The Burden of ‘Pokes’ in Newborns with Hypoglycemia Who Require IV Dextrose

Abstract

Objective:

Neonatal hypoglycemia is one of the most common reasons for NICU admission in otherwise healthy late preterm and term newborns. Routine care requires frequent blood glucose (BG) measurements, typically via painful heel lances, to titrate IV dextrose during treatment. However, little is known about the care for these newborns including the burden of blood glucose measurements.

Patients and Methods:

This was a retrospective chart review (07/2018 to 07/2024) of newborns ≥35 weeks gestation who had hypoglycemia screening in a university-affiliated newborn nursery and subsequently required NICU admission for IV dextrose. IV dextrose was initiated if target glucoses per national guidelines were not achieved with buccal dextrose gel ± oral supplementation. BG measurements before feeds typically via heel lance and point of care glucometers continued until euglycemia was achieved with enteral feeds alone. Descriptive statistics of newborn characteristics, number of BG measurements, and length of stay were analyzed.

Results:

Among 7,823 newborns screened for hypoglycemia in the newborn nursery, 298 (3.8%) required IV dextrose and NICU admission. Newborns who received IV dextrose experienced a median 29 (IQR 19–40) BG measurements while hospitalized. Over 25% of newborns had ≥40 BG measurements and almost half had ≥30 glucose measurements. The median length of stay was 6 (IQR 4–9) days.

Conclusions:

Newborns with hypoglycemia requiring IV dextrose experience a high burden of BG measurements, and most are hospitalized for ≥6 days. Practices that may safely reduce the number of painful ‘pokes’ during clinical care, such as using continuous glucose monitoring, should be investigated.

INTRODUCTION

Neonatal hypoglycemia is one of the most common reasons for neonatal intensive care unit (NICU) admission in otherwise healthy, late preterm and term newborns.^1–3 Among ‘at risk’ newborns (e.g. late preterm, small and large for gestational age, infant of diabetic mother), 4–14% will not respond to buccal dextrose gel and/or enteral supplementation and will require IV dextrose.^4–7 Yet, our understanding of the care for newborns with hypoglycemia who require IV dextrose is limited. Frequent blood glucose (BG) measurements, typically via painful heel lances, are required to titrate IV dextrose to achieve euglycemia and subsequently wean off IV dextrose treatment. However, minimizing painful skin breaking procedures in neonates is highly desired. Painful procedures in neonates lead to physiologic, behavioral, and hormonal disruptions and in preterm neonates are associated with negative long-term neurodevelopmental outcomes.^8–11 The burden of BG measurements in newborns with hypoglycemia requiring IV dextrose is unknown. Our objective was to examine the number of BG measurements and length of stay in newborns with hypoglycemia admitted to the NICU for IV dextrose.

METHODS

This was a retrospective study (07/2018–07/2024) of newborns ≥35 weeks gestation born at a university-affiliated children’s hospital (~4,500 deliveries annually) who had hypoglycemia screening in the newborn nursery and subsequently required NICU admission and IV dextrose. Hypoglycemic newborns were typically admitted to our level II NICU staffed by pediatric hospitalists unless needing invasive respiratory support or a central line in which case they were admitted to our level III/IV NICU staffed by neonatologists. The local institutional review board approved this study.

A standardized clinical care pathway for neonatal hypoglycemia screening aligned with national guidelines¹² (target glucoses ≥35 mg/dl for <4 hours of life; ≥45 mg/dl for >4 hours of life) was followed in the newborn nursery and previously described.⁴ Criteria for hypoglycemia screening included newborns small or large for gestational age (SGA, LGA), infants of diabetic mothers (IDM), or late preterm infants (<37 weeks gestation). BG was obtained 30 minutes after the initiation of the first feed and before each subsequent feed (AC) until 3 consecutive BGs were ≥45 mg/dL. SGA and late preterm infants also had a 24-hour glucose measurement. BGs were measured typically via heel lance using point-of-care (POC) devices (Precision Xceed Pro Blood Glucose or Freestyle Precision Pro Blood Glucose Monitoring System, Abbott Diabetes Care Ltd, Witney, UK).

