Seol mar théacs é seo: Defining the fracture population in a pragmatic multicentre randomised controlled trial