Abstract
Manual dexterity has strongly predicted functional independence for daily life activities among children with cerebral palsy (CP). The Jebsen-Taylor Hand Function Test (JTHFT) is the most widely used assessment tool for exploring manual dexterity in the CP population, though no research has yet examined its psychometric properties for this use. This cross-sectional study explored the validity and internal consistency of the JTHFT in an Italian sample of inpatient and outpatient children with CP aged between 6-18 years (35 girls and 49 boys). We calculated internal consistency with Cronbach’s alpha and tested validity against the Manual Ability Classification System (MACS) using Pearson’s correlation coefficient. To better understand how the JTHFT compares with different levels of the MACS, we performed dominant hand timing variability for each test item. Results showed excellent internal consistency with a Cronbach’s alpha of .944 and .911, respectively, for nondominant and dominant hands. There was also a statistically significant positive linear Pearson’s correlation coefficient between the JTHFT and the MACS (p < .01). We observed high variability in writing performance (Item 1 of the JTHFT) within this sample for each level of the MACS. This study confirms that the JTHFT is a valid assessment tool when used in children with CP aged 6-18 years.
Keywords
Introduction
Cerebral palsy (CP) describes a group of developmental disorders of movement and posture, causing limitations in activity and often accompanied by disturbances of sensation, cognition, communication, and/or behavior (Bax et al., 2005). CP is the most common cause of motor deficiency in young children (Cans, 2000). Limitations in hand function are common in this population, with almost 50% of children with CP presenting an arm–hand dysfunction (Fedrizzi et al., 2007).
To improve manual abilities in the CP population, different approaches include treatments that attempt to reduce hand impairment by increasing hand dexterity and strength (Moon et al., 2017) and others that emphasize hand function with movement constraint or promoting bimanual activities (Tervahauta et al., 2017). According to the International Classification of Functioning Disability and Health, rehabilitation professionals should attend to both body and activity domains. Assessment tools should be consistent with different hand skill domains so that both clinicians and researchers can focus rehabilitation activities on hand grip strength (force applied by the hand to pull objects), hand dexterity (skill and speed in performing tasks), or hand functioning (abilities to use the hand for performing daily activities). Among these methods, manual dexterity and grip strength of both hands are the strongest predictors of manual ability (Arnould et al., 2014) and functional independence for activities of daily living (Golubović & Slavković, 2014). A comprehensive approach to evaluating the assessment process can help rehabilitation professionals, educators, and families better develop children’s hand functioning to integrate these children into a community setting.
The Jebsen-Taylor Hand Function Test (JTHFT; Jebsen et al., 1969) is one of the most widely used assessment tools for investigating hand dexterity, and it has been a recommended tool for the CP population (Castelli et al., 2016; Wagner & Davids, 2012). The Italian version of the JTHFT-IT has been validated on samples of people without health problems (Culicchia et al., 2016), in a nonspecific disease sample (Nobilia et al., 2019) and among patients with rheumatoid arthritis (Savona et al., 2019). All of these studies demonstrated good psychometric (reliability and validity) properties. The JTHFT is very versatile for upper limb assessment, and it is widely used in many countries including Australia (Agnew & Maas, 1982), China (Li-Tsang et al., 2004), and Brazil (Ferreiro et al., 2010).The JTHFT has also been validated for other disorders affecting the upper limbs including hemiparesis (Ferreiro et al., 2010), muscular dystrophies (Artilheiro et al., 2018), stroke (Allgöwer & Hermsdörfer, 2017), Parkinson’s disease (Mak et al., 2015), and carpal tunnel syndrome (Davis Sears & Chung, 2010). Among patients with CP, the JTHFT has been widely used by various researchers (Adler et al., 2015; De Brito Brandão et al., 2010; Gordon et al., 2011). Recently, Araneda et al. (2019) explored reliability and responsiveness of the JTHFT and the Box and Block Test among patients with CP and found evidence of good reliability and the ability to measure change over time. However, as no prior research has examined the JTHFT validity, we performed this study to explore the validity of the JTHFT-IT in an Italian population of children with CP by correlating this test with another measure of manual abilities.
Method
This study was conducted at Bambino Gesù Children’s Hospital, Sapienza University of Rome, and Rehabilitation & Outcome Measures Assessment (ROMA) association. These institutes comprised the institutional review board that approved the current protocol. In the last year, the research group created different validation studies in Italy (Anna et al., 2020; Berardi, Biondillo, et al., 2018; Berardi, Dhrami, et al., 2018; Covotta et al., 2018; Dattoli et al., 2018; Galeoto, Berardi, et al., 2019; Galeoto et al., 2018; Galeoto, Iori, et al., 2019; Galeoto, Scialpi, et al., 2019; Ruggieri et al., 2018).
