Table 2: The included trials scored according to the risk of bias criteria.


Tang and Sun 1996 [22]?+?+++
Yu et al. 2004 [23]++?++++
Wei et al. 2005 [24]?+++
Xu et al. 2007 [25]++?++++
Chen and Jing 2008 [26]++?++++
Yu et al. 2010 [27]+++?++++
Zhang et al. 2011 [28]++?++++
Zhu and Huang 2011 [29]++?++++

A: adequate sequence generation; B: concealment of allocation; C: Blinding (patient); D: blinding (investigator); E: blinding (assessor); F: intention-to-treat analysis (ITT analysis); G: incomplete outcome data addressed(drop-outs); H: free of selective reporting; I: similarity at baseline; J: cointerventions constant; K: compliance acceptable; L: timing outcome assessments similar. +: Yes, −: No, ?: Unclear.