We are grateful for the point Drs Morgan and Sethi raised, and we returned to our initial data collection forms to determine the cause of this problem. We discovered that although the participating neurologists depicted their patients' H&Y stages using some half-point increments (stage 0, 1, 1.5, 2, 2.5, 3, 4, 5) in programming the data for analysis, these scores were inadvertently coded ordinally with values of 1, 2, 3, 4, 5, 6, 7, and 8, respectively. This produced the effect noted by Drs Morgan and Sethi, an artificial expansion of the upper end of the scale's range.
Avorn J, Schneeweiss S, Sudarsky LR, Glynn RJ. Sudden Onset of Somnolence in Parkinson Disease—Reply. Arch Neurol. 2006;63(3):467-468. doi:10.1001/archneur.63.3.467-b