Fig 2 Comparison of methods to identify people at 15% or greater 10 year coronary risk without access to measured cholesterol values