Component Scores, or Total Scores over Components?
- PDF / 478,773 Bytes
- 8 Pages / 504 x 720 pts Page_size
- 49 Downloads / 215 Views
0092-86 I5/2002 Copyright @ 2002 Drug Information Association Inc.
COMPONENT SCORES, OR TOTAL SCORES OVER COMPONENTS? WILLIAMH. CHANGAND CHRISTYCHUANG-STEIN Clinical Biostatistics. Pharmacia Corporation, Kalamazoo, Michigan
Patients’ responses to items on a questionnaire are frequently used to assess treatment effects in clinical trials. In this paper; we will use the Unified Parkinson S Disease Rating Scale to address a common question of whether it is more powerful to compare treatments based on the total scores of part 11 (questions 5- I7 on the questionnaire) and part 111 (questions 18-31) combined or the individual component scores (either part II or part III). We will explore the relationships among the powers associated wirh the various test statistics and display the relationships using graphs. As an illustration, we will apply our findings to help decide on the primary endpoint for a clinical trial in Parkinson S disease patients. Key Words: Parkinson’s disease; Components; Unified Parkinson’s Disease Rating Scale
INTRODUCTION FREQUENTLY, THE OBJECTIVE of a clinical trial is to compare treatments based on subjects’ responses to items on a questionnaire. The response to each item is recorded on an ordinal scale such as the zero to four scale used for speech assessment. These five speech assessment categories from zero to four correspond, respectively, to normal, mildly affected (no difficulty being understood), moderately affected (sometimes asked to repeat statements), severely affected (frequently asked to repeat statements), and mostly unintelligible. In this paper, we will assume that the same scale is used for all items on a questionnaire so that responses, when represented by numerical scores (eg, 0 to 4 for the 0-4 scale) can be added up from the data perspective. We will further assume that items on the questionnaire are combined into “components” so that different components reflect the well-being of subjects on different dimensions. An example is part I1 (questions 5-17) and part 111 (questions 18-31) of the Unified Parkinson’s Disease Rating Scale that is used to evaluate the state of Parkinson patients. The part I1 and part I11 components correspond to the daily living and motor function status of Parkinson patients, respectively. For a detailed discussion on Unified Parkinson’s Disease Rating Scale, the reader is referred to Fahn et a]. (1). Using the Parkinson’s disease example, a common question is whether it is more powerful to compare treatments based on the total scores of part I1 and part 111 combined or the individual component scores (either part I1 or part 111). Intuitively, if the comparative effects of the treatments on the two parts are of the same magnitude and are going in the same Reprint address: Christy Chuang-Stein. Pharmacia Corporation, 9161-298-251, 7000 Portage Rd., Kalarnazoo. MI 49001 (e-rnail: [email protected]).
557
Downloaded from dij.sagepub.com at UNIV OF RHODE ISLAND LIBRARY on April 4, 2015
558
William H. Chang and Christy Chuang-Stein
Data Loading...
