In randomized or comparative studies, when the outcomes are binary or dichotomous, the effect of a specific treatment can be reported using the absolute risk reduction (ARR) and the number needed to treat (NNT), which is the reciprocal of the ARR (1/ARR = NNT). The objective of the present study was to realize a review of the different modalities of treatment of primary immune thrombocytopenia (ITP), using as effect measurement the calculation of the ARR and NNT and their confidence intervals (CI 95%). The number needed to harm (NNH) can be calculated with the same formula of NNT, taking in account only the adverse events (CTCAE scale) of the treatment in relation with those in the control group. The results showed the effect of different types of treatment of ITP. The NNT was better in randomized studies than those of inferior design. The NNH calculation showed the safe level of the intervention. It can be observed that age (youth) and no splenectomy condition exhibited some influence in the favorable NNT report. In conclusion, given the advantages of the ARR and the NNT for clinical decision making, it can be suggested that these measurements of effect should also be reported, in addition to other statistical measurements for ITP treatment or any observational study with dichotomous or binary outcomes.