Assessment of Primitive Reflexes in High-risk Newborns

Background Assessment of primitive reflexes is one of the earliest, simplest, and most frequently used assessment tools among health care providers for newborns and young infants. However, very few data exist for high-risk infants in this topic. Among the various primitive reflexes, this study was undertaken particularly to describe the sucking, Babinski and Moro reflexes in high-risk newborns and to explore their relationships with clinical variables. Methods This study is a cross-sectional descriptive study. Sixty seven high-risk newborns including full-term infants required intensive care as well as premature infants were recruited in a neonatal intensive care unit using convenient sampling method. The sucking, Babinski and Moro reflexes were assessed and classified by normal, abnormal and absence. To explore their relationships with clinical variables, birth-related variables, brain sonogram results, and behavioral state (the Anderson Behavioral State Scale, ABSS) and mental status (the Infant Coma Scale, ICS) were assessed. Results The sucking reflex presented a normal response most frequently (63.5%), followed by Babinski reflex (58.7%) and Moro reflex (42.9%). Newborns who presented normal sucking and Babinski reflex responses were more likely to have older gestational age, heavier birth and current weight, higher Apgar scores, shorter length of hospitalization, better respiratory conditions, and better mental status assessed by ICS, but not with Moro reflex. Conclusions High risk newborns presented more frequent abnormal and absence responses of primitive reflex and the proportions of the responses varied by reflex. Further researches are necessary in exploring diverse aspects of primitive reflexes and revealing their clinical implication in the high-risk newborns that are unique and different to normal healthy newborns. Keywords Primitive reflex; High risk infants; Korean; Moro reflex; Sucking reflex; Babinski reflex; The Anderson Behavioral State Scale; Infant Coma Scale


Introduction
The advances in medical technology and improved neonatal care have profoundly increased the survival of high-risk newborns including full-term infants required intensive care as well as premature infants. Clinicians are encountering smaller and sicker newborns with extremely low birth weight (ELBW) and serious health conditions in the neonatal intensive care unit (NICU). Considering that high-risk newborns experience higher mortality and greater risks of various health and developmental problems, early health assessment for their fi rst course of life is critical. Multidisciplinary-based critical care in the NICU requires a variety of assessment tools to evaluate the conditions of these fragile patients and communicate their condition with families and other health care professionals.
Primitive refl exes are brainstem-mediated, automatic movements which may begin as early gestation week 25-26, and which are fully present at birth in term newborns [1,2]. Assessment of these refl exes is one of the earliest, simplest, and most frequently used assessment tools among health care providers for newborns and young infants [2]. Primitive refl exes start to disappear when the central nervous system matures and voluntary motor activities replace them [2]. The normal age-appropriate response is related to the development of normal motor function of newborns or infants. In general clinical settings, the responses of primitive refl exes have been categorized mostly as dichotomous responses such as normal/abnormal or present/absent. However, primitive refl exes often present various degrees of response, which may be meaningful in the neurological capability to stimu-Manuscript accepted for publication September 30, 2011 a Department of Nursing, Inha University, Incheon, South Korea lus. Persistent, vigorous, weak, or unsymmetrical responses are closely-linked with neurological impairment in full term [3] and high-risk newborns [2]. Very limited data is available concerning the assessment of primitive refl exes of highrisk newborns in both aspects of PTB or pathologic conditions. While not all primitive refl exes are uniformly present in preterm newborns [2], how the diversity of responses of refl exes are related to clinical conditions and how clinicians can interpret them in a neurological context are unclear. Of the primitive refl exes, the sucking, Moro and Babinski refl exes are frequently documented in the literature due to their important roles. The sucking refl ex plays a critical role in oral feeding in coordination with breathing and swallowing, and is essential for nutrition intake for survival and growth. The Moro refl ex is an involuntary protective motor response against abrupt disruption of body balance or extremely sudden stimulations [2]. The Babinski refl ex involves the extensor and fl exor of the foot as a nociceptive motor response of pyramidal tract [4]. Understanding these three refl exes might be more valuable in the neuro-behavioral examination of newborns considering their value for survival, protection, and development.
Therefore, the present study was undertaken to evaluate three representative primitive refl exes, the sucking, Moro and Babinski refl exes in Korean high-risk and preterm newborns. The specifi c study purposes were to describe various levels of their responses and to explore relationships among the primitive refl exes and various clinical conditions in highrisk newborns.

