Download presentation

Presentation is loading. Please wait.

Published byCarl Alliston Modified over 2 years ago

1
Issues of Simultaneous Tests for Non-Inferiority and Superiority Tie-Hua Ng*, Ph. D. U.S. Food and Drug Administration Ng@cber.fda.gov Presented at MCP 2002 August 5-7, 2002 Bethesda, Maryland _______ * The views expressed in this presentation are not necessarily of the U.S. Food and Drug Administration.

2
2 Simultaneous Tests for Non-Inferiority and Superiority Multiplicity adjustment is not necessary –Intersection-union principle (IU) Dunnett and Gent (1996) –Closed testing procedure (CTP) Morikawa and Yoshida (1995) Indisputable

3
3 A Big Question Is Multiplicity Adjustment Necessary?

4
4 IsMultiplicityAdjustmentNecessary?

5
5 Outline Assumptions and Notations Switching between Superiority and Non-Inferiority Is Simultaneous Testing Acceptable? Use of Confidence Interval in Hypothesis Testing --- Pitfall Problems of Simultaneous Testing Conclusion

6
6 Assumptions/Notations Normality and larger is better T: Test/Experimental treatment ( t ) S: Standard therapy/Active control ( s ) : Non-Inferiority Margin (> 0) For a given d (real number), define –Null: H 0 (d): T S - d –Alternative: H 1 (d): T > S - d Non-Inferiority: d = Superiority: d = 0

7
7 Non-Inferiority (d = ) H 0 ( ): T S - against H 1 ( ): T > S - H0()H0() H1()H1() ° T Boundary WorseBetter Mean Response S

8
8 Superiority (d = 0) H 0 (0): T S against H 1 (0): T > S H 0 (0) H 1 (0) ° T Boundary WorseBetter Mean Response S

9
9 Switching between Superiority and Non-Inferiority CPMP (Committee for Proprietary Medicinal Products), European Agency for the Evaluation of Medicinal Products Points to Consider on Switching Between Superiority and Non-Inferiority, 2000. http://www.emea.eu.int/htms/human/ewp/ewpptc.htm

10
10 Switching between Superiority and Non-Inferiority (2) Non-Inferiority Trial –If H 0 ( ) is rejected, proceed to test H 0 (0) –No multiplicity issue, closed testing procedure Superiority Trial –Fail to reject H 0 (0), proceed to test H 0 ( ) –No multiplicity issue –Post hoc specification of

11
11 Switching between Superiority and Non-Inferiority (3) Non-inferiority Trial –Intention-to-treat (ITT) –Per protocol (PP) Superiority Trial –Primary: Intention-to-treat (ITT) –Supportive: Per protocol (PP) Assume ITT = PP

12
12 Simultaneous Testing One-sided 100(1 - )% lower Confidence Interval for T - S Test is worse Test is better Mean Difference (T – S) 0 -- Superiority Non-inferiority Neither

13
13 Simultaneous Testing (2) Multiplicity adjustment is not necessary –Dunnett and Gent (1996) Intersection-Union (IU): Superiority: Both H 0 ( ) and H 0 (0) are rejected –Morikawa and Yoshida (1995) Closed Testing Procedure (CTP): Test H 0 (0) when H 0 ( ) H 0 (0) is rejected

14
14 Simultaneous Testing (3) Discussion Forum (October 1998) –London –PSI (Statisticians in Pharmaceutical Industry) Is Simultaneous Testing of Equivalence [Non- Inferiority] and Superiority Acceptable? –Superiority trial: Fail to reject H 0 (0) No equivalence/non-inferiority claim –Ok: Morikawa and Yoshida (1995) Ref: Phillips et al (2000), DIJ

15
15 Is Simultaneous Testing Acceptable?

16
16 Use of Confidence Interval in Hypothesis Testing H 0 (d): T S - d (at significance level ) One-sided 100(1- )% lower CI for T-S Reject H 0 (d) if and only if the CI excludes -d Test is worseTest is better Mean Difference (T – S) -d Reject H 0 (d) Do not reject H 0 (d)

17
17 Use of Confidence Interval in Hypothesis Testing (2) If CI = (L, ), then H 0 (d) will be rejected for all -d < L. A Tricky Question –Suppose CI = (-1.999, ), L = -1.999 H 0 (2): T S - 2 is rejected (d=2) since -d < L Can we conclude that T > S - 2? Yes, if H 0 (2) is prespecified. No, otherwise.

