- Email: [email protected]
Colostrum Versus Formula Supplementation for Glucose Stabilization in Newborns of Diabetic Mothers
JOGNN
PRINCIPLES & PRACTICE
Colostrum Versus Formula Supplementation for Glucose Stabilization in Newborns of Diabetic Mothers Pamela K. Tozier
Correspondence Pamela Kinney Tozier, RNC, BSN, IBCLC, CCE, East Tower Labor & Delivery, Maine Medical Center, 22 Bramhall Street, Portland, ME 04102. [email protected]
ABSTRACT The purpose of this article is to describe practice change designed to facilitate breastfeeding while maintaining glucose stabilization in infants born to diabetic mothers. Postpractice change outcomes of newborn blood glucose levels, formula supplementation, and colostrum feeds are specifically addressed. There were no significant differences between glucose values for infants given formula supplementation versus those fed colostrum. Postpractice change, admissions to the neonatal intensive care unit (NICU) for glucose stabilization decreased and exclusive breastfeeding increased.
JOGNN, 42, 619-628; 2013. DOI: 10.1111/1552-6909.12260 Accepted August 2013
Keywords colostrum diabetes newborn hypoglycemia glucose stabilization exclusive breastfeeding
ewborns of diabetic mothers are an at-risk group and should receive routine blood glucose monitoring (Wight, Marinelli, & the Academy of Breastfeeding Medicine Protocol Committee, 2006). U.S. hospital practices often follow traditional care models of obtaining blood glucose values, typically drawn as point of care, capillary samples within the first hour after birth. This early blood glucose testing often leads to unnecessary treatment of blood glucose values with formula supplementation. Hospital routines such as providing water, glucose water, or formula supplementation reduce the success rate of exclusive breastfeeding (American Academy of Pediatrics [AAP], 2012). The purpose of this article is to describe the results of an evidence-based practice change for glucose monitoring and promoting breastfeeding in this at-risk group by comparing newborns who were fed colostrum versus those who were fed formula. Postpractice change outcomes of newborn blood glucose levels, formula supplementation, and colostrum feeds are described.
N Pamela Kinney Tozier, RNC, BSN, IBCLC, CCE, is a registered nurse in labor and delivery and a lactation consultant at Maine Medical Center, Portland, ME.
The author reports no conflict of interest or relevant financial relationships.
http://jognn.awhonn.org
Background and Significance The Healthy People 2020 goal is to decrease the proportion of newborns who receive formula
during the first 2 days of life from 24.6% to 14.2% (U.S. Department of Health and Human Services [U.S. DHHS], 2013). The Joint Commission (2009) added exclusive breastfeeding at discharge as a Perinatal Core Measure and mandated that hospitals report when formula is being given, especially when it is not medically indicated. When a breastfeeding infant is fed formula soon after birth, he or she often demonstrates a lack of interest in feeding at the breast due to prior satiation from an overfull stomach. The mother’s milk production subsequently diminishes, which may lead to poor milk supply and contribute to early discontinuation of breastfeeding (Holmes, 2013). Colostrum contains many properties that ensure optimal health and is genetically designed to be the perfect food for a newborn (World Health Organization [WHO], 2013). Infants born to diabetic parents have an accelerated risk of becoming diabetic themselves (Florez, Hirschhorn, & Altshuler, 2003). Formulafed infants are at greater risk of developing type 2 diabetes in later life compared to breastfed infants (Owen, Martin, Whincup, Smith, & Cook, 2006). In overweight or obese children, formula feeding is associated with reduced insulin sensitivity and
C 2013 AWHONN, the Association of Women’s Health, Obstetric and Neonatal Nurses
619
PRINCIPLES & PRACTICE
increased insulin secretion (Manco et al., 2011). In a landmark study, Kostraba et al. (1993) postulated that the sensitization and development of immune memory to cow’s milk protein in infant formulas could be an initial step in the etiology of type 1 diabetes. This may happen in susceptible infants when exposed to cow’s milk formula during the first days of life (Holmberg, Wahlberg, Vaarala, Ludvigsson & the All Babies in Southeast Sweden [ABIS] Study Group, 2007). Beta cell autoimmunity can be triggered through ingesting cow’s milk–based formula and this increases the risk for developing type 1 diabetes up to age 5 (Holmberg et al.). Autoantibodies are associated with beta cell destruction and reduced insulin production found in Type 1 diabetes. Holmberg and colleagues reported that short duration of total breastfeeding, short duration of exclusive breastfeeding, and early formula feeding were associated with increased autoantibodies in the sample of 5-year-old nondiabetic children. In healthy, breastfed, term infants blood glucose levels were lower at one hour after birth than any other time up to 96 hours after birth (Hoseth, Joergensen, Ebbesen & Moeller, 2000). There were no significant differences in blood glucose values for gender, delivery mode, gestational age, or Apgar score. Croke et al. (2009) conducted a meta-analysis of studies reporting neonatal blood glucose levels in the first 3 hours of life. The pooled mean blood glucose was significantly lower at one hour than at 2 or 3 hours after birth, regardless of the feeding method. In at-risk neonates defined as either small or large for gestational age, 1- to 3-hour blood glucose levels did not significantly differ from those of healthy newborns. Croke et al. performed a clinical audit of at-risk newborns that also confirmed the findings of the meta-analysis. In general, breastfed infants have lower blood glucose values than formula-fed infants (Rozance & Hay, 2010). Furthermore, the definition of hypoglycemia is controversial and may not be represented by a single value but rather by a low percentile threshold associated with hours since birth (Alkalay et al., 2006; Rozance & Hay). It is difficult to determine the exact blood glucose value and time period that would ultimately cause neurological impairment in a newborn (Hay, Raju, Higgens, Kalhan, & Devaskar, 2009; Rozance & Hay). Clinical symptoms that may reflect newborn hypoglycemia include alternating tremulousness and lethargy and convulsions responsive to glucose treatment. More severe symptoms of hypoglycemia include “coma, seizures, respiratory
620
JOGNN, 42, 619-628; 2013. DOI: 10.1111/1552-6909.12260
Colostrum Versus Formula Supplementation
depression and/or apnea with cyanosis, hypotonia or limpness, high pitched cry, hypothermia, and poor feeding after initially feeding well” (Rozance & Hay, p. 279). Although asymptomatic, transient episodes of hypoglycemia during the first hours of life are reported to confer a very low risk of long-term neurologic sequelae (Rozance & Hay; Wight et al., 2006), studies defining the glucose concentrations attributable to adverse neurologic development are lacking (Boluyt, van Kempen, & Offringa, 2006). In one study, Burns, Rutherford, Boardman, and Cowan (2008) reported a variety of brain injury patterns and subsequent neurodevelopmental delay in neonates who experienced symptomatic episodes of hypoglycemia (<1.5 mmol/L= 27 mg/dl).Thus, aggressive treatment of symptomatic hypoglycemia is currently recommended (Chantry & Howard, 2013). For the asymptomatic newborn, the high protein content of colostrum helps with glucose stabilization and promotion of ketogenesis (Marinelli, Stellwagen, & MacEnroe, 2012), which protects the immature brain from low or fluctuating glucose levels.
Former Newborn Hypoglycemia Practice In the study hospital, a tertiary care center in the northeastern United States as well as many American hospitals, the standard of care for newborns at risk for hypoglycemia included early testing of blood glucose values, sometimes as soon as 15 minutes after birth. The target value of 45 to 50 mg/dl was mainstream practice for many years. Another established hospital practice was the supplemental feeding of formula to breastfeeding newborns with the expectation of elevating blood glucose levels. These routine hospital practices lead to a significant decrease in the success of exclusive breastfeeding (American Academy of Family Physicians [AAFP], 2013). Exclusive breastfeeding for the first six months of a child’s life is well documented to be best practice for the entire population (AAFP; AAP, 2012; Academy of Breastfeeding Medicine [ABM], 2008; American College of Obstetricians and Gynecologists [ACOG], 2007; WHO, 2013). Another common practice was the use of intravenous glucose infusions for initial blood glucose values below 45 mg/dl, which often resulted in infant transfer to the neonatal intensive care unit (NICU). Anecdotal observations associated with some diabetic mothers who were highly motivated to avoid formula suggested that the blood glucose levels of infants who were given early and frequent swallows of hand-expressed colostrum and breastfeeding attempts were stable. With these
http://jognn.awhonn.org
Tozier, P. K.
encouraging observations, a more comprehensive approach for all diabetic mothers planning on breastfeeding their newborns was undertaken.
PRINCIPLES & PRACTICE
The overarching goal was to change the practice of managing newborn hypoglycemia within an interdisciplinary team focused on evidence-based and family-centered care.
Intended Improvement The overarching goal was to change the practice of managing newborn hypoglycemia within an interdisciplinary team focused on evidence-based and family-centered care. The specific aims of the retrospective chart review were (a) to compare blood glucose values of newborns of diabetic mothers from pre- to postpractice change, (b) to compare blood glucose values of newborns of diabetic mothers who were fed formula to those fed colostrum, (c) to compare the proportion of infants who received formula supplementation, despite the mother’s preference for breastfeeding between pre- and post-practice change, and (d) to describe the frequency of NICU admissions for hypoglycemia treatment pre- and postpractice change.
