The General Movement Assessment (GMA) is a reliable method using video recordings of an infant's general movements (GMs) in the supine position to allow trained clinicians to identify typical or atypical motor behaviors, in particular fidgety GMs that are most predictive at 3 to 5 months after term age.1–6 Home videos have been used as a valuable resource for assessments in pediatric rehabilitation, such as social engagement of children with autism7–10 and contingent responsiveness of children with epilepsy,11–13 as well as earlier identification of potential neurological disorders in young infants.14,15 Einspieler et al14 reported atypical GMs in 14 infants with Rett syndrome during their first 4 months of life based on the retrospective review of their home videos. While all videos included in the above study were taken by parents, no additional video aspects were reported. The result of the above study suggests that videos recorded by parents may be useful for the GMA. Conversely, Ricci et al15 investigated 12 infants (4 born at term and 8 born preterm) after discharge from the neonatal intensive care unit (NICU). Out of 42 videos recorded, 28 were obtained by the neonatal neurologist in the NICU during the writhing period and 14 were recorded at home by parents during the fidgety period. In addition to difficulties with uploading the videos, they reported that 9 (21%) could not be interpreted (4 recorded by parents and 5 by the neonatal neurologist) due to infant irritability, parent interaction with the infant during video recording, or infant extremities not being clearly visible (mainly because of clothing). This suggests a need to improve the usability of videos for the GMA, including those recorded by the caregivers.
The usability of videos obtained from caregivers as a resource in the assessment of early infants has been examined previously using the Alberta Infant Motor Scale (AIMS). Boonzaaijer et al16 investigated the agreement between the AIMS scores based on live observation and video observation. Agreement in scores was reported (intraclass correlation coefficient [ICC] > 0.9), suggesting comparability. Furthermore, according to 94% of participating parents, recording their infants' movement was easy to perform, even though parents were provided only verbal instructionals. When recording AIMS videos, the infants' gross motor performance in different positions during daily life is filmed, making it easy for parents. In contrast, recording of GMs' videos is more complex. In order for the video to be useful for GMA interpretation, the following conditions are required: the infant lying in a supine position with an appropriate state of arousal, extremities exposed for a sufficient duration to capture infants' movements, no facilitation or interaction between parent and infant, and a simple background. Verbal instructionals may be insufficient to convey these requirements; therefore, parents may need additional information on filming GMs to ensure video quality for interpretation.
The use of instructional leaflets for encouraging parents to measure their children's weights and heights at home is beneficial.17 Tjiam et al18 demonstrated that compliance with occlusion therapy for amblyopia improved after providing a leaflet to parents of children in low socioeconomic status areas. Providing an instructional leaflet might improve the usability of GMs' videos recorded by parents. A comparison of GMA interpretations between video sources for the same group of infants might be clinically useful.
The 3 primary aims of this study were to (a) develop an instructional leaflet on home video recordings of infants' GMs who reside in Taiwan, (b) investigate the consistency of GMA results from 2 video sources, in the clinical setting by a GMA-certified physical therapist (PT) or at home by parents, and (c) report the concurrent validity of the GMA at a corrected age (CA) of 3 months, and the predictive validity of the GMA at a CA of 6, 9, and 12 months, using AIMS cutoff scores to compare between the video sources.
The data were generated as part of a study that examined postural control both in infants developing typically and in infants with motor delays (the original cohort study). The Institutional Review Board (IRB) of Chang Gung Medical Foundation, Lin-Kao, Taiwan, reviewed and approved the protocol of this original cohort study before recruitment (IRB No. 101-3548C). Infants were recruited from the NICU and outpatient clinics of Lin-Kao Chang Gung Memorial Hospital-Children's Hospital, Taiwan, between 2009 and 2011. Parents provided consent for the original study.
INSTRUCTIONAL LEAFLET ON HOME THE GMA
Two pediatric PTs designed the instructional leaflet and the associated content validity index (CVI) checklist. The CVI was also developed by a certified GMA pediatric nurse, a NICU pediatric nurse, and 2 caregivers.
