Spectrum of Gross Motor Function in Extremely Low Birth Weight Children With Cerebral Palsy at 18 Months of Age
http://www.100md.com
《小儿科》
Women and Infants Hospital, Providence, Rhode Island
University of Chicago Comer Children's and La Rabida Children's Hospitals, Chicago, Illinois
Rainbow Babies and Children's Hospital, Cleveland, Ohio
National Institute of Child Health and Human Development, Bethesda, Maryland
RTI International, Research Triangle Park, North Carolina
ABSTRACT
Objective. The purpose of this study was to evaluate the relationship between cerebral palsy (CP) diagnoses as measured by the topographic distribution of the tone abnormality with level of function on the Gross Motor Function Classification System (GMFCS) and developmental performance on the Bayley Scales of Infant Development II (BSID-II). It was hypothesized that (1) the greater the number of limbs involved, the higher the GMFCS and the lower the BSID-II Motor Scores and (2) there would be a spectrum of function and skill achievement on the GMFCS and BSID-II Motor Scores for children in each of the CP categories.
Methods. A multicenter, longitudinal cohort study was conducted of 1860 extremely low birth weight (ELBW) infants who were born between August 1, 1995 and February 1, 1998, and evaluated at 18 to 22 months' corrected age. Children were categorized into impairment groups on the basis of the typography of neurologic findings: spastic quadriplegia, triplegia, diplegia, hemiplegia, monoplegia, hypotonic and/or athetotic CP, other abnormal neurologic findings, and normal. The neurologic category then was compared with GMFCS level and BSID-II Motor Scores.
Results. A total of 282 (15.2%) of the 1860 children evaluated had CP. Children with more limbs involved had more abnormal GMFCS levels and lower BSID-II scores, reflecting more severe functional limitations. However, for each CP diagnostic category, there was a spectrum of gross motor functional levels and BSID-II scores. Although more than 1 (26.6%) in 4 of the children with CP had moderate to severe gross motor functional impairment, 1 (27.6%) in 4 had motor functional skills that allowed for ambulation.
Conclusions. Given the range of gross motor skill outcomes for specific types of CP, the GMFCS is a better indicator of gross motor functional impairment than the traditional categorization of CP that specifies the number of limbs with neurologic impairment. The neurodevelopmental assessment of young children is optimized by combining a standard neurologic examination with measures of gross and fine motor function (GMFCS and Bayley Psychomotor Developmental Index). Additional studies to examine longer term functional motor and adaptive-functional developmental skills are required to devise strategies that delineate therapies to optimize functional performance.
Key Words: extremely low birth weight neurologic outcome cerebral palsy gross motor function development sequelae
Abbreviations: CP, cerebral palsy ELBW, extremely low birth weight GMFCS, Gross Motor Function Classification System BSID-II, Bayley Scales of Infant Development II MDI, Mental Developmental Index PDI, Psychomotor Developmental Index
Cerebral palsy (CP) remains a major sequela in extremely low birth weight (ELBW) infants, affecting 9% to 17% of survivors.1–8 The traditional categorization of CP is dependent on the nature of the neurologic abnormality (eg, spasticity, dyskinetic movements, dystonia) and topography of limb involvement (number of limbs involved). The traditional ranking of neurologic severity from greatest (4 limbs) to least (1 limb) impaired, on the basis of topography, includes the following ranking: spastic quadriplegia, triplegia, diplegia, hemiplegia, and monoplegia. More recently, attempts have been made to evaluate the severity of CP by how the child functions in gross motor activities rather than by the specific tone and reflex findings in the limbs.
The primary objective of this study was to evaluate the relationships among the severity of neurologic impairments as measured by the topographic type of CP and 2 measures of motor performance: the Gross Motor Classification System (GMFCS)9 level of function and the Bayley Scales of Infant Development (BSID-II)10 Motor Score. The subjects were ELBW survivors who were evaluated at 18 to 22 months' corrected age. The children with hypotonic ± athetotic CP or "other" neurologic findings were placed in separate categories. It was hypothesized that (1) a greater number of limbs involved in CP would correlate with higher GMFCS and lower BSID-II scores and (2) there would be a spectrum of function and skill achievement on the GMFCS and BSID Motor Scores for children in each of the CP categories.
METHODS
Prospective Neonatal Information
Fourteen participating centers of the National Institute of Child Health and Human Development Neonatal Research Network collected pregnancy and delivery data on all live-born infants who weighed 401 to 1000 g and were born between August 1995 and February 1998. Neonatal outcome data were abstracted from hospital records by trained study coordinators at 120 days after birth, at discharge, or at the time of death, which ever came first. Data on transferred infants were collected until discharge to home or to a chronic care facility.
Cohort Recruitment and Attrition
Families at each center were invited to participate in their center follow-up programs after discharge for a comprehensive assessment at 18 to 22 months' corrected age. Participation was considered standard of care at the 14 centers, and informed consent was not required. Contact to schedule the follow-up evaluation was made by telephone call, postcard, or certified letter. Families were provided with transportation, if needed. Attrition rates differed by center, and although 9 of 14 sites had compliance rates of >80% (range: 81.1–95.4), 5 centers had rates of 78.1%, 77.1%, 75.4%, 55.5%, and 52.7%. The overall follow-up rate was 80.0%.
The cohort consisted of all 2385 ELBW survivors of 3606 infants who were admitted to the neonatal nurseries at the network centers. Fifty-one (2%) infants died after discharge. Of 2334 long-term survivors, 409 were lost to follow-up and 57 had partial follow-up, resulting in 1868 infants. In addition, 155 children did not have the GMFCS recorded and were not included in the analyses. Ninety-one percent of the cohort were inborn. There were no mean differences in birth weight between infants who were lost to follow-up and those who were evaluated. However, infants who were lost to follow-up had a higher gestational age (26.5 vs 26.3 weeks; P < .009), less grades 1 to 4 intraventricular hemorrhage, (35% vs 41%; P < .025), and less chronic lung disease (oxygen supplementation at 36 weeks; 37% vs 43%; P < .01) than infants who were followed.
