Is Meta-Analysis the Same as Data Mining?
On Thursday morning, Arena announced data from meta-analyses of three of the trials for lorcaserin. This data was presented at the European Congress on Obesity and indicated that almost half of patients taking two 10mg doses a day saw more than 5% weight loss, more than double the rate of response in the placebo group. More than one-fifth of those same patients saw better than 10% weight loss; nearly triple the response seen in the placebo group.
Unfortunately, most observers are going to regard this data with a shrug. Clauses like "Modified intent-to-treat with last observation carried forward" are tantamount to data mining in many people's eyes, and the FDA has been very aggressive in rejecting such analyses. This is not to say that lorcaserin doesn't work; rather it just seems unlikely that the FDA is going to revise its viewpoint that lorcaserin offers "marginal efficacy" on the basis of a new look at old data (as opposed to a new study showing better outcomes).
To continue, follow the link:
http://stocks.investopedia.
No comments:
Post a Comment