Application of survival analysis techniques on lung cancer patient data in the Philippines with censored observations

• Main Authors: Arthur, John Alexander O., Chua, Tyrone Justine S.
• Format: text
• Language: English
• Published: Animo Repository 2010
• Subjects: Lungs—Cancer > Philippines > Mathematical models; Mathematics
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/5340
• Institution: De La Salle University

Lung cancer is the leading cause of cancer-related deaths in the Philippines, though little is known about the mechanism of death among people inflicted by the disease in the Philippines. Making inferences from a sample of lung cancer patients will be helpful in studying the timing and occurrence of death among newly-admitted lung cancer patients. A problem with sample data for longitudinal studies in the field of medicine is the loss to follow-up by some patients under observation which is a form of "censoring", but censoring does not hinder researchers from coming up with statistics for estimating and making inferences about survival probabilities and hazards. Survival analysis techniques that handle censored times of death were applied to a sample of newly-admitted lung cancer patients in the period 2000 to 2005 from the Lung Center of the Philippines. Kaplan-Meier estimation, one of these survival analysis techniques, provided quantile values for the survival time from admission of lung cancer patients. Complementary to the Kaplan-Meier estimation is the long rank test which tests the significance of survival distributions between strata (i.e. by sex). Cox regression analysis provided the magnitude and direction of covariates in contributing to the hazard of death for newly-admitted lung cancer patients in terms of hazard ratios. From these survival analysis techniques, inferences regarding the survival of newly-admitted lung cancer patients were made.