Almost all of the studies blinded the participants in the outcome assessment. That avoids bias when asking patients about their pain.

It's difficult to imagine how you would blind the acupuncturists performing the treatment. I suppose you could have one acupuncturist mark the needle location and another blinded one stick the needle in. But a skilled acupuncturist would probably be unblinded by their own knowledge of where it makes sense for the needle to go. So there is definitely risk of performance bias, but they're sticking needles into patients at predefined points - I don't think it's a very large risk.