Establishing an early identification score system for cerebral palsy based on detailed assessment of general movements.

Introduction

Assessment of general movements (GMs) is widely applied to clinical cerebral palsy (CP) screening for high-risk infants because of the advantage of a simple operation, good stability, and non-invasiveness.[1] Examining the presence of transient cramped-synchronized (CS) GMs is often used to assess the predictive validity of CP in Chinese domestic research. In the clinic, infants with poor repertoire (PR) GMs can either develop CP or recover as normal infants. Therefore, distinguishing development of different movements is especially important for determining the prognosis of high-risk infants. Detailed assessment of GMs can be used to assess differences in various movements according to a score of motion to allow for the deficiency in general assessment of GMs. Because of a lack of a detailed control score, further distinguishing different movements cannot be conducted for the degree of risk of neural developmental outcome.

This study aimed to investigate the results of use of a detailed score in the writhing movement period according to different developmental outcomes for high-risk infants. We also aimed to establish an early identification score for CP based on GMs and provide early assessment and early intervention for different degrees of risk.

Materials and methods

Patients

High-risk infants in the Rehabilitation Department of our hospital during January 2016 to January 2017 were retrospectively screened as study subjects. High risks were defined as prematurity, low body weight, asphyxia, hypoxic–ischemic encephalopathy, intracranial hemorrhage, and convulsion.[2] Inclusion criteria for the study were as follows: (1) infants underwent assessment of GMs and received a detailed score at the age of 4 to 6 weeks; (2) infants had the 52-neuromotor examination performed at 0 to 1 year old (age of 6–8 weeks); and (3) there was a follow-up outcome at the 12th month (age of 44–52 weeks) (CP or non-CP[3]) and the Gesell development scale was applied to measure the development quotient (DQ). Ethical approval was obtained from the Ethics Committee of Xuzhou Children’s Hospital. Informed consent for participation in the study was obtained by the patients’ guardians.

Methods

Assessment of GMs

Light in the video collection room of GMs was gentle, temperature was approximately 27°, and interference and noise were avoided.[4] Infants with a bib and diaper were placed in a tailor-made room for sleep and play. A camera was positioned approximately 1 cm above the bed, with an angle of approximately 25°, which ensured that the camera could record the infant’s facial expressions. Additionally, a video serial number card was placed on the right side of the bed. The infant was in the supine position and their awake, calm, and delighted times were recorded. If a cry or hiccup occurred, which did not allow an effective GM recording time (10 minutes), the video was recorded at the right time.

Doctors collected the GM video in accordance with recruiting standards in this study and all data were stored in an appointed computer. Doctors read and assessed the material daily, and then wrote a report. All doctors that performed assessment of GMs participated in the European GM Trust training and obtained a certificate. For assessment of GMs, the GM video was played, the acoustic signal was turned off, and overall assessment of the principle of Gestalt vision was performed. When writhing movements occurred, GMs were divided into (1) normal GMs, (2) poor repertoire (PR) GMs, (3) cramped-synchronized (CS) GMs, (4) and chaotic GMs.

Detailed assessment of GMs in the writhing movement period

Appropriate GM videos of 2 to 3 minutes were extracted and detailed assessment of GMs in the writhing movement period was performed in accordance with Preterm and Term General Movements,[5] including (1) the GM score and (2) detailed score, including the neck score, body score, and score of upper and lower extremity movement. Different parts were scored separately and the whole score was calculated. The highest score was 42 and the lowest score was 5. A higher score indicated a better performance.

The 52-neuromotor examination in 0- to 1-year-olds

Passive muscular tension, motor activity, and reflection were studied according to different styles and characteristics of movement.[6] Normal was recorded as “0” and abnormal was recorded as “1 to 8” separately. The score of every dimension was the sum of abnormal items. The measured results were converted as follows: a score of 8 was “0”, a score of 7 was “1”, a score of 6 was “2”, a score of 5 was “3”, a score of 4 was “4”, a score of 3 was “5”, a score of 2 was “6”, a score of 1 was “7”, and a score of 0 was “8”. A higher score indicated a better performance of GMs.

Gesell development scale

At 12 months old (age of 44–52 weeks), different developmental outcomes were determined according to the diagnostic criteria for CP, including motor behavior, object behavior, language behavior, and personal social behavior. The Gesell development scale was applied to measure the DQ.[7] Non-CP infants with a DQ <85 were classified as the motor retardation group.[8]

Establishment of an early identification score for CP

The detailed score based on GMs was divided in accordance with empirical data from health statistics with 95% confidence intervals in the writhing movement period.

Predictive validity

According to epidemiology, predictive validity includes the predictive sensitivity, predictive specificity, positive predictive value, negative predictive value, false positive rate, and false negative rate. These details can be found in the textbook Epidemiology (4th edition).[9]

Statistical methods

All data were input into a computer using Excel (Microsoft, Redmond, WA, USA) and were analyzed with SPSS version 17.0 (SPSS Inc., Chicago, IL, USA). Scores are shown as mean ± standard deviation (SD). All data had a normal distribution. Measurement data were analyzed by variance analysis, post-hoc multiple comparisons were analyzed by q analysis, and correlations were analyzed by Pearson correlation analysis. P < 0.05 was regarded as statistically significant.