Study design and participants
This is a cross sectional explorative study with high-risk newborns recruited from September 2008 to June 2009 at the NICU of a university hospital located in Korea. Since primitive refl exes are usually present beginning at and after least 25 weeks of gestational age (GA) [2], newborns who were born at 25 weeks or older of GA were included for the study. Newborns were excluded if they had been transferred from other hospitals, to exclude possible missing information that was important to understand the clinical course. Newborns were also excluded if they had congenital, genetic or skeletal disorders; had received sedatives or the use of restraints; or had a maternal history of substance abuse or alcoholism, since these histories might infl uence the motor or neurologic responses.

Data collection
Before data collection, the director of the NICU approved the study with the exemption of the informed consent since neurologic assessments particularly primitive refl exes, mental status assessment and behavioral status are part of the standard nursing procedures in the NICU. As well demographic and clinical information were anonymously obtained from existing medical records involving less than minimal risk without any additional data collection procedure. The researchers followed principles in the Declaration of Helsinki during the whole process of this study. The assessment of the sucking, Moro and Babinski refl exes were performed 48 hours after birth, to permit the newborns to become stable in the extrauterine environment after the usually necessary critical care had been provided. Newborns were examined while they were awake and lying comfortably on their backs. Examinations were avoided within 1 hour of feeding or direct contact with NICU staff or parents. Behavioral and mental status was assessed as a baseline state of newborns immediately before examining primitive refl exes.

Measurements
The study variables consisted of clinical information, primitive refl exes, behavioral and mental status. Clinical information included demographics, birth history, and medical conditions such as respiratory condition, length of hospitalization (LOH), type of risk and brain sonogram. Respiratory status was categorized as completely self (independent), self with assistance such as oxygen therapy and continuous positive airway pressure (CPAP), and ventilator dependent. LOH was determined following discharge through medical record review. The risk group was categorized into; physiological risk group for simple preterm newborns without any pathologic conditions and pathologic risk group if pathological conditions were evident in addition to prematurity. Brain sonogram results were categorized as grade 0-IV (grade 0: normal result; grade I: hemorrhage limited to the germinal matrix; grade II: intraventricular hemorrhage without ventricular dilatation; grade III: intraventricular hemorrhage with acute ventricular dilatation; and grade IV: intraventricular hemorrhage extending into adjacent brain parenchyma) [5].
Evaluation of primitive refl exes for this study was based on the primitive refl ex profi les Capute et al published [3], but did not follow their 5 point scale of 0 for absence to 4+ for pathologic prolonged response. Since Capute et al published the primitive refl ex profi les in 1970s, high risk newborns of those days are quite different from those of these days. In addition, no further study was identifi ed on the discriminative accuracy of 5-rating response. Therefore, we modifi ed the 5 point scale to the 3 point scale indicating 0 for absence, 1 for abnormal and 2 for normal. Each refl ex was examined up to fi ve times if the newborn showed no response or if response was ambiguous to obtain best positive responses.
The newborns' behavioral status was assessed using the Anderson Behavioral State Scale (ABSS) by observing fi ve areas including the extent of eye openness, patterns of respiration, body movement, muscle tension and crying [6], and has been applied to high-risk newborns previously [7]. The score ranges from 1 to 12. Higher scores indicate a more alert state and 6 is a cut-off for an optimal alert state in infants [6]. The Cronbach's alpha was 0.97 with Korean highrisk newborns previously [7]. The newborns' mental status was evaluated using the Infant Coma Scale (ICS), which was developed by modifi cation of the pediatric Glasgow Coma Scale (GCS), is a reliable tool to evaluate the mental status of high-risk newborns [8]. ICS evaluates the mental status of high-risk newborns in eye openness, verbal response and motor response. The total score of ICS can range from 3-15, and is interpreted just like GCS (i.e., higher score corresponding to a more alert state). Reliability and validity of ICS with Korean newborns displays a Cronbach's alpha of 0.78 [8]. The performance and coding rule for each refl ex, as well as the ABSS and ICS protocols, were confi rmed before and during data collection to assure the validity and the reli-ability of the measurements.

