Using Confidence Intervals to Quantify Statistical and Clinical Evidence for the Treatment Effect in a Comparative Study-Moving Beyond P Values.