Hypoglycemia screening of asymptomatic newborns on the 2nd day of life

1 Background

Hypoglycemia is one of the most common met-abolic disturbances during the neonatal period. Physiologically low serum glucose concentrations after birth are part of the transition period, which may last hours to days. Increased insulin levels, exaggerated metabolic demands, and lower glucose resources are believed to contribute to the occurrence of hypoglycaemia in the neonatal phase [1].

Presently, there is no definite consensus regarding the screening and treatment threshold for neonatal hypoglycemia despite the 10% incidence among healthy term neonates [2]. In 2011, the American Academy of Pediatrics Committee on Fetus and Newborn published a clinical report suggesting management guidelines for late-preterm and term infants with associated risk factors, targeting infants of diabetic mothers (IDM), large for gestational age (LGA) and small for gestational age (SGA) infants; healthy, asymptomatic full-term infants without risk factors do not require screening for hypoglycemia [3]. Of note, this report provides guidelines only for the initial 24 hours of life and recommends screening IDM and LGA infants for a period of 12 hours and screening SGA and late preterm infants for 24 hours. In their position statement, the Canadian Paediatric Society recommends a similar screening algorithm based on risk factors, additionally stressing the importance of adequate feeding in LPT and SGA neonates [4]. Moreover, the Pediatric Endocrine Society issued recommendations for the evaluation and management of persistent hypoglycemia in neonates, infants, and children covering the period beyond the initial 24 hours of life [5].

Given that neonatal hypoglycemia is more com-mon during the first 24 hours, the use of hypogly-cemia proposed management plans for infants at risk during the first day of life has been widely studied (level of evidence B). In contrast, data targeting neonatal hypoglycemia beyond the age of 24 hours are generated from lower levels of evidence (C and D), predominantly reflecting observational studies and expert opinions [6]. As a result, pediatricians may encounter difficulties when managing infants at risk for hypoglycemia on the second day of life with discharge planning in progress given that the current AAP clinical report does not specifically address this period. In contrast, the PES suggests a plasma glucose value of 50 mg/dL (2.8 mmol/L) as a cut-off value for the diagnosis of hypoglycemia.

In our study, we compared the incidence of hypoglycemia during the 24–48 hours of life interval among healthy, asymptomatic neonates with and without risk factors using the PES hypoglycemia threshold of 50 mg/dL (2.8 mmol/L). To address concerns regarding safe discharge for high-risk neonates on the second day of life, we preprandially assessed rates of hypoglycemia in relationship to predefined risk factors (prematurity, IDM, LGA, and SGA status) in newborn nursery settings emphasizing early breastfeeding strategies.

2 Materials and methods

2.1 Study design

This prospective observational cohort study was conducted at the newborn nursery of Jack D. Weiler Hospital of the Albert Einstein College of Medicine/Children’s Hospital at Montefiore, Bronx, New York from March 2018 to November 2019. The subjects were recruited using a convenience sampling method.

We included healthy, asymptomatic neonates in our baby-friendly newborn nursery who were > 24 hours and < 48 hours old, rooming with mothers without risk factors for hypoglycemia and those with the following risk factors: IDM and LGA, SGA and late preterm neonates.

Based on our practice guidelines (Fig. 1), neonates without risk factors did not undergo screening for hypoglycemia. IDM and LGA neonates were screened for 12 hours after birth, whereas SGA and late preterm neonates were screened for 24 hours. Late preterm gestation was defined as≥35 and < 37 weeks, SGA status < 10th percentile weight for gestational age and LGA status > 90th percentile weight for gestational age using 2013 Fenton growth curves. Neonates with multiple and/or overlapping risk factors (e.g., IDM late preterm neonate) were screened for 24 hours.

OPEN IN VIEWER

Fig. 1

Newborn nursery hypoglycemia management.

We excluded neonates undergoing sepsis evaluation, known polycythemia (defined as central venous hematocrit > 65%), neonatal abstinence syndrome (at risk for feeding intolerance), and neonates with hypoglycemia in the first 24 hours who failed the newborn nursery hypoglycemia management algorithm and required admission to the neonatal intensive care unit (NICU).

On the second day of life, the preprandial glucose concentration was assessed using a point-of-care test (POCT) bedside device (Abbott Precision XceedPro, Alameda, CA, USA). Neonates with POCT concentrations < 50 mg/dl were considered hypoglycemic, and a clinical care team different from the investigators was promptly notified.

