The assisting hand assessment : continued development, psychometrics and longitudinal use

Abstract

For most people, using both hands together is a natural part of everyday activities. But for children with a unilateral disability in the arm and hand, activities requiring two hands are often a challenge. In order to evaluate interventions that optimize the use of the affected or assisting hand, valid and reliable measures are needed. The Assisting Hand Assessment (AHA) is a newly developed test with the expressed purpose of measuring and describing how effectively children with a unilateral disability in arm and hand use their assisting hand during bimanual tasks. An AHA assessment involves a video-taped play session, which elicits spontaneous use of the hands by using toys that require bimanual handling. The play session is scored using a 4-point rating scale for 22 items which are describing object-related hand actions. The AHA was initially validated for children aged 18 months to 5 years with unilateral cerebral palsy (CP) or obstetric brachial plexus palsy (OBPP) (Small Kids AHA). A test kit for older children containing two board games suitable for children aged 6-12 years has been developed (School Kids AHA). The aim of this thesis was to evaluate the psychometric properties of the AHA including the expanded age group and to describe the longitudinal development of assisting hand use in children with unilateral CP.

The validity of the AHA for ages ranging from 18 months to 12 years was investigated. Using a Rasch measurement model, 409 AHA s of children with unilateral CP or OBPP were analyzed. Results showed that the items measure a unidimensional construct and that the four-point rating scale overall functions well. An excellent targeting between the item difficulties and the children s abilities was found. A high person separation indicates that the AHA can effectively separate between children of different ability into seven distinct ability strata. Differential item functioning between diagnostic groups was identified in some items and discussed. The reliability of the AHA as regards the inter- and intrarater, retest and alternate forms reliability was investigated. In all trials high reliability coefficients (Intraclass Correlation Coefficients) were found, ranging from 0.97 to 0.99. The retest evaluation indicated that the smallest detectable difference measurable with the AHA is a change of 4 points (6% of the scale). It was further found that all tested versions, the Small Kids vs. School Kids AHA and the two board games in School Kids AHA, produce equal results and measure the same construct. The development of assisting hand use between the ages 18 months and 8 years was investigated among 43 children with unilateral CP. The children were regularly assessed using the AHA for on average 4.5 years. All children increased their assisting hand use during the course of the study. A non-linear mixed model was used to create separate average development curves for children at different levels on the Manual Ability Classification System (MACS). The maximum level (limit) of development differed between children in MACS levels I-III. The rate of change was similar in levels I and II and significantly slower in children at level III. The children were also divided into two groups based on their AHA-score at 18 months. The limit and rate of change differed significantly between the groups. The AHA score at 18 months can be used for approximate prediction of future development of assisting hand use.