The assessment at 18 months consisted of an interim medical history, updated social and demographic information, developmental evaluation, neurologic assessment, and the GMFCS.9,11–13 Hearing status information was obtained from parent report or follow-up audiologic test results when available. Severe hearing impairment was defined as hearing loss requiring amplification for both ears. Visual status information included ophthalmologist or parent report of postdischarge eye examinations as well as ophthalmologic procedures. In addition, a standard eye examination was completed by the neurodevelopmental examiner to evaluate the presence of strabismus, pupillary responses, red reflex, nystagmus, or roving eye movements. Visually impaired categories included (1) lack of normal vision demonstrated by use of corrective lenses, (2) legally blind with some functional vision, and (3) no useful vision.
A neurologic examination that was based on the Amiel Tison14 neurologic assessment was performed by certified developmentalists who had been trained to reliability in the examination procedure in a 2-day hands-on workshop. The neurologic assessment included an evaluation of tone, strength, reflexes, joint angles, and posture. Children were scored as normal when no abnormalities were observed. CP was defined as a nonprogressive central nervous system disorder characterized by abnormal muscle tone in at least 1 extremity and abnormal control of movement and posture. For example, hypertonicity with tight Achilles and toe walking was scored as diplegia, whereas hypertonicity with normal gait was scored as "other." Coding for CP in this study required 2 of 3 components: (1) abnormalities observed in the classical neuromotor examination such as tone and reflexes, (2) a delay in motor milestones, and (3) aberrations in primitive or postural reflexes. Each child was classified in 1 of the following categories: spastic quadriplegia (increased tone with motor impairment in all 4 extremities), triplegia (increased tone with motor impairment in 3 extremities), diplegia (increased tone with motor impairment primarily in the lower extremities), hemiplegia (increased tone and motor impairment on 1 side of the body), monoplegia (increased tone and motor impairment in 1 extremity), hypotonic (generalized hypotonia with ataxia), athetotic CP (varying tone with abnormal extrapyramidal movement patterns), other neurologic findings, or normal. The other neurologic findings category included a variety of findings, including hypertonicity, hypotonicity, seizures, dyspraxia, absent pincer, or facial palsy without a specific diagnosis of CP. When a child did not fit a classic CP category and had "mixed findings," a decision was made to place the child in the category with the most dominant findings. Children with chromosomal disorders, multiple congenital malformations, or central nervous system dysgenesis were also placed in the "other" category.
A GMFCS skill level derived from the work of Russell et al and Palisano et al9,11–13 was completed. Table 1 shows the scoring system with skill levels of 0 to 5 developed by Palisano and colleagues.9,11–13 The objective of the GMFCS is to assess the child's gross motor function by observation. A normal category is achieved when the child is able to walk 10 steps independently and fluently at 18 months. The GMFCS does not assess fine motor skills. It should be emphasized that the GMFCS describes the level of a child's motor performance, not the way he or she looks when he or she does it.
The BSID-II,10 including the Mental and Motor Scales, was administered by testers who were trained to reliability. The Motor Scales include both gross motor and fine motor tasks. BSID-II scores of 100 ± 15 represent the mean ± 1 SD. A score <70 is 2 SDs below the mean. The purpose of the BSID-II examination is to describe the child's abilities compared with same-age peers. The primary caregiver or adult who brought the child for the visit stayed with the child during the BSID-II examination, which was administered early in the clinic visit before the medical/neurologic assessment and interviews. Examiners were not able to administer successfully 1 or both parts (Mental Developmental Index [MSI] and Psychomotor Developmental Index [PDI]) of the BSID-II to 195 children. The following reasons were given: acute illness (n = 9), language barrier (n = 6), behavior problem (n = 58), severe developmental delay (n = 52), other (n = 52), sensory impairment and seems to have mild or moderate delay for age (n = 16), and sensory impairment but seems normal (n = 2). The 52 infants with severe delay and too much impairment to be tested were assigned a BSID-II score of 49 for categorical analyses only. Although every effort was made to test children within the window of 18 to 22 months' corrected age, 121 (6.5%) infants were evaluated outside the window because of illness or tracking issues. These data were included because the BSID-II is age adjusted.
Statistical Analyses
Group differences on continuous measures were examined using analysis of variance. Differences on categorical measures were analyzed with 2 analyses and Fisher exact test. Spearman rank-order correlations were also used to measure the association among the neurologic diagnosis, GMFCS, and PDI. All data were analyzed at RTI International.
RESULTS
Characteristics of the study sample by type of CP are shown in Table 2. CP was identified in 15.2% (n = 282) and other abnormal neurologic findings in 7.7% (n = 144) of the sample. The most common type of CP was spastic diplegia (39%) followed by quadriplegia (27.3%), hemiplegia (13.8%), hypotonic CP (9.9%), triplegia (6.0%), and monoplegia (3.9%). Children with CP were more likely to have the following characteristics compared with children with a normal neurologic examination: outborn, received surfactant, lower birth weight, lower gestational age, necrotizing enterocolitis, chronic lung disease, intraventricular hemorrhage 3 to 4, periventricular leukomalacia, and received postnatal steroids. Children with a normal neurologic examination were more likely to have received antenatal steroids than children with CP.
Additional neurosensory morbidities identified at the 18- to 22-month visit are shown in Table 3. Seizures were reported significantly more often in children with a diagnoses of CP than children with a normal neurologic examination. Shunting for posthemorrhagic hydrocephalus ranged from a high of 29.4% for children with triplegia to a low of 0.4% of children with a normal neurologic examination. Children with quadriplegia and hypotonic CP were most likely to be blind (6.7% and 7.1%, respectively). Hearing loss requiring amplification (9.5%) and multiple handicaps (100%) were most common among children with quadriplegia.
The relationship between the neurologic diagnoses and the GMFCS levels are shown in Table 4. The percentage of children who were rated at level 3 to 5 on the GMFCS scores was 0% for children with monoplegia, 7.8% for diplegia, 11.4% for hemiplegia, 31.3% for triplegia, 76.1% for quadriplegia, and 29.2% for hypotonic CP; the rate was 3.9% for children with other neurologic abnormalities and 0% for children with normal neurologic examinations. A substantial majority of children in all CP categories, except quadriplegia, achieved a normal, possible level 1, or level 1 rating (100% of monoplegias, 68.6% of hemiplegias, 78.4% of diplegias, 37.5% of triplegias, 92.2% of abnormal other, and 45.8% of hypotonic CP), indicating that a significant percentage of the children with these diagnoses were walking and had relatively mild involvement. Children in the category of "possible level 1" either were toe walking or had an asymmetric gait. The only children who had CP and scored "normal" on the GMFCS were 4 children with upper extremity monoplegia and 1 child who had hemiplegia and was able to walk symmetrically but had significant upper extremity involvement.
