Thursday, June 21, 2007 8:29:47 AM
On the 2nd CRO and blown data.
Generally, when a company announces the data has issues and needs to be examined more closely it just means the trial failed and they are trying to mine it for some hopefull data to justify continuing on.
But this story is starting to sound a little like the might really have screwed up the data, unbelievably enough.
An important point is that if the data is wrong, it can not be repaired after unblinding. The reason is that bias can slip in any only find issues that favour the trial arm. Thus, the FDA will not accept fixing the data.
But in at least one case I have seen a company solve this problem. Take the raw data before the problem occurs and give it to a new CRO to analyse without letting them know what the first CRO did (and of course, having them clean and lock the data before before re-unblinding). Does raise some eyebrows, but has a slight chance.
If this was just the typical "trial failed due to mild allergy season" type crap there would be no logic in all of getting a second CRO. Additionally, a second full analysis of the raw data would explain the blinded vs unblinded oddity that Beachgal brought up, the problem was found in unblinded data but the new CRO is working blinded.
IF this really is the case, then the new CRO would need a few months to through the cleaning/locking/unblinding/analysis cycle.
Maybe it really would be a good time to save money on the electric bill and have everybody take a long summer vacation
Generally, when a company announces the data has issues and needs to be examined more closely it just means the trial failed and they are trying to mine it for some hopefull data to justify continuing on.
But this story is starting to sound a little like the might really have screwed up the data, unbelievably enough.
An important point is that if the data is wrong, it can not be repaired after unblinding. The reason is that bias can slip in any only find issues that favour the trial arm. Thus, the FDA will not accept fixing the data.
But in at least one case I have seen a company solve this problem. Take the raw data before the problem occurs and give it to a new CRO to analyse without letting them know what the first CRO did (and of course, having them clean and lock the data before before re-unblinding). Does raise some eyebrows, but has a slight chance.
If this was just the typical "trial failed due to mild allergy season" type crap there would be no logic in all of getting a second CRO. Additionally, a second full analysis of the raw data would explain the blinded vs unblinded oddity that Beachgal brought up, the problem was found in unblinded data but the new CRO is working blinded.
IF this really is the case, then the new CRO would need a few months to through the cleaning/locking/unblinding/analysis cycle.
Maybe it really would be a good time to save money on the electric bill and have everybody take a long summer vacation
Join the InvestorsHub Community
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
