Usefulness of statistical implicative analysis to identify prognostic factors in patients with breast cancer

Abstract

Background: identifying prognostic factors in patients with breast cancer continues to be a necessity.Objective: to assess the usefulness of statistical implicative analysis to identify the prognostic factors which contribute to the progress of the patients with breast cancer in the province of Santiago de Cuba.Methods: an analytical observational study was conducted with women of the aforementioned province, with the clinical and histological diagnosis of breast cancer, treated at the "Conrado Benítez" Oncological Hospital, from September 2014 to April 2018. Two groups were formed, balanced this way: the cases were the deceased and the controls were the ones alive at the moment of the selection of the sample, by simple random sampling. Patients with concomitant disease that could interfere with the validity of the results were excluded. As a dependent variable the living or deceased state was taken and as explanatory variables the prognostic factors indicated by the literature and by the consulted experts. The analysis was carried out by logistic regression and statistical implicative analysis.Results: the logistic regression identified as factors of poor prognosis: nuclear grade III-IV, TNM classification III-IV, presence of metastasis and absence of histiocytosis. Premenopause was identified as a good prognostic factor. Those identified by SIA were: metastasis, nuclear grade of poor prognosis, absence of surgical treatment, TNM of poor prognosis and not receiving hormone therapy.Conclusions: the study showed that the statistical implicative analysis is a useful tool in the identification of the prognostic factors that affect the progress of patients with breast cancer.

Keywords

• prognosis
• breast neoplasms
• logistic models
• data interpretation, statistical