首先，來看一下某個(gè)項(xiàng)目的主要目標(biāo)和分析終點(diǎn)。
Primary Objective:
Demonstrate the efficacy of XXX versus placebo on induction of clinical remission in adults.
Efficacy Endpoints:
The efficacy endpoints will be formally examined with statistical hypothesis tests conducted
on the efficacy results obtained from adult patients randomized and dosed in Cohort 1.
Cohort 2 is open-label and does not contain a control group, therefore all of the efficacy endpoints will be summarized and described without statistical hypothesis testing.
這里的Efficacy Results，我列舉幾個(gè)：Complete Mayo score、Partial Mayo score、9-point Mayo score、Pediatric Ulcerative Colitis Activity Index (PUCAI)、Clinical Remission......
主要分析出的表格樣式如下：
如Efficacy Endpoints所說，假設(shè)檢驗(yàn)只對(duì)Cohort 1人群開展，Cohort 2人群只對(duì)分析變量進(jìn)行統(tǒng)計(jì)描述。這里的分析指標(biāo)是臨床緩解率(Clinical Remission)，這里的輸出統(tǒng)計(jì)量有Odds ratio及可信區(qū)間、兩組差異及可信區(qū)間、P值。從這張Table上的Notes看，需要的統(tǒng)計(jì)量已經(jīng)在CMH數(shù)據(jù)集計(jì)算出來，出表的時(shí)候只有引用后，調(diào)整好模板就好。我們接下來看看，CMH數(shù)據(jù)集是如何生成的。
沒有展示前面人群、分組的設(shè)置代碼，這個(gè)分析的代碼基本涵蓋了這個(gè)試驗(yàn)所需要的所有分組的主要統(tǒng)計(jì)分析結(jié)果，下面我們SPEC上面的篩選條件(POPN=1, PARAMCD=‘RMW10D1’ and DTYPE=‘NRI’)篩選出相關(guān)數(shù)據(jù)集的結(jié)果。
