A scoping review of indirect comparison methods and applications using individual patient data

Table 2 Properties of methods to derive indirect and network meta-analysis estimates using individual patient data

	Adjusted indirect comparison (or Bucher method)	Mixed comparison	Meta-regression model	Bayesian hierarchical NMA model	MAIC [48]	STC [49]
No. of empirical studies applying method (n = 33)	2 (6 %) [20, 52]	1 (3 %) [28]	4 (12 %) [19, 21, 25, 26]	17 (52 %) [6, 10, 18, 22–24, 27, 29–33, 42–45, 51]	8 (24 %) MAICs [34–38, 48, 57] and 1 (3 %) extended MAIC [46]	0 (0 %)
Properties
1-stage or 2-stage process	2-stage	2-stage	Both can be applied	Both can be applied	NA	NA
Format of data	IPD+AD/IPD only	IPD+AD/IPD only	IPD+AD/IPD only	IPD+AD/IPD only	IPD+AD	IPD+AD
Avoids selective use of indirect evidence from a network of trials	No	No	Yes	Yes	No	No
Can compare >2 treatments at a time for efficacy/safety	No	No	Yes	Yes	No	No
Preserves within-trial randomization	Yes	Yes	Yes	Yes	No	No
Study-specific true treatment effects can be assumed as fixed or random with common mean effect for each pairwise comparison	Yes	Yes	Yes	Yes	No	No
May account for potential clinical and methodological differences across trials	Yes	Yes	Yes	Yes	No	No
Does not require assessment for transitivity assumption	No	No	No	No	Yes	Yes
Mean treatment effects expressed via consistency equations	No	Yes	Yes	Yes	No	No
Can rank all competing treatments for same condition	No	No	Yes	Yes	No	No
Enables adjustment for predefined set of patient characteristics	No	No	Yes	Yes	Yes	Yes
Can be applied even in disconnected network of trials	No	No	No	No	Yes	Yes

AD aggregated data, IPD individual patient data, MAIC matching adjusted indirect comparison, NA not applicable, NMA network meta-analysis, STC simulated treatment comparison

ISSN: 1471-2288