• Posted On: January 30, 2025
• Posted By: Improzo

The oncology landscape is rapidly changing, with a significant surge in the number of approved therapies. In the past five years alone, 60+ new oncology drugs have been launched, intensifying competition among biopharmaceutical companies to effectively identify and engage specific patient segments that could benefit from these treatments. However, the challenge remains: how can companies accurately pinpoint the right patients at the right time amidst narrow drug labels and small patient populations? To navigate this complexity, it is essential to establish a robust analytics capability that predicts key events in the oncology patient journey. This capability not only assists your sales and marketing teams in targeting oncologists likely to treat eligible patients but also empowers patient support programs to anticipate potential drop-offs and proactively assist those patients. Additionally, it enables medical teams to identify optimal clinical trial sites that align with patient eligibility criteria.

Developing a predictive capability for oncology patient journeys can be structured into four essential steps:

1. Mapping the Oncology Patient Journey
   The first step involves carefully mapping out the oncology patient journey while prioritizing critical events that impact brand success. For instance, predictive modeling should focus on identifying patients who are transitioning between therapy lines, especially for drugs indicated for first or subsequent lines of treatment. Utilizing established guidelines like those from the National Comprehensive Cancer Network (NCCN) can provide a solid foundation for this mapping process. Engaging internal experts and conducting research within oncology practices will further refine your understanding of patient pathways.
2. Integrating Real-World Patient Data
   The next step is to acquire and integrate various real-world data sources, such as claims data, electronic medical records (EMR), lab test results, and specialty pharmacy information. A systematic data strategy is crucial for curating and connecting these sources to effectively support predictive modeling capabilities. Your team should evaluate factors such as data richness, granularity, capture rates, refresh rates, acquisition costs, and connectivity with other data sources. Since a single comprehensive patient-level data source may not be available, it may be necessary to make trade-offs across these dimensions while initially developing predictive capabilities using one or two primary data sources.
3. Developing Predictive Algorithms
   The third step involves creating and testing algorithms designed to predict key events along the oncology journey. Your team can utilize retrospective modeling on real-world data to develop various algorithms while paying close attention to changes in therapy lines across different cancer types. Considerations such as evolving treatment pathways, variability in oncologist practices, and gaps in real-world data must be considered during this process. Collaborating with sales, marketing, and medical teams will enhance algorithm accuracy and foster buy-in for adopting these predictive models. Increasingly, machine learning algorithms are being employed alongside robust feedback mechanisms to create adaptive predictive models that improve over time.
4. Establishing a Scalable Analytics Platform
   The final step is implementing a scalable analytics platform that facilitates both the development of predictive models and the dissemination of insights to relevant stakeholders. Key features of an effective analytics platform include a comprehensive patient data management module for integrating real-world data sources, an intelligent business rules engine for creating longitudinal patient cohorts with diverse eligibility criteria, an analytics workbench for model development using machine learning techniques, and a business insights consumption module for sharing predictive insights across sales, marketing, patient support, and medical teams.

By strategically combining these four components—mapping the patient journey, integrating real-world data, developing predictive algorithms, and establishing a scalable analytics platform—your biopharma team can significantly enhance its capabilities in predicting oncology patient journeys. Organizations that successfully implement these strategies are likely to receive positive feedback from their field sales teams regarding the timely availability of predictive insights. This capability not only allows sales teams to connect with busy oncologists at critical moments but also ensures that effective therapies reach the right cancer patients when they need them most. In an era where precision medicine is paramount, developing an oncology patient journey prediction capability presents a transformative opportunity for your emerging biopharma team to optimize market strategies and improve patient outcomes.