Underestimation of Risk or Overascertainment of Events in Assessing Risk Prediction Equations?