The quorum is the count of the number of members present in the EC meeting. EC consists of five members and I think one of the members should be a Biostatistician. Many of the scientific and technical standards of a research are related to statistics and can be controlled such as sample size calculation, data collection & management and statistical methods.
What do you think?
I agree very strongly with the the idea it is good practise to have a biostatistician contributing to any ethics approval process. I think there are several grey areas or challenges though. First, what is a biostatistician? Most working scientists have had both formal training and applied experience in statistical analysis. Second, what should the formal criteria on which the biostatistician should comment (like statistical power, the analysis methods themselves, the sampling, all these things...?)
I also agree that an ethical community should have a qualified biostatistician. Often there are some people on those communities who know some statistics but not at a satisfactory level and whose primary qualification is not related to biostatistics.
As an answer to Edwin's reply, yes biostatisticians can help in those aspects you named.
I have a different idea:
experimental design and data analysis are tighly related. A good experimental design is mandatory to get data that allows you to adress the relevant scientific questions in the analysis. Formulating these questions, particularily relating these question to a possible experimental design and data analysis requires a lot of statistical expertise. This all is expremely important in designing a study in the first place. This is the time where a bio-statistician would be extremely helpful. A bio-statistician in an ethics commitee only sees the final plan, and to do a good job she would have a hard time and a lot of work eruding if questions were formulated to match some known ways of analysis or some known experimental designs (it should actually be the other way around), and if the really scientifically interesting questions have been adressed (e.g. I often see that studies analyse all pair-wise mean differences in 2x2 designs instead of the interaction, what often would actually adress the biological key question).
In short:
A study that requires an ethics vote should be designed together with a bio-statistician "on board", and this statistician should write a statement for the ethics commitee. The commitee does not neccesarily need a statistician. However, if the commitee feels that the design and analysis part is critical and they have concerns with the statistics statement, they certainly may involve another statistician.
Jochen, yes I agree with you. What you describe is the ideal that is expected to happen but which actually is not always confronted neither.
That is a great point Jochen, that ideally a good statistical design with ethical implications would be fully informed fully-informed from the beginning to, e.g., justify the sample size (not too small to have low statistical power to detect a real effect, not too high to waste resources and impact participants unnecessarily). A related issue is explicit consideration of "effect size".
My experience (as a biologist with formal statistical training and experience as a statistical consultant inside and outside of academia) is that both ethical and practical statistical standards in different academic fields vary widely. With or without a biostatistician (however this may be be defined), I feel that articulating the objectives of any ethics exercise is important (e.g., has the sample size been clearly justified; has the sample been minimised where there is risk; etc.).
Yes... Biostatistician is very important role/person in the Ethics Committee.