Leaderboard

We used a third party vendor called Korchek Technologies because FDA recommends independence of review in section 4.9 of "General Principles of Software Validation; Final Guidance for Industry and FDA Staff". I've attached the document below FDA Validation Guidlines.pdf

I believe the FED recommends you do your own validation. I concur. It really gives you an indepth knowledge of your system DON"T use the scripts provided by the vendor. FDA recommends you make your own so that you validate only the portions of the system you will utilize. I recently managed an upgrade (in progress when I arrived). They used the scripts provided and validated the entire system. Took hundreds of man hours and 60% of it was wasted time. The FDA provides a guidance on validating BBISs.

I agree that, in the case of an emergency, of course group O blood should be given - I have never argued against that and would be completely mad to so do. No, my argument was purely that, in a normal situation, a second sample should be taken from ALL patients.

I suggest you take a moment and determine exactly why you want a second type. Is it simply to be able to meet some outside requirement? Is it to detect the possibility of a testing or clerical error at the bench or is it to determine if the blood in the tube did not come from the patient the test was ordered for? While each of these is a worthy goal, by my way of thinking only one way will achieve all 3 and that is a second sample collected at a different draw. Now you can get even more complicated if you want depending on your level of paranoia. Must that second sample be collected specifically for this purpose or can you "borrow" one from hematology? Must that sample be collected by a second individual or does that matter? Must that 2nd collecting individual be left handed because the original collector was right handed? Lastly you must determine what are the actual limitations imposed by your facility for each shift. Are you a small facility without the staff on all shifts to actually have a second sample drawn by a second person? Take the time to really consider all of this and then keep it as simple as you possibly can. No one ever improved anything by making it more complicated.