Decreasing NICU admissions of asymptomatic infants of women with pregestational diabetes mellitus improves breastfeeding initiation rates

1 Background

Infants born to women with pregestational diabetes mellitus (PGDM) are at high risk for significant morbidities, including macrosomia, prematurity, birth trauma, respiratory distress, and hypoglycemia that often requires NICU care [1–5].

Admission to the NICU may prolong hospitalization, disrupts maternal-infant interaction [6, 7] and can affect breastfeeding (BF) initiation and duration [8–15]. The management of infants of women with PGDM (IDM) should maximize early and prolonged maternal-infant contact, which is best accomplished for asymptomatic IDM in the well baby nursery (WBN) and for those who are symptomatic in the NICU. Recent publications support admission of asymptomatic IDM to transitional or intermediate care facilities to promote bonding, BF and attachment [10, 17]. A decrease in the number of NICU admissions may also improve observed low BF initiation rates of women with PGDM [8–10, 13–15]. At present in our institution, most asymptomatic late preterm (34–36 w GA) and term IDM are admitted to the WBN while symptomatic IDM or those unable to orally feed are admitted directly to the NICU [4, 18]. Changing hospital and medical practices over time have resulted in improvements in BF initiation rates for the general population [19–21] and for women with complex pregnancies [10, 22–24]. Our extensive experience with the care of women with PGDM provided an opportunity to evaluate whether changes in NICU admissions over time have been associated with improvements in BF initiation rates.

Objective: To determine if changes in the number of NICU admissions of asymptomatic IDM born to women with PGDM in two different epochs affected BF initiation rates.

2 Subject and methods

This retrospective cohort investigation was approved by the Institutional Review Board of Wexner Medical Center at The Ohio State University. Hard copy (2008-11) and electronic maternal and neonatal records (2013-16) were reviewed. The transition from hard copy to electronic medical records occurred in 2012 thus, data from 2012 were not collected. Women were diagnosed with PGDM (Class B-C-DRFH) according to clinical and laboratory criteria using the modified White’s classification [7]. Obese was defined by a body mass index (BMI) of 29–34 kg/m², and very obese by a BMI ≥ 35 kg/m². ¹¹ The study population consisted of women with PGDM and their infants, if delivered at ≥ 34 weeks, a point in gestation in which most infants are able to orally feed. Pregnancies affected by major or fatal malformations were excluded. In cases of twin gestation, only the firstborn infant was reported. Upon arrival to labor and delivery, each woman declared her infant feeding preference, i.e., BF, FF or both. Since infant feeding at discharge was similar following delivery by primary or by repeat cesarean in either epoch, all cesarean deliveries were combined for analysis.

Depending on the condition of mother and infant following delivery, interactions (holding, skin to skin contact, BF) were encouraged. In the delivery room and postpartum, these interactions were observed and recorded by the nursing staff. Our institutional guidelines for care of the IDM recommend that any symptomatic infant regardless of the mothers’ class of diabetes be directly transferred from the delivery room to the NICU. Indications for NICU admission included respiratory distress [respiratory distress syndrome (RDS) and transient tachypnea of the newborn (TTNB)], prematurit y (apnea, severe hypotonia and poor sucking behavior), perinatal depression, and birth trauma. NICU admission for prevention of hypoglycemia was also an option for IDM whose mothers had poorly controlled diabetes or severe illness (i.e., classes DRFH). Asymptomatic IDM able to feed are transferred to WBN for routine care and glucose monitoring. Our family-centered care system has full-time lactation consultants and rooming-in available.

Screening for neonatal hypoglycemia (blood glucose <40 mg/dl) was done via serial point of care testing (Accu-Chek ^®) starting within the first hour of life after the first feeding and repeated every 2–4 hours as clinically needed [25, 26]. First BF or first FF was considered early when given within 120 minutes from birth. Asymptomatic IDM in the WBN with hypoglycemia were promptly BF or FF and those with recurrent hypoglycemia were treated with intravenous (IV) dextrose (4–6 mg/kg/min). All IDM, on admission to the NICU were either FF or started on IV dextrose on arrival.

Given that women who BF only 1-2 times per day during their hospitalization were more likely to stop BF soon after delivery, [27] we defined BF as initiated if, at the time of hospital discharge, the infant was exclusively BF or partially BF (when receiving >50% of the feedings directly from the breast or by expressed breast milk). Exclusive BF combined with partial BF was considered any BF. Due to the study design, no follow-up information was available on infant feeding practices beyond hospital discharge.

