RT-HaND_C: A Multi-Source, Validated Real-world Head and Neck Cancer Dataset for Research

Area of Science:

• Oncology
• Data Science
• Medical Informatics

Background:

• Real-world data (RWD) offers valuable insights into head and neck cancer (HNC) patient outcomes, particularly for diverse and comorbid populations often excluded from clinical trials.
• Challenges in RWD quality necessitate rigorous evaluation for reliable real-world evidence generation.
• Developing comprehensive HNC oncology datasets is crucial for advancing research.

Purpose of the Study:

• To develop a large-scale, high-quality HNC oncology dataset integrating multiple data sources.
• To establish a robust evaluation framework for RWD in HNC research.
• To assess the usability of the developed dataset by investigating long-term weight trends post-radiotherapy.

Main Methods:

• The RT-HaND_C dataset was created by integrating structured and unstructured Electronic Health Record (EHR) data, alongside manually curated information.
• Utilized a validated AI-driven Natural Language Processing tool for extracting unstructured EHR data.
• Incorporated extensive demographic, disease, treatment, outcome, and radiotherapy dosimetry data from 2010-2023, with rigorous data quality assessments.

Main Results:

• The RT-HaND_C dataset comprises 2,895 HNC patients with over 1.9 million data points and >2000 data categories.
• Achieved >98% accuracy for most variables, with high data completeness and consistency across key categories.
• Demonstrated statistically significant, persistent weight loss in HNC patients up to 5 years post-radiotherapy, peaking at 6 months.

Conclusions:

• RT-HaND_C is a novel, high-quality real-world data resource with an integrated evaluation framework for HNC research.
• The dataset facilitates multi-modal research through virtual linkage with imaging data.
• RT-HaND_C is available for research collaborations, with ongoing efforts to enhance its completeness and incorporate prospective data.