| The current consensus in rehabilitation is that functional activity in the life situation is the most important outcome to pursue and measure (Keith, 1995). Constraint-Induced Movement (CI) therapy (Taub et al., 1993) is a new treatment for chronic upper extremity hemiparesis that consists of a family of behavioral techniques that induce stroke patients to greatly increase the use of their affected arm for many hours a day over a period of 2 to 3 weeks. The intervention has been demonstrated not only to improve the movement of the impaired upper extremity as measured by laboratory motor tests but also to increase the tendency of patients to use their arm in the real world (Duncan, 1997; Taub et al., 1993; Taub, Crago, & Uswatte, 1998; Taub, Uswatte, & Pidikiti, 1999). The need in CI therapy research for instruments that accurately measure changes in the amount of arm use in the real world has led directly to the work presented in this dissertation.;Arm use in the home situation is currently measured by a structured interview called the Motor Activity Log (MAL; Taub et al., 1993; Uswatte & Taub, 1998). As with other self-report measures, the MAL has some drawbacks; patients' answers maybe influenced by experimenter bias, demand characteristics, or inaccurate recall.;The purpose of this dissertation is to bypass the measurement problems associated with the MAL by developing an objective measure of upper extremity movement in the home situation. The measure proposed is recordings from Computer Science Applications Inc. accelerometers worn by patients on each forearm, the chest, and the ankle. The properties of this instrument were examined in 2 studies. In the first study, 15-min segments of synchronous accelerometer and videotape data were collected from 9 individuals with stroke and 1 healthy person engaged in their regular activities at home or at a rehabilitation clinic. The videotapes were coded by a panel of observers for the duration of more-impaired arm movement and amount of functional activity. The relationships between the observer and accelerometer data were examined using correlational and regression analyses. In the second study, 3- to 6-day periods of accelerometer data were collected from 11 upper and 9 lower extremity CI therapy patients and 4 persons with stroke from the community before and after treatment or a no-treatment period. Ann use in the home was also measured using the MAL. The reliability, stability, sensitivity, and specificity of the accelerometer recordings were examined, as well their relationship to the MAL.;The results indicate that accelerometer recordings provide a highly reliable, stable, and accurate measure of the duration of impaired arm movement that has good sensitivity and specificity with regard to the treatment effect of CI therapy. This new measure makes a start at meeting the need in rehabilitation for objective measures of real-world treatment outcome. |
