Table 5.

Methodological details of fMRI studies on BOLD responses for de-qi.

Studies	One group analysis method	P value of a one group analysis	Two group analysis method	P value of a two group analysis
Studies that compared central BOLD responses between de-qi and sharp pain groups
Hui et al., 2000 [54]	Kolmogorov-Smirnov test	P < 0.001 or P < 0.00001 (cor)
Hui et al., 2005 [33]	group-averaged	P < 0.001 (unc), minimum 3 voxels
Hui et al., 2009 [55]	group-averaged	P < 0.001 (unc), minimum 3 voxels	two-sample unpaired t-tests	P < 0.05 (unc)
Sun et al., 2012 [40]	random effect model	P < 0.0001 (unc), minimum 5 voxels	two-sample unpaired t-tests	P < 0.0001 (unc)
Studies where no subjects reported sharp pain/excluded subjects who experienced sharp pain from fMRI analysis
Napadow et al., 2005 [34]	Fixed effect model	P < 0.0001 (unc), minimum 3 voxels
Jeun et al., 2005 [78]	random effect model	P < 0.01 (unc),
Kong et al., 2007 [79]	random effect model	P < 0.05 (cor), minimum 10 voxels
Zhou and Jia 2008 [66]	random effect model	P < 0.001 (cor) 50 voxels
Na et al., 2009 [80]	random effect model	P < 0.001 (unc), minimum 100 voxels
Bai et al., 2010 [81]	random effect model	P < 0.005 (unc), minimum 5 voxels
Qiu et al., 2010 [82]	random effect model	P < 0.001 (unc), minimum 3 voxels
Fang et al., 2012 [85]	random effect model	P < 0.05 (cor), minimum 5 voxels
Sun et al., 2012 [40]	random effect model	P < 0.0001 (unc), minimum 5 voxels