Upper limb and hand patterns in cerebral palsy: Reliability of two new classifications.

AIM: To evaluate the inter- and intra-rater reliability of two previously developed classifications of upper limb and hand patterns.

METHOD: Two hundred and twelve films of patients with CP (118 of UL postures and 94 of hand tasks; median age 14, 3-46 years) were viewed by 18 examiners from 2 different rehabilitation centers, and one expert who had participated in the design of the classifications. They classed upper limb (3 patterns with sub-types) and hand patterns (2 patterns with subtypes) twice, at 2 months' interval. Inter- and intra-rater reliability were analysed.

RESULTS: Intra-rater and inter-rater reliability were very high for upper limb and hand patterns (0.87 < k < 0.92), and high for the subtypes (0.58 < k < 0.68). Examiners stated that both classifications were useful and feasible in clinical practice.

INTERPRETATION: Despite the single, short training session on use of the classifications, agreement between the examiners and the expert examiner was good to high, confirming that these classifications are easy to use and reliable. The classifications proposed here provide homogenous terminology for use in both clinical practice and research, to describe, evaluate and follow-up changes in upper limb and hand patterns in patients with cerebral palsy, particularly those with dyskinesia.