Abstract
Background:
Parents may consider interrupting breastfeeding to manage neonatal jaundice (NJ). Our aims were to determine correlations of breastfeeding with NJ by examining infants' manifestations in the first week after birth and to understand parents' perceptions toward NJ in relation to breastfeeding.
Materials and Methods:
This prospective cross-sectional study was conducted in a tertiary medical center by examining infants and administering a questionnaire survey to their parents. All healthy infants admitted to the well-baby nursery were eligible for enrollment. A 16-item questionnaire was distributed to parents of enrolled infants from October 2017 to February 2019. Items of the questionnaire included perceptions and knowledge of NJ. In addition, clinical information of enrolled infants was obtained from medical records. Hyperbilirubinemia was defined as a peak transcutaneous bilirubinometer value ≥15 mg/dL.
Results:
In total, 449 parents completed the consent form and participated in the study. Results showed that exclusive breastfeeding was more common in infants with a vaginal delivery (p < 0.001), who were nonprimiparous (p = 0.004) and who had weight loss of >7% (p < 0.001). There was no significant correlation of exclusive breastfeeding with hyperbilirubinemia (p = 0.414). Approximately two-thirds of parents were worried about NJ occurring in their child. Most parents were aware of phototherapy as management of NJ. However, their knowledge of risk factors, complications, and assessments of NJ was relatively deficient. Overall, 29.6% of parents rated breastfeeding as a risk factor for NJ, and 24% of parents indicated that cessation of breastfeeding was a management option for NJ.
Conclusions:
The results indicated that NJ in the first few days after birth poses a significant barrier to breastfeeding. Our findings provide critical information for plotting strategies to enhance parents' willingness to continue breastfeeding.
Introduction
Neonatal jaundice (NJ) is associated with a variety of physiologic and pathologic conditions. Breastfeeding, blood group incompatibility, and a glucose-6-phosphate dehydrogenase (G6PD) deficiency are common factors associated with NJ.1,2 Severe NJ may cause bilirubin encephalopathy, including acute manifestations, long-term complications, and even death. 3 Phototherapy (PT) is the mainstay of traditional treatment, with an exchange transfusion held in reserve for neonates with a risk of bilirubin encephalopathy. 4 With the advent of management, incidences of adverse outcomes decline. 5
NJ in relation to breastfeeding is classified into two categories: breastfeeding jaundice and breast milk jaundice. Insufficient mother's milk is the leading cause of breastfeeding jaundice, which always occurs within the first week after birth. 6 Optimal management of breastfeeding jaundice is early initiation of breastfeeding, followed by frequent breastfeeding. 7 When the intake of human milk is adequate, the correlation of breastfeeding with significant NJ might not exist.8,9 In addition, breast milk jaundice is characterized by prolonged jaundice and may last until 2–3 months of age with no other identifiable cause. 10 Nevertheless, breast milk jaundice causes no risk to infants. However, interruption of breastfeeding to treat NJ has inappropriately been recommended by clinical staff and neonatologists. 11
It is a challenge to promote breastfeeding among mothers of nursing infants with jaundice. Formula use at this critical time often leads to premature interruption of breastfeeding. 12 Breastfeeding should not be interrupted to prevent significant NJ in the first few days after birth. 13 Instead, frequent breastfeeding is encouraged to avoid an inadequate amount of human milk. In this study, we examined the correlation of breastfeeding with NJ through infants' clinical manifestations in the first week after birth. In addition, we conducted a questionnaire survey to investigate parents' perceptions toward NJ in relation to breastfeeding. Our data provide clinical evidence for development of strategies to promote breastfeeding.
Materials and Methods
Design
This was a single-center prospective study, involving exploratory research conducted by examining infants and administering questionnaire interviews to their parents. The Institutional Review Board of Chang Gung Memorial Hospital approved the study protocol (No. 201601443B0). All healthy neonates born in Chang Gung Memorial Hospital at Taipei were eligible for enrollment in this retrospective study. In addition, infants with a gestational age (GA) of <35 weeks, a birth weight of <2000 g, multiple births, an illness (such as a significant congenital anomaly or infection) were excluded.
Setting/sample
Chang Gung Memorial Hospital at Taipei is a tertiary medical center with ∼1500 births per year. It was appraised as being a baby-friendly hospital by the Taiwan Joint Commission on Hospital Accreditation (TJCHA). Breastfeeding and rooming-in are highly encouraged. However, healthy infants are not routinely nursed alongside their mothers. Parents can decide whether their baby stays with them in the maternal ward or stays in the well-baby nursery. For example, parents may choose to stay in the maternal ward for 2–7 days while their baby stays in the well-baby nursery. If not rooming-in, mothers breastfeed their babies in the well-baby nursery. When an infant stays in the nursery, hospital nurses help the parents feed their baby with either breast milk or formula. In addition to helping mothers breastfeed, nurses encourage mothers to breastfeed more frequently if the infant's weight loss exceeds 7% of the birth weight. Nevertheless, formula feeding is allowed if parents insist.
Between October 2017 and February 2019, we approached parents of healthy infants admitted to the well-baby nursery. A trained research assistant interviewed potential participants to explain the study protocol. Informed consent was obtained before the research assistant administered the questionnaire to participating parents. Only one parent was recruited per infant. An introductory letter stating the purpose of the study and promising confidentiality accompanied the questionnaire. Return of the consent form with completed questionnaire was considered enrollment for valid analysis.
