Bloomberg reports that Intense Diabetes Therapy Tied to Higher Death Rates.

"All things in moderation" would seem to be the lesson from many of these trials. It's interesting that the trial would be set up in a way that doesn't reflect current treatment protocols.

In The Death of the Blockbuster Drug, I noted we are in a transitional period. One of the signposts I identified in the progress towards personalized medicine was a need to focus more on therapeutic index.

The current "drug safety" craze is starting to create enough revenue-line damage for big pharma blockbuster drugs that we should be moving closer to clinical trial designs incorporating patient targeting. My current guess is that will begin with pharma adding enrollment criteria (and label exclusions) ruling out patients who won't respond or who will respond badly.

The FDA needs to do a better job of streamlining the regulatory process for such trials, though. Every company I've spoken with who asks for guidance from the FDA on screening gets such a huge laundry list of tasks that they simply avoid it.

A criticism of clinical trials sponsored by pharma companies is the design and enrollment criteria don't mimic real-world conditions. Pharma chooses design and patient populations in order to make it more likely the drug is successful.

What we have been seeing lately is non-pharma trials that make the same "error". They also choose trial designs and patient populations, but with a goal of making the drugs fail. Of course those are not the stated goals of the trials, but that is the hidden baseline hypothesis.

The major problem is the media is having trouble understanding what's going on, and patients are caught in the middle. A good example of the potential harm in this is the swift increase in suicides in adolescents after many parents and doctors pulled them off CNS drugs due to the new warning label.

Pharma runs a drug trial in a preferred patient population subset, and the FDA grants them a label for the entire population.

Academics run the same drug in a suboptimal patient population subset, and generate data like these. This pressures the FDA to narrow the label to exclude the suboptimal patient population.

Patients in the middle ground who should be receiving the drug, get worried and stop taking it.

The pharma and academic approaches are wasteful. They are simply two sides of the same bad penny -- made worse because neither the FDA nor pharma seems to be interested in enabling truly targeted drug trials.