Skip to main content

Table 3 Bayesian hierarchical IPD-NMA models described in the identified methodological articles

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

Model

1-stage or 2-stage process

Format of data

Study design

Type of data

Effect size

Assumptions for treatment by covariate interactions

Donegan et al. [6]

1-stage

IPD only

RCTs

Dichotomous

Odds ratio

No interactions; exchangeable treatment by covariate interactions; common treatment by covariate interactions

Donegan et al. [42]

1-stage

IPD+AD and IPD only

RCTs

Dichotomous

Odds ratio

No interactions; independent treatment by covariate interactions; exchangeable treatment by covariate interactions; common treatment by covariate interactions

Hong et al. [43]

1-stage

IPD only

RCTs

Continuous

Mean difference

Exchangeable treatment by covariate interactions

Jansen [41]

1-stage

IPD+AD and IPD only

RCTs

Dichotomous

Odds ratio

Exchangeable treatment by covariate interactions; common treatment by covariate interactions

Saramago et al. [10]

1-stage

IPD+AD and IPD only

Cluster and individual allocation trials

Dichotomous

Odds ratio

Independent treatment by covariate interactions; exchangeable treatment by covariate interactions; common treatment by covariate interactions

Saramago et al. [44]

1-stage

IPD+AD

RCTs

Time-to-event

Hazard ratio

Common treatment by covariate interactions

Thom et al. [45]

1-stage

IPD+AD

RCTs and single-arm observational studies

Continuous

Mean difference

Independent treatment by covariate interactions

  1. AD aggregated data, IPD individual patient data, IPD-NMA individual patient data network meta-analysis, RCT randomized controlled trial