“…After removing duplicates and screening the titles and abstracts, 27 potentially relevant records were retrieved for further review. Of these, 12 studies were excluded for following reasons: 6 did not have usable data, [ 22 – 27 ] 1 did not report any outcomes of interest, [ 28 ] and 5 used microarray data. [ 29 – 33 ] We retrieved no additional studies from reference lists.…”