Spearman rank-order correlations were run to assess the relationships among the topography classification for spastic CP, the GMFCS scores, and the PDI scores. The strongest correlation was between topography and the GMFCS (r = 0.894; P < .0001), and the weakest correlation was between topography and the PDI scores (r = –0.510; P < .0001). The correlation between the GMFCS and the PDI score was r = –0.584 (P < .0001).
DISCUSSION
The objective of this study was to evaluate the severity of motor and functional involvement in ELBW survivors and the associations among 3 indicators of motor impairment: the topography of CP, the GMFCS, and the BSID-II Motor Scores. When comparing methods of assessment, the standardization, reliability, and validity of the methods are important. We previously reported the methods used in the National Institute of Child Health and Human Development Network to establish interrater reliability on the BSID-II, the GMFCS, and the neurologic examination.8 Each of these assessments provides unique information about a child's neurologic findings, developmental motor status, and gross motor function. Our ability to identify close correlations among them has the following limitations. The traditional spastic CP diagnosis is based on findings of abnormal increased tone, posture, and function and is classified by the number of limbs involved.15 In contrast, the GMFCS reflects the examiner's observation of the child's gross motor functional performance. A normal score indicates that the child can walk 10 steps independently and fluently. The PDI score is a composite of gross motor and fine motor skills. For example, a child may have age-appropriate fine motor skills yet be unable to walk. The BSID-II motor assessment, despite its limitations, has been used as the gold standard for motor performance in the majority of outcome studies of ELBW infants.
One of the challenges in reporting neuromotor outcomes in ELBW children is the identification of clinically relevant operational definitions for the severity of CP. Several strategies have been proposed, including the limb-by-limb classification recommended by the Oxford group.15 The importance of this classification is to highlight the severity by the number of limbs involved. We used this classification to establish our primary study groups. However, limitations of this classification include a lack of operational definitions for the severity of motor impairments and no mechanism to account for variability of tone or function among children with mixed neurologic involvement (spastic and extrapyramidal movement disorders).
A second strategy for obtaining a diagnosis has been to de-emphasize the neurologic classification and observe motor performance with a discriminative instrument, such as the Bayley Motor subtest or the Peabody Developmental Motor Scales.16 It is assumed that the majority of children with motor disability will have PDI scores of <70 or 50 and Peabody Developmental motor standard skills of 60 to 79 or <60. However, this system does not take into account a differential response for fine motor and gross motor tasks and the identification of children who do not have CP but have motor delay. The percentage of low BSID-II Motor Scores in our study increased linearly with the number of limbs involved. A similar relationship was seen for MDI scores, with the exception of the children with hemiplegia, who tended to have higher scores. For children in all of the CP categories, the MDI scores were higher than the PDI scores. In the current study, 7.7% of the children had neurologic findings other than CP. Fifty-six percent of these children had a BSID-II score of <70, and 22.1% had a score of <50. This is a heterogeneous group of children who have a variety of non-CP neurologic findings that are not well addressed in most studies but exhibit a significant amount of developmental morbidity.
A third strategy has been to classify children with respect to ultimate prognosis for independent walking. This strategy pays particular attention to gross motor postural skills. For example, several authors have highlighted that all children with sitting balance by their second birthday become independent walkers.17–19 Thus, attention to this skill is important in young children who are at risk for motor disability.20 This suggests that many children who have PDI scores of <50 at age 2 and, however, have sitting balance will go on to walk. Several new methods, including a revised Gross Motor Functional Measure-66 or -88,12 have been developed to examine more systematically the functional skills in prone, rolling, sitting, crawling, standing, walking, jumping, and running.21–23 In the current study, 65.7% of children with a diagnosis of hemiplegia and 78.4% of children with diplegia were at possible level 1 or level 1, indicating an ability to ambulate, although not independently or fluently.
The classification of the type of CP on the basis of topography has been linked to ultimate motor prognosis. Children with hemiplegia have had excellent motor prognosis for independent ambulation. The majority of children with diplegia are able to attain functional mobility, whereas children with triplegia and quadriplegia struggle with daily motor performance. It is in the gradations of gross motor performance (mobility) that are embedded in the GMFCS that the range and the severity of neurologic impairment among children with a specific CP diagnosis can best be shown.20,24–26 Our data indicate that the majority of children with a diagnosis of either hemiplegia or diplegia were already ambulatory at 18 months.
Small numbers of children were in GMFCS levels 3 to 5, indicating more severe impairment (hemiplegia: 4 of 35; diplegia: 8 of 102; triplegia: 5 of 16). Only among children with quadriplegia did a significant percentage (51 of 67) have GMFCS of 3 to 5.
We did identify a stronger correlation between the topography classification and the GMFCS (r = 0.894) than with the PDI (r = –0.510). That a stronger association with the PDI scores was not identified is not surprising, because the PDI reflects a combination of gross motor and fine motor tasks, whereas the GMFCS reflects only gross motor tasks. Despite that there was a significant decrease in functional ability (increase in GMFCS scores) as the diagnoses progress from monoplegia to hemiplegia, diplegia, triplegia, and quadriplegia, a subgroup of children with a diagnosis of monoplegia, diplegia, and hemiplegia had gross motor function in the "near normal range." This emphasizes the importance of providing a composite of information to parents, including the CP diagnosis, gross motor functional abilities, and Bayley motor performance. In fact, providing a CP diagnosis independently is a disservice for families.
The strength of this study is its prospective nature, its large population base (1868 children), and its attempt to describe systematically the spectrum of CP in ELBW children using multiple measures. A limitation of the study is the endpoint of 18 to 22 months, which is too early for the complete development of preschool motor skills. However, this age has the advantage of maintaining cohort compliance and facilitating the identification of children for comprehensive early intervention services. Additional studies to examine the type of CP and its neuromotor severity, its developmental and functional impact at different ages,20 and the changing patterns of function over time are needed.