Neonates were fed on demand by their mother or by the nurse; the exact time of enteral feeding completion recorded by previously instructed mothers and nurses was verified and noted to assure POCT sampling immediately prior to the next feeding. For each neonate, we prospectively recorded the preprandial glucose value, chronological age in hours at the time of bedside glucose test, type of the last feeding, time interval from the completion of the feeding to the bedside glucose test, neonatal feeding administration (mother vs nursing staff) immediately before the glucose test, and clinical signs of hypoglycaemia prior to and following this feeding (no symptoms, sleepiness/poor feeding, jitteriness). Other clinical information collected from health records included gestational age, birth weight, sex, race, mode of delivery, singleton vs multiple gestations, Apgar scores, SGA or LGA status, maternal diabetes mellitus, maternal body mass index (BMI) at booking, delivery room resuscitation, hypoglycemia incidence and intervention with early feeding and dextrose gel administration during the initial 24 hours of life.

3 Results

Over the study period, we approached 163 mothers of eligible neonates. In total, 12 declined consent, and one neonate was excluded from the study because the feeding timing prior to glucose sampling was missing. The study was completed after glucose samples of 150 neonates (no risk factors = 50, LGA + IDM = 50, SGA + LPT = 50) were obtained.

The expected difference among the 3 groups in gestational age and birth weight was noted in our ethnically diverse study population (Table 1). None of the neonates displayed clinical signs of hypoglycaemia, and all but one neonate was fed by their mothers prior to the glucose test on day of life 2. There was a trend towards a higher maternal BMI in the LGA + IDM group, not reaching statistical significance (P = 0.07).

Primary and secondary study outcomes are dep-icted in Table 2. The incidence of hypoglycemia on the second day of life using the < 50 mg/dL (2.8 mmol/L) threshold was 10% (15/150) across all groups. The SGA and LPT groups comprised the majority of the hypoglycemic cases (53.3%) with an incidence of hypoglycaemia of 16% followed by the no risk factors (8%) and LGA + IDM (6%) groups. The difference between the incidence of hypoglycemia between these groups was not significant (P = 0.21). Similarly, the SGA + LPT group had the lowest serum glucose recorded followed by the no-risk and LGA + IDM groups (61.2±11.8 vs 64.9±11.0 vs 65.9±11.1 mg/dL, P = 0.09). Neonates in this study were fed on demand with the time from last feed completion to glucose check not different between the groups (P = 0.69). The incidence of hypoglycemia was not altered by the feeding type just prior to the test (P = 0.12). Forty-three infants (28.7%) out of 150 enrolled were hypoglycemic (P < 0.00001) and needed 46 therapeutic interventions with early feedings and/or dextrose gel during the first 24 hours after birth (P < 0.0001), and these infants were predominantly in at-risk groups (Table 2).

Fifteen neonates who were hypoglycemic in our study were promptly evaluated and managed by their clinical care teams; 12 responded to additional dextrose gel administration and feeding intervention with hypoglycemia resolution and subsequent discharge from newborn nursery. The remaining 3 neonates (2 from the no-risk group and 1 from the SGA + LPT group) were transferred to the NICU and started on intravenous dextrose therapy; they did not receive additional hypoglycemia evaluation with the exception of plasma glucose concentration verification and point-of-care monitoring and were discharged home after 3 (2 infants) and 4 (1 infant) days in the NICU.

The hypoglycemia threshold for the period from 24 to 48 hours of life is not well established and is influenced by reports of adverse neurodevelopmental outcomes in preterm neonates associated with the hypoglycemia threshold < 47 mg/dL (2.6 mmol/L) suggested by World Health Organization [7–9]. These reports had either “difficulty of providing causation” or were not supported by prospectively collected data [10, 11]. The natural desire to better screen, prevent and treat hypoglycemia is counterbalanced by valid concerns regarding increased NICU admission rates for asymptomatic patients, increased separation, delayed mother-baby bonding, decreased breastfeeding rates, and excessive blood draws that have been raised when implementing these guidelines [12]. Furthermore, screening a large population of otherwise healthy neonates negatively impacts rates of exclusive breastfeeding [13].

In our study, we prospectively compared preprandial plasma glucose concentrations in healthy asymptomatic neonates on the second day of life between neonates with and without risk factors in a baby-friendly newborn nursery setting. We found that the overall incidence of hypoglycemia among healthy asymptomatic neonates between 24 and 48 hours of life was 10% regardless of the risk factors based on the PES recommendation for maintaining blood glucose levels of greater than 50 mg/dl (2.8 mmol/L) in the first 48 hours of life. The study population included neonates who underwent hypoglycemia screening based on the AAP screening guidelines on the first 24 hours of life or no screening in the absence of risk factors.