The Practice Change The Newborn Committee consisted of neonatologists, pediatricians, a family practice physician, a nursing manager and director, unit-based educators, the outreach perinatal nurse, an International Board Certified Lactation Consultant (IBCLC)/Labor and Delivery nurse, and a Newborn Nursery nurse. The committee met for one year from 2010 to 2011 and reviewed the literature to identify evidence-based guidelines regarding newborns’ ages, timing of feeds, and recommended blood glucose values. The Newborn Nursery Hypoglycemia Clinical Algorithm was revised based on the literature review. The revised hypoglycemia algorithm was adopted in June 2011 at which time the practice was changed to include early colostrum feeds, delayed glucose testing, and skin-to-skin contact (see Figure 1).
Newborn Glucose Levels In the new practice, term newborns of diabetic mothers have their glucose levels checked for a minimum of 12 hours after birth, with the first value obtained by 90 minutes of age. Standard practice at this medical center is to collect blood samples by heel prick and analyze immediately with the Precision-Xceed Pro (Abbott Diabetes Care) glucometer, using blood glucose test strips compatible with the glucometer. Point of care quality controls are performed daily with the test strips and glucometers by trained nurses. The changes in the revised algorithm included a lower acceptable threshold of glucose values dur-
JOGNN 2013; Vol. 42, Issue 6
ing the first four hours of life. Instead of checking a newborn’s glucose right after birth (prior to the first feeding), the revised recommendation is to feed first, then obtain the first glucose by 90 minutes of age in an asymptomatic newborn. All newborns have an opportunity to receive colostrum drops by small spoon, syringe, or pipette (in addition to breastfeeding) to achieve glucose stabilization before any formula is given. As long as the newborn is asymptomatic, the initial glucose between 25 mg/dl and 40 mg/dl is acceptable. The newborn is then fed more colostrum drops along with breastfeeding attempts and prolonged skin-to-skin contact. The second glucose value is checked one hour later usually around 21/2 to 3 hours of age. As long as the newborn remains asymptomatic, the newborn has until 4 hours of age to achieve a blood glucose value of 40 mg/dl. If the infant remains asymptomatic, the glucose values can then be checked every 3 hours until 12 hours of age for an infant of a diabetic mother. The target values used in the algorithm (after the first four hours of life) are 45 mg/dl until 24 hours of age, and 50 mg/dl after 24 hours of age.
Pre- and Postbirth Hand Expression Prior to the practice change anecdotal observations suggested the need to collect colostrum prebirth so that having colostrum available would preclude the habit of reaching for formula. A prebirth hand expression guideline was developed to complement the hypoglycemia algorithm. Prebirth hand expression is usually limited to mothers of newborns at risk for hypoglycemia who are scheduled for a cesarean section or during the early phase of labor induction. Women presenting in spontaneous labor may collect prebirth colostrum if their contractions are irregular and labor progression is slow. Prebirth hand expression is done for 10 to 15 minutes, and at subsequent intervals, while the mother is on an external fetal monitor to observe for any tachysystole or fetal decelerations. The drops of colostrum are stored in feeding syringes placed in the patient’s refrigerator or on ice. These colostrum drops are then fed to the baby during the first hour after birth, along with encouragement to breastfeed and maintain continuous skin to skin contact.
621
PRINCIPLES & PRACTICE
Colostrum Versus Formula Supplementation
Figure 1. Revised Newborn Nursery Hypoglycemia Algorithm. Used with Permission of Maine Medical Center.
622
Postbirth hand expression is performed, regardless of method of delivery, so a breastfeeding baby at risk for hypoglycemia will receive colostrum drops every couple hours in addition to normal breastfeeding efforts. Holmes (2013) recommends that direct breastfeeding be attempted in the first 3 hours after birth for newborns with glucose levels between 28 and 39 mg/dl. If the newborn’s latch is ineffective or a repeat measure indicates persistent hypoglycemia, 3 to 10 ml of expressed breast milk should be considered. In addition to hand expression of colostrum, early and prolonged skin-to-skin contact is recommended. Skin-to-skin contact helps blood glucose stabilization by calming the newborn, thereby decreasing the release of stress hormones that increase glucose utilization (Moore, Anderson, & Bergman, 2007). In order to promote skin to skin contact and thereby breastfeeding, the infant’s initial bath is delayed for at least 12 hours. Following
All nurses and physicians involved in the care of newborns were educated about the benefits of colostrum feeds, the algorithm revisions, and the related policies from June to December 2011. Techniques for hand expression were taught to all nurses as well as the at-risk mothers. Staff in-services were held, flyers placed on walls, and champions identified. Continuous skinto-skin contact was encouraged as much as possible after birth, and a skin-to-skin contact policy was written and implemented. More than 50 posters of a mother holding her newly delivered infant skin to skin were displayed in all of the labor/delivery and postpartum rooms highlighting to parents and families its significant effect on stabilizing a newborn’s temperature, heart rate,
JOGNN, 42, 619-628; 2013. DOI: 10.1111/1552-6909.12260
http://jognn.awhonn.org
the bath, the newborn is placed skin to skin for rewarming.
Education
PRINCIPLES & PRACTICE
Tozier, P. K.
respirations, and glucose levels. Childbirth Educators were instructed to teach this new standard to expectant parents. Staff were educated to perform all point of care glucose draws and medication injections (such as vitamin K) on newborns while they were held skin to skin with their mothers whenever feasible.
Methods
practice change was monitored by the principle investigator with regular, systematic chart audits and subsequent feedback to staff to clarify the algorithm, reeducate staff, and reinforce the practice change. Hand expression to obtain colostrum, skin-to-skin contact, and point-of-care blood glucose levels were audited from June through December 2011. This practice is now ongoing with clear communication to and expectations of all staff and patients.
Design A retrospective chart review of all term infants (≥37 weeks) born to mothers with type 1 or gestational diabetes was conducted during a 4-month period before and after the practice change. Infants in the prepractice change group were born from October 2010 through January 2011, and postpractice change from September through December 2011. The setting was a large hospital in a northeastern U.S. city with approximately 2,600 births per year. Because method of delivery does not influence neonatal blood glucose levels (Hoseth et al., 2000), the sample included infants delivered by cesarean section, induced labor, or spontaneous vaginal birth.
Revisions of the hypoglycemia algorithm and the hand expression and skin-to-skin contact policies followed internal procedures including a librarianassisted literature search, the literature review, interdisciplinary collaboration, and administrative approval. The data collection proposal was approved and received a waiver of informed consent by the Institutional Review Board (IRB #4005x). Data were retrieved using medical record numbers that were stored on password-protected computers at the clinical site. Data were deidentified during analysis, and all results are reported as aggregate data.
Data Collection and Analysis
Results
Data about the mother’s preferred choice of feeding method (i.e., breastfeeding or formula feeding), whether colostrum or formula supplementation was given, newborn nursery or NICU admission, and the first three blood glucose values were collected.
The convenience sample consisted of 163 infants born to mothers with type 1 or gestational diabetes, with 86 infants in the prepractice change group and 77 infants in the postpractice change group. For the total sample, 85.3% (n = 139) of the mothers intended to breastfeed compared to 14.7% (n = 24) who intended to use formula or a combination of breastfeeding and formula supplementation. There was no significant difference in mother’s preferred method of feeding from pre- to postpractice change (χ2 = .022, df = 1, p = .881); with 73 (84.8%) versus 66 (85.7%) intending to breast feed at pre- and postpractice change, respectively.
Blood glucose levels of the 163 newborns were examined and tabulated through the first three measurements that coincided with up to the first 5 to 6 hours of life. Data were collected by the principle investigator who is an IBCLC/Labor and Delivery nurse. Data analysis was done using SPSS (version 17). Frequencies were inspected for the nominal, grouping variables. Distributions were inspected for the continuous variables (blood glucose values). The grouping variables were practice change (pre/post), feeding method (colostrum/formula fed), formula supplementation (yes/no), and admission to the NICU (yes/no). Nominal data were analyzed using chi-squared tests. Mean blood glucose values were compared by grouping variables using independent t tests.