Initial treatment for hypoglycemia included 40% buccal dextrose gel 0.2 mg/kg (Glutose^™ 15 Gel or Sweet Cheeks) along with supplemental feeding (breastfeed/expressed human milk, donor human milk, formula). If hypoglycemia persisted after 3 doses of dextrose gel, newborns were admitted to the NICU for IV dextrose. There was no standardized clinical pathway guiding IV dextrose treatment for hypoglycemia, and IV dextrose titration, glucose measurements, and target glucose levels were determined per provider discretion. Typically, 10% IV dextrose was started at 60 to 80 ml/kg/day (glucose infusion rate [GIR] 4.2 to 5.6 mg/kg/min) and increased by 10–20 ml/kg/day to achieve target glucoses >55–60 mg/dL. Per provider discretion, some infants received a 2 ml/kg IV bolus of 10% dextrose for very low glucoses (<20 mg/dL). Higher dextrose containing fluids (i.e. ≥12.5%) were used for GIRs ≥8–10 mg/kg/min. IV dextrose was weaned by 0.5–1 mg/kg/min for each pre-prandial BG >60 mg/dL. Newborns continued ad lib oral feeds during IV dextrose. Pre-prandial BG checks continued until IV dextrose was stopped and euglycemia was achieved with enteral feeding alone.

Newborn characteristics, number of BG measurements during hospitalization, and length of stay were captured from the medical record as part of a limited dataset available from a newborn nursery quality improvement (QI) project on neonatal hypoglycemia. In addition to the primary study group (i.e. newborn who received IV dextrose), newborns who were screened for hypoglycemia in the newborn nursery but did not require IV dextrose served as a comparison group. Between group comparisons were made using Fisher exact test for categorical data and Wilcoxon Rank sum test for continuous data. Significance was set at p<0.05. Descriptive statistics and between group comparisons were performed using Stata 13 (StataCorp, College Station, Texas).

RESULTS

Over the 73-month study period, 7,823 at-risk newborns were screened for hypoglycemia in the newborn nursery, of whom, 298 (3.8%) had hypoglycemia and required IV dextrose in the NICU. Patient characteristics are shown in Table 1. Newborns requiring IV dextrose had a lower gestational age and birthweight and were more likely to be late preterm, SGA, and IDM compared to those who did not require IV dextrose.

Table 1.

Patient Characteristics of ‘at Risk Newborns’ Screened for Hypoglycemia

	Newborns who required IV Dextrose (n = 298)	Newborns who did not require IV Dextrose (n = 7,525)	P valuea
Gestational Age, median (IQR), weeks	37.9 (36.7–39.3)	39.0 (37.3–39.4)	<0.001
Late Preterm, n (%)	91 (30.5%)	1,400 (18.6%)	<0.001
Birthweight, median (IQR), kg	2.85 (2.44–3.51)	3.18 (2.73–3.72)	<0.001
LGA, GA ≥37 weeks, n (%)	34 (11.4%)	1,515 (20.1%)	<0.001
SGA, GA ≥37 weeks, n (%)	52 (17.5%)	935 (12.4%)	0.01
ICD-10 Infant of Diabetic Mother b, n (%)	63 (21.1%)	1,100 (14.6%)	0.003
Length of Stay, median (IQR), days	6 (4–9)	2 (2–3)	<0.001
Number of Glucose Measurements during Hospitalization, median (IQR),	29 (19–40)	5 (4–6)	<0.001

Abbreviations: GA, gestational age, ICD-10, International Classification of Diseases, Tenth Revision; IV, intravenous, LGA, large for gestational age; SGA, small for gestational age. Data represent median (interquartile range) or counts (%).

^aFisher exact or Wilcoxon rank-sum tests.

^bICD-10 P70.0 and P70.1; from prior quality improvement work, newborns of mothers with gestational diabetes were routinely not coded P70.0 and P70.1 in the medical record and therefore our data may be under capture.

Newborns who received IV dextrose experienced a median of 29 (IQR 19–40) glucose measurements. Over 25% (n=75) of newborns had ≥40 BG measurements and almost half (n=144) had ≥30 BG measurements (Figure 1). The median length of stay was 6 (IQR 4–9) days for newborns who received IV dextrose. The number of BG measurements and length of stay were higher in newborns who received IV dextrose compared to those who did not receive IV dextrose (p<0.05, Table 1).

