Cancer Models and Real-world Data: Better Together

Area of Science:

• Health Services Research
• Biostatistics
• Public Health Policy

Background:

• Decision-analytic models are crucial for health policy, synthesizing data to predict long-term outcomes.
• Real-world screening practices in diverse US health systems are complex and challenging to model.
• Existing models often operate under ideal conditions, limiting their applicability to current screening practices.

Purpose of the Study:

• To describe the synergy between decision-analytic models and health care utilization data.
• To explore opportunities for enriching cancer screening models with real-world data.
• To improve the projection of harms and benefits of current cancer screening practices.

Main Methods:

• Leveraging health care utilization data from electronic medical record systems.
• Integrating real-world data to ground and enrich decision-analytic models.
• Utilizing the Population-based Research Optimizing Screening through Personalized Regimens (PROSPR) consortium as a case study.

Main Results:

• Real-world data can enrich cancer screening models, improving their accuracy and applicability.
• Models grounded in real-world data can evaluate new technologies and identify areas for improvement.
• The PROSPR consortium demonstrates a collaborative approach to harmonizing and analyzing screening data.

Conclusions:

• Pairing decision-analytic models with real-world data creates more robust models for policy and practice.
• Improved models can better inform health systems on optimizing cancer screening delivery.
• This approach enhances the ability to project harms, benefits, and value of cancer screening interventions.