PURPOSE: To validate a predictive model for survival of patients attending a palliative radiotherapy clinic. METHODS AND MATERIALS: We described previously a model that had good predictive value for survival of patients referred during 1999 (1). The six prognostic factors (primary cancer site, site of metastases, Karnofsky performance score, and the fatigue, appetite and shortness-of-breath items from the Edmonton Symptom Assessment Scale) identified in this training set were extracted from the prospective database for the year 2000. We generated a partial score whereby each prognostic factor was assigned a value proportional to its prognostic weight. The sum of the partial scores for each patient was used to construct a survival prediction score (SPS). Patients were also grouped according to the number of these risk factors (NRF) that they possessed. The probability of survival at 3, 6, and 12 months was generated. The models were evaluated for their ability to predict survival in this validation set with appropriate statistical tests. RESULTS: The median survival and survival probabilities of the training and validation sets were similar when separated into three groups using both SPS and NRF methods. There was no statistical difference in the performance of the SPS and NRF methods in survival prediction. CONCLUSION: Both the SPS and NRF models for predicting survival in patients referred for palliative radiotherapy have been validated. The NRF model is preferred because it is simpler and avoids the need to remember the weightings among the prognostic factors.
PURPOSE: To validate a predictive model for survival of patients attending a palliative radiotherapy clinic. METHODS AND MATERIALS: We described previously a model that had good predictive value for survival of patients referred during 1999 (1). The six prognostic factors (primary cancer site, site of metastases, Karnofsky performance score, and the fatigue, appetite and shortness-of-breath items from the Edmonton Symptom Assessment Scale) identified in this training set were extracted from the prospective database for the year 2000. We generated a partial score whereby each prognostic factor was assigned a value proportional to its prognostic weight. The sum of the partial scores for each patient was used to construct a survival prediction score (SPS). Patients were also grouped according to the number of these risk factors (NRF) that they possessed. The probability of survival at 3, 6, and 12 months was generated. The models were evaluated for their ability to predict survival in this validation set with appropriate statistical tests. RESULTS: The median survival and survival probabilities of the training and validation sets were similar when separated into three groups using both SPS and NRF methods. There was no statistical difference in the performance of the SPS and NRF methods in survival prediction. CONCLUSION: Both the SPS and NRF models for predicting survival in patients referred for palliative radiotherapy have been validated. The NRF model is preferred because it is simpler and avoids the need to remember the weightings among the prognostic factors.
Authors: Gerald M Y Quan; Jean-Marc Vital; Nicholas Aurouer; Ibrahim Obeid; Jean Palussière; Abou Diallo; Vincent Pointillart Journal: Eur Spine J Date: 2011-06-26 Impact factor: 3.134
Authors: Birgitt van Oorschot; Michael Schuler; Anke Simon; Ursula Schleicher; Hans Geinitz Journal: Strahlenther Onkol Date: 2011-07-22 Impact factor: 3.621
Authors: Vincent Pointillart; Jean-Marc Vital; Rachid Salmi; Abou Diallo; Gerald M Quan Journal: J Cancer Res Clin Oncol Date: 2010-09-04 Impact factor: 4.553
Authors: Maryam Dosani; Scott Tyldesley; Brendan Bakos; Jeremy Hamm; Tim Kong; Sarah Lucas; Jordan Wong; Mitchell Liu; Sarah Hamilton Journal: Support Care Cancer Date: 2018-02-01 Impact factor: 3.603