n=12 sounds small to me, but I don't have the statistics background to really understand its significance.

Does it basically mean that the confidence intervals are way wider than we'd expect with a larger study size? If so, by how much?

also has anyone independently reproduced the results?