Using high-energy proton fluence to improve risk prediction for consequences of solar particle events