Alberto Caballero-Vázquez1, José Luis Romero-Béjar2, Luis Albendín-García3, Nora Suleiman-Martos4, José Luis Gómez-Urquiza3, Gustavo Raúl Cañadas5, Guillermo Arturo Cañadas-De la Fuente3. 1. Diagnostic Lung Cancer Unit, Broncopleural Techniques and Interventional Pulmonology Departament, Hospital Universitario Virgen de las Nieves, 18014 Granada, Spain. 2. Department of Statistics and Operational Research, University of Granada, Avda Fuentenueva, E-18071 Granada, Spain. 3. Faculty of Health Sciences, University of Granada, 18016 Granada, Spain. 4. Faculty of Health Sciences, Campus Universitario de Ceuta, University of Granada, 51001 Ceuta, Spain. 5. Department of Didactics of Mathematics, Faculty of Educational Sciences, University of Granada, 18011 Granada, Spain.
Abstract
BACKGROUND: Lung cancer is typically diagnosed in an advanced phase of its natural history. Explanatory models based on epidemiological and clinical variables provide an approximation of patient survival less than one year using information extracted from the case history only, whereas models involving therapeutic variables must confirm that any treatment applied is worse than surgery in survival terms. Models for classifying less than one year survival for patients diagnosed with lung cancer which are able to identify risk factors and quantify their effect for prognosis are analyzed. METHOD: Two stepwise binary logistic regression models, based on a retrospective study of 521 cases of patients diagnosed with lung cancer in the Interventional Pneumology Unit at the Hospital "Virgen de las Nieves", Granada, Spain. RESULTS: The first model included variables age, history of pulmonary neoplasm, tumor location, dyspnea, dysphonia, and chest pain. The independent risk factors age greater than 70 years, a peripheral location, dyspnea and dysphonia were significant. For the second model, treatments were also significant. CONCLUSIONS: Age, history of pulmonary neoplasm, tumor location, dyspnea, dysphonia, and chest pain are predictors for survival in patients diagnosed with lung cancer at the time of diagnosis. The treatment applied is significant for classifying less than one year survival time which confirms that any treatment is markedly inferior to surgery in terms of survival. This allows to consider applications of more or less aggressive treatments, anticipation of palliative cares or comfort measures, inclusion in clinical trials, etc.
BACKGROUND:Lung cancer is typically diagnosed in an advanced phase of its natural history. Explanatory models based on epidemiological and clinical variables provide an approximation of patient survival less than one year using information extracted from the case history only, whereas models involving therapeutic variables must confirm that any treatment applied is worse than surgery in survival terms. Models for classifying less than one year survival for patients diagnosed with lung cancer which are able to identify risk factors and quantify their effect for prognosis are analyzed. METHOD: Two stepwise binary logistic regression models, based on a retrospective study of 521 cases of patients diagnosed with lung cancer in the Interventional Pneumology Unit at the Hospital "Virgen de las Nieves", Granada, Spain. RESULTS: The first model included variables age, history of pulmonary neoplasm, tumor location, dyspnea, dysphonia, and chest pain. The independent risk factors age greater than 70 years, a peripheral location, dyspnea and dysphonia were significant. For the second model, treatments were also significant. CONCLUSIONS: Age, history of pulmonary neoplasm, tumor location, dyspnea, dysphonia, and chest pain are predictors for survival in patients diagnosed with lung cancer at the time of diagnosis. The treatment applied is significant for classifying less than one year survival time which confirms that any treatment is markedly inferior to surgery in terms of survival. This allows to consider applications of more or less aggressive treatments, anticipation of palliative cares or comfort measures, inclusion in clinical trials, etc.