Validation of the ICF-CY-based MobiScreen for Primary School Children: An Update

Introduction: A new mobility screening based on the International Classification of Functioning, Disability and Health for children and youth ICF-CY for primary school children is validated. Children with a motor development disorder need to be identified early in order to initiate appropriate therapies. The construct of mobility in the sense of the ICF-CY seems to be an ideal basis for a screening that allows different disciplines to communicate in a unified language.
Hypothesis: The new MobiScreen for primary school children meets all psychometric properties and is able to classify children into conspicuous and not conspicuous.
Methods: In total, 608 primary school children (53 % boys, 47 % girls, age 7.14
.70 years) of the first and second grades of primary school participated to undergo the MobiScreen 6-8 and a further motor test (MOBAK 1-4 or DMT 6-18). In total, 568 primary school children (56 % boys, 44 % girls, age 9.36
.79 years) of third and fourth grades of primary school participated to undergo the MobiScreen 8-10 and a further motor test (MOBAK 1-4 or DMT 6-18). Item split times including penalty seconds for performance errors as well as total time including penalty seconds are used to evaluate Inter-Rater-Reliability, Internal Consistency, Criterion Validity, Construct Validity and Diagnostic Accuracy. The significance level was set at p<.05.
Results: MobiScreen 6-8: Inter-Rater-Reliability coefficients range from r=.891 to r=.998. Cronbach's
reached a value of
=.722, inter item correlations range from r=.227 to r=.620. Results of the explorative factor analysis show a single factor model with an eigenvalue of 2.645 with a variance explanation of 53% for this factor. Coefficients for Criterion Validity are all significant and range from r=-.446 (MOBAK 1-4) to r=.553 (DMT 6-18). Discriminant analysis shows significant differences in all tasks between healthy children and children with a developmental disorder. For potential cutoff values between 24 and 28 seconds, values for sensitivity range from 1.00 to .86. The AUC reaches a value of .82. MobiScreen 8-10: Inter-Rater-Reliability coefficients range from r=.996 to r=1.000. Cronbach's
reached a value of
=.540, inter item correlations range from r=.161 to r=.367. Results of the explorative factor analysis show a single factor model with an eigenvalue of 2.074 with a variance explanation of 41 % for this factor. Coefficients for Criterion Validity are all significant and range from r=-.386 (MOBAK 1-4) to r=-.528 (DMT 6-18). Discriminant analysis shows significant differences in all tasks between healthy children and children with a developmental disorder. For potential cutoff values between 28 and 32 seconds, values for sensitivity range from .89 to .80. The AUC reaches a value of .79.
Conclusions: The given criteria were achieved in a medium to high degree and the screening is able to classify children into conspicuous and not conspicuous. Further studies should follow to enlarge the sample size for improving psychometric properties and definitively fix the cutoff values for both versions.