Reliability of Speech Intelligibility Ratings Using the Unified Parkinson Disease Rating Scale

The Unified Parkinson Disease Rating Scale (UPDRS) is widely used by medical, nursing, and allied-health professionals in the management and study of patients with Parkinson disease (PD), who sometimes present with decreased speech intelligibility. As such, examining the reliability of speech intelligibility ratings using the UPDRS is critical in evaluating its utility as an instrument for identifying and documenting the nature of the speech deficits. Previous research has suggested that the reliability of speech intelligibility ratings made by neurologists using the UPDRS is questionable, perhaps due to vague written instructions for obtaining a valid nonstandardized speech sample. Also, there is no literature to date reporting the reliability of speech intelligibility ratings made by clinicians live at the time of patient examination versus ratings later made from audiotape recordings. Furthermore, there is no literature to date that reports the reliability of expert speech-language pathologists' ratings of speech intelligibility using the UPDRS nor how those ratings might compare to those of medical experts. The purpose of this study, therefore, was to determine the interrater and intrarater reliability of speech intelligibility ratings made by medical and nonmedical experts using a standardized speech protocol and modified rating criteria for the Speech item of the Motor Examination subscale of the UPDRS. Three medical experts (a neurologist with expertise in movement disorders, a neurology chief resident, and a registered nurse with experience with movement-disordered patients) and three certified speech-language pathologists with expertise in motor speech disorders rated the tape-recorded speech of 19 subjects with mild PD. For the sample of mildly dysarthric speech obtained for this study, interrater reliability for both the medical experts and the speech-language pathologists was greater than .70 (p. < .05), and for the two physicians it was greater than .90 (p. < .05). Mean intrarater reliability for all listeners was .90, for the medical experts it was .86, and for the speech-language pathologists it was .93 (p. < .05). For both the movement disorders specialist and the speech-language pathologist present during examination of the patient, intrarater reliability for in-clinic versus audiotape ratings was greater than .80 (p. < .05). These findings suggest that with the use of a standardized speech protocol, the Speech item in the Motor Examination subscale of the UPDRS can be used reliably by medical experts and speech-language pathologists to assess the speech intelligibility of patients with mild PD.