Investigation of Prognostic Factors of Survival in Cancer using Cox Proportional Hazard Model and Fitting a Parametric Distribution

Bangladesh is a developing nation that has several difficulties, particularly in the medical field. Several quantitative studies on the key elements influencing cancer patients' survival times. This research describes the present cancer incidence situation in Khulna and how it affects the population in relation to many possible variables. The goal of the current research is to determine the distribution of lifespan and the significant variables that influence cancer patients' chances of surviving. A sample of 263 cancer patients was selected at random, and questions on demographic traits, past medical histories, and clinical features were posed to them. The primary statistical methods used to determine how the variables in this research affected the survival time of the cancer patients were the Cox Proportional Hazard model, Kaplan-Meier test, log-rank test, verification of the PH assumption, and optimal fitting of a parametric distribution. Our investigation revealed that the following factors significantly affect survival time: gender, smoking habit, and cancer stage. In addition to increased hazards, smoking increased the probability of mortality for patients who advanced to stage IV. Additionally, we discovered that the best-fitting model for a cancer patient's survival time was the log-normal distribution.

• Supervision