CONCLUSIONS
CP remains a significant neurodevelopmental sequela that occurs in disproportionately high numbers of ELBW survivors. Overall, 15.2% of our cohort had CP. For each of the CP diagnostic categories except quadriplegia, there was a spectrum of gross motor functional levels and PDI scores that ranged from significant to near-normal function and ability, providing a more positive view of outcomes of children with CP. In fact, 78 (27.7%) of 282 children with any type of CP were ambulatory at 18 to 22 months, lending a much more encouraging view for the motor outcomes of ELBW children with CP than previously described. We emphasize that providing composite information to parents about the CP diagnosis, gross motor function, and BSID-II developmental skills provides a more comprehensive picture of the spectrum of CP. Future studies to examine factors that optimize functional skills in mobility, manipulation, self-care, and communication are required.
APPENDIX. National Institute of Child Health and Human Development Neonatal Research Network (1996–2000)
ACKNOWLEDGMENTS
This study was supported by the National Institute of Child Health and Human Development through Cooperative Agreements U10HD27904, U10 HD27856, U01 HD19897, U10 HD27853, U10 HD27851, U10 HD21364, U10 HD21373, U10 HD21397, U10 HD21385, U10 HD21415, U10 HD27880, U10 HD27881, and U10 HD27871 and CRC grants M01 RR 00750, M01 RR 08084, M01 RR 00070, M01 RR 00997, and M01 RR 06022.
FOOTNOTES
Accepted Nov 16, 2004.
This work was presented in part at the annual meeting of the Society for Pediatric Research; May 1–5, 2000; Boston, MA.
No conflict of interest declared.
REFERENCES
Hagberg B, Hagberg G. The changing panorama of cerebral palsy—bilateral spastic forms in particular. Acta Paediatr Suppl. 1996;416 :48 –52
Powell B, Hagberg G, Olow I. The changing panorama of cerebral palsy in Sweden, VI: prevalence and origin during the birth year period 1983–1986. Acta Paediatr. 1993;82 :387 –393
Stanley F, Watson L. The cerebral palsies in western Australia: trends, 1968 to 1981. Am J Obstet Gynecol. 1988;158 :89 –93
Piecuch RE, Leonard CH, Cooper BA, Kilpatrick SJ, Schlueter MA, Sola A. Outcome of infants born at 24–26 weeks' gestation: II. Neurodevelopmental outcome. Obstet Gynecol. 1997;90 :809 –814
Surman G, Newdick H, Johnson A. Cerebral palsy rates among low-birthweight infants fell in the 1990s. Dev Med Child Neurol. 2003;45 :456 –462
Pinto-Martin JA, Riolo S, Cnaan A, Holzman C, Susser MW, Paneth N. Cranial ultrasound prediction of disabling and nondisabling cerebral palsy at age two in a low birth weight population. Pediatrics. 1995;95 :249 –254
Prevalence and characteristics of children with cerebral palsy in Europe. Dev Med Child Neurol. 2002;44 :633 –640
Vohr BR, Wright LL, Dusick AM, et al. Neurodevelopmental and functional outcomes of extremely low birth weight infants in the National Institute of Child Health and Human Development Neonatal Research Network, 1993–1994. Pediatrics. 2000;105 :1216 –1226
Palisano R, Rosenbaum P, Walter S, Russell D, Wood E, Galuppi B. Development and reliability of a system to classify gross motor function in children with cerebral palsy. Dev Med Child Neurol. 1997;39 :214 –223
Bayley N. Bayley Scales of Infant Development-II. San Antonio, TX: Psychological Corporation; 1993
Russell DJ, Rosenbaum PL, Cadman DT, Gowland C, Hardy S, Jarvis S. The gross motor function measure: a means to evaluate the effects of physical therapy. Dev Med Child Neurol. 1989;31 :341 –352
Russell DJ, Avery LM, Rosenbaum PL, Raina PS, Walter SD, Palisano RJ. Gross Motor Function Measure (GMFM-66 and GMFM-88) User's Manual. London, United Kingdom: Mackeith Press; 2002
Russell DJ, Avery LM, Rosenbaum P, Raina PS, Walter SD, Palisano R. Improved scaling of the Gross Motor Function Measure for children with cerebral palsy: evidence of reliability and validity. Phys Ther. 2000;80 :873 –885
Amiel-Tison C. Neuromotor status. In: Taeusch HW, Hogman MW, eds. Follow-up Management of the High-Risk Infant. Boston, MA: Little, Brown & Company; 1987:115 –126
Evans P, Johnson A, Mutch L, Alberman E. Report of a meeting on the standardisation of the recording and reporting of cerebral palsy. Dev Med Child Neurol. 1986;28 :547 –548
Folio MR, Fewell RR. Peabody Developmental Motor Scales and Activity Cards. Allen, TX: Developmental Learning Materials Resource; 1983
Crothers B, Paine R. The Natural History of Cerebral Palsy. London, United Kingdom: Mackeith Press; 1998
de Paz AC Jr, Burnett SM, Braga LW. Walking prognosis in cerebral palsy: a 22-year prospective analysis. Dev Med Child Neurol. 1994;36 :130 –134
Molnar GE, Gordon SU. Cerebral palsy: predictive value of selected clinical signs for early prognostication of motor function. Arch Phys Med Rehabil. 1976;57 :153 –158
Msall ME, Tremont MR. Functional outcomes in self-care, mobility, communication, and learning in extremely low-birth weight infants. Clin Perinatol. 2000;27 :381 –401
Wood E, Rosenbaum P. The gross motor function classification system for cerebral palsy: a study of reliability and stability over time. Dev Med Child Neurol. 2000;42 :292 –296
Nordmark E, Jarnlo GB, Hagglund G. Comparison of the Gross Motor Function Measure and Paediatric Evaluation of Disability Inventory in assessing motor function in children undergoing selective dorsal rhizotomy. Dev Med Child Neurol. 2000;42 :245 –252
Rosenbaum PL, Walter SD, Hanna SE, et al. Prognosis for gross motor function in cerebral palsy: creation of motor development curves. JAMA. 2002;288 :1357 –1363
O'Shea TM, Dammann O. Antecedents of cerebral palsy in very low-birth weight infants. Clin Perinatol. 2000;27 :285 –302
Wheater M, Rennie JM. Perinatal infection is an important risk factor for cerebral palsy in very-low-birthweight infants. Dev Med Child Neurol. 2000;42 :364 –367
Palta M, Sadek-Badawi M, Evans M, Weinstein MR, McGuinnes G. Functional assessment of a multicenter very low-birth-weight cohort at age 5 years. Newborn Lung Project. Arch Pediatr Adolesc Med. 2000;154 :23 –30(Betty R. Vohr, MD, Michae)
University of Chicago Comer Children's and La Rabida Children's Hospitals, Chicago, Illinois
Rainbow Babies and Children's Hospital, Cleveland, Ohio
National Institute of Child Health and Human Development, Bethesda, Maryland
RTI International, Research Triangle Park, North Carolina
ABSTRACT
Objective. The purpose of this study was to evaluate the relationship between cerebral palsy (CP) diagnoses as measured by the topographic distribution of the tone abnormality with level of function on the Gross Motor Function Classification System (GMFCS) and developmental performance on the Bayley Scales of Infant Development II (BSID-II). It was hypothesized that (1) the greater the number of limbs involved, the higher the GMFCS and the lower the BSID-II Motor Scores and (2) there would be a spectrum of function and skill achievement on the GMFCS and BSID-II Motor Scores for children in each of the CP categories.