Pediatric PTs created the instructional leaflet in traditional Chinese according to the GMA manual (see Supplemental Digital Content 1, available at: http://links.lww.com/PPT/A276).1 A total of 10 items in 3 domains were provided: camera and environmental settings for video recording (6 items), infant preparation (2 items), and scenes to stop recording (2 items). Each item included text descriptions and photographs to show the parents correct and incorrect settings if possible. The custom-made CVI checklist was designed in part to determine (a) reasonable clarity and simplicity of text description, (b) clarity and simplicity of photographs and simplicity of text regarding demonstration of correct procedures, and (c) clarity and simplicity of photographs and text with regard to demonstrating incorrect procedures. The CVI checklist responses were scored using a 4-point Likert scale from 1 (not relevant) to 4 (highly relevant). The CVI for each item and domain was determined by calculating the proportion of items that received a rating of at least 3 by the 6 experts. Following Waltz and Bausell,19 the CVI of a measure evaluated by 6 experts needed to be at least 83% to be considered acceptable (Table 1 and Supplemental Digital Content 2, available at: http://links.lww.com/PPT/A277).
The appropriateness of the instructional leaflet was examined by the CVI.20 The item-level CVI (I-CVI) was used to determine the number of item agreements in the instructional leaflet. The scale-level CVI (S-CVI) was used for the number of domain agreements.
CONSISTENCY AND CONCURRENT VALIDITY OF THE GMA BETWEEN VIDEOS
The inclusion criteria were (1) provision of the instructional leaflet on home GMs' video recordings for GMA interpretation to parents, which was used to record and then provide at least 1 home GMs' video, and (2) an infant postmenstrual age (PMA) of 49 to 60 weeks. The exclusion criteria were (1) high risk infants with known congenital malformations or chromosomal abnormalities, and (2) infants who were unable to lie in the supine position due to medical devices, such as a ventilator. The infants' chronological ages, birth weights, gestational ages, and risk factors for neurological impairments are shown in Table 2. A total of 29 GMs videos of the 116 in the original cohort were provided by parents and those children received follow-up with the AIMS at a CA of 3, 6, 9, and 12 months (Figure 1).
The examiner was a PT who had completed the basic training course and passed the certification examination on Prechtl's GMA in 2011. The intrarater reliability of the GMA has been established.21 The same examiner reexamined the videos 1 year later to establish the test-retest reliability (percentage agreement: 98.3%; κ: 0.931). Final GMA results were based on the second GMA interpretation during test-retest reliability examination.
A GMA-certified PT filmed the infants while conducting the clinical evaluation. At the same time, parents were invited to video their infants' GMs at home according to the instructional leaflet. Only parents who provided a home video were included in the study. The GMs' videos had to have recorded at least 1 fidgety stage (PMA at 49-60 weeks). If the parent filmed more than 1 GMs, we selected the video with the time interval closest to that recorded by the PT. Based on the videos, the observer used the GMA to classify the infants' GMs,1,22–25 as the target assessment, for the period of fidgety movements.
The gross motor outcomes of the infants were determined using the AIMS, as the criterion measure, around a CA of 3, 6, 9, and 12 months. Interrater reliability was established prior to the start of the study by 2 PTs from the original cohort study team (ICC(2,1) = 0.981). Past studies have suggested that the 5th and 10th-percentile cutoffs for infants at different months of age produce different sensitivities and specificities in the AIMS.26 Thus, the sensitivities and specificities in 5th- and 10th-percentile cutoffs, respectively, were compared in this study.
SPSS version 19.0 was used for statistical analyses. Descriptive statistics represented the infants' assessment results. The consistency between the 2 videos was determined using Cohen's κ,27 where a κ value less than 0.20 is often considered “poor” agreement, 0.21 to 0.40 “fair” agreement, 0.41 to 0.60 “moderate” agreement, 0.61 to 0.80 “good” agreement, and 0.81 to 1.00 “excellent” agreement.28 The concurrent validity of the GMA was represented using the same metrics. The sensitivity, specificity, positive predictive value, negative predictive value, false-positive rate, false-negative rate, positive likelihood ratio, negative likelihood ratio, and power were determined.
The average agreement of the text description, correct demonstration, and incorrect demonstration in the instructional leaflet was excellent, and all items were retained as achieving CVI scores more than 83% (Table 1 and Supplemental Digital Content 2, available at: http://links.lww.com/PPT/A277).
Consistency and Concurrent Validity of the GMA Between Videos
The GMA results from GMs' videos filmed by the PT or parents were identical except for subject 8 (Table 2). Consistency of GMA results between the 2 videos was very good (κ: 0.869; P < .05). Concurrent validity between the AIMS and the GMA for both the 5th- and 10th-percentile cutoffs was fair to moderate based on the κ value in infants with a CA of 3 months (κ: 0.266-0.525; Table 3). Predictive validity between the AIMS and the GMA for both the 5th- and 10th-percentile cutoffs indicated good to high correlation in infants with a CA of 6, 9, and 12 months, and the sensitivity and the specificity of the GMA were both more than 75% (Table 4). The motor trajectories at all ages are graphed in Figure 2.
