To optimize palliative care in patients with cancer who are in their last year of life, timely and accurate prognostication is needed. However, available instruments for prognostication, such as the surprise question and various prediction models using clinical variables, are not well validated or lack discriminative ability. We prospectively included cancer patients from six hospitals in the Netherlands to develop a clinical prediction model to predict 1-year mortality in patients with advanced cancer. In a relatively large cohort of 867 patients, we showed that an extended prediction model that combines the surprise question, clinical characteristics (age, cancer type, visceral metastases, brain metastases, performance status, weight loss, pain, and dyspnea), and laboratory values (hemoglobin, C-reactive protein, and serum albumin) has better discrimination (c-statistic 0.78) in predicting 1-year mortality than the surprise question (c-statistic 0·69), clinical characteristics (c-statistic 0·70), or laboratory values (c-statistic 0·71) alone. Additionally, our prediction model also showed better discrimination than other models in literature. We developed a nomogram and web-based calculator to calculate the 1-year mortality risk for individual patients with advanced cancer.

Date Submitted: 2022-03-12