Hi StockObserver – I have never looked at MXC or this trial before. So my off the top of my head comments are going to be pretty superficial.
I suppose when I look at something like this I try to keep things in proportion.
If MSB (with a MC of $3b) had produced this I would say it was the greatest load of rubbish. But if SUD (with a MC of $10m) had produced this I would consider it a minor miracle and outstanding work.
From this angle the question is whether they are punching above their weight.
Pretty debatable I think. The English in the results announcement is terrible and the results come apart pretty easily.
Ten subjects – 3 received placebo – 7 the active. Everyone went to a NEWS score of zero at EOT. Except for one placebo patient who went to a score of 16 at EOT.
As everyone spotted in this thread - you don’t need to be a biostatistician – take out this one outlier – the whole effect disappears.
But for the sake of argument – leave the outlier in and try to replicate the ANOVA analysis. I inputted the numbers from the announcement table – and get the same means as they do.
When I do the ANOVA I get p=0.051 which is not statistically significant.
They claim they got p<0.005. Which is a bit silly because no-one says this. People say p<0.05 or p<0.01 – they never report p<0.005.
The ANOVA they performed doesn’t meet the assumption for equal variances so shouldn’t have been done in the first place.
So inappropriate test – and then they make a blooper on the p value and claim something is statistically significant when its not.
The table 3 is incorrect – the placebo group 2.6 average point increase (got worse). The treatment group should be negative 7.14 average point decrease (got better).
Now of-course I could have made a blooper in all this.
Someone check with the company. If my anova is wrong or my test of homogeneity of variances is incorrect I’ll donate $50 to the charity of the posters choice with the receipt posted on HC.
But I do give them a couple of marks for trying – and at least trying to make the results transparent. Its just the execution that they need some help with.
At $40m MC I remember OBJ hiring a hall and getting some blokes with dodgy knees running around with a fridge magnet and some magic rubbing oil on their knees; and calling the clinical trial a great success. This is better .. but maybe don't expect a Lancet article to come out of all this.
View attachment 2410412