Data Analyses
The Statistical Packages for the Social Sciences version 18.0 was used for data analysis. Descriptive statistics and variable distributions were evaluated for data quality assessment. To describe clinical information and primitive refl exes, frequencies with percentages and means with standard deviations (SD) and ranges are presented. To determine relationships between primitive refl exes and other research variables, χ 2 , analysis of variance (ANOVA) and Bonferroni post hoc analysis was used with α = 0.05 in a two-tailed test.

Results
A total of 63 newborns were included in the data analysis. Thirty three newborns (52.4%) were male and 11 (17.5%) were full-term newborns. Their mean GA of 33.6 (± 3.4) weeks (range: 25.9-40.7 weeks) and a mean birth weight (BW) of 2041.7 (± 706.3) grams (range: 806-3910 grams). At the time of assessment, the mean current age and current weight was 2.9 (± 1.5) days (range: 1-7 days) and 1983.8 (± 750.3) grams (range: 639-3900 grams), respectively. The sucking refl ex presented a normal response most frequently (63.5%), followed by Babinski refl ex (58.7%) and Moro refl ex (42.9%). While one-third of newborns presented an abnormal Moro refl ex (38.1%) or abnormal Babinski refl ex (33.3%), only 11% of newborns presented an abnormal response of the sucking refl ex. Absence responses were most frequent at the sucking refl ex (25.4%), followed by the Moro refl ex (19.0%) and Babinski refl ex (7.9%). Most of the newborns were preterm births and displayed fair Apgar scores, good respiration, and minimal brain injuries in sonogram. The majority of the newborns had pathologic conditions, stayed long in the hospital, low ABSS scores and moderate ICS scores (Table 1). We also explored relationships among primitive refl exes and clinical variables using χ 2 analyses for categorical variables and ANOVA analyses for continuous variables. Babinski refl ex showed different responses by brain sonogram results (χ 2 = 16.56, P = 0.035). The sucking refl ex (χ 2 = 26.96, P < 0.001) and Babinski refl ex (χ 2 = 16.23, P = 0.003) were associated with respiratory condition, while the Moro refl ex was associated only with the risk type (χ 2 = 8.91, P = 0.012). Table 2 presented ANOVA analyses among the refl exes and continuous variables. While the sucking and Babinski refl exes concurrently presented signifi cant associations with seven out of nine variables, the Moro refl ex showed no relation with any of variables except ICS score (F = 8.37, P = 0.001). Newborns presenting a normal response of the sucking and Babinski refl exes were more likely to have older GA, heavier birth and current weight, shorter LOH, and higher Apgar scores and ICS scores. Current age and ABSS did not display signifi cant patterns along with any response of the three primitive refl exes.

