Pharmaceutical trial data – what we do not know

Elliott Spitzer is indeed breaking new ground with this lawsuit. There has not been anything like it.

As a general principle, there ought not to be a dispute about the need for full disclosure of data in clinical research, including
the negative. Failure to do so would not only mislead physicians and injure patients, but would do irreparable harm to the integrity of a research area which has been generally impeccable.

In this particular case, SKF executives denied suppressing data, and also distanced themselves from that internal memo which appears like a smoking gun.

Too bad this lawsuit comes at a time when charges of incestuous relationships between clinical investigators and drug companies have placed these companies under a shadow.

Perhaps this suit will do for clinical research as Spitzer’s previous forays on corporate scandals did for Wall Street.

We might as well have all of the data to look at, even in cases where the trial is flawed. Let the people who want to research it research it. There are a lot of problems with what we are doing now. This is just one area that greatly needs to change.

Other areas that need to change: The FDA’s reporting system for adverse drug events needs to be arranged so it’s avilable online, searchable, made anonymous to patient name. The current system is a complete mess. If you ever order the data it’s not even organized in any useful way, you spend months with grep just trying to get it sorted. This IS intentional in my opinion.

We need funding for *outcome* based studies. How are people doing 1 year…5 years..10 years down the line on this? Right now “long term” is considered 1 year. How many people are on drugs for 10-20-30 years?

It’s one thing to say an agent is better than nothing, but how does it fair against known agents? I think insurance companies (being the primary $$) would have a huge interest in this. Yeah, you can say Prilosec is cheaper than Prevacid, but lets look at them head to head and see which one actually prevents NSAID induced ulcers, and if it’s cheaper and safer to say keep a person on a drug like naproxen + prilosec/prevacid than simply switch them to Vioxx/Celebrex/Bextra? This kind of research desperately needs to be done, and it shouldn’t be based purely on the decisions of accountants the way it is now who will often look at this situation and say okay this one is 80 and this one is 9 dollars, we will pay for the one that is 9, even though the patients ends up needing 1-2 more drugs to tolerate it. Every new drug is a risk.

There are a lot of subtle ways to manipulate drug trial data as well. It raises a red flag with me when I see high dropout rates, and some researchers simply exclude entire groups of patients from the final report and fail to mention it, nor do they make the most important part of it easily available – the raw data – which is what you need to spot manipulation. If you allow kicking out based on results to get the numbers you want, you know the numbers aren’t honest. This happens far too often, in particular with antidepressants where the evidence of them having any benefit is very weak to begin with.

I wish there was a way to reign in the PR machine as well, but it’s just another layer of regulation that wont help much & often backfires. The PR hits that appear for new drugs in the paper have claims that are so divorced from the trial data that you have to wonder if they are even talking about the same agent. Witness the recent PR assault for Cymbalta being touted off label for Fibromyalgia. There was a similar PR hit for Paxil for the same condition not long ago that was an even worse gross distortion of the trial data.