"# PowerSimulation"

Code written by Tim Dawes, August 2018

The model assumes the use of a linear mixed-effect model (90% power, type 1 error rate of 5%)

Assumptions relating to the data for analysis

RVEF deteriorates annually by 1% RVEF [14], and >1% in patients with pathological variants (22)

Age and gender have independent effects on RV function (23)

Allelic variants occur in BRIDGE-PAH cohort as homozygotic wildtype (66-88% patients), heterozygotic variant (9-27%) and homozygotic variant (2-6%), frequency varying by locus*

RV function measurement reproducibility data is spatially non-uniform but has mean variability of ~2%*

Collaborating centres will provide data on 20 further patients each

Missing data (scan intolerance / poor image quality) will occur in 2-3% scans [1/43 intolerant, to date]

*unpublished data