Types of statistical analysis in prognostic studies




Mario E. Rendón-Macías, Faculty of Health Sciences, School of Medicine, Universidad Panamericana, Mexico City, Mexico
Ana S. Castillo-Ivón, Faculty of Health Sciences, School of Medicine, Universidad Panamericana, Mexico City, Mexico
Lorena Orozco-Díaz, Faculty of Health Sciences, School of Medicine, Universidad Panamericana, Mexico City, Mexico


Prognostic studies may have a descriptive exploratory objective on an outcome or a comparative objective in the search for factors associated with it. A second objective is explanatory to determine the effect of a particular prognostic factor adjusted for its confounders, with or without the intention of establishing causality. The third objective is the construction of a predictive prognostic scale. For each of these objectives, there are recommended statistical methods for clarification and validity. In this article, the methods and application examples are presented. The proper selection of analytical methods allows for clear and valid communication of the results of a prognostic study.



Keywords: Statistical analysis. Prognosis. Validity.