This is central to quality assurance. First in order to evaluate a process it must be determined that it is being consistently performed, usually to a quantifiable level of 5 nines (99.999% of the time). If it's not consistent, then you don't know if a test measurement reported is valid or the result of not doing the test right. These kinds of errors can skew the result distribution. If the distribution is not bell-shaped, it calls into question the consistency of technique. This is especially important when the sample size is small. A common error identified in process audits is failure to document details, for example recording the calibration of the instrument. Technicians do the calibration but don't log it. This shows up as a difference between the amount of calibration reagents used and the number of log entries made.
An audit can uncover problems with a process if it discovers a rater doesn't know the standard, or can't correctly point out where the data used in rating or ranking activities is kept. Also the correct version of any standards documentation must be identified by the person being interviewed. So even though there is inherent subjectivity in a process, I don't think that is what is being called out; rather the process followed was found to have inconsistencies which in turn leads to lack of confidence in the data.
robust is not simply a synonym for "good". http://en.wikipedia.org/wiki/Robustness
I would call this a controversy. The question is what is the best way to quantify dystrophin. There needs to be consensus on this in the scientific community before the FDA can use it as a surrogate for clinical improvement. The issue as I see it is whether SRPT purposefully used an assay that would cast eteplirsen in a favorable light to hide inadequate production based on animal models and becker's phenotype data.
It does not surprise me that FDA wants more info. After all that's why there's a p3 underway. And asking for more dystrophin assay data lends support to approval on the marker. This is looking like rolling submissions to me even though it has not been characterized as such.
Under the circumstances, selling the shelf at a premium would have motivated the buyers to hedge by shorting against the box. Letting them in cheap encouraged them to hold. It provided them with a 20% gain for safety.
Just a hunch, but I think contract renewal is in the works and the updraft is someone with ears on in the government front running the news. The thinking used to be that you should have t least two countermeasures per threat, and to get them you should plant quite a few seeds and then water them all.
I only see 3 EUA's for ebola on the FDA web site. Nothing else in the federal rigister,They are all for PCR assays. I take this to mean that any drugs given to patients this far is under compassionate use, not emergency authorization.