Jackson, Barbara J. PhD; Needelman, Howard MD; Roberts, Holly PhD; Willet, Sandy PT; PT, MS, PCS; McMorris, Carol BS
INTRODUCTION AND PURPOSE
In 2008, 12.3% of all live births in the United States were preterm.1 With advances in medical care and technology, survival rates of infants born preterm have steadily increased, up 6% from 1998 to 2008.1 This growing population of infants remains at high risk for adverse developmental outcomes.2,3 Marks et al4 found that infants born preterm were 2 times more likely than their age-matched peers born at term to be deemed eligible for early intervention services, and, in one recent study completed in the Netherlands, 75% of children born prior to 30 weeks gestation had at least 1 identified disability by the age of 5 years compared with only 27% of the control group born at term.5
Medical professionals who provide follow-up for infants who are at high risk for developmental delays play a crucial role in early identification so that preventative interventions might be implemented and supports established before school age.6 The most frequent early, observable delay in this high-risk population is in the area of gross motor development.6 Early identification of gross motor delays is, therefore, essential to maximize the child's potential for positive developmental and functional outcomes. Learning and development are in their most rapid period during the first 5 years. Early identification takes advantage of this critical developmental period and provides a window of opportunity to maximize the benefits of early intervention programs. If gross motor delays are the first observable sign of developmental concerns, screening tools that reliably assist in identifying early gross motor difficulties are crucial for professionals who provide systematic follow-up of infants who are at high risk for developmental delays.
Nebraska has developed a statewide system of neurodevelopmental follow-up, called Developmental Tracking Infants Progress Statewide (TIPS), which supports the identification of infants with a neonatal intensive care unit (NICU) experience and provides a mechanism to enhance referrals for early intervention programs (see Jackson and Needelman7 for a complete description of the program). Infants enrolled in this program are followed for the first 3 years of life. All infants in the state's 10 major NICUs, except those who stay in the NICU less than 48 hours, are eligible to be enrolled and are then evaluated by an interdisciplinary team at 6, 16, and 24 months' corrected age, as well as 36 months' chronological age. Infants whose screening results indicate delays or areas of developmental concern can then be referred to local early intervention programs for further assessment to determine eligibility for IDEA Part C services.8 Follow-up data from this project show that most children are referred for services at 6 months due to identification of gross motor delays (69% for gross motor compared with 11% for language delays) whereas infants at 16 and 24 months are more likely to qualify because of language delays (50% for language delays) (B. J. Jackson, unpublished document, 2010). Early referral is important as high-quality, intensive early intervention has been shown to result in the child requiring fewer special education and other habilitative services later in life.9
One of the challenges of implementing a system of child-find is identifying appropriate screening assessments, assessments with strong psychometric properties that are cost-effective, comprehensive, and easy to administer in a clinic setting. During the past 10 years, an increasing number of screening assessments have become available to assist practitioners in identifying children who may need further assessment. The Bayley Scales of Infant and Toddler Development, Third Edition (BSID-III), Screening Test.10 provides an assessment that differentiates the areas of concern by domain (ie, cognitive, language, fine, and gross motor). The majority of the gross motor items can be completed through observation with minimal handling of the infant and can be completed within 5 to 10 minutes. The BSID-III Screening Test can be administered by any member of a transdisciplinary team who has experience and training in assessment and test interpretation and experience with young children birth through 3.
The Alberta Infant Motor Scale (AIMS) is a well-established gross motor diagnostic tool that was chosen as a comparison reference in this study.11 Because the BSID-III Screening Test is a screening tool with a minimal number of gross motor items (eg, only 8–12 observed items dependent on the individual child) scored at 6 months of age, it is important to compare results with an established gross motor tool that has a large representative, normative sample of gross motor behaviors across age bands. If the 2 tools correlate in their ability to sensitively and accurately identify children whose gross motor development is delayed at 6 months of age, use of the more quickly administered screening tool would be beneficial for practitioners in a clinical, transdisciplinary setting. The purpose of this study is to determine the efficacy of the gross motor items on the BSID-III Screening Test-Gross Motor Subtest (GMS) for identifying gross motor delays in young infants who might benefit from more in-depth assessment to determine eligibility for early intervention. Specifically, the questions addressed were as follows: (1) Do the scores from the GMS of the BSID-III Screening Test correlate with the AIMS? and (2) Can BSID-III Screening Test–GMS results be used to predict acceptance into early intervention services?