Results

General information

A total of 114 subjects were included in this study. There were 80 (70.2%) boys and 34 (29.8%), girls, with a mean gestational age at birth of 36.53 ± 2.66 weeks and a birth weight of 2880 ± 610 g.

Detailed score

According to GMs, 36 (31.6%) infants were in the normal group, 58 (50.9%) were in the PR group, and 20 (17.5%) were in the CS group.

Detailed scores in the writhing movement period were examined by normal distribution K-S analysis and plotted in a histogram. GZ values in the normal, PR, and CS groups were 0.868, 1.220, and 0.810, respectively (P > 0.05), which are consistent with a normal distribution. Detailed scores in the writhing movement period were significantly different among the groups (F = 208.186, P < 0.001). Post-hoc multiple comparisons showed that the mean detailed score in the normal group was significantly higher than that in the PR and CS groups (q = 18.240 and 32.010, respectively, both P < 0.01). Additionally, the mean detailed score in the PR group was significantly higher than that in the CS group (q = 13.770, P < 0.01) (Table 1).

Table 1.

Detailed scores among the groups with different forms of general movements.

Group	Infants(n)	Detailed score(mean ± SD)	F
Normal	36	38.58 ± 1.96	208.186***
PR	58	23.72 ± 6.38
CS	20	12.50 ± 2.44

SD: standard deviation; PR: poor repertoire; CS: cramped-synchronized. ***P < 0.001.

The 52-neuromotor examination in 0- to 1-year-olds

The 52-neuro-motorial examination scores were examined by normal distribution K-S analysis and plotted in a histogram. Z values in the normal, PR, and CS groups were 1.318, 1.347, and 0.608, respectively (P > 0.05), which is consistent with a normal distribution.

The scores of the 52-neuromotor examination in the writhing movement period were significantly different among the groups (F = 140.914, P < 0.001). Post-hoc multiple comparisons showed that the mean score of the 52-neuromotor examination in the normal group was significantly higher than that in the PR and CS groups (q = 11.830 and 27.400, respectively, both P < 0.01). Additionally, the mean score of the 52-neuromotor examination in the PR group was significantly higher than that in the CS group (q = 15.5570, P < 0.01) (Table 2).

Table 2.

Scores of the 52-neuromotor examination among groups with different forms of general movements.

Group	52-neuromotor examination score (mean ± SD)	F
Normal	23.00 ± 0.83	140.914***
PR	18.59 ± 2.79
CS	12.90 ± 1.74

SD: standard deviation; PR: poor repertoire; CS: cramped-synchronized. ***P < 0.001.

Correlations between detailed scores in the writhing movement period and 52-neuromotor examination scores

The associations of detailed scores with 52-neuromotor examination scores were analyzed by Pearson correlation analysis. Detailed scores were positively related to the separate scores of passive muscular tension, motor activity, and reflection in the 52-neuromotor examination, with correlation coefficients of 0.951, 0.913, and 0.917, respectively (all P < 0.001). Detailed scores were positively related to 52-neuromotor examination scores, with a correlation coefficient of 0.218 (P < 0.001).

Detailed score in the writhing movement period among groups with different motor development outcomes

At 12 months, infants were divided into different groups in accordance with the diagnostic criteria or/and DQ of the Gesell development scale. There were 46 (40.4%) infants in the normal group, 44 (38.6%) in the poor development (non-CP) group, and 24 (21.1%) in the CP group.

Detailed scores in the writhing movement period were examined by normal distribution K-S analysis and plotted in a histogram. Z values in the normal, non-CP, and CP groups were 1.218, 1.224, and 0.882, respectively (P > 0.05), which is consistent with a normal distribution.

Detailed scores in the writhing movement period were significantly different among the groups (F = 118.351, P < 0.001). Post-hoc multiple comparisons showed that the mean detailed score in the normal group was significantly higher than that in the non-CP and CP groups (q = 10.800 and 23.000, respectively, both P < 0.01). Additionally, the mean score of the 52-neuromotor examination in the non-CP group was significantly higher than that in the CP group (q = 12.200, P < 0.01) (Table 3).

Table 3.

Detailed scores in the writhing movement period among groups with different motor development outcomes.

Group	Infants(n)	Detailed score at the 12th month(mean ± SD)	F
Normal	46	35.33 ± 4.57	118.351***
Non-CP	44	24.66 ± 8.13
CP	24	12.71 ± 2.20

SD: standard deviation; CP: cerebral palsy. ***P < 0.001.

Establishing an early identification score from the detailed score in groups with different motor development outcomes

The detailed score was divided according to 95% CIs of the detailed score in the writhing movement period. Different early identification scores were based on the Integrated Management of Childhood Illness using three colors of red, yellow, and green.[9] The normal group was in the green area. The mean GM detailed score in normal infants was >−1.96 SD. The mild group was in the green area, between the normal group and severely abnormal group. The severely abnormal group was in the red area. The mean GM detailed score in abnormal infants was <−1.96 SD.

