統計問題（28）：關於交叉設計臨床試驗的問題（多選）
下列哪些陳述適用於交叉設計臨床試驗（crossover clinicaltrials?）？Which of the following statements apply to crossover clinicaltrials?a) They are not subject to confounding by age, sex, or most personal characteristicsb) Randomisation is not possiblec) They are useful in studies using mortality as the endpointd) They need fewer people than parallel group trialsAnswer
In a crossover trial each person receives the two treatments to be compared: first one, then the other. The order in which the treatments are given can be randomised and it is good practice to avoid a systematic advantage to one treatment and to improve the prospects of preserving blinding when this is part of the design.If the disease being studied is a stable long term condition then this approach has several advantages. Because the effect is compared within each person, the usual problems of confounding caused by differences between people who receive one treatment or the other do not exist. The comparison within each person alsoreduces variability and a control group is not needed, so fewer people are needed for the study. However, if the condition changes over time, or if the effect of one treatment persists into the time the other treatment is given, then theresults can be difficult to disentangle.中文解釋：
在交叉試驗中，每個人都接受兩種要比較的治療措施：先第一種，然後另一種。給予治療的順序可以是隨機的，這種做法的優勢是避免對一種治療產生系統性作用，並改善在設計中盲法的原則。如果所研究的疾病是穩定的長期病狀，則該方法具有幾個優點。因為在每個人中比較效果，所以不存在由接受一種治療或另一種治療的人之間的差異引起的常見混淆問題。每個人之間的比較也減少了變異性，並且不需要對照組，因此研究所需的人員更少。但是，如果病情隨時間變化，或者如果一種治療的效果持續到另一種治療的持續時間，那麼結果可能難以解開。所以答案是選擇 A,D
如果您對交叉設計不是很理解，請看一下視頻：