520 likes | 766 Views
(C)Stephen Senn. 2. Basic Thesis. Design theory has great potential in drug developmentBut this potential is unrealisedThose working in so-called optimal design are so ignorant of application realities that where their influence is not zero it is harmfulOn the other hand the understanding of design theory by biostatisticians is pitifully inadequateWe must cooperate properly to cure this parlous state of affairs.
E N D
1. (C)Stephen Senn 1 Lost Opportunities for Design Theory in Drug Development Stephen Senn
2. (C)Stephen Senn 2 Basic Thesis Design theory has great potential in drug development
But this potential is unrealised
Those working in so-called optimal design are so ignorant of application realities that where their influence is not zero it is harmful
On the other hand the understanding of design theory by biostatisticians is pitifully inadequate
We must cooperate properly to cure this parlous state of affairs
3. (C)Stephen Senn 3 Outline Quick tutorial on cross-over trials
I shall then give two introductory examples of nonsense
By leading design theoreticians
By leading biostatisticians
I shall then consider ‘design nonsense’ further
Some conclusions
After lunch a case-study
4. (C)Stephen Senn 4 Warning I am a biostatistician
We are used to thinking of data matrices with rows as subjects and columns as measurements
That means that we write sequences for designs with rows representing subjects and columns representing periods
5. (C)Stephen Senn 5
6. (C)Stephen Senn 6
7. (C)Stephen Senn 7
8. (C)Stephen Senn 8
9. (C)Stephen Senn 9 Simple Carry-over Carry-over lasts for exactly one period
It depends only on the engendering treatment and is unmodified by the perturbed treatment
There is a huge literature proposing ‘optimal’ designs for this model
There is no empirical evidence that any of this has been useful
10. (C)Stephen Senn 10 Three Period Bioequivalence Designs Three formulation designs in six sequences common.
Subjects randomised in equal numbers to six possible sequences.
For example, 18 subjects, three on each of the sequences ABC, ACB, BAC, BCA, CAB, CBA.
A = test formulation under fasting conditions,
B = test formulation under fed conditions
C = reference formulation under fed conditions.
11. (C)Stephen Senn 11
12. (C)Stephen Senn 12 Properties of these weights Sum 0 in any column,
eliminates the period effect.
Sum 0 in any row
eliminates patient effect
Sum 0 over cells labelled A
A has no part in definition of contrast
Sum to 1 over the cells labelled B and to -1 over the cells labelled C
Estimate contrast B-C
13. (C)Stephen Senn 13
14. (C)Stephen Senn 14
15. (C)Stephen Senn 15 Properties of These Weights As before
Estimates B-C contrast
Eliminates, period and patient effect
Eliminates A
Sum to zero over cells labelled a,b, and c
Eliminate simple carry-over
16. (C)Stephen Senn 16 Have We Got Something for Nothing? Sum of squares weights of first scheme is 1/3 (or 4/12)
Sum of squares of weights of second scheme is 5/12
Given independent homoscedastic within- patient errors, there is thus a 25% increase in variance
Penalty for adjusting is loss of efficiency
17. (C)Stephen Senn 17 First ExampleSome Design Theory Nonsense
18. (C)Stephen Senn 18 What’s wrong here?
19. (C)Stephen Senn 19
20. (C)Stephen Senn 20
21. (C)Stephen Senn 21 Multi-Story
22. (C)Stephen Senn 22 Conclusion Multi-dose trials real scope for design theory.
These will employ active wash-out
Design problem is trade-off between exploiting correlation and eliminating carry-over.
Short vs long active wash-out periods
23. (C)Stephen Senn 23 Second ExampleSome Biostatistics Nonsense
24. (C)Stephen Senn 24 Second ExampleSome Biostatistics Nonsense
25. (C)Stephen Senn 25 What is wrong 1. It’s not correct design theory As any design expert knows residual degrees of freedom are (nearly) irrelevant to efficiency
It is the impact of adjustment on the degree of orthogonality of the design matrix that is important
26. (C)Stephen Senn 26 What is wrong 2.It’s not realistic biostatistics In fact as any biostatistician who has had to think about it will know from a practical point of view far from being optimal Balaam’s design is simply inadmissible
The reasons is that only half of the resources are devoted to actually measuring the treatment
The rest are devoted to providing an adjustment for a form of carry-over that is itself implausible
27. (C)Stephen Senn 27 Allocation of patients for two designs
28. (C)Stephen Senn 28 Investigation of the real efficiency of Balaam’s design
29. (C)Stephen Senn 29
30. (C)Stephen Senn 30
31. (C)Stephen Senn 31 Dose Response:The Statistician’s Version
32. (C)Stephen Senn 32 The Models Which use Simple Carry-over are Inconsistent
33. (C)Stephen Senn 33
34. (C)Stephen Senn 34 The Rhinoceros
35. (C)Stephen Senn 35 The Phoenix Bioequivalence Trials Analysed by D’Angelo & Potvin
20 drug classes
1989-1999
12 or more subjects
96 three period designs
324 two period designs
36. (C)Stephen Senn 36
37. (C)Stephen Senn 37
38. (C)Stephen Senn 38
39. (C)Stephen Senn 39 Conclusions Distribution of P-values uniform
no evidence of carry-over
Carry-over a priori implausible
presence testable by assay
No point is testing for it
leads to bias
Or adjusting for it
increased variance
40. (C)Stephen Senn 40 Do Bayesians do Better? In principle the Bayesian approach ought to allow us to be more flexible about nuisance parameters such as carry-over
However, the Bayesian track record is not impressive here
Realistic models have not been employed
41. (C)Stephen Senn 41 Classic analysis ignoring carry-over
Accumulated analysis of variance
Change d.f. s.s. m.s. v.r. F pr.