Practice Change After the hypoglycemia algorithm was revised and staff completed the initial education, the practice change began in June 2011. Integrity of the
JOGNN 2013; Vol. 42, Issue 6
Ethics
Blood Glucose Visual examination of the blood glucose distributions at the three time points appeared to be normal. There was a significant Kolmogorov-Smirnov value for the first blood glucose distribution (p = .001) but not the second (p = .20) or the third distribution (p = .20). These statistical tests indicate that one of the three distributions were non-normal (Pett, 1997). Parametric and nonparametric analyses were conducted on the data, and the results were the same. Because mean blood glucose values may be more understandable in a clinical setting than median
623
PRINCIPLES & PRACTICE
Infants who were exclusively fed colostrum showed no difference in blood glucose levels compared to formula-fed infants.
values, the more traditional parametric results are reported here. Each blood glucose distribution had four or fewer outliers, which were examined for data entry accuracy. There were few missing data points for the three blood glucose distributions with 3 (1.8%), 5 (3%), and 6 (3.6%) missing items, respectively. For Aim 1, blood glucose values were compared from pre- to postpractice change, without consideration for the type of feeding. The first three mean blood glucose values prior to the practice change were 68.2 (SD = 21.7), 58.3 (SD = 18.2), and 58.8 (SD = 14.5) mg/dl, respectively. After the practice change, the mean blood glucose values were 53.4 (SD = 20.2), 56.5 (SD = 15.5), and 53.2 (SD = 13.1) mg/dl, respectively. There were significant differences found between the pre- and postpractice change blood glucose values for the first, t(158) = 4.47, p = .000, and third, t(155) = 2.52, p = .013 blood draws (see Table 1). Blood glucose variability was lower in the postpractice change group compared to the prechange group at each successive blood draw (see Table 1, Figure 2).
Breastfed versus Formula-fed Groups For the second aim, mean blood glucose values were first examined by type of feeding without consideration of the practice change. For the colostrum-fed group of newborns, the first three mean blood glucose values were 57.6 (SD = 18.8), 56.2 (SD = 15.0), and 54.0 (SD = 11.2) mg/dl, respectively. For those receiving formula supplementation, the first three mean blood glucose values were 64.6 (SD = 24.8), 57.4 (SD = 16.7), and 57.3 (SD = 13.5) mg/dl, respectively. There were no significant differences between blood glucose levels by type of feeding at each of the three data collection times.
624
Colostrum Versus Formula Supplementation
found in each of the three blood glucose values from pre- to postpractice change (see Table 1). There was less variability in postpractice change blood glucose values in the colostrum-fed group compared to the formula-fed group (see Table 1, Figure 3). Fifty-eight percent (50/85) of newborns received formula supplementation prior to the practice change. After the practice change it decreased to 27.3% (21/77), reflecting an improvement of 53.6%. Prior to the practice change 18.8% (16/85) of the term infants born to diabetic mothers were transferred to the NICU for blood glucose stabilization as compared to 6.5% (5/77) postpractice change, a 68% improvement. The postpractice change group was much more likely to avoid NICU admission than the prepractice change group (χ2 = 5.44, df = 1, p .020).
Discussion These results demonstrate that the first three blood glucose values of infants born to mothers with type 1 or gestational diabetes who were fed colostrum are no different than those of infants who receive formula supplementation. There were significant differences in two of the three mean blood glucose levels in the colostrum-fed group when examined from pre- to postpractice change. Higher mean blood glucose levels occurred in the prepractice change group. Although postpractice change mean blood glucose values were lower for colostrum-fed infants, these levels remained within the normal range (≥40 mg/dl) for asymptomatic infants in the first 4 hours of life. Importantly, these results show that blood glucose variability was reduced with the first two blood draws in the postpractice change group of breastfed infants; and blood glucose variability stabilized with the third blood draw. Conversely, there was greater variability in the blood glucose levels of infants who received formula supplementation in the postpractice change group.
Next, mean blood glucose values were examined by feeding type and by pre- and postpractice change conditions. In the colostrum-fed group, there were significant differences in the first, t(74) = 4.42, p = .000, and third blood glucose values, t(74) = 3.68, p = .000, from pre- to postpractice change (see Table). For the formula-supplemented group, no differences were
The revised hypoglycemia algorithm and related policies resulted in a practice change that supports exclusive breastfeeding in this hospital. The number of babies who received formula (n = 50) compared to the number who received colostrum (n = 25) prior to the practice change was reversed after the practice change (n = 21 formula vs. n = 52 colostrum). This reduction in formula supplementation serves as evidence of practice change adoption. Furthermore, there were
JOGNN, 42, 619-628; 2013. DOI: 10.1111/1552-6909.12260
http://jognn.awhonn.org
PRINCIPLES & PRACTICE
Tozier, P. K.
Table 1: Comparison of the First Three Blood Glucose Values Between Pre- and Postpractice Change Groups Prepractice change
Postpractice change
n
Mean
SD
n
Mean
SD
Time 1 BG
84
68.2
21.7
76
53.4
20.2
Time 2 BG
82
58.3
18.2
76
56.5
Time 3 BG
82
58.8
14.5
75
Time 1 BG
25
69.7
20.3
Time 2 BG
25
60.0
Time 3 BG
25
Time 1 BG Time 2 BG Time 3 BG
t
df
p
4.47
158
.000
15.5
0.68
156
.495
53.2
13.1
2.52
155
.013
51
51.6
14.8
4.42
74
.000
15.2
52
54.4
14.7
1.53
75
.130
60.3
10.3
51
51.0
10.4
3.68
74
.000
50
67.9
22.8
21
56.7
28.0
1.77
69
.081
50
57.4
17.3
20
57.5
15.3
−0.02
68
.987
50
58.7
14.5
20
53.9
10.3
1.35
68
.180
Total sample
Breastfed only
Formula-fed
Note. BG = blood glucose (mg/dl); SD = standard deviation.
significantly fewer NICU transfers to treat hypoglycemia in the postpractice change group. Prior to the practice change 11.8% (10/85) of infants received intravenous treatment for hypoglycemia compared to 5.2% (4/77) of newborns who received intravenous treatment postpractice change.
Mean (+SD) BG Pre- and Postchange -
100 90
Limitations Limitations of this study include the use of a convenience sample and a retrospective chart review. There is no distinction in the chart for determining if the blood glucose values were obtained by pointof-care heel stick or by laboratory analysis. However, blood glucose values are routinely obtained
(+21.7) (+20.2) -
80
(+18.2) (+- 15.2) -
(+14.5) -
(+13.1) -
mg/dl
70 60 50
Prechange
40
Postchange
30 20 10 0
68.2
53.4
58.3
1
56.5 2
58.8
53.2 3
Blood Draw Time Points Figure 2. The First Three Mean Blood Glucose Values and Variability Comparing the Pre-and Postpractice Change Groups. BG = blood glucose (mg/dl), SD = standard deviation.
JOGNN 2013; Vol. 42, Issue 6
625
PRINCIPLES & PRACTICE
Colostrum Versus Formula Supplementation
The practice change empowered mothers to exclusively breastfeed, which may significantly affect the health and well-being of their children.
by point-of-care heel sticks in this institution, and abnormally low values are laboratory confirmed. Timing of the blood glucose draws (e.g., exact number of minutes or hours since birth) was not examined, except that the first three lab draws after birth were compared from pre- to postpractice change. The practice change called for delaying the drawing of the first blood glucose until after the first feeding. The meta-analysis by Croke et al. (2009) included studies in which feeding varied from not fed to breastfed or formula-fed newborns. These authors noted similar limitations and suggested future studies consider measuring time of blood glucose draws and feeding status. In retrospective chart reviews, this type of limitation is inevitable because the principle investigator cannot control some of the available data. The length of time infants spent in skin-to-skin contact, as well as the amount of colostrum or formula were not analyzed. These data were collected from birth to 6 hours of life and are not intended to suggest the trajectory of blood glucose values beyond this time frame. Generalizability may be limited to mothers who reside in a northeastern metropolitan U.S. city who gave birth in a major tertiary medical center.
(+28.0) -
80.0
mg/dl
70.0
The revised hypoglycemia algorithm and practice change provided a first step toward empowering mothers to practice exclusive breastfeeding that may significantly affect the health and well-being of their children. According to the Center for Disease Control (2013), diabetes has increased by 176% over the last 30 years. It is projected to increase another 64% between 2010 and 2025, affecting over 53 million people (Rowley & Bezold, 2012). The AAP (2012) stated that exclusive breastfeeding for 6 months will decrease a child’s risk of type 2 diabetes by 40% and type 1 diabetes by 30% and cited the positive effect of exclusive breastfeeding on weight control and feeding self-regulation. Obesity has also been associated with formula feeding due to the ease and tendency to overfeed beginning at birth (Koletzko et al., 2009). Breast milk, on the other hand,
Mean (+SD) BG Postpractice Change -
100.0 90.0
Implications for Practice Several challenges were identified with the practice change. When the practice of teaching hand expression for colostrum drops was first introduced, it was met with a myriad of reactions. Some staff stated that they were uncomfortable assisting mothers with hand expression. Occasionally some staff indicated a lack of time to fully engage in assisting with the hand expression process. Some physicians did not recognize or accept the importance of colostrum. Some patients were hesitant to ask for privacy or were uncomfortable with hand expression of colostrum while visitors were present in the room.
(+14.7) -
(+14.8) -
(+15.3) -
(+10.4) -
(+10.3) -
60.0 50.0
Colostrum
40.0
Formula
30.0 20.0 10.0
51.6
56.7
54.4
57.5
51.0
53.9
0.0 1
2 Blood Draw Time Points
3
Figure 3. The First Three Mean Blood Glucose Values and Variability Comparing Colostrum and Formula-Fed Infants in the Postpractice Change Group. BG = blood glucose (mg/dl), SD = standard deviation.
