“…Therefore, meta-analyses were conducted with a random effect model, using Comprehensive Meta-Analysis version 2 (Biostat, Englewood, NJ). However, only 4 studies 38,42,44,46 were subjected to subgroup analyses for disease sites because the other studies reported the HRs of patients with different affected sites together. Most of the included studies dichotomized patients into 2 groups according to the platelet count or PLR.…”