This has been previously discussed (https://www.researchgate.net/post/How_we_can_do_a_Meta-analysis_of_ordinal_outcomes_and_measurement_scales )
Authors have implemented Wilcoxon-Mann-Whitney generalised odds ratio methods originally in Stata (Cumming et al. 2015). The method could also be implemented with genodds R package (I guess).
Cumming TB, Churilov L, Sena ES. The Missing Medians: Exclusion of Ordinal Data from Meta-Analyses. PLoS One. 2015 Dec 23;10(12):e0145580. doi: 10.1371/journal.pone.0145580. PMID: 26697876; PMCID: PMC4689383.