Second, in some studies, both performance and response biases are possible since the comparisons between massage and no-massage treatment or conventional care cannot be blinded. On the other
hand, results did not change when either blinded studies or nonblinded studies were used in the analysis.