This is our most recent data where you see there's a gradual increase over time from 1988 to 2013, that randomized trials are increasing and that's great news. We're up to over 40 or 50 or even 60 published in a year in some years in the Journal of Bone and Joint Surgery, our single biggest journal, which I think is really encouraging. The other encouraging thing is that the quality of the randomized trial seems to be increasing as well.

This is where the trends that are a bit disturbing come in, they were less often collaborative. If we look at the 1988 to 2000 era, they were multi-centered in 67 percent of the randomized trials. So, 67 percent of the 72 randomized trials published between 1988 and 2000 were multi-centered and only 30 percent of the 280 plus that were done in 2001 to 2013 were multi-centered. That's really unfortunate when you think about it, these trials are going to be smaller, these trials are definitely going to be smaller if they're not multi-centered and they're not likely to be positive.

Why are we spending so much money to do conduct studies that ultimately are not moving the needle? It's a really big issue.

It's a huge issue actually that we're not advancing science and I think that's particularly the most important point. But on a, perhaps smaller and more focused level, if we look at what it does to the patient's confidence in us, that patients are being enrolled in trials and perhaps subjected to treatments or subjected to the vagaries of doing research filling out forms, etc. when really when you look at the studies that they're involved in those studies are destined to not be successful in determining an outcome or an appropriate outcome. Moreover, one of the bigger problems we're facing in health care systems is that perhaps payers and people who help to reimburse us for what we do are going to look at say you know, you guys are really not showing us that the treatments you're providing are helping people.

[…] If they were to include another center they wouldn't get the prestige of having done and designed and completed that trial so they were resistant to including other centers because they wanted, they needed that for themselves and the institution wanted the credit for doing the randomized trial.

[In reference to meta-analysis] […] We are starting to see research moving to the combination of results of case series because there aren't enough randomized trials on many topics. We need more trials. We need more people doing trials so we can actually use meta-analysis more powerfully.

Even within our country there are randomized trials that go on in one center that is 200 kilometers from me and some of the people I work with are resistant to get involved in that because they say, you know ‘well we didn't design that and that's not that's not our work, we should we be able to do something too’. The truth is, well we're not and they're doing a good job of it so why would we not collaborate with that group, and that's the barrier we're having right now.

The real problem with small randomized trials is that they can lead to misleading estimates. People are more likely to believe a randomized trial of 25 patients randomized and discount a large observational study of 30 000 patients, saying well -- that's not level one evidence.