Participants and Procedures
We recruited our participant sample during the period from September 2018 to April 2019 from inpatient and outpatient services of two different hospitals in Rome, Bambino Gesù Children’s Hospital, and Policlinico Umberto I University Hospital. We estimated the needed sample size by assuming an ideal ratio of respondents to items of 10:1 (Tabachnick & Fidell, 2007); considering that the JTHFT-IT has seven items, the minimum sample size required for validation is 70 participants. Furthermore, considering a 95% confidence interval (α = .05) and a dropout rate of 30%, our analysis suggested a need to approach at least 91 children with CP. This convenience sample of 91 children met inclusion criteria of having a diagnosis of CP and being aged between 6-18 years. Exclusion criteria were children with other medical complications (e.g., cerebral visual impairment and ventilator-dependent). Recruitment strategies included hospital meetings with children and families to explain the research, its purposes and data management strategies. As seven children opted not to participate, our data analysis was based on 84 children enrolled into the study, 35 girls and 49 boys. We obtained informed written consent from parents/guardians of all participating children (Galeoto et al., 2015, 2016), and we obtained assent from children when appropriate (according to age and cognitive ability level). Before initiating the study, a medical doctor completed a study-specific clinical record for each child that described sociodemographic information and patient diagnosis. Then, three trained rehabilitation professionals (two occupational therapists and a physical therapist) administered the JTHFT-IT (Nobilia et al., 2019) following the administration of Manual Ability Classification System (MACS; Eliasson et al., 2006); the clinician who administered the JTHFT-IT was blinded from data regarding the MACS, administered by others.
Instruments
The JTHFT-IT (Nobilia et al., 2019) consists of seven unimanual items that are administered using standardized procedures and verbal instructions and performed first with the nondominant hand (NDH) and then with the dominant hand (DH). The functional tasks on the JTHFT-IT include (a) writing a 24-letter, third-grade reading difficulty sentence; (b) turning 300-500 (7.62 cm by 12.7 cm) cards in simulated page turning; (c) picking up small common objects (including pennies, paper clips, and bottle caps) and placing them in a container; (d) stacking checkers; (e) simulated feeding; and (f) moving light cans and heavier (1 pound) cans. The subtests are scored by recording with a stopwatch the number of seconds required to complete each task. We gave participants ample time to complete the test in relation to their decreased hand function. The maximum time allowed to perform tasks was 120 seconds. The JTHFT-IT demonstrated good reliability (Nobilia et al., 2019; Savona et al., 2019) and good validity (Berardi et al., 2019). Specifically for CP population, the test showed excellent reliability in short time (intraclass coefficient > .900; Araneda et al., 2019).
The MACS (Eliasson et al., 2006) aims to classify how children with CP use both hands when handling objects in their daily activities and what impact environmental and personal factors have on the children’s performance. MACS levels are based on children’s ability to initiate handling objects by themselves and their need for assistance or adaptation to perform manual, daily, age-appropriate life activities. The MACS covers the age-group between 4-18 years in levels ranging from I-V (Jeevanantham et al., 2015). Children at Level I are able to handle objects easily, while those at Level II handle most objects, but with mildly reduced quality or speed, and those at Level III handle objects with difficulty and need help to prepare or change activities. At Level IV, children handle a limited number of objects and require continuous support to partially conclude the activities, and at Level V, children are unable to handle objects (Eliasson et al., 2006). As a classification system, the MACS demonstrated good validity and excellent agreement between therapists (.97) and between parents and therapists (.96; Eliasson et al., 2006).
Data Analysis
All statistical analyses were performed using IBM SPSS Statistics (Version 20.0; IBM Corp., Armonk, NY). Data were summarized and analyzed using frequency tables, means, and standard deviations. Internal consistency is typically based on correlations between different items on the same test (Tang et al., 2014). We used Cronbach’s alpha to calculate internal consistency or reliability, and we considered these coefficients indicative of acceptable reliability when they exceeded 0.7 (Nunnally, 1979). Conceptually, α is the mean of all possible split-half correlations for a set of items. Validity is a second important parameter in determining an instrument’s psychometric quality. The term validity refers to whether or not the test measures what it claims to measure (Lakshmi & Akbar Mohideen, 2013). We calculated the relationship (i.e., Pearson’s correlation coefficients) between the JTHFT-IT and MACS scores for the entire study sample and considered correlation coefficients significant at p < .05. Pearson’s correlation coefficients range from 0 (indicating no linear relationship between variables) to 1 (indicating a perfect linear relationship) and we interpreted the obtained values as follows: (a) < 0.3 = a weak linear relationship, (b) 0.3 to 0.69 = a moderate linear relationship, and (c) values ≥ 0.7 = a strong linear relationship (Ratner, 2009).
Results
We recruited 91 children who met the exclusion/inclusion criteria. Seven children decided not to participate, and 84 children were enrolled in the study. Table 1 shows the participant characteristics.
