Background

: cancer is one of the leading causes of death worldwide. The identification of the prognostic factors involved in its evolution is of great interest for the secondary and tertiary prevention of this frequent entity.

Objective

: to identify the prognostic factors of cervical cancer mortality and evaluate the usefulness of implicative statistical analysis in the clinical-epidemiological investigations of causality.

Methods

: study of cases and controls in 106 women with the clinical and histological diagnosis of cervical cancer treated at the Oncology Hospital of the province Santiago de Cuba from January 2016 to January 2018. The implicative statistical analysis was applied together with the binary logistic regression, which was considered as a gold standard.

Results

: binary logistic regression identified the presence of metastases as the only poor prognostic factor (OR=2.995) and chemotherapy (OR=0.151) and event-free interval (OR=0.602) as good prognostic factors. Implicative statistical analysis detected these same factors, but also identified other poor prognostic factors such as complications, advanced stage, and relapse; all with a 99% of implication intensity.

Conclusions

: this technique reached a high sensitivity and specificity when compared with the regression and was useful for the identification of prognostic factors.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )