On Predicting Survival Opportunities for Prostate Cancer by COX Regression in PSU Patients Data

Normally, prostate cancer disease causes no symptoms and grows slowly. The patient does not feel any abnormalities in the body along with their routine activities. The severity of the incidence, the number of new patients and the mortality rate which is likely to increase gradually. For this reason, survival analysis of prostate cancer patients is helpful to be used as a guideline to improve for a doctor regarding medical decision making by data science model in the medical treatment process. Therefore, this research study is interested to study the survival rate and its related factors of prostate cancer patients. Prostate cancer features such as age, family history, Prostate-Specific Antigen (PSA), Gleason group, Gleason score, Gleason pattern, and pathology report are suitable for predicting survival opportunities. This research is a cohort study of patients aged 40-89 years old. The data is collected from Songkhlanagarind Hospital between 2015-2018 following the Good Clinical Practice (GCP) protocols. COX regression is used to analyzed patients' survival opportunities from prostate cancer. The result shows from 78 prostate cancer patients by 22 patients in 2015, of which 63.636% were alive and 36.364% dead. In 2016 there were 18 patients who were still alive. 77.778% and 22.222% dead. In 2017, there were 20 patients who were still alive. 75.000% and 25.000% dead. And in the year 2018, there were 18 patients who were still alive. 88.889% and 11.111% dead. This research will be developed to add specific treatment features or add real-time events to show accurate forecasting results with the graphic user interface which is friendly to the doctor.