Trial by media: dangers of misinterpretation of medical statistics