Drug labels (package inserts) are the direct way of communicating prescribing information to physicians- but they are written by the drug companies and then negotiated or approved by the FDA. Drug approval requires that preclinical studies and the results of at least two phase 3 studies in patients - are submitted to highly qualifed FDA experts which may take a year to evaluate.
The FDA produce review documents recording the reasoning behind approval. These are lengthy and can be complex - the problem seems to be that important information is getting lost - and never makes it onto a drug label. The authors report a number of possible harms - which fail to make it onto the drug labels, but also efficacy data that can be useful, but again is lost from the final drug sheets.
The FDA has recognised some problems with drug labels and in 2006 introduced a highlights section to emphasize drug warnings and indications. It has also issued guidance about reporting trial results in labels - but in at least 2 drugs studied, labels remained unimproved.
It seems that a better way of communicating information that is already known, but not passed into the clinicians hands must be expedited.