Eleven boys and 14 girls, visiting a mainstream school, aged 4–12 years, were tested neurologically by 3 investigators. Inter- and intra-assessor reliability were based on videotapes of the assessments. To determine test–retest reliability children were re-assessed after about 1 month. The various forms of reliability were calculated for neurological classification (normal, simple MND, complex MND), clusters of dysfunction and single items.
Twelve girls and 7 boys showed a normal neurological condition; 2 girls and 1 boy were classified as having simple MND and 3 boys as having complex MND. The 3 forms of reliability for neurological classification were good (κ=0.71–0.83). Reliability for the majority of cluster scores was good. Test–retest agreement was moderate for the clusters reflexes and coordination and poor for fine manipulation; inter-assessor agreement was moderate for the clusters coordination and fine manipulation; intra-assessor agreement moderate for fine manipulation. Reliability for the majority of items was good.
In conclusion, the Touwen examination has a moderate to good reliability when applied in a relatively healthy population. Whether reliability is similarly good in populations of children with minor developmental disorders has to be determined.