"In some situations, especially if it is believed the intervention will significantly affect the intracluster correlation coefficient, publishing the intracluster correlation coefficient from both the control and the intervention arm may be useful." . . . .