Methods. A multicenter, longitudinal cohort study was conducted of 1860 extremely low birth weight (ELBW) infants who were born between August 1, 1995 and February 1, 1998, and evaluated at 18 to 22 months' corrected age. Children were categorized into impairment groups on the basis of the typography of neurologic findings: spastic quadriplegia, triplegia, diplegia, hemiplegia, monoplegia, hypotonic and/or athetotic CP, other abnormal neurologic findings, and normal. The neurologic category then was compared with GMFCS level and BSID-II Motor Scores.
Results. A total of 282 (15.2%) of the 1860 children evaluated had CP. Children with more limbs involved had more abnormal GMFCS levels and lower BSID-II scores, reflecting more severe functional limitations. However, for each CP diagnostic category, there was a spectrum of gross motor functional levels and BSID-II scores. Although more than 1 (26.6%) in 4 of the children with CP had moderate to severe gross motor functional impairment, 1 (27.6%) in 4 had motor functional skills that allowed for ambulation.
Conclusions. Given the range of gross motor skill outcomes for specific types of CP, the GMFCS is a better indicator of gross motor functional impairment than the traditional categorization of CP that specifies the number of limbs with neurologic impairment. The neurodevelopmental assessment of young children is optimized by combining a standard neurologic examination with measures of gross and fine motor function (GMFCS and Bayley Psychomotor Developmental Index). Additional studies to examine longer term functional motor and adaptive-functional developmental skills are required to devise strategies that delineate therapies to optimize functional performance.
Key Words: extremely low birth weight neurologic outcome cerebral palsy gross motor function development sequelae
Abbreviations: CP, cerebral palsy ELBW, extremely low birth weight GMFCS, Gross Motor Function Classification System BSID-II, Bayley Scales of Infant Development II MDI, Mental Developmental Index PDI, Psychomotor Developmental Index
Cerebral palsy (CP) remains a major sequela in extremely low birth weight (ELBW) infants, affecting 9% to 17% of survivors.1–8 The traditional categorization of CP is dependent on the nature of the neurologic abnormality (eg, spasticity, dyskinetic movements, dystonia) and topography of limb involvement (number of limbs involved). The traditional ranking of neurologic severity from greatest (4 limbs) to least (1 limb) impaired, on the basis of topography, includes the following ranking: spastic quadriplegia, triplegia, diplegia, hemiplegia, and monoplegia. More recently, attempts have been made to evaluate the severity of CP by how the child functions in gross motor activities rather than by the specific tone and reflex findings in the limbs.
The primary objective of this study was to evaluate the relationships among the severity of neurologic impairments as measured by the topographic type of CP and 2 measures of motor performance: the Gross Motor Classification System (GMFCS)9 level of function and the Bayley Scales of Infant Development (BSID-II)10 Motor Score. The subjects were ELBW survivors who were evaluated at 18 to 22 months' corrected age. The children with hypotonic ± athetotic CP or "other" neurologic findings were placed in separate categories. It was hypothesized that (1) a greater number of limbs involved in CP would correlate with higher GMFCS and lower BSID-II scores and (2) there would be a spectrum of function and skill achievement on the GMFCS and BSID Motor Scores for children in each of the CP categories.
METHODS
Prospective Neonatal Information
Fourteen participating centers of the National Institute of Child Health and Human Development Neonatal Research Network collected pregnancy and delivery data on all live-born infants who weighed 401 to 1000 g and were born between August 1995 and February 1998. Neonatal outcome data were abstracted from hospital records by trained study coordinators at 120 days after birth, at discharge, or at the time of death, which ever came first. Data on transferred infants were collected until discharge to home or to a chronic care facility.
Cohort Recruitment and Attrition
Families at each center were invited to participate in their center follow-up programs after discharge for a comprehensive assessment at 18 to 22 months' corrected age. Participation was considered standard of care at the 14 centers, and informed consent was not required. Contact to schedule the follow-up evaluation was made by telephone call, postcard, or certified letter. Families were provided with transportation, if needed. Attrition rates differed by center, and although 9 of 14 sites had compliance rates of >80% (range: 81.1–95.4), 5 centers had rates of 78.1%, 77.1%, 75.4%, 55.5%, and 52.7%. The overall follow-up rate was 80.0%.
The cohort consisted of all 2385 ELBW survivors of 3606 infants who were admitted to the neonatal nurseries at the network centers. Fifty-one (2%) infants died after discharge. Of 2334 long-term survivors, 409 were lost to follow-up and 57 had partial follow-up, resulting in 1868 infants. In addition, 155 children did not have the GMFCS recorded and were not included in the analyses. Ninety-one percent of the cohort were inborn. There were no mean differences in birth weight between infants who were lost to follow-up and those who were evaluated. However, infants who were lost to follow-up had a higher gestational age (26.5 vs 26.3 weeks; P < .009), less grades 1 to 4 intraventricular hemorrhage, (35% vs 41%; P < .025), and less chronic lung disease (oxygen supplementation at 36 weeks; 37% vs 43%; P < .01) than infants who were followed.
The assessment at 18 months consisted of an interim medical history, updated social and demographic information, developmental evaluation, neurologic assessment, and the GMFCS.9,11–13 Hearing status information was obtained from parent report or follow-up audiologic test results when available. Severe hearing impairment was defined as hearing loss requiring amplification for both ears. Visual status information included ophthalmologist or parent report of postdischarge eye examinations as well as ophthalmologic procedures. In addition, a standard eye examination was completed by the neurodevelopmental examiner to evaluate the presence of strabismus, pupillary responses, red reflex, nystagmus, or roving eye movements. Visually impaired categories included (1) lack of normal vision demonstrated by use of corrective lenses, (2) legally blind with some functional vision, and (3) no useful vision.
