Modeling the Survival of Stomach Cancer Patients in Meru County using The Stratified Cox Model

Abstract

Cancer is a disease that can affect anyone regardless of age, social-economic status, or sex. Research has shown there are over 28,000 new cancer cases in a year in Kenya, with a mortality of 22,000, implying a 78.5% chance that the victims do not survive. If not detected early, treated on time, and the right treatment chosen, cancer treatment is less likely to succeed, reducing the chances of survival. One of the most common types of cancer is stomach cancer. It is also the most prevalent cancer in Meru County. The purpose of the study was to find the relationship between the various treatment methods and the survivorship of stomach cancer patients. By doing so, patients and health workers can select the best treatment for cancer patients at different stages. The study modeled the survival of stomach cancer patients using the Stratified Cox model in the case of Meru County, Kenya. The study's general objective was to model the survival of stomach cancer patients in Meru County using the Stratified Cox model. The data was first fitted in a Stratified Cox model to do this. Then hazard functions were determined. From the hazard functions, hazard rates were calculated using R version 4.3.1. Chemotherapy was used as a reference category. The study used secondary data obtained from Meru General Hospital between 2017 and 2021. Different treatment methods: radiotherapy, chemotherapy, hormone therapy, and surgery are compared for each stage while considering several demographic characteristics such as age and sex. The research investigated the hazard rates that, in turn, helped find the survival of patients with stomach cancer based on the treatment method used. Hazard ratios were obtained from the collected data to determine and recommend the best treatment method at a particular stage of stomach cancer. After analysis, results showed that surgery is the best treatment for stage 1 and 2 cancer, while radiotherapy and chemotherapy are the best for stage 3 and 4, respectively. Notably, patients below 50 have higher survival rates than those above 50. It was also noted that women have higher survival rates than men. The three objectives were met, where the first objective involved fitting the data into the model. Hazard functions were formed, and the hazard rates were calculated using the coefficients from the hazard functions.   Based on the objectives, it was recommended that modeling data after combining several treatments should be done. Also, the survivorship of patients after combining treatments should be found and compared with the survivorship after using one treatment at a time. Lastly, since herbal treatment is becoming a common treatment, enough data should be corrected and the treatment compared with other treatment methods.

Keywords: Stomach Cancer, Stratified Cox model, Hazard ratio/rates, chemotherapy, radiotherapy, hormone therapy, surgery