The incidence of hypoglycemia on the second day of life in at-risk neonates was reported by Harris et al., who evaluated neonates enrolled in the Sugar Babies Study and reported an incidence of 20% for SGA/late preterm infants and 17–20% for IDM/LGA neonates compared to 12% and 6%, respectively, for these groups in our study [14, 15]. This difference may be due to the higher proportion of breastfed neonates in the Sugar Babies study and due to the fact that our enrolment criteria did not include neonates failing newborn nursery management followed by NICU admission during the first 24 hours.

Importantly, a recent GLOW study published by the same investigators reported glucose concentration in the first 5 days of life among healthy asymptomatic neonates without risk factors and compared it to the cohort of neonates with risk factors for hypoglycemia from the Sugar Babies Study, utilizing both continuous glucose monitoring and intermittent glucose testing for glycemia determination [16]. Using the cut-off of 50 mg/dL (2.8 mmol/L), the incidence of hypoglycemia on the second day of life was 22% (using Epoc blood gas analyser) among neonates without risk factors and as high as 58% using continuous monitoring data. The incidence of hypoglycemia in our group of 50 healthy term neonates without risk factors was 8%, which is less than that reported by Harris et al. This difference may be attributed to the fact that the GLOW study performed intermittent glucose measurements twice daily after the first 24 hours of life, and most of the neonates (85%) were exclusively breastfed.

Consistent with the Glow study, we showed no difference in hypoglycemia incidence between neonates with and without risk factors on the second day of life. Importantly, neonates with risk factors were more likely to be hypoglycemic (P < 0.0001) and require intervention to achieve normoglycemia during the first 24 hours of life (P < 0.0001); it is unclear whether this difference would persist beyond the initial 24 hours with our increased study sample size. Our results are supportive of the 2019 Canadian Paediatric Society position paper (using 47 mg/dl or 2.6 mmol/L hypoglycemia threshold) statement: “it is reasonable to screen once or twice on day 2 when there has been more than one glucose reading < 2.6 mmol/L in the first 24 hours, to ensure levels remain at or above this level” [4].

The time elapsed from the last feeding prior to POCT sampling with a mean value of approximately 2.5 hours was not predictive of hypoglycemia, which is consistent with a previous report evaluating term, appropriate size for gestation and exclusively breastfed neonates [17]. Additionally, the type of last feeding prior to glucose testing, including breastfeeding, formula feeding, or a combination of both, was not associated with the development of hypoglycemia. While the Heck and Erenberg study re-ported a decreased incidence of asymptomatic hypoglycemia among exclusively breastfed neonates, others reported a greater immediate posttreatment increased in glucose concentration in formula-fed infants; obviously, it is important that these findings are interpreted in a broader context of fewer repeat episodes of hypoglycemia associated with breastfeeding with the prospects of long-term benefits [18–20].

One limitation of our study was hypoglycemia management at the discretion of clinical teams after concentration < 50 mg/dL (2.8 mmol/L) detection as well as lack of dedicated follow-up beyond routine postdischarge visits with family pediatricians. Another limitation of the study is that the glucose concentration was determined by a point-of-care test, not by a glucose oxidase method. Point-of-care glucose metres utilize whole blood samples with results typically internally converted to plasma values because these devices are calibrated in their manufacturing and validation processes to report equivalent plasma values [21].

Three healthy, asymptomatic neonates (2%) were admitted to the NICU based on 24–48 hours of hypoglycemia screening and failed hypoglycaemia management in the newborn nursery. All patients subsequently had short, uncomplicated NICU stays ending with rapid hypoglycemia resolution, suggesting transitional hypoglycemia physiology. Ten percent hypoglycemia rates in our study approximate the findings recently reported by Dani et al.: “in our centre, up to 10% of healthy term infants have glycaemia of 45 to 49 mg/dl at 24 to 47 hours of life” [22]. Importantly, our additional random preprandial hypoglycemia screening of asymptomatic healthy neonates at 24- to 48-hour intervals beyond the traditional 24-hour AAP screening guidelines was not suggestive of more serious hypoglycemia etiologies; there is uncertain, if any, clinical significance of more pronounced neonatal transitional hypoglycemia.

5 Conclusion

The ten percent hypoglycemia incidence detection using a single preprandial time point on the second day of life among asymptomatic, healthy newborns who have already passed hypoglycemia screening during the first 24 hours of life is high; the significance of these findings remains unclear in the absence of long-term neurodevelopmental follow-up of these infants.