Objective. Of the various measures developed for studying persons with rheumatoid arthritis, only one that focuses on joint protection has undergone extensive testing, the Joint Protection Behavior Assessment (JPBA).The purpose of the present study was to examine the interater and intrarater reliability of the JPBA.
Method. Six healthy participants performed the JPBA under three test conditions (uninformed, informed, completely guided joint protection behavior). The 18 test performances were videotaped and scored by nine independent raters.
Results. Analysis of these data showed that interrater reliability (intraclass correlation coefficient [ICC]) was .90 or higher, and intrarater reliability was .95 or higher (ICC). The correlation between the JPBA and its two shortened versions was .95 or higher (ICC). Internal consistency was also high, with a coefficient alpha of 0.95 for the complete JPBA. Kappa values showed that for most subtasks, there was fair to excellent agreement between raters and consistency of raters over time.
Conclusion.Our data suggest that the complete JPBA has excellent clinimetric properties and that the shortened versions are adequate for clinical situations. Some improvements in the test manual suggested by the present study may further improve the measure. A repeat of this study under real-world circumstances would provide an estimate of JPBA reliability in clinical practice.