626
JOGNN, 42, 619-628; 2013. DOI: 10.1111/1552-6909.12260
http://jognn.awhonn.org
PRINCIPLES & PRACTICE
Tozier, P. K.
contains maternal hormones such as leptin that signals satiety and decreases the sensation of hunger in the infant (Savino, Liguori, Fissore, & Oggero, 2009). Breastfeeding may reduce childhood obesity by lowering plasma insulin levels, decreasing fat storage, and preventing adipocyte development (Oddy, 2012).
An audit of Canadian guidelines. Pediatric Child Health, 14(4), 238–244. Florez, J. C., Hirschhorn, J., & Altshuler, D. (2003). The inherited basis of diabetes mellitus: Implications for the genetic analysis of complex traits. Annual Review of Genomics and Human Genetics, 4, 257–291. Hay, W. W., Raju, T. N. K., Higgens, R. D., Kalhan, S. C., & Devaskar, S. U. (2009). Knowledge gaps and research needs for understanding and treating neonatal hypoglycemia: Workshop re-
Adopting hospital practices that facilitate exclusive breastfeeding is in line with the U.S. Surgeon General’s Call to Action (U.S. DHHS, 2011) and other professional and scientific associations’ position statements (AAFP, 2013; AAP, 2012; ABM, 2008; ACOG, 2007; WHO, 2013). The traditional approach of feeding formula to an asymptomatic breastfeeding term newborn at risk for hypoglycemia needs to be reevaluated in the context of evidence for safety and breastfeeding outcomes. This practice change demonstrated that colostrum stabilized newborn glucose levels as effectively as formula in the first 6 hours after birth and supported exclusive breastfeeding for mothers and newborns.
port from Eunice Kennedy Shriver National Institute of Child Health and Human Development. Journal of Pediatrics, 155(5), 612–617. Holmberg H., Wahlberg J., Vaarala O., Ludvigsson, J., & the All Babies in Southeast Sweden Study Group (ABIS). (2007). Short duration of breastfeeding as a risk factor for beta-cell autoantibodies in 5 year old children in the general population. British Journal of Nutrition, 97, 111–116. Holmes, A. V. (2013). Establishing successful breastfeeding in the newborn period. Pediatric Clinics of North America, 60, 147–168. doi:10.1016/j.pcl.2012.09.013 Hoseth, E., Joergensen, A., Ebbesen, F., & Moeller, M. (2000). Blood glucose levels in a population of healthy, breastfed, term infants of appropriate size for gestational age. Archives of Disease in Childhood: Fetal and Neonatal Edition, 83, F117–F119. Kostraba, J. N., Cruickshanks, K. J., Lawler-Heavner, J., Jobim, L. F., Rewers, M. J., Gay, E.C, . . . Hamman, R. F. (1993). Early exposure to cow’s milk and solid foods in infancy, genetic predisposition, and risk of IDDM. Diabetes, 42(2), 288–295.
REFERENCES Alkalay, A. L., Sarnat, H. B., Flores-Sarnat, L., Elashoff, J. D., Farber,
Koletzko, B., von Kries, R., Closa, R., Escribano, J., Scaglionis, S., Gio-
S. J., & Simmons, C. F. (2006). Population meta-analysis of low
vannini, M., . . . Grote, V. (2009). Can infant feeding choices
plasma glucose thresholds in full-term normal newborns. Amer-
modulate later obesity risk? American Journal of Clinical Nutri-
ican Journal of Perinatology, 23(2), 115–119.
tion, 89(5), 1502S-1508S. doi:10.3945/ajcn 27113D
American Academy of Family Physicians. (2013). Breastfeed-
Manco, M., Alterio, A., Bugianesi, E., Ciampalini, P., Mariani, P., Finoc-
ing (policy statement). Retrieved from http://www.aafp.org/
chi, M., . . . Nobili, V. (2011). Insulin dynamics of breast- or
about/policies/all/breastfeeding.html
formula-fed overweight and obese children. Journal of the Amer-
American Academy of Pediatrics. (2012). Breastfeeding and the use of human milk. Pediatrics, 129, e827–e841. Academy of Breastfeeding Medicine. (2008). ABM statements: Position on breastfeeding. Breastfeeding Medicine, 3(4), 267–270. doi:10.1089/bfm.2008.9988
ican College of Nutrition, 30(1), 29–38. Marinelli, K., Stellwagen, L., & MacEnroe, T. (2012). Enriching hospital support of breastfeeding [Audio podcast]. Retrieved from: http://www.ilca.org/i4a/pages/index.cfm?pageid=4131 Moore, E. R., Anderson, G. C., & Bergman, N. (2007). Early skin to skin
American College of Obstetricians and Gynecologists. (2007). Com-
contact for mothers and their healthy newborn infants. Cochrane
mittee opinion: Breastfeeding: Maternal and infant aspects.
Database of Systematic Reviews, Issue 3. Art. No: CD003519.
Retrieved
from
http://www.acog.org/∼/media/Committee%
20Opinions/Committee%20on%20Health%20Care%20for% 20Underserved%20Women/co361.pdf?dmc=1&ts= 20130708T0927582672
doi:10.1002/14651858.CD003519.pub2. Oddy, W. H. (2012). Infant feeding and obesity risk in the child. Breastfeeding Review, 20(2), 7–12. Owen, C. G., Martin, R. M., Whincup, P. H., Smith, G. D., & Cook, D.
Boluyt, N., van Kempen, A., & Offringa, M. (2006). Neurodevelop-
G. (2006). Does breastfeeding influence the risk of type 2 dia-
ment after neonatal hypoglycemia: A systematic review and de-
betes in later life? A quantitative analysis of published evidence.
sign of an optimal future study. Pediatrics, 117(6), 2231–2243. doi:10.1542/peds.2005-1919 Burns, C. M., Rutherford, M. A., Boardman, J. P., & Cowan, F. M. (2008). Patterns of cerebral injury and neurodevelopment outcomes af-
American Journal of Clinical Nutrition, 84, 1043–1054. Pett, M. A. (1997). Nonparametric statistics for health care research. Thousand Oaks, CA: Sage. Rowley, W. R., & Bezold, C. (2012). Creating public awareness: State
ter symptomatic neonatal hypoglycemia. Pediatrics, 122(1), 65–
2025 diabetes forecasts. Population Health Management, 15,
74. doi:10.1542/peds.2007-2822
194–200.
Centers for Disease Control. (2013). Crude and age-adjusted percent-
Rozance, P. J., & Hay, W. W. (2010). Describing hypoglycemia – Defi-
age of civilian, noninstitutionalized population with diagnosed
nition or operational threshold? Early Human Development, 86,
diabetes, United States 1980–2011. Retrieved from http://www. cdc.gov/print.do?url=http%3A//www.cdc.gov/diabetes/ statistics/prev/national/figage.htm Chantry, C. J., & Howard, C. R. (2013). Clinical protocols for management of breastfeeding. Pediatric Clinics of North America, 60, 75–113. doi:10.1016/j.pcl, 2012.10.005 Croke, J., Sullivan, M., Ryan-Drover, A., Randell, E., Andrews, W., & Aziz, K. (2009). Two hour blood glucose levels in at-risk babies:
JOGNN 2013; Vol. 42, Issue 6
275–280. doi:10.1016/j.earlhumdev .2010.05.002 Savino, F., Liguori, S. A., Fissore, M. F., & Oggero, R. (2009). Breast milk hormones and their positive effect on obesity. International Journal of Pediatric Endocrinology, 327505. doi:10.1155/2009/327505 The Joint Commission. (2009). Perinatal core measure set selection form. Retrieved from: http://www.jointcommission.org/ perinatal_care/
627
PRINCIPLES & PRACTICE
U.S. Department of Health and Human Service. (2011). The sur-
Wight, N., Marinelli, K. & the Academy of Breastfeeding Medicine
geon general’s call to action to support breastfeeding. Wash-
Protocol Committee. (2006). ABM Clinical Protocol #1: Guide-
ington, DC: U.S. Department of Health and Human Services,
lines for glucose monitoring and treatment of hypoglycemia
Office of the Surgeon General. Retrieved from http://www.
in breastfed neonates. Breastfeeding Medicine, 1, 178–
surgeongeneral.gov/initiatives/breastfeeding/
628
Colostrum Versus Formula Supplementation
184.
