Neuro-Developmental follow up of high-risk newborns using hammersmith-dubowitz method

Background and Objectives: High risk infants are prone to delay in neurological development due to perinatal damage sustained by the brain and nervous tissue. Despite being on regular follow-up, delayed neurological development is often missed even by experienced examiners as most of the examination and history are subjective. We aim to validate an objective scoring systemthe Hammersmith-Dubowitz neurological exam which was extensively researched and developed taking into consideration normal responses from a normal cohort. Materials and Methods: Over a period of two years, 112 infants were categorized as high-risk newborns and were followed up for a period of one year of age. The children underwent detailed milestone assessment, physical and neurological examination by two independent examiners along with objective scoring with Hammersmith infant neurological exam. Any abnormality in these tests was considered delayed development. Results: Of the 112 infants, 102 came for at least one follow up and 81 completed one year follow up with a mean follow up duration of 9.42 months. On the combined assessment scale of developmental, physical and complete neurological examination 7.84% infants were found to be abnormal. However, on objective assessment with Hammersmith neurological examination 16.67 % of the infants were found to have abnormal neurological development. Conclusion: Hammersmith neurological examination is a useful objective scoring tool to identify delayed neurological development early.


Introduction
The two important aspects of neonatal care are the fragility of a neonate and the consequences which can result from management of such a patient which greatly impacts his/her future. A second's delay might lead to years of morbidity and dependence especially when the damage is Neurological. High-risk newborns (HRNB) are especially vulnerable to neurological damage and its complications [1,2]. Early identification of these children can help in better management of their disability [3,4].
The neurological assessment of newborns has evolved through several stages to reach a modern era. The current standard of examination is using one of the many scoring scales to objectively assess the neurological status and express it in numbers or other The neonatal behavioral assessment scale (NBAS); Neurobehavioral assessment of the preterm infants (NAPI); the Assessment of preterm infant's behaviour (APIB); the Neonatal intensive care unit network neurobehavioral scale (NNNS); and Dubowitz scale. Of these HINE is proved in multiple studies to be more comprehensive and superior in identifying delayed neurological development and milestone progress [5,6].

Original Research Article
Pediatric Review: International Journal of Pediatric Research Available online at: www.medresearch.in 278|P a g e The use of scales like HINE has enabled early identification of infants with delayed neurological development and early behavioral changes. These scales are specifically designed as screening tests to identify children at risk for these conditions. These include various components like motor, sensory, behavioral and milestones.
Most of these scales requires prior training and experience whether by formal or informal means for their use. The main criticism against these scales are the fact that administration of such an objective examination is time-consuming, requires expertise and children may not be available for follow-up.
Moreover, these scales are effective only when done serially and well documented. The clinical effectiveness of these scales considering the above said factors in a clinical environment that exists in India is not proven to date. Thus the questions arise -Are these scales effective in our clinical setting? Will the follow-up compliance be good in our country, where there is high prevalence of poverty and illiteracy? Will these scales reduce the need for unnecessary investigations in the infant population at risk?
Since the purpose of infant follow-up in first year of age is to identify impairments as early and broadly as possible and to provide guidance for families. Their choice of standardized assessment is heavily influenced by feasibility and prognostic considerations. In particular, examinations need to balance the demands of clinical imperatives and time constraints.
We selected the HINE for implementation in this study because it is a well-studied neurological examination in healthy or high-risk infants and is an objective scale and results are expressed in numbers easily understandable and without controversy.
The HINE is an easily performed and relatively brief standardized and scorable clinical neurological examination for infants aged between 2 and 24 months, accessible to all clinicians, with good inter-observer reliability even in less experienced staff. It has no associated costs such as lengthy certifications or proprietary forms. The use of the HINE optimality score and cutoff scores provides prognostic information on the severity of motor outcome.
The HINE can further help to identify those infants needing specific rehabilitation programs. It includes 26 items assessing cranial nerve function, posture, quality, and quantity of movements, muscle tone, and reflexes and reactions. Each item is scored individually (0, 1, 2, or 3), with a sum score of all individual items (range 0 to 78). A questionnaire with instructions and diagrams is included on the scoring sheet, similar to the Dubowitz neonatal neurological examination. Optimality scores for infants three to 18 months are based on the frequency distribution of neurological findings in a typical infant population: when an item is found in at least 90% of infants, it is considered optimal [7].
Sequential use of the HINE allows the identification of early signs of cerebral palsy and other neuromotor disorders, whereas individual items are predictive of motor outcomes. For example, in preterm infants assessed between six-and 15-months corrected age, scores greater than 64 predict independent walking with a sensitivity of 98% and specificity of 85% [8].
Conversely, scores less than 52 were highly predictive of cerebral palsy and severe motor impairments [9].