A neurologic examination that was based on the Amiel Tison14 neurologic assessment was performed by certified developmentalists who had been trained to reliability in the examination procedure in a 2-day hands-on workshop. The neurologic assessment included an evaluation of tone, strength, reflexes, joint angles, and posture. Children were scored as normal when no abnormalities were observed. CP was defined as a nonprogressive central nervous system disorder characterized by abnormal muscle tone in at least 1 extremity and abnormal control of movement and posture. For example, hypertonicity with tight Achilles and toe walking was scored as diplegia, whereas hypertonicity with normal gait was scored as "other." Coding for CP in this study required 2 of 3 components: (1) abnormalities observed in the classical neuromotor examination such as tone and reflexes, (2) a delay in motor milestones, and (3) aberrations in primitive or postural reflexes. Each child was classified in 1 of the following categories: spastic quadriplegia (increased tone with motor impairment in all 4 extremities), triplegia (increased tone with motor impairment in 3 extremities), diplegia (increased tone with motor impairment primarily in the lower extremities), hemiplegia (increased tone and motor impairment on 1 side of the body), monoplegia (increased tone and motor impairment in 1 extremity), hypotonic (generalized hypotonia with ataxia), athetotic CP (varying tone with abnormal extrapyramidal movement patterns), other neurologic findings, or normal. The other neurologic findings category included a variety of findings, including hypertonicity, hypotonicity, seizures, dyspraxia, absent pincer, or facial palsy without a specific diagnosis of CP. When a child did not fit a classic CP category and had "mixed findings," a decision was made to place the child in the category with the most dominant findings. Children with chromosomal disorders, multiple congenital malformations, or central nervous system dysgenesis were also placed in the "other" category.
A GMFCS skill level derived from the work of Russell et al and Palisano et al9,11–13 was completed. Table 1 shows the scoring system with skill levels of 0 to 5 developed by Palisano and colleagues.9,11–13 The objective of the GMFCS is to assess the child's gross motor function by observation. A normal category is achieved when the child is able to walk 10 steps independently and fluently at 18 months. The GMFCS does not assess fine motor skills. It should be emphasized that the GMFCS describes the level of a child's motor performance, not the way he or she looks when he or she does it.
The BSID-II,10 including the Mental and Motor Scales, was administered by testers who were trained to reliability. The Motor Scales include both gross motor and fine motor tasks. BSID-II scores of 100 ± 15 represent the mean ± 1 SD. A score <70 is 2 SDs below the mean. The purpose of the BSID-II examination is to describe the child's abilities compared with same-age peers. The primary caregiver or adult who brought the child for the visit stayed with the child during the BSID-II examination, which was administered early in the clinic visit before the medical/neurologic assessment and interviews. Examiners were not able to administer successfully 1 or both parts (Mental Developmental Index [MSI] and Psychomotor Developmental Index [PDI]) of the BSID-II to 195 children. The following reasons were given: acute illness (n = 9), language barrier (n = 6), behavior problem (n = 58), severe developmental delay (n = 52), other (n = 52), sensory impairment and seems to have mild or moderate delay for age (n = 16), and sensory impairment but seems normal (n = 2). The 52 infants with severe delay and too much impairment to be tested were assigned a BSID-II score of 49 for categorical analyses only. Although every effort was made to test children within the window of 18 to 22 months' corrected age, 121 (6.5%) infants were evaluated outside the window because of illness or tracking issues. These data were included because the BSID-II is age adjusted.
Statistical Analyses
Group differences on continuous measures were examined using analysis of variance. Differences on categorical measures were analyzed with 2 analyses and Fisher exact test. Spearman rank-order correlations were also used to measure the association among the neurologic diagnosis, GMFCS, and PDI. All data were analyzed at RTI International.
RESULTS
Characteristics of the study sample by type of CP are shown in Table 2. CP was identified in 15.2% (n = 282) and other abnormal neurologic findings in 7.7% (n = 144) of the sample. The most common type of CP was spastic diplegia (39%) followed by quadriplegia (27.3%), hemiplegia (13.8%), hypotonic CP (9.9%), triplegia (6.0%), and monoplegia (3.9%). Children with CP were more likely to have the following characteristics compared with children with a normal neurologic examination: outborn, received surfactant, lower birth weight, lower gestational age, necrotizing enterocolitis, chronic lung disease, intraventricular hemorrhage 3 to 4, periventricular leukomalacia, and received postnatal steroids. Children with a normal neurologic examination were more likely to have received antenatal steroids than children with CP.
Additional neurosensory morbidities identified at the 18- to 22-month visit are shown in Table 3. Seizures were reported significantly more often in children with a diagnoses of CP than children with a normal neurologic examination. Shunting for posthemorrhagic hydrocephalus ranged from a high of 29.4% for children with triplegia to a low of 0.4% of children with a normal neurologic examination. Children with quadriplegia and hypotonic CP were most likely to be blind (6.7% and 7.1%, respectively). Hearing loss requiring amplification (9.5%) and multiple handicaps (100%) were most common among children with quadriplegia.
The relationship between the neurologic diagnoses and the GMFCS levels are shown in Table 4. The percentage of children who were rated at level 3 to 5 on the GMFCS scores was 0% for children with monoplegia, 7.8% for diplegia, 11.4% for hemiplegia, 31.3% for triplegia, 76.1% for quadriplegia, and 29.2% for hypotonic CP; the rate was 3.9% for children with other neurologic abnormalities and 0% for children with normal neurologic examinations. A substantial majority of children in all CP categories, except quadriplegia, achieved a normal, possible level 1, or level 1 rating (100% of monoplegias, 68.6% of hemiplegias, 78.4% of diplegias, 37.5% of triplegias, 92.2% of abnormal other, and 45.8% of hypotonic CP), indicating that a significant percentage of the children with these diagnoses were walking and had relatively mild involvement. Children in the category of "possible level 1" either were toe walking or had an asymmetric gait. The only children who had CP and scored "normal" on the GMFCS were 4 children with upper extremity monoplegia and 1 child who had hemiplegia and was able to walk symmetrically but had significant upper extremity involvement.
