IMS America Hospital Supply Index This database contains estimates of medical surgical products in 1500 product categories purchased by 350 hospitals in the United States. This data can be used to derive comprehensive estimates of product purchases and device sales. However, the data do not allow analysis of user characteristics, and devices purchased but not used are not accounted for.
Total hospitals in the US is what? 5,000-10,000? IMS is polling about 3.5-7.0%% of the US hospitals.
As for the comment that hospital sales are different from pills sold in drugstores, but in many ways they present an easier task for data-tracking companies because there are fewer sites to be checked
Hospital sales for a competitive drug like recothrom will be much more lumpy than retail - they either convert a hospital or they don't. The hospital, as polled center, is not a statistically averaging device like a pharmacy. Unlike a retail pharmacy they don't get 0.5-4% of sales at each pharmacy. Given that I would have expected at most(!) perhaps 5-10 hospital conversions by Mid Feb (in time to actually accept a bulk delivery by end of Feb) there is even a moderately good chance that they polled none of the hospitals which converted. (0.93^10=~0.5). And this all assumes that IMS has no time lag in their access to the hospital databases. Point being not that the data is wrong per se - but that it has huge error bounds. It isn't meant to be used for drug launches that are all or nothing and to hospitals only.
Market Data 
Markets 
AI