This retrospective study included 93 infants who participated in a 6-month follow-up visit at 1 of the Developmental TIPS follow-up clinics. Table 1 provides descriptive information on this population including descriptive data on those infants who were accepted and those who were not accepted for services. All of the children had experienced an NICU stay for at least 48 hours. At the time of discharge, the NICU medical team assigned a level of follow-up: follow-up by survey for children at low risk for developmental delay, referral to a clinic for follow-up of children at high risk for developmental delay, and for those children who had known disabilities, referral to their local early intervention services. Children at low risk for developmental delay were those with a birth weight of more than 1500 g and limited medical problems. Children who were identified to be at high risk (see Table 2 for a list of risk factors), and who participated in a clinic, were included in this study. To be identified to be at high risk for developmental delay, they had to meet at least 1 of the listed criteria. The infants participated in the clinic visit at approximately 6 to 8 months' corrected age, which is the first of 3 scheduled clinic visits.
Institutional Review Board approval was obtained from the University of Nebraska Medical Center and Children's Hospital and Medical Center (Omaha, Nebraska) for this study. Children were assessed in the clinic using an interdisciplinary approach. After the 6-month visit, children were tracked to determine their acceptance rate into early intervention services. The clinic is staffed by a developmental/behavioral pediatrician, an occupational and a physical therapist, and a psychology graduate student at the postmasters level. In 2006, the clinic adopted the newly released BSID-III Screening Test. The BSID-III Screening Test has been found to have consistent results, with both high reliability and validity. High test-retest reliability was found across domains (r = 0.80–0.83). Studies with special groups supported the validity of the assessment with relatively low number of children who were developing typically classified as at risk for developmental delay, whereas a relatively high number of children from clinical groups known to be at risk for delay were classified appropriately.12 On the BSID-III Screening Test–GMS, the clinician identifies the appropriate subset of gross motor items to complete based on the child's chronological age. The child must successfully complete the first item of the subset of items to obtain a basal level and a ceiling is established by obtaining a zero on 4 consecutive items. Sample items include the following: (1) makes crawling movements, (2) elevates trunk while prone, (3) sits with support for 30 seconds, and (4) rolls from back to stomach. Typically a 6-month-old child would be observed on a range of 8 to 12 items.
On the BSID-III Screening Test, the child's performance is scored on the basis of a risk category classification: competent, emerging, or at risk. These risk categories were determined on the basis of comparison of scores with the BSID-III. Raw scores on both the BSID-III and BSID-III Screening Test are converted to age-corrected z scores in order to have a common measure. Table 3 describes the scaled score equivalent to the BSID-III and the definition for each classification. Although the Bayley-III Screening Test screens 4 areas (cognitive, language, fine, and gross motor) for purposes of this study, only the results of the GMS were used.
At the clinic visit as part of the gross motor evaluation the AIMS was also administered. The AIMS is used to identify infants with gross motor delay from birth to 18 months. It measures gross motor development from both a quantitative and a qualitative perspective. The AIMS is reported as a percentile rank on the basis of a raw score. The AIMS is considered a reliable and valid instrument for measuring infant motor development.11 The interrater reliability coefficient was high (r = 0.9891). Concurrent validity of the AIMS was reported for both the Bayley Scales of Infant Development GMS (r = 0.93) and the Peabody Development Motor Scale (r = 0.95) using a population of infants identified as normal and abnormal.11 In addition, the AIMS has been shown to have high levels of intrarater and interrater reliability and high concurrent validity when used with infants born preterm.11
SPSS statistical package version 16.0 was used to perform the data analyses. To answer the question about the relationship (ie, correlation) between the BSID-III Screening Test–GMS and the AIMS, the scaled scores of the BSID-III Screening Test–GMS were compared with the percentile scores of the AIMS using a point biserial correlation method. To answer the question about the best predictors of acceptance into early intervention services, a binary logical regression analysis was used. The predictor variables for this regression analysis included the scaled scores on the BSID-III Screening Test and the percentile scores on the AIMS. The criterion variable used for the analysis was acceptance of participants into early intervention services.