U.S. Department of Health and Human Services. (2013). Healthy
World Health Organization. (2013). Programmes and projects: Ma-
People 2020: Maternal, infant, and child health objec-
ternal, newborn, child and adolescent health. Breastfeeding.
tives: MICH-23. Retrieved from http://www.healthypeople.gov/
Retrieved from http://www.who.int/maternal_child_adolescent/
2020/topicsobjectives2020/objectiveslist.aspx?topicid=26
topics/child/nutrition/breastfeeding/en/index.html
JOGNN, 42, 619-628; 2013. DOI: 10.1111/1552-6909.12260
http://jognn.awhonn.org
PRINCIPLES & PRACTICE
Colostrum Versus Formula Supplementation for Glucose Stabilization in Newborns of Diabetic Mothers Pamela K. Tozier
Correspondence Pamela Kinney Tozier, RNC, BSN, IBCLC, CCE, East Tower Labor & Delivery, Maine Medical Center, 22 Bramhall Street, Portland, ME 04102. [email protected]
ABSTRACT The purpose of this article is to describe practice change designed to facilitate breastfeeding while maintaining glucose stabilization in infants born to diabetic mothers. Postpractice change outcomes of newborn blood glucose levels, formula supplementation, and colostrum feeds are specifically addressed. There were no significant differences between glucose values for infants given formula supplementation versus those fed colostrum. Postpractice change, admissions to the neonatal intensive care unit (NICU) for glucose stabilization decreased and exclusive breastfeeding increased.
JOGNN, 42, 619-628; 2013. DOI: 10.1111/1552-6909.12260 Accepted August 2013
Keywords colostrum diabetes newborn hypoglycemia glucose stabilization exclusive breastfeeding
ewborns of diabetic mothers are an at-risk group and should receive routine blood glucose monitoring (Wight, Marinelli, & the Academy of Breastfeeding Medicine Protocol Committee, 2006). U.S. hospital practices often follow traditional care models of obtaining blood glucose values, typically drawn as point of care, capillary samples within the first hour after birth. This early blood glucose testing often leads to unnecessary treatment of blood glucose values with formula supplementation. Hospital routines such as providing water, glucose water, or formula supplementation reduce the success rate of exclusive breastfeeding (American Academy of Pediatrics [AAP], 2012). The purpose of this article is to describe the results of an evidence-based practice change for glucose monitoring and promoting breastfeeding in this at-risk group by comparing newborns who were fed colostrum versus those who were fed formula. Postpractice change outcomes of newborn blood glucose levels, formula supplementation, and colostrum feeds are described.
N Pamela Kinney Tozier, RNC, BSN, IBCLC, CCE, is a registered nurse in labor and delivery and a lactation consultant at Maine Medical Center, Portland, ME.
The author reports no conflict of interest or relevant financial relationships.
http://jognn.awhonn.org
Background and Significance The Healthy People 2020 goal is to decrease the proportion of newborns who receive formula
during the first 2 days of life from 24.6% to 14.2% (U.S. Department of Health and Human Services [U.S. DHHS], 2013). The Joint Commission (2009) added exclusive breastfeeding at discharge as a Perinatal Core Measure and mandated that hospitals report when formula is being given, especially when it is not medically indicated. When a breastfeeding infant is fed formula soon after birth, he or she often demonstrates a lack of interest in feeding at the breast due to prior satiation from an overfull stomach. The mother’s milk production subsequently diminishes, which may lead to poor milk supply and contribute to early discontinuation of breastfeeding (Holmes, 2013). Colostrum contains many properties that ensure optimal health and is genetically designed to be the perfect food for a newborn (World Health Organization [WHO], 2013). Infants born to diabetic parents have an accelerated risk of becoming diabetic themselves (Florez, Hirschhorn, & Altshuler, 2003). Formulafed infants are at greater risk of developing type 2 diabetes in later life compared to breastfed infants (Owen, Martin, Whincup, Smith, & Cook, 2006). In overweight or obese children, formula feeding is associated with reduced insulin sensitivity and
C 2013 AWHONN, the Association of Women’s Health, Obstetric and Neonatal Nurses
619
PRINCIPLES & PRACTICE
increased insulin secretion (Manco et al., 2011). In a landmark study, Kostraba et al. (1993) postulated that the sensitization and development of immune memory to cow’s milk protein in infant formulas could be an initial step in the etiology of type 1 diabetes. This may happen in susceptible infants when exposed to cow’s milk formula during the first days of life (Holmberg, Wahlberg, Vaarala, Ludvigsson & the All Babies in Southeast Sweden [ABIS] Study Group, 2007). Beta cell autoimmunity can be triggered through ingesting cow’s milk–based formula and this increases the risk for developing type 1 diabetes up to age 5 (Holmberg et al.). Autoantibodies are associated with beta cell destruction and reduced insulin production found in Type 1 diabetes. Holmberg and colleagues reported that short duration of total breastfeeding, short duration of exclusive breastfeeding, and early formula feeding were associated with increased autoantibodies in the sample of 5-year-old nondiabetic children. In healthy, breastfed, term infants blood glucose levels were lower at one hour after birth than any other time up to 96 hours after birth (Hoseth, Joergensen, Ebbesen & Moeller, 2000). There were no significant differences in blood glucose values for gender, delivery mode, gestational age, or Apgar score. Croke et al. (2009) conducted a meta-analysis of studies reporting neonatal blood glucose levels in the first 3 hours of life. The pooled mean blood glucose was significantly lower at one hour than at 2 or 3 hours after birth, regardless of the feeding method. In at-risk neonates defined as either small or large for gestational age, 1- to 3-hour blood glucose levels did not significantly differ from those of healthy newborns. Croke et al. performed a clinical audit of at-risk newborns that also confirmed the findings of the meta-analysis. In general, breastfed infants have lower blood glucose values than formula-fed infants (Rozance & Hay, 2010). Furthermore, the definition of hypoglycemia is controversial and may not be represented by a single value but rather by a low percentile threshold associated with hours since birth (Alkalay et al., 2006; Rozance & Hay). It is difficult to determine the exact blood glucose value and time period that would ultimately cause neurological impairment in a newborn (Hay, Raju, Higgens, Kalhan, & Devaskar, 2009; Rozance & Hay). Clinical symptoms that may reflect newborn hypoglycemia include alternating tremulousness and lethargy and convulsions responsive to glucose treatment. More severe symptoms of hypoglycemia include “coma, seizures, respiratory
620
JOGNN, 42, 619-628; 2013. DOI: 10.1111/1552-6909.12260
Colostrum Versus Formula Supplementation
depression and/or apnea with cyanosis, hypotonia or limpness, high pitched cry, hypothermia, and poor feeding after initially feeding well” (Rozance & Hay, p. 279). Although asymptomatic, transient episodes of hypoglycemia during the first hours of life are reported to confer a very low risk of long-term neurologic sequelae (Rozance & Hay; Wight et al., 2006), studies defining the glucose concentrations attributable to adverse neurologic development are lacking (Boluyt, van Kempen, & Offringa, 2006). In one study, Burns, Rutherford, Boardman, and Cowan (2008) reported a variety of brain injury patterns and subsequent neurodevelopmental delay in neonates who experienced symptomatic episodes of hypoglycemia (<1.5 mmol/L= 27 mg/dl).Thus, aggressive treatment of symptomatic hypoglycemia is currently recommended (Chantry & Howard, 2013). For the asymptomatic newborn, the high protein content of colostrum helps with glucose stabilization and promotion of ketogenesis (Marinelli, Stellwagen, & MacEnroe, 2012), which protects the immature brain from low or fluctuating glucose levels.
Former Newborn Hypoglycemia Practice In the study hospital, a tertiary care center in the northeastern United States as well as many American hospitals, the standard of care for newborns at risk for hypoglycemia included early testing of blood glucose values, sometimes as soon as 15 minutes after birth. The target value of 45 to 50 mg/dl was mainstream practice for many years. Another established hospital practice was the supplemental feeding of formula to breastfeeding newborns with the expectation of elevating blood glucose levels. These routine hospital practices lead to a significant decrease in the success of exclusive breastfeeding (American Academy of Family Physicians [AAFP], 2013). Exclusive breastfeeding for the first six months of a child’s life is well documented to be best practice for the entire population (AAFP; AAP, 2012; Academy of Breastfeeding Medicine [ABM], 2008; American College of Obstetricians and Gynecologists [ACOG], 2007; WHO, 2013). Another common practice was the use of intravenous glucose infusions for initial blood glucose values below 45 mg/dl, which often resulted in infant transfer to the neonatal intensive care unit (NICU). Anecdotal observations associated with some diabetic mothers who were highly motivated to avoid formula suggested that the blood glucose levels of infants who were given early and frequent swallows of hand-expressed colostrum and breastfeeding attempts were stable. With these
http://jognn.awhonn.org
Tozier, P. K.
encouraging observations, a more comprehensive approach for all diabetic mothers planning on breastfeeding their newborns was undertaken.
PRINCIPLES & PRACTICE
The overarching goal was to change the practice of managing newborn hypoglycemia within an interdisciplinary team focused on evidence-based and family-centered care.
Intended Improvement The overarching goal was to change the practice of managing newborn hypoglycemia within an interdisciplinary team focused on evidence-based and family-centered care. The specific aims of the retrospective chart review were (a) to compare blood glucose values of newborns of diabetic mothers from pre- to postpractice change, (b) to compare blood glucose values of newborns of diabetic mothers who were fed formula to those fed colostrum, (c) to compare the proportion of infants who received formula supplementation, despite the mother’s preference for breastfeeding between pre- and post-practice change, and (d) to describe the frequency of NICU admissions for hypoglycemia treatment pre- and postpractice change.
