Many thanks to Paul Milligan, Dr. Robert Nemeth, Prof. Stephen Senn, Dr. 
Lada Smoljanovic, James R. McDermott and
Dr. D.J. Wright  who replied to my posting;

I summarised the replies for those who are interested.

Thanks, again.



Dr. Luigi Santoro
Head of Statistical Service Unit
Mediolanum farmaceutici S.p.A.
Via S. Giuseppe Cottolengo, 15 - 20143 Milano - Italy
Tel.: +39 0289132343
Fax: +39 0289132375
e-mail: [log in to unmask]
Web site Mediolanum farmaceutici S.p.A.: http://www.mediolanum-farma.com



My Query was the following:

Dear Allstatters,
I have got the job, from my director, of studying the differencies between
equivalence trial and non-inferiority trial, especially concerning the
sample size calculation.
Unfortunately I've no experience with non-inferiority trial so I need to
begin by reading some good paper.
Is there anyone who can briefly explain me the difference or, at least,
knows some referencies ?

Thanks in advance.





Paul Milligan wrote:

Equivalence/Noninferiroty references:

WC Blackwelder, Similarity/Equivalence Trials for combination vaccines, 
Annals New York Academy of Sciences 1995,754:321-328

CP Farrington and G Manning, Test statistics  and sample size formulae for 
comparative binomial trials, Statistics in Medicine 1990, 9:1447-1454

these papers discuss both cases, non-inferiority, where you want to show 
that one treatment is at least as good as another, within some margin, and 
so you are interested in whether the upper confidence limit for the risk 
difference is less than some specified value,  and equivalence, where both 
confidence limits are of interest.




Dr. Robert Nemeth wrote:

Dear Dr. Santoro,

your question makes me guess that you are familiar with equivalence trials.
It makes things easier: non-inferiority trials are nothing more than simple
comparative trials with a shifted null-hypothesis. Such trials sometimes
call
one-sided equivalence trial, as you specify an equivalence range for only
one
direction (depending on the definition of inferiority).

The following example makes this concept clear:
Let us suppose that it should be shown that a new drug (test) is
non-inferior to another
marketed one (reference) concerning e.g. the decrease of the pulse rate in
patients
with heart failure. Supposing  an expected decrease of 10 beats per minute
(bpm) for
the reference drug, you want to show the the new drug is not inferior, i.e.
it generates a
clinically similar effect. Now you have to define the relevant difference
(or equivalence
range). Let this value 5 bpm. It means that the new drug should lead to a
decrease of
at least 5=10-5 bpm.

In a bit more formal way it can be expressed with the following hypotheses:
 H_0:  mu_test <= mu_ref-delta         H_1: mu_test > mu_ref-delta
 where mu denotes the expected value and delta the equivalence range.

Concerning the sample size, the situation is equivalent to a one-sided
testing procedure.

I hope that it helps and kind regards

Robert Nemeth

_____________________________________________
Robert Nemeth, PhD
Senior Statistician
Focus Clinical Drug Development GmbH

Email: [log in to unmask]
Tel: +49 (0) 2131 155 315    Fax: +49 (0) 2131 155 378
Web-site Focus Clinical Drug Development GmnH: http://www.fcdd.de
_____________________________________________




Prof. Stephen Senn wrote:

You might like to have a look at chapter 15 of my book, Statistical
Issues in Drug Development, Wiley 1997.

Also note that the CPMP have a position paper out.

Regards

Stephen Senn

Professor Stephen Senn
Department of Statistical Science &
Department of Epidemiology and Public Health

University College London
Room 316, 1-19 Torrington Place
LONDON WC1E 6BT

Tel: +44 (0) 207 679 1698
Fax: +44 (0) 207 813 0280
Email: [log in to unmask]
webpage: http://www.ucl.ac.uk/~ucaksjs/
-------------------------------------------------




Dr. Lada Smoljanovic wrote:

Dear Dr. Santoro,

If you visit http://www.eudra.org/ema.html <http://www.eudra.org/ema.html>
(search for the main keywords), you might get some documents from the
European Agency for the Evaluation of Medicinal Products that might help.

Best regards,

Dr Lada Smoljanovic
Napp Pharmaceutical Ltd
Cambridge Science Park
Cambridge, UK




James R. McDermott wrote:

Dear Dr. Santoro

As a member of the PSI (Statisticians in the Pharmaceutical Industry)
Training subcommittee I am aware of a course that we are giving called
Equivalence and Non-inferiority Trials presented by Dr. Richard Kay.  It is
on the 7th Feb 2000 and is in the UK it should be advertised on the psi web
page (http://www.psiweb.org/events.htm).  Recently Professors Machin and
Campbell presented a Sample Size course for us that contained a relevant
section.  If you need a more urgent answer I could review their notes for
you.  Please let me know what response you get.

Yours Sincerely,

	James


James R. McDermott
Managing Director
Oxford Pharmaceutical Sciences Limited
tel: +44 (0)1491 833338
fax: +44 (0)1491 833334

e-mail: [log in to unmask]
www.ops-web.com



Dr. D.J. Wright wrote:

	 Dear  Dr. Luigi Santoro,

		You may find the following references useful

	1.	Jones, Jarvis, Lewis and Ebbut. 'Trials to assess
equivalence: the importance of regorous methods',
	British Medical Journal, 313, 36-39 (1996).

	2.	Lewis. Statistical Principles for Clinical Trials: An
introductory note on an international guideline. Statistics in Medicine, 
18,
1903-1904 (1999).

	3.	ICH E9 Expert Working Group. Statistical Principles for
Clinical Trials: ICH Harmonised Tripartite Guideline. Statistics in
Medicine, 18, 1905-1942 (1999).
>
	Yours sincerely,

		Dr. D.J. Wright
>



%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%