The management of IDMs have evolved over the last decades, however some medical practices and institutional changes incorporated during epoch 2 included: adoption of the 2011 AAP and 2014 ABM guidelines for management of infants at risk for hypoglycemia and early feeding for prevention and treatment of hypoglycemia [25, 26]. Development of guidelines for the management of asymptomatic infants of women with PGDM [10–12]. Expansion of lactation services (from 2–4 FTE) to cover weekends and holidays and extension of smoking cessation programs to antepartum and postpartum services. Emphasis was placed on the education of staff nurses of the perinatal services on promotion and support of BF.

3 Results

Our study population consisted of 386 mother-infant dyads during epoch 1 and 457 during epoch 2. Comparisons of the demographic and clinical characteristics of women in each epoch are presented in Table 1. The distribution of White’s classes was similar across time. Women in epoch 2 were slightly older, had lower average hemoglobin A1C values and delivered more often by repeat cesarean and less often by primary cesarean. A decrease in the number of whites and an increase in the number of Hispanics was observed over time. Rates of very obese women remain unchanged. The number of women who smoked during their pregnancy decreased significantly from epoch 1 to epoch 2 (Table 1). There were no maternal or neonatal deaths.

3.4 Infant feeding at discharge from the hospital

For all infants admitted to the NICU and to the WBN combined, at discharge, exclusive BF and any BF increased across the two epochs (from 13 to 24% and 45 to 60% , p 0.0001 respectively). Regression analysis showed that women with PGDM were more likely to initiate any BF during epoch 2 (aOR 1.9, 95% CI 1.4–2.7, p 0.0003). Conversely, a significant decline (55 to 40%)* in FF at discharge was noted. Including only women who intended to BF (254 from epoch 1 and 298 from epoch 2) in the analysis, infant feeding rates at discharge increased further (exclusive BF from 18 to 35% and any BF from 62% to 80% , p 0.0001).

Among IDM admitted to the NICU, at the time of discharge, exclusive BF and any BF increased across time, while FF decreased (Table 3). Including only women who intended to BF (127 from epoch 1 and 115 from epoch 2) in the analysis (Fig. 1) exclusive BF (17 to 27%)* and any BF (56% to 75%)* increased further. Conversely, FF declined (44 to 25%)*. Regression analysis showed that after controlling for potential confounders women with PGDM whose infants were admitted to the NICU were less likely to initiate any BF (a OR 0.6, 95% , CI 0.4–0.9, p 0.005).

OPEN IN VIEWER

Fig.1

Feeding at discharge of IDM whose mothers intended to breastfeed. Number of infants by place of admission: NICU 157 (Epoch 1) and 115 (Epoch 2), WBN 97 (Epoch 1) and 185 (Epoch 2) ^*p < 0.001.

Among IDM admitted to the WBN, at the time of discharge, exclusive BF and any BF increased across time, while FF decreased (Table 3). Including only women who intended to BF (97 from epoch 1 and 185 from epoch 2) in the analysis (Fig. 1) exclusive BF (17 to 27%)* and any BF (56% to 75%)* increased further. Conversely, FF declined (30 to 16%)*.

4 Discussion

The incidence of diabetes complicating pregnancy continues to rise [5]. Thus, it may be anticipated that NICU admissions may increase, especially for IDM born to women with PGDM [1, 23]. The number of NICU admissions has been considered a global marker of undesirable perinatal outcomes for women with complex pregnancies such as diabetes [4, 24]. BF initiation failure deserves to be considered an adverse perinatal outcome especially for women with high risk obstetrical conditions who could also benefit from lactation [8, 29]. Germaine to our study, the association of NICU admissions with low BF initiation and BF duration have been well-recognized [4, 23].

The significant decrease in NICU admissions observed during epoch 2 coincided with increased rates in exclusive BF and in any BF at the time of discharge. These findings are more impressive considering they involve women with PGDM well known for low BF initiation rates [4, 24]. The improvement of BF initiation rates for IDM in the NICU and the WBN is likely a reflection of multidisciplinary efforts and institutional changes made during the last decades [5, 24].