The Practice Change The Newborn Committee consisted of neonatologists, pediatricians, a family practice physician, a nursing manager and director, unit-based educators, the outreach perinatal nurse, an International Board Certified Lactation Consultant (IBCLC)/Labor and Delivery nurse, and a Newborn Nursery nurse. The committee met for one year from 2010 to 2011 and reviewed the literature to identify evidence-based guidelines regarding newborns’ ages, timing of feeds, and recommended blood glucose values. The Newborn Nursery Hypoglycemia Clinical Algorithm was revised based on the literature review. The revised hypoglycemia algorithm was adopted in June 2011 at which time the practice was changed to include early colostrum feeds, delayed glucose testing, and skin-to-skin contact (see Figure 1).
Newborn Glucose Levels In the new practice, term newborns of diabetic mothers have their glucose levels checked for a minimum of 12 hours after birth, with the first value obtained by 90 minutes of age. Standard practice at this medical center is to collect blood samples by heel prick and analyze immediately with the Precision-Xceed Pro (Abbott Diabetes Care) glucometer, using blood glucose test strips compatible with the glucometer. Point of care quality controls are performed daily with the test strips and glucometers by trained nurses. The changes in the revised algorithm included a lower acceptable threshold of glucose values dur-
JOGNN 2013; Vol. 42, Issue 6
ing the first four hours of life. Instead of checking a newborn’s glucose right after birth (prior to the first feeding), the revised recommendation is to feed first, then obtain the first glucose by 90 minutes of age in an asymptomatic newborn. All newborns have an opportunity to receive colostrum drops by small spoon, syringe, or pipette (in addition to breastfeeding) to achieve glucose stabilization before any formula is given. As long as the newborn is asymptomatic, the initial glucose between 25 mg/dl and 40 mg/dl is acceptable. The newborn is then fed more colostrum drops along with breastfeeding attempts and prolonged skin-to-skin contact. The second glucose value is checked one hour later usually around 21/2 to 3 hours of age. As long as the newborn remains asymptomatic, the newborn has until 4 hours of age to achieve a blood glucose value of 40 mg/dl. If the infant remains asymptomatic, the glucose values can then be checked every 3 hours until 12 hours of age for an infant of a diabetic mother. The target values used in the algorithm (after the first four hours of life) are 45 mg/dl until 24 hours of age, and 50 mg/dl after 24 hours of age.
Pre- and Postbirth Hand Expression Prior to the practice change anecdotal observations suggested the need to collect colostrum prebirth so that having colostrum available would preclude the habit of reaching for formula. A prebirth hand expression guideline was developed to complement the hypoglycemia algorithm. Prebirth hand expression is usually limited to mothers of newborns at risk for hypoglycemia who are scheduled for a cesarean section or during the early phase of labor induction. Women presenting in spontaneous labor may collect prebirth colostrum if their contractions are irregular and labor progression is slow. Prebirth hand expression is done for 10 to 15 minutes, and at subsequent intervals, while the mother is on an external fetal monitor to observe for any tachysystole or fetal decelerations. The drops of colostrum are stored in feeding syringes placed in the patient’s refrigerator or on ice. These colostrum drops are then fed to the baby during the first hour after birth, along with encouragement to breastfeed and maintain continuous skin to skin contact.
621
PRINCIPLES & PRACTICE
Colostrum Versus Formula Supplementation
Figure 1. Revised Newborn Nursery Hypoglycemia Algorithm. Used with Permission of Maine Medical Center.
622
Postbirth hand expression is performed, regardless of method of delivery, so a breastfeeding baby at risk for hypoglycemia will receive colostrum drops every couple hours in addition to normal breastfeeding efforts. Holmes (2013) recommends that direct breastfeeding be attempted in the first 3 hours after birth for newborns with glucose levels between 28 and 39 mg/dl. If the newborn’s latch is ineffective or a repeat measure indicates persistent hypoglycemia, 3 to 10 ml of expressed breast milk should be considered. In addition to hand expression of colostrum, early and prolonged skin-to-skin contact is recommended. Skin-to-skin contact helps blood glucose stabilization by calming the newborn, thereby decreasing the release of stress hormones that increase glucose utilization (Moore, Anderson, & Bergman, 2007). In order to promote skin to skin contact and thereby breastfeeding, the infant’s initial bath is delayed for at least 12 hours. Following
All nurses and physicians involved in the care of newborns were educated about the benefits of colostrum feeds, the algorithm revisions, and the related policies from June to December 2011. Techniques for hand expression were taught to all nurses as well as the at-risk mothers. Staff in-services were held, flyers placed on walls, and champions identified. Continuous skinto-skin contact was encouraged as much as possible after birth, and a skin-to-skin contact policy was written and implemented. More than 50 posters of a mother holding her newly delivered infant skin to skin were displayed in all of the labor/delivery and postpartum rooms highlighting to parents and families its significant effect on stabilizing a newborn’s temperature, heart rate,
JOGNN, 42, 619-628; 2013. DOI: 10.1111/1552-6909.12260
http://jognn.awhonn.org
the bath, the newborn is placed skin to skin for rewarming.
Education
PRINCIPLES & PRACTICE
Tozier, P. K.
respirations, and glucose levels. Childbirth Educators were instructed to teach this new standard to expectant parents. Staff were educated to perform all point of care glucose draws and medication injections (such as vitamin K) on newborns while they were held skin to skin with their mothers whenever feasible.
Methods
practice change was monitored by the principle investigator with regular, systematic chart audits and subsequent feedback to staff to clarify the algorithm, reeducate staff, and reinforce the practice change. Hand expression to obtain colostrum, skin-to-skin contact, and point-of-care blood glucose levels were audited from June through December 2011. This practice is now ongoing with clear communication to and expectations of all staff and patients.
Design A retrospective chart review of all term infants (≥37 weeks) born to mothers with type 1 or gestational diabetes was conducted during a 4-month period before and after the practice change. Infants in the prepractice change group were born from October 2010 through January 2011, and postpractice change from September through December 2011. The setting was a large hospital in a northeastern U.S. city with approximately 2,600 births per year. Because method of delivery does not influence neonatal blood glucose levels (Hoseth et al., 2000), the sample included infants delivered by cesarean section, induced labor, or spontaneous vaginal birth.
Revisions of the hypoglycemia algorithm and the hand expression and skin-to-skin contact policies followed internal procedures including a librarianassisted literature search, the literature review, interdisciplinary collaboration, and administrative approval. The data collection proposal was approved and received a waiver of informed consent by the Institutional Review Board (IRB #4005x). Data were retrieved using medical record numbers that were stored on password-protected computers at the clinical site. Data were deidentified during analysis, and all results are reported as aggregate data.
Data Collection and Analysis
Results
Data about the mother’s preferred choice of feeding method (i.e., breastfeeding or formula feeding), whether colostrum or formula supplementation was given, newborn nursery or NICU admission, and the first three blood glucose values were collected.
The convenience sample consisted of 163 infants born to mothers with type 1 or gestational diabetes, with 86 infants in the prepractice change group and 77 infants in the postpractice change group. For the total sample, 85.3% (n = 139) of the mothers intended to breastfeed compared to 14.7% (n = 24) who intended to use formula or a combination of breastfeeding and formula supplementation. There was no significant difference in mother’s preferred method of feeding from pre- to postpractice change (χ2 = .022, df = 1, p = .881); with 73 (84.8%) versus 66 (85.7%) intending to breast feed at pre- and postpractice change, respectively.
Blood glucose levels of the 163 newborns were examined and tabulated through the first three measurements that coincided with up to the first 5 to 6 hours of life. Data were collected by the principle investigator who is an IBCLC/Labor and Delivery nurse. Data analysis was done using SPSS (version 17). Frequencies were inspected for the nominal, grouping variables. Distributions were inspected for the continuous variables (blood glucose values). The grouping variables were practice change (pre/post), feeding method (colostrum/formula fed), formula supplementation (yes/no), and admission to the NICU (yes/no). Nominal data were analyzed using chi-squared tests. Mean blood glucose values were compared by grouping variables using independent t tests.