Figure 1.

Number of blood glucose measurements in newborns ≥35 weeks gestation who had hypoglycemia screening in a university-affiliated newborn nursery and subsequently required NICU admission for IV dextrose due to persistent hypoglycemia (n=298).

DISCUSSION

In this study of 298 late preterm and term newborns who required NICU admission and IV dextrose after persistent hypoglycemia in the newborn nursery, we found a high burden of BG measurements, with almost half experiencing ≥30 ‘pokes’. Newborns requiring IV dextrose had prolonged length of stay with a median hospitalization of 6 days. Our study highlights the high resource utilization and potential opportunity for less invasive, more baby-friendly approaches for neonatal hypoglycemia, one of the most common conditions requiring NICU admission in late preterm and term newborns.

Recent research and QI efforts around neonatal hypoglycemia have focused on standardized screening approaches and first-line enteral treatments (i.e. early breastfeeding initiation and buccal dextrose gel) that can be implemented in the newborn nursery and reduce need for NICU admission and IV dextrose.^{4–7,12–14} However, there is limited understanding about the care needs for those who fail enteral measures and require IV dextrose. Key knowledge gaps remain around best practices, including strategies for BG monitoring, IV dextrose initiation, and IV dextrose titration.¹⁵ The lack of standardized clinical pathways to guide IV dextrose treatment likely leads to unwarranted variation in care, including at our own hospital. Future research is needed to guide the development of evidence-based practices to advance high-value care for neonates with hypoglycemia who require IV dextrose.

Glycemic response to IV dextrose is variable between neonates. Therefore, IV dextrose titration guided by close glucose monitoring is critical for achieving euglycemia. Current routine practice for glucose monitoring in newborns utilizes intermittent POC testing. Advantages of POC devices include ease-of-use, low blood volume requirements, and rapid turnaround time. However, POC devices have lower precision at lower BG levels typically seen in the newborn period, which may lead to diagnostic errors impacting hypoglycemia management.^16,17 POC devices also require separate ‘pokes’ (i.e. heel lances or venipuncture) for each measurement, which can result in an excessive number (>30–40) of ‘pokes’ during IV dextrose treatment as shown in our study. This burden of ‘pokes’ is rationalized by the concern of missing sustained hypoglycemia, placing a newborn at risk for poor neurologic outcomes, or hyperglycemia, indicating the need for a lower GIR. However, minimizing painful procedures should be a goal in newborns whenever possible. As physicians we are entrusted to “First, do no harm.” Parents worry about the pain their baby experiences during painful procedures, and these worries are associated with parental stress levels.¹⁸ There is also growing concern about the potential long-term implications of repeated painful procedures in newborns. Accumulating evidence in preterm infants has demonstrated painful skin-breaking procedures in the neonatal period are associated with long-term alterations in brain structure and poorer motor, cognitive, and behavioral outcomes as children.^8–11 While equivalent studies in term neonates have not been performed, concerns about the negative consequences of repeated painful procedures in this population are valid.

Technological advances in continuous glucose monitoring (CGM) offer opportunities to advance hypoglycemia treatment approaches in newborns. Randomized controlled trials in very low birthweight preterm neonates have shown the use of CGM improved glycemic control and reduced the time of hypoglycemia compared to intermittent POC glucose monitoring alone.^19–21 CGM also reduced the number of ‘pokes’ needed for intermittent POC testing by 25%.¹⁹ CGMs may similarly benefit late preterm and term neonates requiring IV dextrose treatment for persistent hypoglycemia. However, additional research and development will first be needed to account for the unique considerations and special needs of newborns, as current FDA-approved CGM devices developed for children and adults might not translate readily to neonates.^22,23

A limitation of the current study includes being single center, which may limit its generalizability. However, our practices for neonatal hypoglycemia are in-line with published guidelines and other children’s hospitals. An additional limitation was data were collected originally as part of a QI initiative and therefore limited in scope. For example, we could not examine other clinical conditions and outcomes relevant to hypoglycemia treatment, including need for central line, enteral nutritional support, and comorbidities. Nonetheless, our data provide one of the first and largest inquires in newborns with hypoglycemia requiring IV dextrose and serves as a starting point for future work aimed at understanding the needs and best care practices for this population.