Then an ascendant hierarchical clustering was used to determine the number of subgroups on the basis of coordinates of the main components retained by MCA. The 4 clustered subgroups were numbered according to the order of the selection for the classification. The validity of this method was measured by the “cubic clustering criterion” with a cut-off of 2. After subgroups were selected, χ2 tests were performed to compare and highlight parameters defining each distinct profile of patients.