Practice Change After the hypoglycemia algorithm was revised and staff completed the initial education, the practice change began in June 2011. Integrity of the
JOGNN 2013; Vol. 42, Issue 6
Ethics
Blood Glucose Visual examination of the blood glucose distributions at the three time points appeared to be normal. There was a significant Kolmogorov-Smirnov value for the first blood glucose distribution (p = .001) but not the second (p = .20) or the third distribution (p = .20). These statistical tests indicate that one of the three distributions were non-normal (Pett, 1997). Parametric and nonparametric analyses were conducted on the data, and the results were the same. Because mean blood glucose values may be more understandable in a clinical setting than median
623
PRINCIPLES & PRACTICE
Infants who were exclusively fed colostrum showed no difference in blood glucose levels compared to formula-fed infants.
values, the more traditional parametric results are reported here. Each blood glucose distribution had four or fewer outliers, which were examined for data entry accuracy. There were few missing data points for the three blood glucose distributions with 3 (1.8%), 5 (3%), and 6 (3.6%) missing items, respectively. For Aim 1, blood glucose values were compared from pre- to postpractice change, without consideration for the type of feeding. The first three mean blood glucose values prior to the practice change were 68.2 (SD = 21.7), 58.3 (SD = 18.2), and 58.8 (SD = 14.5) mg/dl, respectively. After the practice change, the mean blood glucose values were 53.4 (SD = 20.2), 56.5 (SD = 15.5), and 53.2 (SD = 13.1) mg/dl, respectively. There were significant differences found between the pre- and postpractice change blood glucose values for the first, t(158) = 4.47, p = .000, and third, t(155) = 2.52, p = .013 blood draws (see Table 1). Blood glucose variability was lower in the postpractice change group compared to the prechange group at each successive blood draw (see Table 1, Figure 2).
Breastfed versus Formula-fed Groups For the second aim, mean blood glucose values were first examined by type of feeding without consideration of the practice change. For the colostrum-fed group of newborns, the first three mean blood glucose values were 57.6 (SD = 18.8), 56.2 (SD = 15.0), and 54.0 (SD = 11.2) mg/dl, respectively. For those receiving formula supplementation, the first three mean blood glucose values were 64.6 (SD = 24.8), 57.4 (SD = 16.7), and 57.3 (SD = 13.5) mg/dl, respectively. There were no significant differences between blood glucose levels by type of feeding at each of the three data collection times.
624
Colostrum Versus Formula Supplementation
found in each of the three blood glucose values from pre- to postpractice change (see Table 1). There was less variability in postpractice change blood glucose values in the colostrum-fed group compared to the formula-fed group (see Table 1, Figure 3). Fifty-eight percent (50/85) of newborns received formula supplementation prior to the practice change. After the practice change it decreased to 27.3% (21/77), reflecting an improvement of 53.6%. Prior to the practice change 18.8% (16/85) of the term infants born to diabetic mothers were transferred to the NICU for blood glucose stabilization as compared to 6.5% (5/77) postpractice change, a 68% improvement. The postpractice change group was much more likely to avoid NICU admission than the prepractice change group (χ2 = 5.44, df = 1, p .020).
Discussion These results demonstrate that the first three blood glucose values of infants born to mothers with type 1 or gestational diabetes who were fed colostrum are no different than those of infants who receive formula supplementation. There were significant differences in two of the three mean blood glucose levels in the colostrum-fed group when examined from pre- to postpractice change. Higher mean blood glucose levels occurred in the prepractice change group. Although postpractice change mean blood glucose values were lower for colostrum-fed infants, these levels remained within the normal range (≥40 mg/dl) for asymptomatic infants in the first 4 hours of life. Importantly, these results show that blood glucose variability was reduced with the first two blood draws in the postpractice change group of breastfed infants; and blood glucose variability stabilized with the third blood draw. Conversely, there was greater variability in the blood glucose levels of infants who received formula supplementation in the postpractice change group.
Next, mean blood glucose values were examined by feeding type and by pre- and postpractice change conditions. In the colostrum-fed group, there were significant differences in the first, t(74) = 4.42, p = .000, and third blood glucose values, t(74) = 3.68, p = .000, from pre- to postpractice change (see Table). For the formula-supplemented group, no differences were
The revised hypoglycemia algorithm and related policies resulted in a practice change that supports exclusive breastfeeding in this hospital. The number of babies who received formula (n = 50) compared to the number who received colostrum (n = 25) prior to the practice change was reversed after the practice change (n = 21 formula vs. n = 52 colostrum). This reduction in formula supplementation serves as evidence of practice change adoption. Furthermore, there were
JOGNN, 42, 619-628; 2013. DOI: 10.1111/1552-6909.12260
http://jognn.awhonn.org
PRINCIPLES & PRACTICE
Tozier, P. K.
Table 1: Comparison of the First Three Blood Glucose Values Between Pre- and Postpractice Change Groups Prepractice change
Postpractice change
n
Mean
SD
n
Mean
SD
Time 1 BG
84
68.2
21.7
76
53.4
20.2
Time 2 BG
82
58.3
18.2
76
56.5
Time 3 BG
82
58.8
14.5
75
Time 1 BG
25
69.7
20.3
Time 2 BG
25
60.0
Time 3 BG
25
Time 1 BG Time 2 BG Time 3 BG
t
df
p
4.47
158
.000
15.5
0.68
156
.495
53.2
13.1
2.52
155
.013
51
51.6
14.8
4.42
74
.000
15.2
52
54.4
14.7
1.53
75
.130
60.3
10.3
51
51.0
10.4
3.68
74
.000
50
67.9
22.8
21
56.7
28.0
1.77
69
.081
50
57.4
17.3
20
57.5
15.3
−0.02
68
.987
50
58.7
14.5
20
53.9
10.3
1.35
68
.180
Total sample
Breastfed only
Formula-fed
Note. BG = blood glucose (mg/dl); SD = standard deviation.
significantly fewer NICU transfers to treat hypoglycemia in the postpractice change group. Prior to the practice change 11.8% (10/85) of infants received intravenous treatment for hypoglycemia compared to 5.2% (4/77) of newborns who received intravenous treatment postpractice change.
Mean (+SD) BG Pre- and Postchange -
100 90
Limitations Limitations of this study include the use of a convenience sample and a retrospective chart review. There is no distinction in the chart for determining if the blood glucose values were obtained by pointof-care heel stick or by laboratory analysis. However, blood glucose values are routinely obtained
(+21.7) (+20.2) -
80
(+18.2) (+- 15.2) -
(+14.5) -
(+13.1) -
mg/dl
70 60 50
Prechange
40
Postchange
30 20 10 0
68.2
53.4
58.3
1
56.5 2
58.8
53.2 3
Blood Draw Time Points Figure 2. The First Three Mean Blood Glucose Values and Variability Comparing the Pre-and Postpractice Change Groups. BG = blood glucose (mg/dl), SD = standard deviation.
JOGNN 2013; Vol. 42, Issue 6
625
PRINCIPLES & PRACTICE
Colostrum Versus Formula Supplementation
The practice change empowered mothers to exclusively breastfeed, which may significantly affect the health and well-being of their children.
by point-of-care heel sticks in this institution, and abnormally low values are laboratory confirmed. Timing of the blood glucose draws (e.g., exact number of minutes or hours since birth) was not examined, except that the first three lab draws after birth were compared from pre- to postpractice change. The practice change called for delaying the drawing of the first blood glucose until after the first feeding. The meta-analysis by Croke et al. (2009) included studies in which feeding varied from not fed to breastfed or formula-fed newborns. These authors noted similar limitations and suggested future studies consider measuring time of blood glucose draws and feeding status. In retrospective chart reviews, this type of limitation is inevitable because the principle investigator cannot control some of the available data. The length of time infants spent in skin-to-skin contact, as well as the amount of colostrum or formula were not analyzed. These data were collected from birth to 6 hours of life and are not intended to suggest the trajectory of blood glucose values beyond this time frame. Generalizability may be limited to mothers who reside in a northeastern metropolitan U.S. city who gave birth in a major tertiary medical center.
(+28.0) -
80.0
mg/dl
70.0
The revised hypoglycemia algorithm and practice change provided a first step toward empowering mothers to practice exclusive breastfeeding that may significantly affect the health and well-being of their children. According to the Center for Disease Control (2013), diabetes has increased by 176% over the last 30 years. It is projected to increase another 64% between 2010 and 2025, affecting over 53 million people (Rowley & Bezold, 2012). The AAP (2012) stated that exclusive breastfeeding for 6 months will decrease a child’s risk of type 2 diabetes by 40% and type 1 diabetes by 30% and cited the positive effect of exclusive breastfeeding on weight control and feeding self-regulation. Obesity has also been associated with formula feeding due to the ease and tendency to overfeed beginning at birth (Koletzko et al., 2009). Breast milk, on the other hand,
Mean (+SD) BG Postpractice Change -
100.0 90.0
Implications for Practice Several challenges were identified with the practice change. When the practice of teaching hand expression for colostrum drops was first introduced, it was met with a myriad of reactions. Some staff stated that they were uncomfortable assisting mothers with hand expression. Occasionally some staff indicated a lack of time to fully engage in assisting with the hand expression process. Some physicians did not recognize or accept the importance of colostrum. Some patients were hesitant to ask for privacy or were uncomfortable with hand expression of colostrum while visitors were present in the room.
(+14.7) -
(+14.8) -
(+15.3) -
(+10.4) -
(+10.3) -
60.0 50.0
Colostrum
40.0
Formula
30.0 20.0 10.0
51.6
56.7
54.4
57.5
51.0
53.9
0.0 1
2 Blood Draw Time Points
3
Figure 3. The First Three Mean Blood Glucose Values and Variability Comparing Colostrum and Formula-Fed Infants in the Postpractice Change Group. BG = blood glucose (mg/dl), SD = standard deviation.