18
18 Use of Confidence Interval in Hypothesis Testing (3) Post hoc specification of H 0 (d) is a No

19
19 Simultaneous Testing: Problems H 0 (d 1 ) and H 0 (d 2 ), for d 1 > d 2 One-sided (1 - )100% lower CI for T - S Test is worse Test is better Mean Difference (T – S) -d 2 -d 1 Reject H 0 (d 2 ) Reject H 0 (d 1 ) Neither

20
20 Simultaneous Testing: Problems (2) H 0 (d 1 ), H 0 (d 2 ) and H 0 (d 3 ), for d 1 > d 2 > d 3 One-sided (1 - )100% lower CI for T - S Test is worse Test is better Mean Difference (T – S) -d 3 -d 1 Reject H 0 (d 3 ) Reject H 0 (d 2 ) None -d 2 Reject H 0 (d 1 )

21
21 Simultaneous Testing: Problems (3) H 0 (d 1 ), H 0 (d 2 ),…, H 0 (d k ), for d 1 > d 2 > … > d k One-sided (1 - )100% lower CI for T - S Test is worse Test is better Mean Difference (T – S) -d k -d 1 Reject H 0 (d k ) Reject H 0 (d 2 ) None -d 2 Reject H 0 (d 1 ) … …...... -d 3

22
22 Simultaneous Testing: Problems (4) Choose k large enough Pr[-d 1 < Lower limit < -d k ] close to 1 Max |d k - d k-1 | < a given small number Simultaneous testing of H 0 (d i ), i = 1,…, k Post hoc specification of H 0 (d)

23
23 Simultaneous Testing: Problems (5) Number of Nested hypotheses Exploratory (many H 0 (d)) Confirmatory (one H 0 (d)) 1 2 3 4 …………. k ………… Simultaneous H 0 ( ) and H 0 (0)

24
24 Simultaneous Testing: Problems (6) What is wrong with IU and CTP? Nothing Pr[Rejecting at least one true null] What kind of problems?

25
25 Simultaneous Testing: Problems (7) Post hoc specification of H 0 (d) Let -d 0 = 100(1 - )% lower limit - Reject H 0 (d 0 ), since -d 0 < lower limit Repeat the same trial independently Pr[Rejecting H 0 (d 0 )] = 0.5 +

26
26 Simultaneous Testing: Problems (8) Simultaneous testing of many H 0 (d) –Repeat the same trial independently –Low probability of confirming the finding 1 st trial: Reject H 0 (d j ) but not H 0 (d j+1 ) 2 nd trial: Pr[Rejecting H 0 (d j )] is low (e.g., 0.5+)

27
27 Simultaneous Testing: Problems (9) Simultaneous testing of H 0 ( ) and H 0 (0)? Confirm the finding = 2 Known variance Let T - S Significance level = 0.025 80% power for H 0 ( ) (at = 0)

28
28 Simultaneous Testing: Problems (10) f ( ) = Pr[Rejecting H 0 ( ) | ] f 0 ( ) = Pr[Rejecting H 0 (0) | ]

29
29 Simultaneous Testing: Problems (11) Test one null hypothesis H 0 ( ) Suppose that H 0 ( ) is rejected Repeat the same trial independently Pr[Rejecting H 0 ( ) again] = f ( )

30
30 Simultaneous Testing: Problems (12) Test H 0 ( ) and H 0 (0) simultaneously Suppose that H 0 ( ) or H 0 (0) is rejected Repeat the same trial independently Pr[Rejecting the same null hypothesis again] = [1 - w( )] · f ( ) + w( ) · f 0 ( ) = f ( ) - f 0 ( ) [1 – f 0 ( )/f ( )], where w( ) = f 0 ( )/f ( )

31
31 Simultaneous Testing: Problems (13) [1 - w( )] · f ( ) + w( ) · f 0 ( ) where w( ) = f 0 ( )/f ( ) Simultaneous tests in the 2 nd trial

32
32 Simultaneous Testing: Problems (14) Ratio: 1 – [f 0 ( )/f ( )] [1 – f 0 ( )/f ( )] Ratio may be as low as 0.75

33
33 Conclusion Many H 0 (d): Problematic Not type I error rate H 0 ( ) and H 0 (0): Acceptable? If “zero tolerance policy”: No If 25% reduction cannot be tolerated: No If 25% reduction can be tolerated: Yes

34
34 Is Simultaneous Testing of H 0 ( ) and H 0 (0) Acceptable?

35
35 You be the judge

36
36 References Dunnett and Gent (1976), Statistics in Medicine, 15, 1729-1738. Committee for Proprietary Medicinal Products (CPMP; 2002). Points to Consider on Switching Between Superiority and Non-Inferiority. http://www.emea.eu.int/htms/human/ewp/ewpptc.htm http://www.emea.eu.int/htms/human/ewp/ewpptc.htm Morikawa T, Yoshida M. (1995), Journal of Biopharmaceutical Statistics, 5:297-306. Phillips et al., (2000), Drug Information Journal, 34:337-348.

37
37

Similar presentations

Presentation is loading. Please wait....

OK

1..

1..

© 2017 SlidePlayer.com Inc.

All rights reserved.

Ads by Google

Ppt on object-oriented technologies Ppt on bluetooth based smart sensor networks conferences Download ppt on heritage of india Ppt on education system in india during british rule Memory games for kids ppt on batteries Full ppt on electron beam machining application Ppt on touch screen technology download Ppt on db2 introduction to logic Ppt on latest technology in mobile Ppt on sectors of indian economy class 10