According to different early identification scores, the detailed score in the writhing movement period in groups with different motor development outcomes was determined. In the normal motor development group (green area), scores ranged from 42 to 27 (i.e., scores were ≥27). In the motor retardation group (yellow area), scores ranged from 26 to 18. In the CP group (red area), scores ranged from 17 to 5 (i.e., scores were ≤17).

Predictive validity of the red area

Detailed scores <17 were regarded as the red area and the diagnostic criteria at the 12th month were regarded as the gold standard. In this study, there were 33 infants in the red area and 24 were diagnosed with CP at the 12th month. The predictive validity was 92.11%, sensitivity was 100%, specificity was 90.00%, the positive predictive value was 72.73%, the negative predictive value was 100%, the false positive rate was 10, and the false negative rate was 0.

Discussion

Detailed scores in the writhing movement period reflect basic features of GMs

Assessing GMs is based on visual perception, while the detailed score of GMs is used to assess details, such as the sequence, extent, velocity, location where GMs are assessed, and exercise start and end times of GMs. These details not only complement assessment of GMs, but are also useful for differentiating between different types of movements. In this study, infants in the normal group had the highest detailed scores, while those in the CS group had the lowest scores. This finding suggested that the GM detailed score was consistent with assessment of GMs. Therefore, a better assessment of GMs is reflected by a higher detailed score. CS is the most abnormal movement. The detailed score combined with assessment of GMs is useful in high-risks infants.[10] The combination of early abnormal GMs and detailed analysis in infants is correlated with the severity of CP in later life.[11,12]

Consistency between 52-neuromotor examination scores and detailed scores

The 52-neuromotor examination in 0- to 1-year-olds contains many aspects, such as head circumference, vision and acoustic response, passive muscular tension, motor activity, and reflection.[9] This assessment can comprehensively assess fetal motor development, obtain clinical confirmation, and is an effective and reliable measurement method.[13]

In this study, different GMs had different characteristics. A better form of GMs reflects better development of muscle tension, movement, and reflection. In the 52-neuromotor examination, infants with CS GMs had the lowest detailed score, which was significantly lower than that of infants with normal GMs and PR GMs. Therefore, high-risk infants with CS have the worst damage to nerve function. Although GMs are assessed by visual perception, there is consistency between the detailed score and traditional and connected 52-neuromotor examination scores, which is consistent with the predictive results.[9] Different detailed scores can reflect different forms of GMs. Compared with assessment of traditional GMs, the detailed score is more comprehensive and accurate.

The detailed score is important for determining different developmental outcomes

CP is a disorder of nonprogressive injury of the fetal or infant brain, which leads to continuous and non-progressive disturbance of development of movement and posture.[14] Early diagnosis and early treatment are prerequisites for ameliorating the prognosis of CP. Therefore, distinguishing the severity of CP is especially important in clinical treatment.

CS GMs are widely used in early screening of CP and their outcome is varied. In this study, high-risk infants with different developmental outcomes had significantly different detailed scores. In the CP group, the detailed score was lowest in the writhing movement period and the normal group had the highest detailed score. These results suggest that the detailed score is important for determining different developmental outcomes.

Further study is required for detailed scores with different developmental scores. Establishing an early identification score of CP can identify high-risk infants with different risks of nerve development to provide early diagnosis and early intervention.

Establishing an early identification score and its clinical application

In this study, the number of high-risk infants with CP was higher than that of infants with CS. With assessment of GMs, high-risk infants can recover after intervention. Abnormal brain development in high-risk infants is correlated with different degrees of brain injury and brain development.

CP is often complicated by other nerve dysplasia, such as epilepsy. Many infants stop rehabilitation treatment for poor control of epilepsy, and this then develops into CP. In follow-up of high-risk infants, early diagnosis of CP is vital. Appropriate treatment for CP should be provided in a timely manner, the potential prognosis should be informed to parents, complications should be noted, and a feasible rehabilitation goal should be set. Furthermore, infants with mild abnormality should also be paid close attention regarding their rehabilitation. Positive, individual, and systematic rehabilitation can help high-risk infants recover. Therefore, establishment of an early identification score of CP is an important step in high-risk infants.

In our study, the false negative rate was 0 for the early identification score. Therefore, this score was able to identify all positive cases, avoiding missed diagnosis. The false positive rate was 10%, which was correlated with early rehabilitation. Therefore, the early identification score has good sensitivity and specificity, which indicate its usefulness.

Infants in the yellow area according to the early identification score provide the most useful information, and the presence of infants in the yellow area is correlated with flexibility and compensation of the infants’ brain. Timely intervention can decrease the risk of disability. Therefore, timely follow-up and appropriate rehabilitation are vital, so that these infants can recover their function.

Click here for additional data file.

Supplemental material, IMR902579 Supplemental Material for Establishing an early identification score system for cerebral palsy based on detailed assessment of general movements by Yuqing Wang, Ping Zhu, Zhongxiu Yang and Guixiong Gu in Journal of International Medical Research