SEMI-PARAMETRIC ANALYSIS OF CERVICAL CANCER DATA: A SINGLE-CENTRE STUDY

Abstract

In women, cervical cancer ranks fourth in terms of mortality and is the third most prevalent disease. It remains to be a great concern among clinicians in Malaysia, yet published works on the survival of cervical cancer patients are somewhat scarce. Thus, this study aims to identify the prognostic factors that significantly affect the risk of death of cervical cancer patients using the semi-parametric Cox proportional hazards regression analysis. This study commenced with the univariate and multivariate Cox proportional hazards regression analyses, followed by the proportional hazards assumption test for the preliminary final model. Data on cervical cancer patients treated at Hospital Universiti Sains Malaysia (HUSM) between 2013 and 2017 was utilised. In the univariate analysis, stage at diagnosis and primary treatment were found to be statistically significant at the 5% level of significance. In the multivariate analysis, histologic type, stage at diagnosis, and distant metastasis were found to be statistically significant. The proportional hazards assumption for each variable in the preliminary final model is tested based on the scaled Schoenfeld residuals. Accordingly, this study showed that patients with stage III–IV adenocarcinoma-type cervical cancer treated at HUSM have the highest likelihood of death from the disease.