Both the BSID-III Screening Test and AIMS were completed on 93 children at their first Developmental TIPS clinic visit at 6 to 8 months' corrected age. The first analysis evaluated the relationship of the 2 gross motor assessments. The results of the point biserial correlation revealed that the BSID-III Screening Test–GMS and the AIMS had a weak to moderate correlation in this population (r = 0.397).
The second analysis evaluated to what degree the results of the 2 gross motor assessments predicted acceptance into early intervention. A total of 14% of the children (n = 13) were accepted for early intervention. Of those accepted for early intervention, 38.5% (n = 5) scored in the “at risk” range, 23.1% (n = 3) scored in the “emergent” range, and 38.5% (n = 5) scored in the “competent” range on the BSID-II Screening Test–GMS. Of those who did not receive early intervention (n = 80), both referred and not referred, 5% (n = 4) scored in the “at risk” range, 18.8% (n = 15) scored in the “emergent” range, and 76.2% (n = 61) scored in the “competent” range on the BSID-II Screening Test–GMS.
On the AIMS, children who were accepted for early intervention (n = 13) had a mean percentile score of 16% compared with those children (n = 80) who were not receiving service and who had a mean percentile score of 28.2%. The regression analysis revealed that the BSID-III Screening Test–GMS accounted for a significant portion of the variance in acceptance to early intervention services, χ2(1) = 7.871, P = .005. The BSID-III Screening Test–GMS was the only variable in the model that accounted for a significant portion of the acceptance category variance (Cox and Snell, R2 = 0.093; Nagelkerke, R2 = 0.165). The AIMS scores were excluded from the analysis as they did not provide additional variance to acceptance status beyond that provided by the BSID-III Screening Test–GMS.
Because of the increasing number of infants born preterm who are known to be at a greater risk for disability, there is a growing need by community health care providers to identify children at an early age that may benefit from early intervention services. The earlier this group of young infants can receive services, the more likely the effects of prematurity can potentially be ameliorated. This study confirmed previous research that suggests that delays for the young infants are typically in the motor area and highlights the need for an effective screening tool. The BSID-III Screening Test–GMS provides clinicians with a multiuse assessment that can either focus on a single domain or provide a comprehensive screening approach across domains. The results of this study suggest that, for the NICU population, the BSID-III Screening Test–GMS can effectively identify those young infants who need early intervention services. Most importantly, the test provides clinicians a time-efficient means for assessing young children.
As reported earlier, the BSID–III and the AIMS were highly correlated (r = 0.93). So it was surprising the BSID-III Screening Test–GMS and AIMS were not highly correlated (r = 0.397). One explanation may be that each assessment has a distinctly different purpose resulting in a difference in the type and function of the item set selected for each assessment. The Bayley-III Screening Test–GMS was developed to cast a broad net to ensure that children with possible delays are not missed. Therefore, the items on this assessment were selected that highly discriminate those children needing to be referred for further assessment from those who do not need services. In contrast, the AIMS usefulness is 2-fold: to provide information for diagnosis, and for program planning of intervention strategies. Items on the AIMS were selected to provide finer discrimination between motor behaviors in order to provide more detailed information that is useful for determining both a diagnosis and for program planning.
Haastert et al13 found that the early gross motor development profile of infants born preterm was different from typical gross motor development, with the population of premature infants scoring significantly lower compared with infants born at term across all age levels. These results suggested the need to adjust the norms when using the AIMS with infants born preterm to better detect infants with mild gross motor problems or learning difficulties. A future study could determine whether the correlation between the AIMS and the BSID-III Screening Test–GMS would be improved if the adjusted AIMS scores were used for this population of infants born preterm and if the adjusted AIMS scores would better predict acceptance into early intervention than the original AIMS scoring.
The pattern of at-risk status and acceptance was found in the majority of the infants. A small subgroup (5) of the infants scored “competent” on the GMS, yet was accepted for services. Initially this was a surprising finding; however, closer examination of this subgroup found that these children were accepted on the basis of health needs (failed hearing screen, gastrotube feedings, etc) and in those circumstances they would be accepted because of health impairments. This eligibility criterion allows the child to be accepted without demonstrating significant delays. These results suggest that it is important that both health and gross motor factors be considered in making referrals to early intervention services.
