Clark:
I think you are being overly harsh in terms of your classifying this as 'data mining'. The Stroop was one of six tests given, and the results were significant at p=.01. I did not expect that to occur--but if I had to choose one domain where a trend would have been expected (given post-CABG pt complaints of slowed decision making and information processing), this is it.
Had they come out with some third-order 'finding' like (I'm making this up):
'male CABG patients over 62 but under 64 showed improvement on the Stroop at six weeks, significant at .049', where there are a host of covariates thrown in, your criticism would be justified. But they didnt. If it's data mining, they did not have to dig deep to find something of value.
Trailmaking is a paper and pencil test, and was not given. Both Trails and the Stroop tap frontal lobe functions, though Trails adds in a visual scan/motor speed component.
So far as I am aware, no previous CABG trial did pre-post comparisons using the Stroop--in fact I am not aware of any drug producing a positive neuroprotectant effect in CABG--so I consider this an early but promising finding. Certainly it is not ready for PhIII, but Phase IIb is a viable option.
From my perspective, I had hoped for a trend showing a signal of effect. This finding bettered that expectation.
Harry Tracy
NeuroInvestment |
