[Design and analysis of a prognosis study in urological clinical research].

Prognosis is a description of the course of a disease from the beginning. In comparison to risk factors, prognostic factors are relatively frequent and may often be estimated by our personal clinical experience. Nevertheless, the cases of diseases usually cared for at hospitals and described in the medical bibliography are often biased samples and have the tendency to overestimate severity. Cohort studies imply follow-up of the groups of individuals over time. They are the observational studies with the highest acceptance within the scientific community, because they include the target population in the study. If we look for providing solid information, the observation of the members of a cohort, independently of what they have in common, should comply with two criteria: 1. Cohorts should be observed for a significant period of time of the natural history of the event studied. 2. All members in a cohort should be observed for the full follow-up period. As in any cohorts observation, studies comparing prognosis among various groups of patients may be blased if differences appear due to cohort recruitment methods, patients shifting from their initial groups, and unequal results evaluation.