Development of a Clinical Rating Scale for Persons With Parkinson's Disease

To the Editor: Clinical rating scales for persons with Parkinson's disease measure the symptoms of the disease, physical performance, and clinical fluctuations. The information is used to assess the stage of the disease, response to treatment modalities (drug and physical intervention), and impact on physical activities in both clinical practice and research. The available scales are intended for use with people across all stages of disease, from those with minimal deficits to those who are severely disabled. Our goal was to develop a clinical rating scale that would (1) target the early to middle stages of Parkinson's disease, (2) organize the items into subscales differentiating impairments and performance items, (3) be useful among different disciplines, and (4) be expedient to administer. The purpose of this study was to develop and examine the test properties of such a clinical rating scale, the Duke University Parkinson's Rating Scale (DUPRS). To develop a preliminary scale, we reviewed current rating scales and included important items by consensus. Many were subsequently rewritten to expand scoring options for early impairment and to clarify the statements. Step 1: Five subjects, representing Hoehn & Yahr1 stages 2 to 4, were videotaped in standard interviews and examinations. Sixteen raters (four each in nursing, physical therapy, geriatric medicine, and neurology) scored each videotaped subject in random order at time 0 and at 6 weeks. Interrater reliability and test-retest reliability were performed. We used variance component analysis for partition of the effects of individual rater, profession, subject, and time. Virtually none of the variance was attributable to the discipline of the rater nor the time the videotape was rated. Most of the variability was accounted for by the five subjects and reflected that they were in different stages of the disease. Step 2: The DUPRS was administered to 40 subjects (mean age (SD) = 69.2 (7.6) years, 71% male and 29% female, Hoehn & Yahr stage 2-4) by a single rater (physical therapist). A subset (n = 19 subjects) were administered the DUPRS a second time by a second independent rater (geriatrician) to assess interrater reliability. Additional scales were administered to all 40 subjects (Folstein Mini-Mental State Examination,2 NUDS,3 CESD,4 Pearlin,5 Schwab and England6) by a third rater (geriatrician) blinded to the results of the DUPRS. Based on the results of the interrater reliability, several items were revised or deleted. The remaining 24 items included 11 impairment items, 10 performance assessment items, and three cognitive/emotion items; each item was graded 0 (normal) to 4 (worst). These data were used to assess construct validity in Step 4. Step 3: Using the 24 item scale, 38 subjects were then assessed by two raters. Factor analysis (Varimax rotation) was performed yielding the following subscales: impairments (postural stability, posture, rigidity, bradykinesia, falls, clinical fluctuations, dyskinesia and tremor); performance (hygiene, bed mobility, fine and gross motor movement, simple transfer, dressing, mealtime activity, dressing, gait and sit to stand), cognitive/emotion (motivation, depression, and memory). Analyses included tests of interrater reliability and validity. We used weighted kappas for interrater reliability (11 items k > .4; 9 items k > .6; 4 items k < .4). We used the intraclass correlation coefficient (ICC) to determine interrater reliability of the subscales (impairments 0.82; performance assessment 0.89; cognitive/emotion 0.80). Internal consistency (Cronbach's alpha) of the subscales was good: impairments 0.80; performance assessment 0.89; cognitive/emotion 0.56. Step 4: Using the 24-item DUPRS, we assessed construct validity using data from 40 subjects in Step 2. Spearman correlation coefficients for the subscales and independently applied tests shown in the table. We examined the correlation of the final DUPRS scale (Step 3) with the H&Y, r = .73 (P = .0001). The DUPRS is a 24-item scale that assesses impairments and physical performance using a combination of clinical examination, observation, and self report. The scale was designed to target people in early to middle stages of Parkinson's disease, at which time patients may require minimal assistance, but before symptoms become severe. During these stages, pharmacological and physical interventions may be instrumental in preserving or restoring physical independence. The scale takes 10 to 15 minutes to administer. Reliability and validity are good. Healthcare professionals from different disciplines can use the instrument, which is important for both interdisciplinary research and clinical endeavors. The subscales of the DUPRS are organized into categories of impairments, physical performance, and cognitive items. This organization of the DUPRS may allow further examination of the relationships between function and impairments for those in the early and middle stages of this disease. Currently, an investigation is ongoing to test the validity (including the Unified Parkinson's Disease Rating Scale7) of the DUPRS in an independent population and to examine further its clinical utility.