626
JOGNN, 42, 619-628; 2013. DOI: 10.1111/1552-6909.12260
http://jognn.awhonn.org
PRINCIPLES & PRACTICE
Tozier, P. K.
contains maternal hormones such as leptin that signals satiety and decreases the sensation of hunger in the infant (Savino, Liguori, Fissore, & Oggero, 2009). Breastfeeding may reduce childhood obesity by lowering plasma insulin levels, decreasing fat storage, and preventing adipocyte development (Oddy, 2012).
An audit of Canadian guidelines. Pediatric Child Health, 14(4), 238–244. Florez, J. C., Hirschhorn, J., & Altshuler, D. (2003). The inherited basis of diabetes mellitus: Implications for the genetic analysis of complex traits. Annual Review of Genomics and Human Genetics, 4, 257–291. Hay, W. W., Raju, T. N. K., Higgens, R. D., Kalhan, S. C., & Devaskar, S. U. (2009). Knowledge gaps and research needs for understanding and treating neonatal hypoglycemia: Workshop re-
Adopting hospital practices that facilitate exclusive breastfeeding is in line with the U.S. Surgeon General’s Call to Action (U.S. DHHS, 2011) and other professional and scientific associations’ position statements (AAFP, 2013; AAP, 2012; ABM, 2008; ACOG, 2007; WHO, 2013). The traditional approach of feeding formula to an asymptomatic breastfeeding term newborn at risk for hypoglycemia needs to be reevaluated in the context of evidence for safety and breastfeeding outcomes. This practice change demonstrated that colostrum stabilized newborn glucose levels as effectively as formula in the first 6 hours after birth and supported exclusive breastfeeding for mothers and newborns.
port from Eunice Kennedy Shriver National Institute of Child Health and Human Development. Journal of Pediatrics, 155(5), 612–617. Holmberg H., Wahlberg J., Vaarala O., Ludvigsson, J., & the All Babies in Southeast Sweden Study Group (ABIS). (2007). Short duration of breastfeeding as a risk factor for beta-cell autoantibodies in 5 year old children in the general population. British Journal of Nutrition, 97, 111–116. Holmes, A. V. (2013). Establishing successful breastfeeding in the newborn period. Pediatric Clinics of North America, 60, 147–168. doi:10.1016/j.pcl.2012.09.013 Hoseth, E., Joergensen, A., Ebbesen, F., & Moeller, M. (2000). Blood glucose levels in a population of healthy, breastfed, term infants of appropriate size for gestational age. Archives of Disease in Childhood: Fetal and Neonatal Edition, 83, F117–F119. Kostraba, J. N., Cruickshanks, K. J., Lawler-Heavner, J., Jobim, L. F., Rewers, M. J., Gay, E.C, . . . Hamman, R. F. (1993). Early exposure to cow’s milk and solid foods in infancy, genetic predisposition, and risk of IDDM. Diabetes, 42(2), 288–295.
REFERENCES Alkalay, A. L., Sarnat, H. B., Flores-Sarnat, L., Elashoff, J. D., Farber,
Koletzko, B., von Kries, R., Closa, R., Escribano, J., Scaglionis, S., Gio-
S. J., & Simmons, C. F. (2006). Population meta-analysis of low
vannini, M., . . . Grote, V. (2009). Can infant feeding choices
plasma glucose thresholds in full-term normal newborns. Amer-
modulate later obesity risk? American Journal of Clinical Nutri-
ican Journal of Perinatology, 23(2), 115–119.
tion, 89(5), 1502S-1508S. doi:10.3945/ajcn 27113D
American Academy of Family Physicians. (2013). Breastfeed-
Manco, M., Alterio, A., Bugianesi, E., Ciampalini, P., Mariani, P., Finoc-
ing (policy statement). Retrieved from http://www.aafp.org/
chi, M., . . . Nobili, V. (2011). Insulin dynamics of breast- or
about/policies/all/breastfeeding.html
formula-fed overweight and obese children. Journal of the Amer-
American Academy of Pediatrics. (2012). Breastfeeding and the use of human milk. Pediatrics, 129, e827–e841. Academy of Breastfeeding Medicine. (2008). ABM statements: Position on breastfeeding. Breastfeeding Medicine, 3(4), 267–270. doi:10.1089/bfm.2008.9988
ican College of Nutrition, 30(1), 29–38. Marinelli, K., Stellwagen, L., & MacEnroe, T. (2012). Enriching hospital support of breastfeeding [Audio podcast]. Retrieved from: http://www.ilca.org/i4a/pages/index.cfm?pageid=4131 Moore, E. R., Anderson, G. C., & Bergman, N. (2007). Early skin to skin
American College of Obstetricians and Gynecologists. (2007). Com-
contact for mothers and their healthy newborn infants. Cochrane
mittee opinion: Breastfeeding: Maternal and infant aspects.
Database of Systematic Reviews, Issue 3. Art. No: CD003519.
Retrieved
from
http://www.acog.org/∼/media/Committee%
20Opinions/Committee%20on%20Health%20Care%20for% 20Underserved%20Women/co361.pdf?dmc=1&ts= 20130708T0927582672
doi:10.1002/14651858.CD003519.pub2. Oddy, W. H. (2012). Infant feeding and obesity risk in the child. Breastfeeding Review, 20(2), 7–12. Owen, C. G., Martin, R. M., Whincup, P. H., Smith, G. D., & Cook, D.
Boluyt, N., van Kempen, A., & Offringa, M. (2006). Neurodevelop-
G. (2006). Does breastfeeding influence the risk of type 2 dia-
ment after neonatal hypoglycemia: A systematic review and de-
betes in later life? A quantitative analysis of published evidence.
sign of an optimal future study. Pediatrics, 117(6), 2231–2243. doi:10.1542/peds.2005-1919 Burns, C. M., Rutherford, M. A., Boardman, J. P., & Cowan, F. M. (2008). Patterns of cerebral injury and neurodevelopment outcomes af-
American Journal of Clinical Nutrition, 84, 1043–1054. Pett, M. A. (1997). Nonparametric statistics for health care research. Thousand Oaks, CA: Sage. Rowley, W. R., & Bezold, C. (2012). Creating public awareness: State
ter symptomatic neonatal hypoglycemia. Pediatrics, 122(1), 65–
2025 diabetes forecasts. Population Health Management, 15,
74. doi:10.1542/peds.2007-2822
194–200.
Centers for Disease Control. (2013). Crude and age-adjusted percent-
Rozance, P. J., & Hay, W. W. (2010). Describing hypoglycemia – Defi-
age of civilian, noninstitutionalized population with diagnosed
nition or operational threshold? Early Human Development, 86,
diabetes, United States 1980–2011. Retrieved from http://www. cdc.gov/print.do?url=http%3A//www.cdc.gov/diabetes/ statistics/prev/national/figage.htm Chantry, C. J., & Howard, C. R. (2013). Clinical protocols for management of breastfeeding. Pediatric Clinics of North America, 60, 75–113. doi:10.1016/j.pcl, 2012.10.005 Croke, J., Sullivan, M., Ryan-Drover, A., Randell, E., Andrews, W., & Aziz, K. (2009). Two hour blood glucose levels in at-risk babies:
JOGNN 2013; Vol. 42, Issue 6
275–280. doi:10.1016/j.earlhumdev .2010.05.002 Savino, F., Liguori, S. A., Fissore, M. F., & Oggero, R. (2009). Breast milk hormones and their positive effect on obesity. International Journal of Pediatric Endocrinology, 327505. doi:10.1155/2009/327505 The Joint Commission. (2009). Perinatal core measure set selection form. Retrieved from: http://www.jointcommission.org/ perinatal_care/
627
PRINCIPLES & PRACTICE
U.S. Department of Health and Human Service. (2011). The sur-
Wight, N., Marinelli, K. & the Academy of Breastfeeding Medicine
geon general’s call to action to support breastfeeding. Wash-
Protocol Committee. (2006). ABM Clinical Protocol #1: Guide-
ington, DC: U.S. Department of Health and Human Services,
lines for glucose monitoring and treatment of hypoglycemia
Office of the Surgeon General. Retrieved from http://www.
in breastfed neonates. Breastfeeding Medicine, 1, 178–
surgeongeneral.gov/initiatives/breastfeeding/
628
Colostrum Versus Formula Supplementation
184.
U.S. Department of Health and Human Services. (2013). Healthy
World Health Organization. (2013). Programmes and projects: Ma-
People 2020: Maternal, infant, and child health objec-
ternal, newborn, child and adolescent health. Breastfeeding.
tives: MICH-23. Retrieved from http://www.healthypeople.gov/
Retrieved from http://www.who.int/maternal_child_adolescent/
2020/topicsobjectives2020/objectiveslist.aspx?topicid=26
topics/child/nutrition/breastfeeding/en/index.html
JOGNN, 42, 619-628; 2013. DOI: 10.1111/1552-6909.12260
http://jognn.awhonn.org