The Effects of Presenting Clinical Trials Results in Different Formats to Medical Students

pp. 225-230

Abstract

Introduction. Previous studies have demonstrated that presenting treatment benefits in terms of relative risk reduction rather than in terms of absolute risk reduction or number of needed to treat patients, should favor the sense of outcome effectiveness.

Purposes. The purpose of this study was to perform a cognitive evaluation to assess how the form of presenting the risks and benefits of screening methods and treatments affects medical students’ decision-making.

Methods. Sixty-five medical students attending a Biostatistics course answered a questionnaire reporting the results of clinical trials expressed as relative risk reduction (RRR), absolute risk reduction (ARR) and/or number needed to treat (NNT), with or without associated graphs.

Results. Students’ performance was similar when comparing treatment benefits, both in relative and absolute risks presentations (RRR: 57.7% vs. ARR: 51.5%, p=0.319); however, performance was worse when information was expressed as NNT (RRR: 57.7% vs. NNT: 31.3%, p=0.000002). Inclusion of modified-scale graphs was misinterpreted as a real data difference (RRR: 98.5% vs. ARR: 43.1%, p<0.000001).

Conclusions. This study demonstrated the risks related to misinterpretation of statistical results, and the need to improve the students’ training in this type of quantitative analysis, in order to improve the medical decision-making process.