86 to 0.93) using goniometers. In contrast, Bovens et al (1990) reported poor reliability for measurements by physicians of physiological wrist extension using vision. Reliability for measuring physiological thumb abduction was reported to be higher using a pollexograph (ICC 0.59, 95% CI 0.42 to 0.89) than a goniometer (ICC 0.37, 95% CI –0.42 to 0.79). Finally, measuring accessory movements of carpal bones against the capitate bone using a 3-point scale yielded fair to moderate

reliability (weighted Kappa from 0.29 to 0.42) in healthy individuals and fair to almost perfect reliability (weighted Kappa from 0.33 to 0.87) in post-operative patients ( Staes et al 2009). This systematic review included 21 studies investigating inter-rater reliability selleck compound of measurements of passive movements of upper extremity joints, of which 11 demonstrated acceptable reliability (ICC > 0.75). Reliability varied considerably with the method of measurement and ICC ranged

from 0.26 (95% CI –0.01 to 0.69) for measuring the physiological range of shoulder internal rotation using vision to 0.99 (95% CI 0.98 to 1.0) for the physiological range of finger and thumb flexion/extension using a goniometer. In general, measurements of physiological range of motion using instruments were more reliable than measurements using vision. Furthermore, measurements of physiological range of motion were also more reliable than measurements of end-feel or of accessory range Bafilomycin A1 cost of motion. Overall, methodological quality of included studies was poor, although two high-quality studies reported almost perfect reliability (Glasgow et al 2003, Nomden et al 2009). In general, already reliability for measurements of passive movements of upper extremity joints were substantially higher than for measurements of passive

segmental intervertebral and sacroiliac joints which rarely exceed Kappa 0.40 (Van Trijffel et al 2005, Van der Wurff et al 2000). Seffinger et al (2004) attributed these differences in reliability to differences in size of joints. We think, however, that differences may be more linked to a joint’s potential physiological range of motion. For instance, measurement of large joints with limited range such as the sacroiliac joint is associated with poor reliability, whereas measurement of small joints with greater range, such as the atlantoaxial spinal Libraries segment and finger joints, has been shown to be reliable (Cleland et al 2006, Glasgow et al 2003, Ogince et al 2007, Van der Wurff et al 2000). We also found that measuring large physiological ranges of motion, like that in the shoulder and in the wrist, frequently yielded satisfactory levels of reliability and note that these levels were predominantly as a result of using goniometers or inclinometers.