Introduction - Effective symptom management is a critical component of cancer treatment. Computational tools that predict the course and severity of these symptoms can assist oncology clinicians to personalize the patient’s treatment regimen and provide more timely interventions. Cancer-related fatigue (CRF) is the most common symptom associated with cancer and its treatments. CRF has a negative impact on the patients’ ability to tolerate treatments and on their quality of life. One of the limitations to effective treatment of CRF is the availability of a valid and reliable model to predict the severity of CRF.

Objectives - To generate a model using demographic and clinical characteristics and CRF levels collected just prior to CTX administration (T1) to predict evening CRF severity 1 week after chemotherapy (CTX) administration (T2).

Methods - Support vector regression was used to predict the severity of the evening CRF at T2 in 689 patients (n=518 training, n=171 testing). Model fit was assessed with 10-times 10-fold cross-validation root-mean-square error (RMSE).  

Results - Our model achieved high accuracy (mean reported CRF score=5.7, mean predicted CRF score=5.6) and low prediction error (RSME/mean=0.269). Predictors with the highest importance were: morning and evening fatigue at T1, attentional function, sleep disturbance, and performance status. Predictors with the lowest importance were: living alone, caregiver to adult, level of education, cycle length, and number of metastatic sites. 

Conclusions - This type of predictive model can be used to identify high risk patients, educate patients about their symptom experience, and improve the timing of pre-emptive and personalized symptom management interventions.

01 / PREDICTING EVENING CANCER-RELATED FATIGUE SEVERITY IN ONCOLOGY PATIENTS ONE WEEK FOLLOWING CHEMOTHERAPY