70,000 Children are Missing. Well, their Data is...
/Let's imagine you are collecting donations for a worthwhile cause – saving the Slow Loris, for example. Look at that guy.
Save me, children!
Moved to help, 70,000 children empty their piggy banks and give you their hard-earned. Now imagine that instead of giving that to International Animal Rescue, you do nothing with it.
No, you don't spend it yourself. You just, like, leave it somewhere. Forever. Would those kids have a right to be angry with you?
This is precisely analogous to what happens when a clinical trial is performed but never published. According to a study appearing in the journal Pediatrics, between 2008 and 2012, roughly 70,000 children participated in clinical trials that have never seen the light of day.
Remember – participating in a clinical trial is voluntary. These volunteers are often doing this through purely altruistic motives – no expectation of personal benefit – merely to advance science. So what happened to these 70,000 kids?
Researchers examined records from clinicaltrials.gov:
"Where Pharma Trials go to Die"
Since 2000, individuals and companies are mandated to register their trials on this website, presumably before the trials start. Here's an example from yours truly:
What do the Methods Man and Merck have in common? We both have to register our studies on this website.
From 2008 to 2010, the researchers identified 559 pediatric randomized trials registered in clinicaltrials.gov.
104, a bit less than 20%, were discontinued – meaning the study was stopped before enrollment was complete. This is not necessarily a bad thing – around 12% were stopped for "informative" reasons which means they stopped the study early because the intervention was very effective or wasn't working at all. But it also speaks to some poor planning. Fully 36% of the discontinued studies were stopped because of inadequate recruitment.
Of the 455 trials that were actually completed, 136, or about 30% never got published. This is a bad thing. The trial was finished – the data was there. It just needs to be written up.
The researchers tried to figure out what might lead to a trial failing to get published. The most important factor? The primary funding source. Your study was more than 3 times as likely to go unpublished if it was funded by Industry compared to Academia. Making this finding more damning is that industry-sponsored studies were less likely to be terminated early – these guys are good at planning trials – they get 'em done. Not publishing the data is a breach of the ethical obligation you have to your clinical trial volunteers. And needless to say, it biases the published literature. If this situation angers you, as it does me, check out www.alltrials.net which has more information.