School absenteeism data for surveillance purposes: a proxy for acute respiratory infection rates