A note on the use of the intraclass correlation coefficient in the evaluation of agreement between two methods of measurement.