A second subgroup of interest was those infants with emerging or at-risk status who were not accepted for early intervention services. Access to services was not an issue as early intervention services for infants in this state are mandated within the state education system. Typically children who scored in the emerging range are not accepted into services as they are not demonstrating significant delays but may be referred for services on the basis of the recommendations of the clinician. Of interest are those children who were at high risk. It would be expected that a small number of the children at risk for developmental delay (in this study there were 4) would not be eligible for services since the BSID-III Screening Test–GMS is used to screen. Its purpose is to cast a broad net to ensure that children are not missed who may be eligible for services. Of interest was that at their 16-month visit, 2 of these children were within the emerging range for their motor skills but were at risk for developmental delay in language skills and were accepted for services. This pattern of delay over time suggests that those children demonstrating risk for motor skills should have high priority for follow-up assessment to monitor their need for services.
Overall, the BSID-III Screening Test–GMS was found to be an effective tool at 6 to 8 months' corrected age for identifying children who were eligible for early intervention. With minimal training, the BSID-III Screening Test can be administered by members of the health care or early intervention team from any discipline. This cost-effective approach has the additional advantage of allowing the practitioner to screen multiple areas in a brief period of time. This type of screening tool has great applicability for transdisciplinary/interdisciplinary teams as it allows for the use of one tool to identify whether further assessment is needed.
It is essential for community health care providers to have cost-effective, valid assessment tools to assist in making decisions regarding referral for early intervention. The BSID-III Screening Test–GMS is a valuable resource for these practitioners as it allows for (1) completion of a comprehensive, domain-specific screening assessment for evaluating children's development, (2) completion of the assessment in a shorter-time frame than comprehensive evaluations, and (3) practitioners with a variety of training and skills to complete the screening. For the population of infants born premature, the BSID-III Screening Test–GMS effectively helped to appropriately identify those children in need of services, maximizing the benefits to the child.
1. March of Dimes. Peristats, United States, preterm birth. www.marchofdimes.com
. Published 2008. Accessed August 6, 2011.
2. Leversen KT, Sommerfelt K, Rønnestad A, et al. Prediction of neurodevelopmental and sensory outcome in 5 years in Norwegian children born extremely preterm. Pediatrics. 2011;127(3):630–638.
3. Myers E, Ment LR. Long-term outcome of preterm infants and the role of neuroimaging. Clin Perinatal. 2009;36:773–789.
4. Marks K, Hix-Small H, Clark K, Newman J. Lowering developmental screening thresholds and raising quality improvement for pre-term children. Pediatrics. 2009;123(6):1516–1523.
5. Potharst ES, van Wassenaer AG, et al. High incidence of multi-domain disabilities in very preterm children at five years of age. Pediatrics
. Accessed June 29, 2011.
6. Spittle AJ, Orton J, Doyle LW, et al. Early developmental intervention programs post hospital discharge to prevent motor and cognitive impairments in preterm infants. Cochrane Database Syst Rev 2007;2:CD005495.
7. Jackson B, Needelman H. Building a system of child find through a 3-tiered model of follow-up. infants and toddlers: an interdisciplinary. J Spec Care Pract. 2007;20(3):255–265.
8. US Department of Education. Individuals with Disabilities Education Act (IDEA). http://idea.ed.gov/
. Published 2004. Accessed November, 15, 2011.
9. Ramey CT, Ramey SL. Early intervention and early experience. Am Psycholt 1998;53(2):109–120.
10. Bayley N. Bayley Scales of Infant and Toddler Development Screening Test. 3rd ed. San Antonio, TX: Harcourt Assessment; 2006.
11. Piper M, Darrah J, Piper MC. Alberta Infant Motor Scale (AIMS). Maryland Heights, MO: Elsevier Health Science; 1994.
12. Bayley N. Bayley Scales of Infant and Toddler Development: Technical Manual. 3rd ed. San Antonio, TX: Harcourt Assessment; 2006.
13. Van Haastert M, De Vries L, Helders P, Jongmans M. Early gross motor development of preterm infants according to the Alberta Infant Motor Scale. J Pediatr. 2006;49(5):618–622.
early intervention; female; humans; male; motor skills; neuropsychological tests/standards; neuropsychological tests/statistics & numerical data; predictive value of tests; preschool child; preterm infant; risk factors; time factors
© 2012 Lippincott Williams & Wilkins, Inc.