Table 9.

Software for fitting meta-analysis models (full references in bibliography)

Method	Software options	Additional guidance (authors and reference number)
Network Plots	Stata: networkplotfrom the mvmeta package [39]	Chaimani et al. (2013) [40]
	R: netgraph command in netmeta [41], gemtc [42], pcnetmeta [43]
		Rucker and Schwarzer (2016) [44]
Network Meta-Analysis	WinBUGS: flexible modelling framework, model code available from the Univerisity of Bristol website (https://www.bristol. ac.uk/population-health-sciences/ centres/cresyda/mpes/code/).	Welton et al. (2012) [45], Dias and Caldwell (2019) [17]
		Dias et al. [46]
	R2WinBUGS [47], BUGSnet [48]
	R: netgraph command in netmeta [41], gemtc [42], pcnetmeta [43]	[27], Neupane et al. (2014) [49]
	Stata: mvmeta package [50]	Chaimani and Salanti (2015) [51], Chaimani et al. (2013) [40]
	Webtools: MetaInsight [52], MetaDTA [53] and CINeMA [54]
Network Meta-Regression	WinBUGS: can utilise study level covariates from NMA models, model code available from the Univerisity of Bristol website (https://www.bristol. ac.uk/population-health-sciences/ centres/cresyda/mpes/code/).
	Stata: metareg command, mvmeta package
	R: mvmeta package, metafor package [55], GeMTC package [42]
IPD Meta-Analysis	WinBUGS: inclusion of IPD and aggregated data model available in paper by Saramago et al. (2012) [21]	Freeman et al. (2018) [24], Freeman and Carpenter (2017) [56], Riley et al. (2008) [57]
		Debray et al. (2015) [58], Tierney et al. (2015) [59]
		PRISMA-IPD checklist [60]
Component Network Meta-Analysis	WinBUGS: component NMA model code available from the Univerisity of Bristol website (https://www.bristol. ac.uk/population-health-sciences/ centres/cresyda/mpes/code/).	Welton et al. (2009) [11]
	R: The additive model can be implemented using the discomb command in the frequentist netmeta package [41]
		Higgins et al. (2019) [2], Caldwell and Welton (2016) [5]
		Melendez-Torres et al. (2015) [6]
		Inclusion of covariates in component NMA models by Freeman et al. (2018) [24]