The results of this study support the effectiveness of the instructional leaflet in guiding parental home recording of infants' GMs. The consistency of the GMA results between the 2 videos was high. Some precautions must be taken during the GMA assessment process to ensure that video quality is sufficient for interpretation by a certified PT. After the parents read the instructional leaflet, they can determine the most appropriate time for filming the GMs to avoid recording when the infant is crying or sleepy. Before conducting this study, we also surveyed the expert content validity of the instructional leaflet, in which parents responded that photographic instructions were clearer than text instructions only.
Boonzaaijer et al16 found that AIMS scores based on home video recordings were comparable to assessments based on live observations. As our participants were younger than in their study, the parents' ability to record the videos of GMs had to be confirmed before accepting such an assessment. The very high consistency in our study, where all home-recorded GMs videos were suitable, was due to the use of the instructional leaflet, and compares favorably with Ricci et al's15 finding that 21% of videos were not suitable for interpretation.
Further review of the GMs video of subject 8 showed that video recording by the parents was consistent with the instructional leaflet, and the same results were found when both videos were further assessed by another GMA-certified PT. Therefore, the different GMA interpretation between videos recorded by PT and parents cannot be attributed to the compliance of the family in this case.
The second purpose of this study was to examine the feasibility of home-recording GMs for GMA assessment by a clinical certified PT. The results suggest that all interpretations of recorded GMs during the fidgety period, either by the parents at home or by a clinical certified PT, were satisfactory, indicating that home video recording of GMs by parents is feasible. Recently, computer techniques have been developed that can capture and analyze GMs' videos,29–31 such as General Movement Toolbox (GMT).29 With the help of those advanced techniques for the GMA, videos of GMs can be analyzed by clinicians who are not familiar with the GMA, or be used to confirm GMA results. The use of home-recorded videos is likely to facilitate early detection of neurodevelopmental disorders in newborns. In addition, parents could choose the time for recording the GMs at home, overcoming the issue of GMs not appearing as expected in the clinic, resulting in false positives (hypokinesis, no GMs during the recording) or a necessity to extend the recording period, thus increasing time and labor costs. Furthermore, as GMA evaluation requires long-term assessment to track developmental trends in children, the use of such an assessment method is likely to be very helpful for the children and their families.
In contrast to the study by Snider et al32 that only recruited preterm infants, our study enrolled preterm and term newborn infants. The concurrent validity of the GMA between the videos was comparable with the findings of previous studies.33–34 The concurrent validity of home videos was equally high as those from the clinical setting using either 5th or 10th percentiles as the cutoff in the AIMS at a CA of 3 months. This result may depend on recording in adequate locations and suitable timings by parents. Newborns or infants display more relatively stable moods and motor performance at home than in the clinic. In a clinical environment, not only the light but also sound is complex and the timing of recording must coincide with the schedules of clinical staff. Infants are unstable, and their state and motor performance may be affected by hunger or wet diapers. In contrast, parents are able to choose the best timing to record the infant, increasing the usability of the GMs' video, resulting in more accurate interpretation of the results.
Glascoe and Byrne35 suggested a sensitivity and specificity more than 70% in a good children screening tool, a threshold that was reached or exceeded in our study at a CA of 6, 9, and 12 months. As the predictive validity results of our study are comparable with those of other studies,5,33,34,36–42 it can be regarded as satisfactory for the interpretation of GMA results obtained from both video sources. Sensitivity and specificity were better when using 10th and 5th percentiles as the cutoff in the AIMS at a CA of 6 months and 9 months, respectively, a finding is similar to that of Albuquerque et al.43 Overall, it can be concluded that interpretation of GMA results from the home GMs' videos allows early detection of possible motor delay and rapid referral to an intervention program.
Our study has limitations. The study was conducted by a single GMA-certified PT who was aware of the medical history of the infants, which may have biased the interpretation of results. We recommend that future research discuss interrater reliability of the GMA based on home GMs videos determined by masked examiners or GMT.29 It would have been preferable if the numbers of infants developing typically and at-risk infants had been similar. This study was a clinical study and recruited participating parents mainly from the outpatient clinic. While 116 infants were included in the original cohort study, the parents of only 75 infants voluntarily provided GMs' videos. Thirty-six infants were unable to complete each AIMS follow-up, because the time could not be matched, or they lived too far away from the clinic. It is of interest to us how to motivate the parents to use new technology, such as smartphones, to participate in the follow-up studies.
