For the 32 reviews for which statistical heterogeneity may exist, twelve reviews (38%) implemented adequate methods: subgroup analysis (seven reviews), sensitivity analysis (two reviews), or meta-regression (three reviews) to explore clinical and/or methodological factors as a potential explanation of statistical heterogeneity within the trial sets.
Several examples of mechanisms by which a causal relationship may exist have been reviewed [ 12, 14, 17], including the supposition that rapid disintegration could increase oral and esophageal mucosal exposure to active drug, thereby increasing the likelihood of irritation.
For α2, α3, α4 and β2 and β4 subunits, pairwise combinations of α and β (e.g. α3β4 and α4β2) are sufficient to form a functional receptor in vitro, but far more complex isoforms may exist in vivo (reviewed in 217, 218, 229).
Although other ABA sensing pathway(s) may exist (for review [8]), the extreme ABA insensitivity of the triple mutant ost1 snrk2.2 snrk2.3 indicates that protein phosphorylation mediated by ABA-activated SnRK2s is essential to regulate all aspects of ABA signaling [30] [32].
Publication bias may exist in the review of this literature.
In addition, language biases may exist in the review process, from submission to acceptance [ 17].
Alternatively, there may exist remote regulatory elements, as reviewed elsewhere [50], controlling retinoid-mediated transcription of Akap12β from a distance.
However, it is possible that important studies relevant to this review may exist in other languages.
First, the remaining heterogeneity does suggest that additional factors related to efficacy, beyond those considered in this review, may exist.
It is difficult to identify reviews with both parallel-arm and split-mouth trials with usual strategies and we acknowledge that unidentified reviews may exist.
