Comments on the use of multiple roles to maximize the utilization of outcome research