2477 - Predicting Decreased Quality of Life in Cancer Survivors Undergoing Radiotherapy Using a Random Forest Model

Purpose/Objective(s): Radiotherapy aims to prolong survival and alleviate symptoms, with a strong focus on improving the quality of life (QoL) for cancer patients. However, QoL often declines during and after treatment, and there are limited predictive methods to identify this decline. This study aimed to develop a machine learning-based random forest model to predict decreased QoL in cancer survivors undergoing radiotherapy.

Materials/Methods: A total of 1,436 participants were prospectively enrolled in this cross-sectional study at a cancer center in mainland China from May 2023 to October 2023. Participants were randomly divided into training and validation cohorts in an 8:2 ratio. The European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-C30) was used to assess QoL, with a score below 60 points indicating decreased QoL. The M. D. Anderson Symptom Inventory assessed the symptoms experienced by participants. A random forest model was developed to predict the occurrence of decreased QoL in cancer survivors treated with radiation.

Results: The average QoL score for participants was 79.82 ± 0.511. Approximately 19.37% (233 out of 1,203) of participants experienced decreased QoL. Pain and fatigue were the most frequently reported symptoms, while social functioning dysfunction was noted on the functional scales. The top ten factors affecting QoL were pain, fatigue, nausea, distress, depression, work impairment, shortness of breath, type of cancer, insomnia, and numbness. The random forest model demonstrated an area under the curve of 1.0 for the training cohort and 0.814 for the validation cohort.

Conclusion: The random forest model used in this study exhibited a high level of accuracy in predicting decreased QoL in cancer survivors treated with radiotherapy, providing valuable insights for early intervention.