Materials and Methods
Place As a part of the study, all the TERM infants delivered at our hospital were considered for eligibility. As all the infants were delivered at the same institute and all were term only those satisfying criteria for high-risk newborns were included. Infants born out of both normal vaginal delivery and caesarean section were included in the study. The inclusion criteria were as follows: Selection Criteria, Inclusion Criteria: All high-risk term newborns were included in the study with the criteria for high risk being-1. Major morbidities such as chronic lung disease, intraventricular haemorrhage, and periventricular leucomalacia 2. Perinatal asphyxia -Apgar score 3 or less at 5 min and/or hypoxic-ischemic encephalopathy 3 Once the child is recruited, basic data and maternal data were noted including all possible risk factors. After discharge, the children were followed up over the phone and were called for repeat follow up examinations at 3, 6, 9 and 12 months [10], coinciding with their vaccination schedule.
The children were administered the HINE infant scoring system and the scores were documented. Any infants with low scores were referred to a specialist clinic and these children were in close follow up along with regular examinations.
Those children identified with the low score were followed up to note any improvement in the score with intervention. A score of < 60 as a sign of retardation and referred for further management and investigations.
The infants with normal scores were followed up to one year or further for any neurological deficits or significant delay in milestones.
All these infants were followed up for a period of one year with interval assessments at 3, 6, 9 and 12 months of age coinciding with their vaccination schedule. The children underwent detailed milestone assessment, physical and neurological examination by two independent examiners with a minimum two years of experience in newborn assessment along with objective scoring with Hammersmith infant neurological exam.
Any abnormality in these tests was considered delayed development. The follow-up ended after one year from birth. All the data was compiled and statistically analysed.
Statistical Methods: Descriptive and inferential statistical analysis has been carried out in the present study. Significance is assessed at 5% level of significance.
The following assumptions on data are made. Assumptions: 1. Dependent variables should be normally distributed, 2.Samples drawn from the population should be random, Cases of the samples should be independent. Chi-square/ Fisher Exact test has been used to find the significance of study parameters on categorical scale between two or more groups.

Results
Number of Subjects: 112 A total of 2173 potential newborns were screened for recruitment. Of which 112 were highrisk term newborns. All the infants were admitted to NICU with different diagnoses and were treated appropriately.
Eight of the infant's parents were not willing to come for further follow-up or be a part of the study. 22 infants were discharged against medical advice and 19 infants died during treatment due to various causes. Of the 112 infants discharged from the NICU, all were assessed and were called for a repeat assessment at 3, 6, 9, and 12 months coinciding with their vaccination schedule. Of the 112 infants, only 102 presented for one follow up at least.
The infants were stratified based on normal and abnormal HINE. Under both the groups, some infants lost to follow-up.
And the infants with abnormal HINE scores underwent necessary interventions after which some of them showed improved HINE scores on further follow-ups.      Out of the 112 high-risk term newborns include in the study, at 12 months follow up -a total of 72.3% (81 out of 112) babies came for followup of which 66.1% (74 out of 81) had normal HINE scores and 6.3% (7 out of 81) had abnormal HINE scores and 27.7% (31 out of 112) lost to followup. Initially at the beginning of the study, it took around 15 to 20 minutes for the assessment of HINE score each time in a well cooperative infant. Gradually with repetitive assessments and experience, the time taken to examine decreased to a maximum of 5 minutes in a well cooperative infant. Just like Ballards scoring which initially took time for assessment but eventually with practice, it is routinely done easily, similarly HINE can also be performed easily over time. This makes it easier for use in regular clinical practice.

