Reliability and Concurrent Validity of the Assessment of General Movements

Abstract

Background and Purpose: The results by previous studies which examined the reliability and validity of The assessment of general movements (GMsA) were inconsistent Therefore the purpose of this study was to determine the intra- and inter-rater reliability of GMsA in categorical and dichotomous ratings and concurrent validity of GMsA in comparison with the Morgan (MNNE) or Hammersmith Neonatal Neurobehavioral examination (HNNE) the Alberta Infants Motor Scale (AIMS) and the Peabody Developmental Motor Scales 2nd (PDMS-Ⅱ) on infants with term typical development and high risk of developmental disorders Methods: Sixty four high risk (experimental group) and 20 normal infants (control group) were recruited There were a total of 148 and 237 recordings to determine reliability and concurrent validity respectively Spontaneous movements were recorded in preterm writhing and fidgety period respectively The recording time was about 1 hour in preterm period and at least 15 minutes in writhing and fidgety period At the time of recording each infant were assessed with one or two assessment tools to determine concurrent validities At preterm age MNNE were assessed At age of 38 to 42 weeks gestational age (GA) HNNE was assessed From term age to 60 weeks GA AIMS was assessed At age of 49 to 60 weeks GA PDMS-II was assessed The quality of GMs was classified as normal poor repertoire chaotic and cramped-synchronized general movements in preterm and writhing period and normal abnormal and absent fidgety movements in fidgety period All recordings were assessed by categorical rating and further classified to dichotomous rating for analysis The inter-rater reliability examined the agreements between three physical therapists and the intra-rater reliability examined the agreements on the same recording after one month later Percentage of agreement and Cohen’s kappa were used to examine the reliability in categorical rating and the percentage of agreement and Prevalence Adjusted and Biased Adjusted Kappa (PABAK) was used to determine the reliability in dichotomous rating Percentage of agreement and Cramer’s V correlation coefficient were used to determine the correlations between GMsA and concurrent examinations Results and Discussions: The results showed that the intra- and inter-rater reliabilities were fair to moderate (mean percentage: 69%) and low to fair (mean percentage: 54%) in categorical rating respectively and fair to substantial (mean percentage: 78%) and low to fair (mean percentage: 67%) in dichotomous rating respectively The agreements within or between raters in this study were not good due to several possible reasons First the training method by the conception of seed teacher self-learning and discussions were insufficient Second few recordings for practice caused that raters were lack of confidence in using the assessment Third twenty percentage of recordings was in the transition phase of GMs and caused of increasing the difficulty of assessing Fourth the procedure (assessing together time of each assessment and duration between first and second rating) was different from previous studies The concurrent validities of GMsA were various in comparison with different examinations The correlation between GMsA and MNNE or HNNE were low It might be cause by limited reliability of neonatal examinations because of fragile and unstable infants in preterm life and early infancy The correlation between GMsA and the AIMS were moderate and highest in comparison to other examinations The correlation between GMsA and PDMS-II was fair The results also showed that the correlation between GMsA and fine motor quotient of PDMS-II was higher than between GMsA and gross motor quotient of PDMS-II It agreed with the results of previous study which found that fidgety movements related to later fine motor development Conclusions: The results indicated the reliabilities of GMsA were general fair and more discussion and practices for the raters who did not attend training course was needed The concurrent validities with other assessment tools were various The correlations between GMsA and other examinations were various However the age of the participants in each period was various in this study Further research could examine the concordance with other examinations at a certain age

Date of Award	2014 Sept 5
Original language	English
Supervisor	Bih-Jen Hsue (Supervisor)