May 9th, 2013, 6:04 am
I can only speak from experience in the US, so this might be of limited use to you. My impression is that the MHRA has similar attitudes to the FDA though.I think what capafan2 describes is bioinformatics not biostatistics. At least from what I've seen (4+ yrs in pharma), Biostatisticians will generally be using well known methods and using things like Bayesian statistics only when they perceive it will be favorably received by the FDA. Different agencies within the FDA have different opinions on Bayesian methods; usually it's not the standard.SAS is the standard software. It is not mandated by the FDA per se, but basically all drug and device submissions are done with SAS datasets / output / validated programs. You can google for the reasons for this, but, if you were working in biostats in the US, you could not avoid learning SAS.Computational issues do not drive biostats work. Datasets can be somewhat large (hundreds of thousands of records), but in a range that SAS can handle without appealing to the GPU for example. DNA based datasets in bioinformatics are much larger and I do not doubt that Hadoop / parallel computing are used. Since biostats people are basically at a confirmatory part in the process and bioinformatics people are at a data mining part of the process, the biostats issues are more like design of experiments, hypothesis testing, etc. The biostat part of the process is heavily influenced by the regulatory environment.In the US, data manager and biostatistician are separate tracks. A MS in Stats is the necessary condition to work as a biostatistician but basically anyone can work as a data manager as long as they can program SAS. Mobility to senior biostats roles will be limited if you only have the MS, but there are exceptions to the rule. A recent head of Pfizer's biostat department had a MS. A MS in Stats is considered a competitive advantage for getting a data manager role. If you had a PhD, people would question why you wanted to work as a data manager.The job of data manager revolves around the regulatory standards for datasets and TLFs (Tables Listings and Figures). Data will be collected through a paper or electronic CRF (Case Report Form) and that has to be put into datasets that conform to the regulatory standards (in the US, SDTM and ADaM) which in turn have to be put into TLFs that were prescribed by the pre-study SAP (statistical analysis plan) that would have been written by a senior biostatistician.Data manager is pretty good job IMHO. The pay is reasonable but will be less than Finance. There is usually a shortage of people who can do this work, so you have stable employability. There is not very much stress as the basic programming problems are usually straightforward but a lot of your time will be spent in minutia. It is a mix between an IT and quantitative job.Even though my experience is only in the US, I hope you find this helpful. Good luck.
Last edited by
jthomas781 on May 8th, 2013, 10:00 pm, edited 1 time in total.