Discussion
Delayed neurological development is the leading cause of permanent disability in a newborn infant. Not only does it affect the infant alone, but also causes a major social and economic burden on the family who will have to manage a morbidly disabled child. In a developing country like India, the social effects of such a disabled child are multiplied manifold due to the scarce resources available for treatment and the poor financial means available with the family.
Ultimately these children due to lack of adequate care have high mortality rates within five years of life [10] Thus, identifying these children early is of prime importance. On looking up the various means for the assessment of these high-risk infants, we found the Hammersmith Infant Neurological Examination (HINE). Though the HINE score has previously been evaluated in various settings, a complete prospective trail using it has never been done on the Indian population. Moreover, we used the scale exclusively in high-risk term infants as a screening tool in a specified population set to identify early developmental delay which is unique to our study. We have compared our study with various other studies and compared the results.

Demography and maternal factors:
The gender distribution among the infants in our study is 49.1% male and 50.9% females which were not on expected lines with respect to the Indian population. However, given to the health of the female child among the low socioeconomic and low-income groups in Indian population. These findings are unique to our study.
Multiple maternal factors have been implicated in the neurological growth retardation in various studies in the west [1,5]. However when the factor of prematurity has been eliminated and the results analyzed no prenatal or maternal factors including mode of delivery affected the normal development of the nervous system of a newborn. The commonest gestational age was 37 and 38 weeks as seen in many Indian studies [10,11].
The mode of delivery is almost equal with 49.1% of children born out of LSCS and the rest were by NVD. Though these rates are significantly higher than the rest of the Indian population which is 17.2% [12,13], this can be explained by the fact that the study population consists of high-risk infants who in turn would have been high-risk deliveries or had fetal distress in the prenatal period.
The birth weight in our study was on lines with the rest of the Indian population. 16% of these high-risk newborns were ventilated. Considering the fact that those infants who expired during the course of the study were excluded this number might have been higher.
Fetal Outcomes: Perinatal asphyxia, neonatal seizures, RDS and Neonatal jaundice are the most common condition among infants classified as high-risk newborns. This correlates well with the data published by Simeonsson et al [14]. Overall in the study population with the use of the HINE score we were able to detect 15.1% of the infants had a significant delay in neurological development.
Though the number varies significantly in various studies ranging from 6.3% to 33%, [15,16] the relatively lower number in our 89 study though the study population is of high-risk infants can be explained by the following factors: a. This is not a population based study, it is among the infants treated at a tertiary care centre with a good NICU setup and availability of neonatologist throughout. b. The high rate of LSCS (49.1%) reflects the standard of care received by these infants and their mothers in the perinatal period. c. The careful follow-up and detection of these conditions. d.
The follow-up period is only upto one year which is a limitation as conditions like autism are not usually evident by one year of age. However since the main aim of the study was not to calculate the prevalence of the condition these findings need not be attributed major significance. 27.7% of the infants were lost to follow-up at the end of one year with most of them being female infants; The follow-up rate was good compared to other studies in this arena [5,6].
This can be attributed to clubbing the infant examination schedule with that of the vaccination schedule. Since the vaccination compliance in India is upto 85% in semi-urban and urban areas [12] we sought to replicate these numbers and partially succeeded. With combining of the vaccination schedule and infant examination at 3rd months, 6th month, 9th month and one year we were able to achieve high follow up rates and hence recommend the same to be replicated in future studies to maintain high follow up rates at least upto one year of age. Moreover prior education of the patients about the need for regular follow up, maintaining a database of contact details including the telephone numbers and months contact over telephone also resulted in better followup and sustenance rates.

Hammersmith Infant Neurological Examination:
With the use of HINE score, we were able to identify 17(15.1%) infants with abnormal neurological development as early as at 3rd month of age. Of these infants, 6 of the 17(35.5%) improved back to normal HINE score by one year with appropriate referral and early initiation of treatment.
Thus HINE score helped identify these infants early and improved outcomes with early identification. The sensitivity of Hammersmith Score was 90.9% in our study with a specificity of Hammersmith Score nearly 100%. The positive predictive value was 100%. Mean duration of detection of developmental delay using the hammer smith score was 5.1 months. Only one patient was detected with delay in milestones with a normal neurological examination.
The objective scoring system of the HINE assessment gives a clear-cut picture of the neurological development status of the infant and is very useful to identify these children early. Similar findings were noted in two studies by Romeo et al. [17,18] using HINE assessment. Of the various assessments evaluated by multiple authors, [19,20,21,22] the HINE assessment is definitely better in many aspects as it is holistic, comprehensive, short, uncomplicated and requires minimum training.