Spearman rank-order correlations were run to assess the relationships among the topography classification for spastic CP, the GMFCS scores, and the PDI scores. The strongest correlation was between topography and the GMFCS (r = 0.894; P < .0001), and the weakest correlation was between topography and the PDI scores (r = –0.510; P < .0001). The correlation between the GMFCS and the PDI score was r = –0.584 (P < .0001).
DISCUSSION
The objective of this study was to evaluate the severity of motor and functional involvement in ELBW survivors and the associations among 3 indicators of motor impairment: the topography of CP, the GMFCS, and the BSID-II Motor Scores. When comparing methods of assessment, the standardization, reliability, and validity of the methods are important. We previously reported the methods used in the National Institute of Child Health and Human Development Network to establish interrater reliability on the BSID-II, the GMFCS, and the neurologic examination.8 Each of these assessments provides unique information about a child's neurologic findings, developmental motor status, and gross motor function. Our ability to identify close correlations among them has the following limitations. The traditional spastic CP diagnosis is based on findings of abnormal increased tone, posture, and function and is classified by the number of limbs involved.15 In contrast, the GMFCS reflects the examiner's observation of the child's gross motor functional performance. A normal score indicates that the child can walk 10 steps independently and fluently. The PDI score is a composite of gross motor and fine motor skills. For example, a child may have age-appropriate fine motor skills yet be unable to walk. The BSID-II motor assessment, despite its limitations, has been used as the gold standard for motor performance in the majority of outcome studies of ELBW infants.
One of the challenges in reporting neuromotor outcomes in ELBW children is the identification of clinically relevant operational definitions for the severity of CP. Several strategies have been proposed, including the limb-by-limb classification recommended by the Oxford group.15 The importance of this classification is to highlight the severity by the number of limbs involved. We used this classification to establish our primary study groups. However, limitations of this classification include a lack of operational definitions for the severity of motor impairments and no mechanism to account for variability of tone or function among children with mixed neurologic involvement (spastic and extrapyramidal movement disorders).
A second strategy for obtaining a diagnosis has been to de-emphasize the neurologic classification and observe motor performance with a discriminative instrument, such as the Bayley Motor subtest or the Peabody Developmental Motor Scales.16 It is assumed that the majority of children with motor disability will have PDI scores of <70 or 50 and Peabody Developmental motor standard skills of 60 to 79 or <60. However, this system does not take into account a differential response for fine motor and gross motor tasks and the identification of children who do not have CP but have motor delay. The percentage of low BSID-II Motor Scores in our study increased linearly with the number of limbs involved. A similar relationship was seen for MDI scores, with the exception of the children with hemiplegia, who tended to have higher scores. For children in all of the CP categories, the MDI scores were higher than the PDI scores. In the current study, 7.7% of the children had neurologic findings other than CP. Fifty-six percent of these children had a BSID-II score of <70, and 22.1% had a score of <50. This is a heterogeneous group of children who have a variety of non-CP neurologic findings that are not well addressed in most studies but exhibit a significant amount of developmental morbidity.
A third strategy has been to classify children with respect to ultimate prognosis for independent walking. This strategy pays particular attention to gross motor postural skills. For example, several authors have highlighted that all children with sitting balance by their second birthday become independent walkers.17–19 Thus, attention to this skill is important in young children who are at risk for motor disability.20 This suggests that many children who have PDI scores of <50 at age 2 and, however, have sitting balance will go on to walk. Several new methods, including a revised Gross Motor Functional Measure-66 or -88,12 have been developed to examine more systematically the functional skills in prone, rolling, sitting, crawling, standing, walking, jumping, and running.21–23 In the current study, 65.7% of children with a diagnosis of hemiplegia and 78.4% of children with diplegia were at possible level 1 or level 1, indicating an ability to ambulate, although not independently or fluently.
The classification of the type of CP on the basis of topography has been linked to ultimate motor prognosis. Children with hemiplegia have had excellent motor prognosis for independent ambulation. The majority of children with diplegia are able to attain functional mobility, whereas children with triplegia and quadriplegia struggle with daily motor performance. It is in the gradations of gross motor performance (mobility) that are embedded in the GMFCS that the range and the severity of neurologic impairment among children with a specific CP diagnosis can best be shown.20,24–26 Our data indicate that the majority of children with a diagnosis of either hemiplegia or diplegia were already ambulatory at 18 months.
Small numbers of children were in GMFCS levels 3 to 5, indicating more severe impairment (hemiplegia: 4 of 35; diplegia: 8 of 102; triplegia: 5 of 16). Only among children with quadriplegia did a significant percentage (51 of 67) have GMFCS of 3 to 5.
We did identify a stronger correlation between the topography classification and the GMFCS (r = 0.894) than with the PDI (r = –0.510). That a stronger association with the PDI scores was not identified is not surprising, because the PDI reflects a combination of gross motor and fine motor tasks, whereas the GMFCS reflects only gross motor tasks. Despite that there was a significant decrease in functional ability (increase in GMFCS scores) as the diagnoses progress from monoplegia to hemiplegia, diplegia, triplegia, and quadriplegia, a subgroup of children with a diagnosis of monoplegia, diplegia, and hemiplegia had gross motor function in the "near normal range." This emphasizes the importance of providing a composite of information to parents, including the CP diagnosis, gross motor functional abilities, and Bayley motor performance. In fact, providing a CP diagnosis independently is a disservice for families.
The strength of this study is its prospective nature, its large population base (1868 children), and its attempt to describe systematically the spectrum of CP in ELBW children using multiple measures. A limitation of the study is the endpoint of 18 to 22 months, which is too early for the complete development of preschool motor skills. However, this age has the advantage of maintaining cohort compliance and facilitating the identification of children for comprehensive early intervention services. Additional studies to examine the type of CP and its neuromotor severity, its developmental and functional impact at different ages,20 and the changing patterns of function over time are needed.
