Reference datasets 2 [Software]

posted by ElMaestro  – Denmark, 2014-10-11 11:34  – Posting: # 13687
Views: 10,330

Hi all,

now that reference datasets have been produced and published with (hopefully correct) reference results by some BE wackos I wonder about software PQ.

It can be argued that a system -in this case the computer doing the BE analysis- needs to be kept updated on several fronts and protected from undesired pertubations. It quickly becomes the eternal discussion of isolation versus options of updating. Noone really knows and can control well what goes on when you hook up a computer with e.g. Windows to the internet. There is absolutely no good way of telling the system that any traffic that could cause interference with SAS or WinNonlin or Minesweeper is prohibited while internet connections for good cause is required. You never know if the next anti-virus update screws up a dll of some other software without the user having a chance to know. Restricting access thorugh the fireall for some programs doesn't solve it. And so forth.

So I wonder wouldn't it be very solid and a good idea to simply perform some degree of PQ "every day" or "immediately before" a reported analysis; spend 30 secs literally on running the reference datasets through the stats package, compare the output and tick off the results, and then if the results match conclude that the system is PQ that day and run the necessary analysis?

Would be a minute or two of extra work on the day of analysis, and would require an extra form and possibly a sentence or two in a SOP. And we'd all sleep better.


Edit: Category changed. [Helmut]

if (3) 4

x=c("Foo", "Bar")
typeof(b[,1]) ##aha, integer?
b[,1]+1 ##then let me add 1

Best regards,

“(...) targeted cancer therapies will benefit fewer than 2 percent of the cancer patients they’re aimed at. That reality is often lost on consumers, who are being fed a steady diet of winning anecdotes about miracle cures.” New York Times (ed.), June 9, 2018.

Complete thread:

 Mix view
Bioequivalence and Bioavailability Forum |  Admin contact
19,408 posts in 4,123 threads, 1,325 registered users;
online 6 (1 registered, 5 guests [including 4 identified bots]).
Forum time (Europe/Vienna): 10:22 UTC

We must be careful not to confuse data with the abstractions
we use to analyze them.    William James

BEBAC Ing. Helmut Schütz