Sample Characteristics.
We calculated the JTHFT item consistency with Cronbach’s alpha and obtained values of .944 and .911, respectively, for the NDH and DH. Table 2 illustrates that all items contribute positively to the scale’s homogeneity.
JTHFT-IT Internal Item Consistency for the Nondominant and Dominant Hands.
There were also highly significant positive Pearson’s correlation coefficients between the JTHFT-IT and the MACS results for both the DH (values between .557 and .738) and the NDH (values between .529 and .724; p < .01). These results are summarized in Table 3, and, to better understand how the DH timing variability between the JTHFT-IT and the MACS interrelates at the JTHFT item level, see Figure 1A to 1G.
JTHFT-IT Item Correlations With the MACS for Both Nondominant and Dominant Hands.
**<0.01
JTHFT Item Correlations With the MACS. Figures from (A) to (G) represent items from 1 to 7, respectively. MACS = Manual Ability Classification System.
Discussion
This study was the first to investigate the JTHFT validity among children with CP, and we relied on this measure’s positive relationship with manual ability, as measured by the MACS, in this validity effort. We also demonstrated very good internal consistency of the JTHFT-IT for this clinical group, consistent with similarly strong values reported for populations with muscular dystrophy (Artilheiro et al., 2018), rheumatoid arthritis (Savona et al., 2019), stroke (Ferreiro et al., 2010), and among populations without diseases (Nobilia et al., 2019). Internal consistency is a measurement of whether several items that propose to measure the same general construct produce similar scores. Our finding demonstrates homogeneity of the JTHFT-IT items when used in CP population.
Regarding the correlation of the JTHFT-IT with MACS, it is not possible to make a direct comparison with other studies. However, we found a very high, statistically significant correlation between the JTHFT-IT measure of manual dexterity and the MACS measure of manual functions for both the DH and the NDH (p <.01). This finding highlights other investigations’ similar results regarding the interdependence of manual dexterity, manual ability, and abilities to the perform different activities of daily living (Arnould et al., 2014; Golubović & Slavković, 2014). In a recent study, Araneda et al. (2019) investigated the JTHFT’s reproducibility and responsiveness. Repeated measures over time and between different raters were not significantly different. In addition, Araneda et al. showed the JTHFT’s good ability to detect change in the CP population in relation to different levels of the MACS.
This study sought to explain the close relationship between dexterity and manual ability. Specific comparisons between the JTHFT-IT and MACS seem to have delineated these differences and peculiarities. For example, children with Level V skills on the MACS (do not handle objects and have severely limited ability to perform even simple actions, so requiring total assistance) were unable to perform both Item 3 (pick-up small objects and placing them in a container) and Item 4 (stacking checkers), children at Levels III or IV on the MACS have more variability in performing tasks while the same activities are easier for children at Levels I and II. In the same manner, children at Level V of the MACS, with specific difficulties on Items 2 (cards turning), 5 (simulated feeding), and 6 (moving empty cans), showed higher variability in the time required for performing actions than those children at more skilled levels of the MACS. The positive correlations between the MACS classification system and JTHFT indicate that as children demonstrate more skillful levels on the MACS their time required to complete JTHFT items increases.
On Item 1 of the JTHFT-IT (writing), we observed significant variability in time required for children at each level of the MACS. In fact, children at Level I of the MACS can complete Item 1 within 15–75 seconds, while children at Levels II and III show more variability and required more time in performing task (but they can perform it), while children at Level IV seem to have an important difficulty in that they tended not to complete the task in the allotted time. Children at Level V of the MACS cannot perform it at all, probably because of a relationship between their cognitive level and their manual ability. In fact, manual ability in children with CP varies, due to neurologic subtype and cognitive level (Oskoui et al., 2013). The aspect of cognitive ability needs to be further investigated with quantitative neuropsychological measures.
Regarding limitations to the present investigation, our relatively small participant sample did not permit an unequivocal conclusion regarding an apparent strong relationship between manual function and manual dexterity, and it would be risky to try to explain underlying reasons for the JTHFT dexterity performance variability that we uncovered between children grouped at the same MACS level of manual ability. In addition, it would be useful to investigate correlations between stratified manual abilities and JTHFT performances among children with varied neurologic subtype and also to correlate manual JTHFT dexterity with other outcome measures that are specific for different functional activities.
In conclusion, considering that CP causes different impairments of the hand meaning that different rehabilitation approaches are used, different assessment tools are also necessary. Therefore, by comparing tools that measure different aspects of the hand, it is possible to observe how hand abilities influence each other. In fact, this study shows how dexterity and practical functioning are related. In the end, this study lends significant support to the use of the JTHFT-IT as a valid assessment tool for 6-18 year-old children with CP.