CONCLUSIONS
CP remains a significant neurodevelopmental sequela that occurs in disproportionately high numbers of ELBW survivors. Overall, 15.2% of our cohort had CP. For each of the CP diagnostic categories except quadriplegia, there was a spectrum of gross motor functional levels and PDI scores that ranged from significant to near-normal function and ability, providing a more positive view of outcomes of children with CP. In fact, 78 (27.7%) of 282 children with any type of CP were ambulatory at 18 to 22 months, lending a much more encouraging view for the motor outcomes of ELBW children with CP than previously described. We emphasize that providing composite information to parents about the CP diagnosis, gross motor function, and BSID-II developmental skills provides a more comprehensive picture of the spectrum of CP. Future studies to examine factors that optimize functional skills in mobility, manipulation, self-care, and communication are required.
APPENDIX. National Institute of Child Health and Human Development Neonatal Research Network (1996–2000)
ACKNOWLEDGMENTS
This study was supported by the National Institute of Child Health and Human Development through Cooperative Agreements U10HD27904, U10 HD27856, U01 HD19897, U10 HD27853, U10 HD27851, U10 HD21364, U10 HD21373, U10 HD21397, U10 HD21385, U10 HD21415, U10 HD27880, U10 HD27881, and U10 HD27871 and CRC grants M01 RR 00750, M01 RR 08084, M01 RR 00070, M01 RR 00997, and M01 RR 06022.
FOOTNOTES
Accepted Nov 16, 2004.
This work was presented in part at the annual meeting of the Society for Pediatric Research; May 1–5, 2000; Boston, MA.
No conflict of interest declared.
REFERENCES
Hagberg B, Hagberg G. The changing panorama of cerebral palsy—bilateral spastic forms in particular. Acta Paediatr Suppl. 1996;416 :48 –52
Powell B, Hagberg G, Olow I. The changing panorama of cerebral palsy in Sweden, VI: prevalence and origin during the birth year period 1983–1986. Acta Paediatr. 1993;82 :387 –393
Stanley F, Watson L. The cerebral palsies in western Australia: trends, 1968 to 1981. Am J Obstet Gynecol. 1988;158 :89 –93
Piecuch RE, Leonard CH, Cooper BA, Kilpatrick SJ, Schlueter MA, Sola A. Outcome of infants born at 24–26 weeks' gestation: II. Neurodevelopmental outcome. Obstet Gynecol. 1997;90 :809 –814
Surman G, Newdick H, Johnson A. Cerebral palsy rates among low-birthweight infants fell in the 1990s. Dev Med Child Neurol. 2003;45 :456 –462
Pinto-Martin JA, Riolo S, Cnaan A, Holzman C, Susser MW, Paneth N. Cranial ultrasound prediction of disabling and nondisabling cerebral palsy at age two in a low birth weight population. Pediatrics. 1995;95 :249 –254
Prevalence and characteristics of children with cerebral palsy in Europe. Dev Med Child Neurol. 2002;44 :633 –640
Vohr BR, Wright LL, Dusick AM, et al. Neurodevelopmental and functional outcomes of extremely low birth weight infants in the National Institute of Child Health and Human Development Neonatal Research Network, 1993–1994. Pediatrics. 2000;105 :1216 –1226
Palisano R, Rosenbaum P, Walter S, Russell D, Wood E, Galuppi B. Development and reliability of a system to classify gross motor function in children with cerebral palsy. Dev Med Child Neurol. 1997;39 :214 –223
Bayley N. Bayley Scales of Infant Development-II. San Antonio, TX: Psychological Corporation; 1993
Russell DJ, Rosenbaum PL, Cadman DT, Gowland C, Hardy S, Jarvis S. The gross motor function measure: a means to evaluate the effects of physical therapy. Dev Med Child Neurol. 1989;31 :341 –352
Russell DJ, Avery LM, Rosenbaum PL, Raina PS, Walter SD, Palisano RJ. Gross Motor Function Measure (GMFM-66 and GMFM-88) User's Manual. London, United Kingdom: Mackeith Press; 2002
Russell DJ, Avery LM, Rosenbaum P, Raina PS, Walter SD, Palisano R. Improved scaling of the Gross Motor Function Measure for children with cerebral palsy: evidence of reliability and validity. Phys Ther. 2000;80 :873 –885
Amiel-Tison C. Neuromotor status. In: Taeusch HW, Hogman MW, eds. Follow-up Management of the High-Risk Infant. Boston, MA: Little, Brown & Company; 1987:115 –126
Evans P, Johnson A, Mutch L, Alberman E. Report of a meeting on the standardisation of the recording and reporting of cerebral palsy. Dev Med Child Neurol. 1986;28 :547 –548
Folio MR, Fewell RR. Peabody Developmental Motor Scales and Activity Cards. Allen, TX: Developmental Learning Materials Resource; 1983
Crothers B, Paine R. The Natural History of Cerebral Palsy. London, United Kingdom: Mackeith Press; 1998
de Paz AC Jr, Burnett SM, Braga LW. Walking prognosis in cerebral palsy: a 22-year prospective analysis. Dev Med Child Neurol. 1994;36 :130 –134
Molnar GE, Gordon SU. Cerebral palsy: predictive value of selected clinical signs for early prognostication of motor function. Arch Phys Med Rehabil. 1976;57 :153 –158
Msall ME, Tremont MR. Functional outcomes in self-care, mobility, communication, and learning in extremely low-birth weight infants. Clin Perinatol. 2000;27 :381 –401
Wood E, Rosenbaum P. The gross motor function classification system for cerebral palsy: a study of reliability and stability over time. Dev Med Child Neurol. 2000;42 :292 –296
Nordmark E, Jarnlo GB, Hagglund G. Comparison of the Gross Motor Function Measure and Paediatric Evaluation of Disability Inventory in assessing motor function in children undergoing selective dorsal rhizotomy. Dev Med Child Neurol. 2000;42 :245 –252
Rosenbaum PL, Walter SD, Hanna SE, et al. Prognosis for gross motor function in cerebral palsy: creation of motor development curves. JAMA. 2002;288 :1357 –1363
O'Shea TM, Dammann O. Antecedents of cerebral palsy in very low-birth weight infants. Clin Perinatol. 2000;27 :285 –302
Wheater M, Rennie JM. Perinatal infection is an important risk factor for cerebral palsy in very-low-birthweight infants. Dev Med Child Neurol. 2000;42 :364 –367
Palta M, Sadek-Badawi M, Evans M, Weinstein MR, McGuinnes G. Functional assessment of a multicenter very low-birth-weight cohort at age 5 years. Newborn Lung Project. Arch Pediatr Adolesc Med. 2000;154 :23 –30(Betty R. Vohr, MD, Michae)