Reproducibility and validity of patient-rated assessment of speech, swallowing, and saliva control in Parkinson's disease

OBJECTIVE: To report on the development and psychometric evaluation of the Radboud Oral Motor Inventory for Parkinson's Disease (ROMP), a newly developed patient-rated assessment of speech, swallowing, and saliva control in patients with Parkinson's disease (PD). DESIGN: Reliability and validity study. SETTING: Tertiary-care Parkinson center for multidisciplinary assessment. PARTICIPANTS: Consecutive community-dwelling patients with PD (n=129) or atypical parkinsonism (AP; n=49; mean +/- SD age, 64+/-9.8y; mean +/- SD disease duration, 7y; median Hoehn and Yahr [HY] stage, 2.5). INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: To evaluate reproducibility, 60 patients completed the ROMP twice within a mean of 24+/-12 days. To study validity, another cohort of 118 patients who had completed the ROMP was assessed by both a neurologist (HY stage, Unified Parkinson's Disease Rating Scale III) and speech-language pathologist (severity of dysarthria, dysphagia, drooling) who were blinded to ROMP scores. RESULTS: Confirmatory factor analysis identified the 3 a priori-designed ROMP domains of speech, swallowing, and saliva control. Internal consistency was .95 for the total ROMP and .87 to .94 for the 3 domains or subscales. Intraclass correlation coefficients for reproducibility were .94 and .83 to .92 for the subscales. Construct validity was substantial to good with correlations ranging from .36 to .82. The ROMP differentiated significantly (P<.001) between patients indicated for speech therapy (based on independent assessment) and those who were not and between mild, moderate, and severe PD according to HY stage. CONCLUSIONS: The ROMP provides a reliable and valid instrument to evaluate patient-perceived problems with speech, swallowing, and saliva control in patients with PD or AP