.2002.5 . P85-P97.

1 2 3 4 ICH-E9. Statistical Principles for Clinical Trials. 1998.2 5 FDA. Guideline for Format and content of Clinical and Statistical Sections of an Application. 1988.7 6 FDA-Docket No.97D-0188. International Conference on Harmonisation;Draft Guidelines on General Considerations for Clinical Trials;Availbility.1997.5

1

2

(protocol)(case report form, CRF)(data management)(statistical analysis plan)

(principal investigator)(clinical study report)

3

1(effectiveness)(safety)

(1)(primary variable)(target variable)primary endpoint(secondary variable)

(2) (composite variable)(rating scale)

(3) (global assessment variable)(scale of ordered categorical ratings)

2 (bias) (randomization)(blinding)

(1) (randomization) (stratified)(block)

(2) (blind method)(double-blind)(single-blind)(open-label)

(placebo)(double dummy)

e) (emergency envelope)

f (treatment group)(control group)(unblinding)

1 (parallel group design)(crossover design) (factorial design)

Parallel group design Crossover design Sequence,Period22

2(multicenter trial)

(consistency test)

CMH(mixed effect model)

3. (superiority)(equivalence)(non-inferiority)

SuperiorityEquivalenceNon-inferiority1 H0 H

23

-- -- H0 H0

H

=0.05,

4(group sequential design) II(=0.05)(alpha spending function)

5. (sample size) ()

1SDA100200240

0.906300.093700.05801600.886080.113920.05751500.862010.137990.05701400.834640.165360.05651300.801340.198660.05601201- n2 n1 2 71.6%,49.1%,=0.05,

P1=0.60,P2=0.35

Power

n1=n2

=0.05

0.20

0.80362

69

0.10

0.90088

88

=0.01

0.20

0.80137

99

0.10

0.90224

124

0.70.7D0=0.07p10 =0.05=0.20 N1=N2=734=0.10 N1=N2=927

0.60.55 =0.05

Power

Ne=Ns

D0=0.10

0.10

0.90015

185

0.20

0.80157

134

D0=0.05

0.10

0.90025

418

0.20

0.80059

302

6. (case report form)

1. /(Audit or inspection)

2. (interim analysis)

3. /

1.

2/

3

4query list, query form)

5(database)6

7(range check)(logic check)

8()(visual check)

9(blind review)

(transformation)(outlier)(statistical model)

(parametric statistics)(non- parametric statistics)

10

1. (statistical analysis plan)(statistical tables)

2.Analysis setsI

intention-to-treatITTFull Analysis Set ITTFAS

(per protocolPP)

safety set

(Safetey population) (Full-Analysis-Set) (FAS)

(Per protocol) (PP) 1. 2. , , 3. (80%120%)

1. 2. 3.

3(missing value)

LOCFLast observation carry forward

4 (transformation)

5(descriptive statistical analysis)(hypothesis test)(parameter estimation)

. (safety evaluation)(survival analysis)

/

,

(significant level)(statistic)