Evidence-based evaluation of benefits in therapeutic interventions: Methodologically controlled and non-randomly assigned reflections on the number needed to treat

Background. The number needed to treat (NNT), calculated as the reciprocal of the absolute risk reduction, is a parameter that provides quantitative information on the efficacy of therapeutic interventions. The introduction of this parameter was provoked by the demonstration that physicians were not too familiar with the percentage expressions of risk reduction, but preferred, in assessing the efficacy of a treatment in the context of a clinical trial, numbers that directly indicated patients. Description. Although the results of the evolution in the concept of the NNT have been available for various years, the diffusion of such parameters is still poor. The frequency of explicit reporting of the NNT in the published randomized controlled trials of five major biomedical journals has been recently assessed. This review has shown that in 359 eligible papers, the NNT was recorded in only eight articles. This represents a major drawback of biomedical literature, since the NNT allows health operators to understand how much effort is needed to prevent a given event. Conclusions. The NNT represents a measure of immediate and major clinical impact, and should always be reported, when appropriate, as an expression of the results of clinical trials. As a single number, it cannot stand alone as the sole justification for including a new treatment in the therapeutic armamentarium; still, it should be seen as an indicator by means of which the same treatment may be critically considered. It seems therefore appropriate to require that authors publishing results of clinical trials should provide readers with the information necessary to allow them to make responsible decisions; in this case, adopting the useful tool constituted by the NNT.