Concurrent validity and discriminative ability of Dutch performance-based motor tests in 5 to 6 years old children.

Abstract

To assess the concurrent validity and discriminative ability of total, gross and fine motor (TM, GM and FM) scores of Dutch performance-based motor tests, the Baecke-Fassaert Motor Test (BFMT) and the 8- and 4-Skills Scan (SkSc) with the Movement Assessment Battery (MABC) for children at age 5. 116 Dutch children (40.3% boys) were included. Spearman's rho correlations and area under the curves (AUC) were assessed. Correlations between the TM scores of the tests were strong (absolute values from 0.58 to .65); the correlations between the GM scores and the FM scores between and within tests were weaker (absolute values from 0.30 to 0.45). Related to the cut-off (15th percentile) of the MABC, the AUC of the BFMT, 8- and 4-SkSc, the AUC was 0.853 (95% CI: 0.757-0.949), 0.905 (95% CI: 0.837-0.972) and 0.844 (95% CI: 0.730-0.957), respectively. At optimal cut-offs, the sensitivity and specificity of the BFMT, the 8- and 4-SkSc were 78.6 and 78.4%, 92.2 and 73.2%, 78.6 and 76.3%, respectively. All tests had a reasonably high discriminative ability, but validation with the MABC-2 and adaptations are needed to meet the requirements for screening (i.e. sensitivity ≥80% and specificity ≥90%). The relatively weak correlation between GM and FM scores implies that tests should be normalized and validated for GM and FM ability, separately.