The instructional leaflet on home video recording of GMs for the GMA successfully informed parents of the correct way of recording the videos, and the quality of the videos was sufficient for clinicians to interpret. We successfully demonstrated consistency of GMA results between videos recorded by the GMA-certified PT in the clinical setting and parents at home.
The authors would like to thank the parents and their children for their participation. We are appreciative and grateful for the continued support of the physical therapists and physicians from the Department of Physical Medicine and Rehabilitation at the Lin-Kao Chang Gung Memorial Hospital and Tao-Yuan Chang Gung Memorial Hospital of Chang Gung Medical Foundation in Taiwan; we are especially appreciative of the support from Dr Chia-Ling Chen and Dr Yu-Cheng Pei. We would like to thank Anthony Abram (www.uni-edit.net) for editing and proofreading this article.
1. Einspieler C, Bos A, Ferrari F, Cioni G. Prechtl's Method on the Qualitative Assessment of General Movements in Preterm, Term and Young Infants (Clinics in Developmental Medicine No. 167). London, England: Mac Keith Press; 2004.
2. Einspieler C, Cioni G, Paolicelli PB, et al The early markers for later dyskinetic cerebral palsy are different from those for spastic cerebral palsy. Neuropediatrics. 2002;33(2):73–78.
3. Einspieler C, Prechtl HF. Prechtl's assessment of general movements: a diagnostic tool for the functional assessment of the young nervous system. Ment Retard Dev Disabil Res Rev. 2005;11(1):61–67.
4. Prechtl HF. State of the art of a new functional assessment of the young nervous system. An early predictor of cerebral palsy. Early Hum Dev. 1997;50(1):1–11.
5. Prechtl HF, Einspieler C, Cioni G, et al An early marker for neurological deficits after perinatal brain lesions. Lancet. 1997;349(9062):1361–1363.
6. Ferrari F, Cioni G, Prechtl HF. Qualitative changes of general movements in preterm infants with brain lesions. Early Hum Dev. 1990;23(3):193–231.
7. Clifford SM, Dissanayake C. The early development of joint attention in infants with autistic disorder using home video
observations and parental interview. J Autism Dev Disord. 2008;38(5):791–805.
8. Watson LR, Crais ER, Baranek GT, et al Communicative gesture use in infants with and without autism: a retrospective home video
study. Am J Speech Lang Pathol. 2013;22(1):25–39.
9. Maestro S, Muratori F, Barbieri F, et al Early behavioral development in autistic children: the first 2 years of life through home movies. Psychopathology. 2001;34(3):147–152.
10. Ozonoff S, Iosif A-M, Young GS, et al Onset patterns in autism: correspondence between home video
and parent report. J Am Acad Child Psy. 2011;50(8):796–806.
11. Dash D, Sharma A, Yuvraj K, et al Can home video
facilitate diagnosis of epilepsy type in a developing country? Epilepsy Res. 2016;125:19–23.
12. Stephenson J, Breningstall G, Steer C, et al Anoxic‐epileptic seizures: home video
recordings of epileptic seizures induced by syncopes. Epileptic Disord. 2004;6(1):15–19.
13. Johansen JR, Lindahl G, Sandstedt P. Home-video observation of seizures in children with epilepsy-impact on quality of family life. Seizure. 1999;8(6):356–357.
14. Einspieler C, Kerr AM, Prechtl HF. Abnormal general movements in girls with Rett disorder: the first 4 months of life. Brain Dev. 2005;27(suppl 1):S8–S13.
15. Ricci E, Einspieler C, Craig AK. Feasibility of using the general movements assessment of infants in the United States. Phys Occup Ther Pediatr. 2017:1–11.
16. Boonzaaijer M, van Dam E, van Haastert IC, et al Concurrent validity between live and home video
observations using the Alberta Infant Motor Scale
. Pediatr Phys Ther. 2017;29(2):146.
17. Huybrechts I, Beirlaen C, De Vriendt T, et al Validity of instructional leaflets for parents to measure their child's weight and height at home: results obtained from a randomised controlled trial. BMJ open. 2014;4(2): e003768.
18. Tjiam A, Holtslag G, Van Minderhout H, et al Randomised comparison of 3 tools for improving compliance with occlusion therapy: an educational cartoon story, a reward calendar, and an information leaflet for parents. Graef Arch Clin Exp. 2013;251(1):321–329.
