Prostatectomized patients are at risk of resurgence: this is the reason why, during a follow-up period, they are monitored for PSA growth, an indicator of tumor progression. The presence of tumors can be evaluated with an expensive exam, called PET-PSMA (Positron Emission Tomography with Prostate-Specific Membrane Antigen). But, to optimize the benefit/risk ratio, patients should be referred to this exam only when the evidence is strong. The aim is to estimate the optimal time to recommend the exam, based on patients' history and collected data. We build a Hierarchical Bayesian model that describes, jointly, the PSA growth curve and the probability of a positive PET-PSMA. Our proposal is to process all information about the patient in order to give an informed estimate of the optimal time.