While intention to BF has long been recognized as a strong predictor of BF initiation and BF duration, the intention to FF has been largely ignored. It is however important to remember that these women seldom change their original feeding intention postpartum [4, 30]. Additional factors for the low BF initiation observed among women with PGDM include the high incidence of obesity, [7, 11] cesarean delivery [9, 32] complications of labor and delivery, need for specialized care of their infants [11, 15], as well as for delayed lactogenesis II [32, 33].

Maternal and neonatal morbidities that prompted admissions to the NICU often preclude skin-to-skin contact and early BF [6, 35]. The mother-infant separation undoubtedly has an emotional impact on the mother and likely a physiological effect on the infant [6, 36]. Also, low BF initiation among women with PDGM may be due to concerns for their own health or by the life disruptions created by the birth of a healthy or sick newborn [7, 36]. In addition, going home while leaving their infant at the NICU may add another stress to their childbearing experience.

In the past, IDM especially macrosomic whose mothers were on insulin were routinely admitted to NICU for prevention of hypoglycemia [1, 22]. Indeed, during epoch 1 one third of IDM were admitted with that diagnosis. However, few of these IDM became hypoglycemic and most stayed at the NICU for less than 24 hours. Perhaps, many of them could have been BF or partially BF at the WBN or other intermediate care facility [3, 26].

Cesarean birth is known to increase the time before skin-to-skin contact, delay the onset of lactation, reduce the incidence of exclusive BF, and increase the likelihood of formula supplementation [31, 38]. In a recent study of women with high risk obstetrical conditions including 89 with PGDM, initial mother-infant contact occurred within one hour in 38% of the cases following a pre-labor cesarean and in 19% after a primary cesarean, but no differences in BF initiation rates following either mode of delivery were noted [14]. It is possible that the post cesarean delay of mother-infant contact may not be an absolute obstacle to BF initiation among women with complex pregnancies who intended to BF [18]. On the other hand, our findings concurred with the literature in that BF initiation rates are higher following vaginal deliveries [10–12, 37].

The beneficial impact of early BF on BF initiation and duration is no longer in dispute [3, 23]. Due to the almost universal initiation of IV dextrose upon admission to the NICU, early BF or FF was less common during the first epoch. Since mother-infant interactions were limited under these conditions, it is not surprising that BF initiation was lower among IDM in the NICU than those in the WBN [8, 14].

Among infants admitted to WBN, early breastfeeding increased from 15% to 62% and early formula feeding decreased from 85% to 55% from epoch 1 to epoch 2. Concurrently, hypoglycemia affected about a third of the infants in either epoch. Regardless of mothers’ diabetes class, early feeding choice or epochs, oral feedings corrected the hypoglycemia in the majority of cases, avoiding the need for IV dextrose and mother-infant separation [3, 26].

In spite of the harmful effects of smoking on the fetus and neonate approximately ten percent of women continue to smoke during pregnancy and lactation [38–40]. Reduced production of milk, shorter lactation, delayed initiation of the sucking reflex and lower sucking pressure during BF have been reported [40]. Considering that 15 to 30% of women with complex pregnancies including PGDM smoked during pregnancy, [4, 23] the decrease in smoking rate observed here is notable. This decline is encouraging especially among former smokers since it appears that women who quit smoking prior to the delivery date may not smoke during subsequent pregnancies [39].

Limitations of the investigation are those inherent to retrospective designs and include the lack of follow-up regarding infant feeding after discharge. Also, the definition of BF initiation at discharge may only be applicable to women with complex pregnancies for whom earlier mother-infant contact may be delayed. The strength of the investigation rests on the size of the study population, the comparison of two different epochs and the fact that mothers and infants were cared for at a single institution. This study is unique because it provides BF rates for all women with PGDM regardless of infant feeding preferences as well as specific rates for women who intended to BF exclusively.

5 Conclusions

In summary, regardless of place of IDM admission, BF initiation rates increased during epoch 2. Although many factors may have contributed to these changes, the lower rate of admissions to the NICU remained temporally related with the improvement in BF initiation rates. IDM who required specialized care were identified and treated in the NICU while asymptomatic IDM were safely admitted to WBN. Early feeding for prevention or treatment of hypoglycemia was an important characteristic of epoch 2. Although encouraging, the observed increase in BF initiation rates are far from ideal. The discordance between intention to BF and BF initiation remains a challenge and highlights the need for targeted novel interventions.