A big thank you to all those who responded to my original query "Sample
Size/power calculation Software".
Below I have included a quick summary, the orignial request and all the
responses.
Best regards,
Saghir
Summary:
'Review of nQuery and Pass for power and sample size calculations' (vol. 1
no. 2 p135-140) by Peter Lane in Pharmaceutical Statistics is an excellent
review of nQuery and PASS software. It also mentions other software
packages.
Other packages that were recommended include: Stata, R, PS and Sample Power
2.
I was hoping for more on sample size and power calculations for
non-inferiority and equivalence trials. I was looking for ideas to verify
calculations.
Original Message:
> I was wondering what sample size / power calculation software allstat
> readers use. I'm primarily interested in software for clinical trials
> (especially crossover and non-inferiority designs).
>
> I'm interested in your experiences with such software.
>
> Best regards,
> Saghir Bashir
>
Responses:
--------------------------
From: Paul Brown
There was an interesting article in a recent Pharmaceutical Stats
journal titled: 'Review of nQuery and Pass for power and sample size
calculations' (vol. 1 no. 2 p135+). We use nQuery but I like to write a
quick excel spreadsheet to verify the calculations (if time permits).
Paul
--------------------------
From: Steff Lewis
I use something called PS, which is by Dupont and Plummer. It's a
free download from the web. It doesn't do anything very fancy, but
is handy for quick calculations (and it's free).
I have also used the following website for an equivalence study:
http://www.ucalgary.ca/~patten/blackwelder.html
It's also free.
--------------------------
From: Edmund J. Bini, M.D.
I use Sample Power 2. It is a stand alone product made by SPSS. I have
been happy with it. Here is the link.
http://www.spss.com/spssbi/samplepower/
--------------------------
From: Roberto Ferrara
we use NQuery Advisor version 4 (I think a version 5 should be on the market
now). It covers a lot of possible situations, and I think it is becoming
quite an industry standard. It is quite simple to use; you have anyway to
pay careful attention that the figures you are inputing really mean what
they are expected to mean, particularly for equivalence.
--------------------------
From: Peter Lane
I wrote a comparative review of Pass 2002 and nQuery 5 for Pharmaceutical
Statistics. [Modified for list - Saghir] PSI members can view it online at
http://download.interscience.wiley.com/cgi-bin/fulltext?ID=102523138&PLACEBO
=IE.pdf&mode=pdf
This also refers to some of the other packages I have come across. All the
main statistical groups in R&D within GSK are now using Pass, though we
have access to nQuery as well.
--------------------------
From: Neil Shephard
The statistical package Stata (http://www.stata.com) has a number
of functions for calculating power and allows calculation when the
study design compares one out come, be it proportions or means.
It also has the ability to handle sample size/power calculations for
longitudinal studies with similar outcomes.
If you've not used Stata before then this may be a bit of a steep
learning curve if all you want to do is calculate sample-size/power,
but it is a superb statistical package (IMO), and well worth inveting
in (in comparision to other proprietry statisical packages its
relatively cheap).
Another alternative which would be free is the R package
(http://www.r-project.org) (which is basically just a free
implmentation of the S programming language on which S-Plus is
based). I'm not very familiar with this package, but I'm sure it
would have the ability to calculate sample size/power.
Both of these packages allow users to write additional
functions/commands/routines, so there may be even more scope
for power calculations than I have given above.
-------------------------- END --------------------------
---------------------------------------------------------
Legal Notice: This electronic mail and its attachments are intended solely
for the person(s) to whom they are addressed and contain information which
is confidential or otherwise protected from disclosure, except for the
purpose they are intended to. Dissemination, distribution, or reproduction
by anyone other than their intended recipients is prohibited and may be
illegal. If you are not an intended recipient, please immediately inform the
sender and send him/her back the present e-mail and its attachments and
destroy any copies which may be in your possession.
---------------------------------------------------------
|