What helps medical students understand screening statistics?

Research indicates that about one in two physicians misunderstands important cancer screening statistics that are used to evaluate and communicate benefits and harms. For instance, many physicians mistakenly believe that increased detection and survival rates are sufficient proof that cancer screening saves lives. We investigated what factors facilitate or impede the understanding of such important statistics in medical students. We conducted an experiment with 173 medical students (23% final year, 9% interns, 68% residents) from the Cayetano Heredia University in Lima, Peru. Participants completed a questionnaire assessing their a-priori beliefs about screening (e.g., the belief that screening is always the best choice, that foregoing screening is irresponsible). Then they read statistical information about benefits and harms from screening for a hypothetical cancer. Participants were randomly assigned to one of two conditions: in one condition the described screening was effective (i.e., reduced mortality) and in the other it was not effective (i.e., did not reduce mortality). We assessed participants’ comprehension of the information, numeracy, science literacy, statistical education history, and demographics. Results from multiple regression showed that stronger positive beliefs about screening predicted worse comprehension (Beta = -0.17, p = .025). In contrast, numeracy had a similar, independent effect on comprehension in the opposite direction (Beta = 0.19, p = .022). These results held regardless of the effectiveness of screening and having completed courses on methodology/statistics. Medical students who had highly positive attitudes towards screening and had low numeracy misunderstood important screening statistics. Numeracy skills should me emphasized in medical curricula, as they can help counteract common but influential psychological biases when evaluating evidence and making recommendations to patients.