Welcome to the world of clinical trials. You’ve got a great idea that has real potential to improve health. But I have some bad news for you. Are you ready?
Your intervention won’t work. I’m sorry. It won’t. Not even a little.
“But hold on a minute,” you say. “I haven’t even run the trial yet. How can you possibly know it won’t work?” Experience my friend – and as any experienced trialist can tell you, the intervention never works. (Ok, sometimes the intervention works, but it usually doesn’t, even though we almost always think it will.)
With that in mind, let’s go forward in time. The last follow-up was completed a few weeks ago. Now you are standing next to me, the trial statistician, at the front of a large meeting room in some cookie-cutter hotel. I have a clicker in my hand, there is a graph projected onto the large screen behind us, and I just told everyone that the intervention didn’t work.
Who is everyone? The entire study team of course, all sat quietly at a giant U-shaped table with little name placards in front of them. All the site investigators are there. There is someone from the DMSB too. And of course all the other people it takes to run a trial. Someone from the funding body is there as well - it seems they are interested in what you did with their money. And don’t forget the patients’ representative – they gave more than anyone.
Look at their faces. They look so confused. Disappointed. A few even look sad. Oh no, are those tears? And we were having so much fun up to this point, drinking free coffee and catching up since our last meeting. Now they are just blankly staring at us. You have to say something. You have to say something now.
What are you going to say?
You’re new to this, so let me help you. To survive this moment, you need to be able to look everyone in the eye and remind them that the trial was designed to rigorously answer an important question. If this is true, then you have likely just learned something useful, regardless of whether the intervention worked or not. Further, if the study was well designed, other people will be more likely to trust your result and act on it. In other words, there is no such thing as a “negative” trial result – the answer given by a well-designed trial is always useful, whether the intervention worked or not. So you simply remind everyone of this. People will still be disappointed of course - we’d be fools to test treatments if we didn’t think they worked, and it’s natural to hope. But at least we know we ran a good trial and added knowledge to the world - that’s not nothing.
And what if the trial wasn’t well designed? Then the errors in the design might be used to reasonably explain away any result, “positive” or “negative”. Maybe the intervention didn’t appear to work because the trial was too small. Or maybe it only appeared to work because you used a surrogate outcome. And regardless of the explanations you come up with, nobody else will believe the result anyway, so what’s the point? So the best case scenario is that you’ve wasted lots of time and money. The worst case scenario is that you have needlessly put patients at risk. Trust me, there’s not enough perfume in France to cover that kind of stink.
So here is my advice. At the start of any trial, imagine yourself in the situation above. The intervention didn’t work. How are you going to live with yourself? Now act accordingly.