As the paper described, all they did was pull available data from other studies. That's about as low a barrier to repro as you could wish - you could probably do it yourself.