Clustering interval data: an application in the epidemiologic field