Discussion
Assessing primitive refl exes is an important part of high risk newborns. Primitive refl ex responses in premature newborns may often vary in degree of responses. However clinicians have mostly using dichotomous criteria of presence or absence of primitive refl exes without guide or research based evidence in high-risks. It is also unclear how the degree of their response should be described in detail and how much the information could be meaningful clinically. This study was conducted to describe primitive refl exes assessed in 63 Korean high-risk newborns. After review of details in each primitive refl ex, three particularly noteworthy fi ndings warrant further comment.
First, a considerably high number of high-risk newborns (36%) presented an abnormal or absent sucking refl ex. The sucking is a survival refl ex that is observed as early as GA 25-26 week and typically before 28 weeks, although intrauterine swallowing activity begins as early as 12 week of GA [1]. Absence of the sucking refl ex in 25.4% of the newborns was an unexpectedly substantial fi nding considering that their mean GA was 34.6 weeks. A possible explanation is that their clinical condition, such as diffi cult respiration and decreased mental status, may have delayed sucking refl ex response. Premature newborns who are ventilated can exhibit signifi cantly poor sucking ability [9]. The presence of nasal CPAP or endotracheal tube, as well as oro-gastric tube if indicated, could alter perioral activities [10]. Therefore, the newborns in this study may have developed a similar alteration or habituation from strenuous nasal or perioral stimulation, which was responsible for the absence of the sucking response. Furthermore, although the associations between brain sonogram fi ndings and refl exes were not identifi ed, the mental status as assessed by ICS showed a statistically signifi cant association with the sucking refl ex. Particularly, the mean ICS score of the newborns presenting an abnormal or absent sucking refl ex response < 9, which is a cut-off point for the normal mental status of infants [8]. In a previous study, sucking behavior was shown to provide an indirect indicator of maturity of neurological development in premature newborns [11]. Therefore, an abnormal or absent sucking refl ex may imply a neurologic impairment in the high risk newborns in this study. This statement is consistent with their longer LOH.
Secondly, among the three primitive refl exes, the Moro refl ex presented quite different patterns with the clinical conditions of the newborns from the other two refl exes, while very similar response patterns were observed in the sucking and Babinski refl exes. Moro refl ex presented as the least frequent normal response and most frequent abnormal response. These observations are consistent with previous studies indicating that the Moro refl ex is especially weak in preterm newborns because of lower muscle tone, poor resistance to passive movements and slow arm recoil, compared with those of full term newborns at same post-conceptual age [12]. In our study, morbidity-related factors such as LOH, Apgar scores, and ICS total scores were statistically associated with the sucking and Babinski refl exes, but not the Moro refl ex. This data may refer that the Moro refl ex is more likely related to infant development rather than pathologic conditions.
Lastly, through this study we tried to highlight the main issues of whether the presentation of an abnormal response(s) of the primitive refl exes in the high-risk newborns is bad, and, which of the refl exes carries a greater clinical importance in the case of an abnormal or absence response. In this study, there were signifi cance differences of clinical conditions between normal and the other responses. Although most of the posthoc analyses have not presented any differences of clinical condition between abnormal and absent responses, abnormal response seems as much adverse as absence response. This phenomenon is more obvious in the sucking and Babinski refl ex. Therefore, it might be more prudent to assess them separately and search for an explanation through further research. How do we apply this evidence to our practice? Assessing primitive refl exes is a part of standard care in NICUs which is easy to evaluate and feasible to perform in any circumstances without high technology equipment. Although many of clinicians understand the importance and advantages of primitive refl ex assessment and have assessed them to determine if they are present or absent, clinicians often disregard abnormal responses which high-risk newborns often present. Subtle changes of primitive refl ex could be valuable indicators of current medical conditions and potential future health and developmental outcomes while newborns grow up. It would be benefi cial to develop standardized assessment protocol for primitive refl exes in high-risk infants using at least four scales (absence, hypoactive, normal, and hyperactive) as assessment of adults' refl exes. Further study is also recommended in this understudied area. It would be more benefi cial to assess variety of primitive refl exes using more detail assessment criteria. For example, the sucking refl ex may be differently categorized by its presence, frequency, regularity, pressure and coordination with breathing or swallowing. Some refl exes, such as the Babinski refl ex, might present different responses bilaterally. In that case, assessing whether responses are consistent bilaterally may refl ect more differentiated aspects of clinical conditions.