Ok – there are problems with the term ‘clinical significance’ and it is interpreted in different ways. I apologise for not being clearer in my post.
Soo is right in that you need first to decide what reduction is possible / realistic – this is based on existing CS rates in your area. So a 10% reduction may be possible in an area with high CS rates (e.g. some US hospitals have CS rates > 50%) but might be more difficult to achieve if you are talking 25% (English average).
This will feed into the sample size calculation as Carolyn has said. You may decide perhaps that a reduction of 25% to 23% is plausible / realistic for your intervention in your area, and that might require a very large sample in order to demonstrate statistical significance.
However, if you achieved a 2% effect size would your team be happy? Would it be clinically important? I think many might believe a 2% reduction in operative delivery was clinically meaningful. So answering the question – what is the smallest clinically beneficial value of the effect size? – can be very helpful in planning your study.
There are people on this list with a much better understanding of how that figure (clinical significance) can be used in statistical calculations, so I’m not going to try to go there.
Hope that makes sense.
Best wishes
Vanora
However, a reduction in, for example, 25% CS to 15% CS would only require about 600 participants – is this the question being asked? Or is it more how we decide if this is the minimum clinical significance, or, say, 25% to 23% (which, as Vanora says, will require much bigger numbers)?
All the best
Soo
From: A forum for discussion on midwifery and reproductive health
research. [mailto:[log in to unmask]] On Behalf Of Vanora Hundley
Sent: 01 June 2012 14:37
To: [log in to unmask]
Subject: Re: Advice wanted
Hi Cathy,
This is a great question - a statistically significant reduction is likely to require a very large sample size (probably beyond the realm of most RCTs). However, if I understand your email correctly, you want to know what would be meaningful reduction in clinical
terms. I don't have an answer but I look forward to reading the discussion.
Best wishes
Vanora
________________________________________
From: A forum for discussion on midwifery and reproductive health research. [[log in to unmask]] On Behalf Of Cathy Walton [[log in to unmask]]
Sent: 01 June 2012 12:01
To: [log in to unmask]
Subject: Advice wanted
Dear Midwifery research group members,
I am a consultant midwife at King's College Hospital in London and have
been a member of this group for a number of years and rarely contribute
but always find the discussions interesting. Sadly, I don't get to do
much research in my job, as it really isn't easy to have time for it.
However, recently I have been involved in the planning of 2 possible
RCTs and would like to pick the brains of the esteemed members of this list.
Would anyone have any suggestions for what would be a clinically
significant % point reduction in the casaerean section rate if used as a
a primary outcome?
If so, can you direct me to any evidence or reference that might be
useful to base this on?
Many thanks,
Cathy Walton
The University of Stirling is ranked in the top 50 in the world in The Times Higher Education 100 Under 50 table, which ranks the world's best 100 universities under 50 years old.
The University of Stirling is a charity registered in Scotland, number SC 011159.