19. Waltz CF, Bausell BR. Nursing Research: Design Statistics and Computer Analysis. Philadelphia, PA: FA Davis; 1981.
20. Lynn MR. Determination and quantification of content validity. Nurs Res. 1986;35(6):382–385.
21. Yeh KK, Liu WY, Wong AMK, et al Intra-observer reliability of Prechtl's method for the qualitative assessment of general movements in Taiwanese infants. J Phys Ther Sci. 2016;28(5):1588–1594.
22. Bruggink JL, Einspieler C, Butcher PR, et al Quantitative aspects of the early motor repertoire in preterm infants: do they predict minor neurological dysfunction at school age? Early Hum Dev. 2009;85(1):25–36.
23. Ferrari F, Prechtl HF, Cioni G, et al Posture, spontaneous movements, and behavioural state organisation in infants affected by brain malformations. Early Hum Dev. 1997;50(1):87–113.
24. Fjortoft T, Einspieler C, Adde L, et al Inter-observer reliability of the “Assessment of Motor Repertoire—3 to 5 Months” based on video recordings of infants. Early Hum Dev. 2009;85(5):297–302.
25. Fjortoft T, Grunewaldt KH, Lohaugen GC, et al Assessment of motor behaviour in high-risk-infants at 3 months predicts motor and cognitive outcomes in 10 years old children. Early Hum Dev. 2013;89(10):787–793.
26. Darrah J, Piper M, Watt MJ. Assessment of gross motor skills of at‐risk infants: predictive validity of the Alberta Infant Motor Scale
. Dev Med Child Neurol. 1998;40(7):485–491.
27. Cohen J. A coefficient of agreement for nominal scales. Educ Psychol Meas. 1960;20(1):37–46.
28. Altman DG. Practical Statistics for Medical Research. London, England: Chapman & Hall; 1991.
29. Adde L, Helbostad JL, Jensenius AR, et al Using computer-based video analysis in the study of fidgety movements. Early Hum Dev. 2009;85(9):541–547.
30. Marchi V, Hakala A, Knight A, et al Automated pose estimation captures key aspects of general movements at 8‐17 weeks from conventional videos. Acta Paediatr. 2019;108(10):1817–1824.
31. Adde L, Yang H, Sæther R, et al Characteristics of general movements in preterm infants assessed by computer-based video analysis. Physiother Theory Pract. 2018;34(4):286–292.
32. Snider LM, Majnemer A, Mazer B, et al A comparison of the general movements assessment with traditional approaches to newborn and infant assessment: concurrent validity. Early Hum Dev. 2008;84(5):297–303.
33. Cioni G, Ferrari F, Einspieler C, et al Comparison between observation of spontaneous movements and neurologic examination in preterm infants. J Pediatr. 1997;130(5):704–711.
34. Cioni G, Prechtl HF, Ferrari F, et al Which better predicts later outcome in full-term infants: quality of general movements or neurological examination? Early Hum Dev. 1997;50(1):71–85.
35. Glascoe FP, Byrne KE. The accuracy of 3 developmental screening tests. J Early Intervention. 1993;17(4):368–379.
36. Stahlmann N, Härtel C, Knopp A, et al Predictive value of neurodevelopmental assessment versus evaluation of general movements for motor outcome in preterm infants with birth weights <1500 g. Neuropediatrics. 2007;38(02):91–99.
37. Constantinou JC, Adamson-Macedo EN, Mirmiran M, et al Movement, imaging and neurobehavioral assessment as predictors of cerebral palsy in preterm infants. J Perinatol. 2007;27(4):225.
38. Adde L, Rygg M, Lossius K, et al General movement assessment
: predicting cerebral palsy in clinical practise. Early Hum Dev. 2007;83(1):13–18.
39. Paro-Panjan D, Sustersic B, Neubauer D. Comparison of 2 methods of neurologic assessment in infants. Pediatr Neurol. 2005;33(5):317–324.
40. Garcia JM, Gherpelli JLD, Le1 CR. The role of spontaneous general movement assessment
in the neurological outcome of cerebral lesions in preterm infants. J Pediatr (Rio J). 2004;80(4):296–304.
41. Ferrari F, Cioni G, Einspieler C, et al Cramped synchronized general movements in preterm infants as an early marker for cerebral palsy. Arch Pediatr Adolesc Med. 2002;156(5):460–467.
42. Geerdink J, Hopkins B. Qualitative changes in general movements and their prognostic value in preterm infants. Eur J Pediatr. 1993;152(4):362–367.
43. Albuquerque PLD, Guerra MQDF, Lima MDC, et al Concurrent validity of the Alberta Infant Motor Scale
to detect delayed gross motor development in preterm infants: a comparative study with the Bayley III. Dev neurorehabil. 2018;21(6):408–414.