Network meta-analysis combining individual patient and aggregate data from a mixture of study designs with an application to pulmonary arterial hypertension

Table 5 Summary of NMA models used for comparison of treatment combinations for PAH synthesising aggregate data from the literature and IPD from IMPRES study

Model	Data	Description	Section with model details
M1	Aggregate data only	Aggregates the IPD and includes observational data through random effects on change from baseline 6MWD	2.1
M2	Aggregate and IPD	Extends M1 to combine aggregate data and IPD	2.2
M3	Aggregate and IPD	Extends M2 to include covariate adjustments on change from baseline 6MWD	2.3
M4	Aggregate and IPD	Extends M3 to include interactions between treatment effect and covariates	2.4
S1	Aggregate and IPD	Same as M4 but SE in observational studies are inflated by a factor of 10 to downweight their evidence	3.3
S2	Aggregate and IPD	Same as M4 but constructed a control arm for observational studies where all patients assumed to deteriorate by 25 m from baseline in 6MWD	3.4