《英国医学会杂志》（BMJ）自2008年9月开始至2015年由两位流行病与统计学专家不间断地出了300多期statistical question系列。在这个系列中，两位学者每次出一道统计学选择题，进行选择并解释。现在我精选300道Statistical Question，形成中文版，请有兴趣的朋友们进行回答。统计问题（9）：相对危险度的置信区间，你会选择吗?一项用于治疗社区获得性肺炎的抗生素试验显示，相对死亡风险为0.5，P值为0.01。*请大家在看问题答案解析之前，先“投票”，也看看大家的结果QuestionA trial of antibiotics for treating community acquired pneumoniashowed that the relative risk of death was 0.5 with a P value of 0.01. Whichone of the following is a possible 95% confidence interval for the relativerisk?a) (–0.8 to –0.3)b) (0.2 to 1.6)c) (0.3 to 0.8)d) (1.2 to 1.6)AnswerRelative risks are the result of dividing two risksgreater than zero so a relative risk is never negative. Any confidence intervalmust contain the result within it. As the P value is small the 95% confidenceinterval cannot include the measure of “no effect,” which for a relative riskis 1. The only answer satisfying all these conditions is c.中文解释：相对风险是将两个大于零的风险相除的结果，因此相对风险永远不会为负。任何置信区间必须在其中包含结果。由于P值很小，因此95％的置信区间不能包含“无影响”的度量，对于相对风险而言，该值是1。满足所有这些条件的唯一答案是c。所以答案是选择 C原文出处：BMJ杂志
