond traditional metrics, such as bankruptcy rates or out-of-pocket expenses. Instead, the project aims to quantify broader indicators of financial well-being through our surveys, such as changes in employment stability, housing security, food access, and healthcare avoidance behaviors. This comprehensive view reflects the complex interplay between financial stress and social context, particularly for patients dealing with long-term or late-stage cancers. The app simultaneously screens for financial toxicity and SDOH disparities, providing a comprehensive assessment of patient- and caregiver-level risk while complementing tools that focus on only 1 domain.
To enhance coordination between financial navigation (using technology) and clinical teams, we plan to integrate ONCO-NET survey results directly into the institutional EMR. Patients identified with high financial distress will be automatically flagged, triggering notifications to financial navigators and social workers for proactive follow-up. To maximize uptake, the tool will be incorporated into routine clinic workflows, with staff introducing the app during visits and guiding patients—regardless of language preference—through completion. This integration ensures that patients have access to tailored resources while enabling the care team to efficiently monitor engagement and provide timely support, ultimately improving adherence to appointments and continuity of care. The ONCO-NET initiative also has a unique, scalable opportunity. The prospective ability to collect and analyze longitudinal data on patient follow-up with the provided resources, as well as patient cost and quality of life outcomes, creates the potential for real-time program evaluation and continuous quality improvement. As the system matures, it could serve as a prototype for similar interventions in other chronic disease contexts, such as diabetes, cardiovascular disease, or Alzheimer’s care, where financial strain is a known barrier to treatment adherence and access.
Notably, the project aligns with national priorities in value-based care and health equity. Initiatives from organizations such as the National Cancer Institute (NCI), the American Society of Clinical Oncology (ASCO), and the Centers for Medicare & Medicaid Services (CMS) have all called for innovative approaches to mitigate the financial burden of cancer treatment. By combining artificial intelligence, user-centered design, and a rigorous mixed-methods evaluation, this project directly responds to those calls with a scalable, evidence-informed model.
In conclusion, this hybrid technology, combined with financial navigation intervention, offers a first-of-its-kind, integrative solution to a long-standing problem in oncology care. It exemplifies the best of what digital health can offer: timely, personalized, and equitable support for patients facing both medical and financial uncertainty. With its foundation in clinical informatics, artificial intelligence, and human-centered care, the ONCO-NET platform is well-positioned to make a meaningful impact not only on cancer outcomes but also on the broader goal of delivering compassionate, accessible healthcare to all.
Footnotes
Ethical Considerations
This study was approved by the University of Kansas Medical Center Institutional Review Board (Protocol number:
Consent to Participate
All participants provided written informed consent prior to participation, and the study was conducted in accordance with the Declaration of Helsinki and relevant institutional guidelines.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: National Cancer Institute Cancer Center Support Grant P30CA168524 supported this study. It also used the Biostatistics and Informatics Shared Resource and Masonic Cancer Alliance (MCA).
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