+ Patient 28 600.897 21.461 3.99 <.001
+ Period 1 18.776 18.776 3.49 0.073
+ Treatment 1 58.364 58.364 10.84 0.003
Residual 27 145.360 5.384
Total 57 823.397 14.446
Estimates and confidence limits for treatment 2.037 0.768 3.306
Analysis of first period values only
Two sample t
Standard Standard error
Sample Size Mean Variance deviation of mean
Drug 17 8.118 14.74 3.839 0.9310
Placebo 12 7.667 8.97 2.995 0.8646
Difference of means: 0.451
Standard error of difference: 1.327
95% confidence interval for difference in means: (-2.272, 3.174)
Classic analysis ignoring carry-over
Accumulated analysis of variance
Change d.f. s.s. m.s. v.r. F pr.
+ Patient 28 600.897 21.461 3.99 <.001
+ Period 1 18.776 18.776 3.49 0.073
+ Treatment 1 58.364 58.364 10.84 0.003
Residual 27 145.360 5.384
Total 57 823.397 14.446
Estimates and confidence limits for treatment 2.037 0.768 3.306
Analysis of first period values only
Two sample t
Standard Standard error
Sample Size Mean Variance deviation of mean
Drug 17 8.118 14.74 3.839 0.9310
Placebo 12 7.667 8.97 2.995 0.8646
Difference of means: 0.451
Standard error of difference: 1.327
95% confidence interval for difference in means: (-2.272, 3.174)
42. (C)Stephen Senn 42 This is a re-analysis of the Hills and Armitage data by
1. Teather, D, Morrey, GH. Bayesian methods in cross-over trials, Byocybernetics and Biomedical Engineering 1995; 15: 41-52.
This is a re-analysis of the Hills and Armitage data by
1. Teather, D, Morrey, GH. Bayesian methods in cross-over trials, Byocybernetics and Biomedical Engineering 1995; 15: 41-52.
43. (C)Stephen Senn 43 Identical Priors for Treatment and Carryover? Patients treated repeatedly during trial
Fourteen day treatment period
Average time to last treatment plausibly 4 hours
Average time to previous treatment seven days
Saying that it is just as likely that carry-over could be greater than treatment is not coherent
In any case the two cannot be independent
Is negative carry-over as likely as positive carry-over?
Large values of carry-over seem implausible if the treatment effect is small. This implies that the prior distributions cannot be independent.
Negative carry-over would imply that the residual effect of a treatment had an opposite effect to the direct effect of that treatment.Large values of carry-over seem implausible if the treatment effect is small. This implies that the prior distributions cannot be independent.
Negative carry-over would imply that the residual effect of a treatment had an opposite effect to the direct effect of that treatment.
44. (C)Stephen Senn 44 So What are Acceptable Models for Carry-over? Ignoring carry-over altogether (not allowing for it because one believes one has taken adequate steps to eliminate it)
This is always a reasonable strategy
Using an integrated pharmacokinetic pharmacodynamic model (Sheiner et al, 1991)
This may work for dose-finding trials
Very difficult to implement where more than one molecule is involved
45. (C)Stephen Senn 45 The Sheiner model
46. (C)Stephen Senn 46
47. (C)Stephen Senn 47
48. (C)Stephen Senn 48
49. (C)Stephen Senn 49
50. (C)Stephen Senn 50 Advice for Design-Theoreticians Resist the temptation to give advice if you are unfamiliar with the application area
Seek collaborators
Ground your models in pharmacology
Remember that the goal is good medicine not elegant mathematics
Don’t defend the indefensible
51. (C)Stephen Senn 51 Advice for biostatisticians Remember that design theoreticians have many powerful results
It’s just conceivable that some of them may even be useful
52. (C)Stephen Senn 52
53. (C)Stephen Senn 53 References