Measurement
A questionnaire survey was administered to determine knowledge of NJ among parents staying in the nursery during the first week after birth. The recruited parents completed a questionnaire while the baby stayed in the well-baby nursery. We developed a 16-item questionnaire based on previous questionnaires.14–17 This questionnaire included items for measuring awareness of, worry about, and knowledge (sources, risk factors, complications, assessments, and management) of NJ in the first week after birth. Worry about NJ was rated using a 5-point Likert scale (strongly agree, agree, neutral, disagree, and strongly disagree). Background characteristics of parents (including gender, age, educational level, and employment status) were also collected.
The content validity of the questionnaire was examined by five experts to assess the clarity and ease of completion. Test–retest reliability coefficients were used to estimate the internal consistency of all indices. The same questionnaire was pilot-tested twice at an interval of 2 weeks on a sample of 15 nurses working in the nursery. A content validity index of 0.975 and test–retest reliability coefficient of 0.927 indicated excellent validity and reliability of parameters in this questionnaire.
The bilirubin level of enrolled infants was screened daily with a noninvasive transcutaneous bilirubinometer (TcB) BiliCheck device (Spectrx, Norcross, GA). The protocol of PT was modified from 2004 guidelines of the American Academy of Pediatrics, 13 as indicated for infants with a GA of 35–37 weeks (TcB ≥7 mg/dL at <24 hours old, ≥9 mg/dL at 24–35 hours old, ≥10 mg/dL at 36–47 hours old, ≥12 mg/dL at 48–59 hours old, ≥13 mg/dL at 60–71 hours old, ≥14 mg/dL at 72–95 hours old, and ≥15 mg/dL at ≥96 hours old) and infants with a GA of >37 weeks (TcB ≥9 mg/dL at <24 hours old, ≥11 mg/dL at 24–35 hours old, ≥12 mg/dL at 36–47 hours old, ≥14 mg/dL at 48–59 hours old, ≥15 mg/dL at 60–71 hours old, ≥16 mg/dL at 72–95 hours old, and ≥17 mg/dL at ≥96 hours old). The serum total bilirubin level was measured when the TcB was significantly higher and when intensive PT was indicated.
Data collection
A trained research assistant explained the study protocol and obtained informed consent before administering the questionnaire to parents. Birth data and clinical information of enrolled infants were obtained from medical records—including the birth weight, GA, delivery mode, gender, and feeding type. Exclusive breastfeeding was defined as breastfeeding since birth. Partial breastfeeding was defined as at least one meal of breast milk and formula daily. The frequency of stool output was classified into two categories: (1) more than four times per day on average since birth and (2) four times or fewer per day on average since birth.
Data analysis
A commercially available program was used for all statistical analyses (SPSS® vers. 19.0 for Windows, Chicago, IL). To analyze categorical variables, we used a chi-square test or Fisher's exact test when appropriate. Continuous variables were compared using Student's t-test. Outcome variables for analysis included items of the questionnaire survey and clinical information. Comparisons of clinical data with items of the questionnaire survey were analyzed as well. Significance was defined as p < 0.05 in two-tailed tests.
Results
Demographic data
We approached 488 eligible parents, and 449 parents agreed to participate in the study (for a participation rate of 92%). Specifically, 222 fathers and 227 mothers completed the questionnaire survey during 1–7 days after the birth of their child. Table 1 shows the demographic information of enrolled parents and infants. The average ages of fathers and mothers were 34.5 ± 7.7 and 33.9 ± 4.9 years, respectively. Most parents were well educated. About three-quarters of participants were employed.
Demographic Information of Enrolled Parents and Infants (n = 449)
GA, gestational age; SD, standard deviation.
As for infant information, 261 (58.1%) infants were firstborn. During their stay in the nursery, 139 (31%) infants were exclusively breastfed. In addition, 28 (6.2%) infants were preterm. The average age of enrolled infants at the time of questionnaire survey was 2.3 ± 1.2 (1–7) days.
Clinical characteristics
Table 2 illustrates the clinical features of infants according to whether they were exclusively breastfed. Weight loss of >7% was more common in exclusively breastfed infants (p < 0.001). Similarly, weight loss of >10% was more common in exclusively breastfed infants (p < 0.001). In addition, being a firstborn child (p = 0.004) and having been delivered by cesarean section (p < 0.001) were less common in infants who were exclusively breastfed. Other information—including the peak TcB level, use of PT, stool passage, gender, and GA—revealed no significant differences between infants who were and those who were not exclusively breastfed.
Correlations of Exclusive Breastfeeding with Infant Clinical and Birth Information (n = 449)
PT, phototherapy; TcB, transcutaneous bilirubinometer.
Perceptions toward NJ
Of the 449 participating parents, one mother and two fathers were unaware of NJ. Table 3 shows the perceptions toward NJ among 446 parents who were aware of NJ. The most common source of knowledge about NJ was from the internet, followed by friends, relatives, nurses, doctors, experiences, and books.
Parents' Perceptions of Neonatal Jaundice Between Their Infants With and Without Phototherapy (n = 446)
G6PD, glucose-6-phosphate dehydrogenase; NJ, neonatal jaundice.
Approximately one-half of parents did not know the risk factors for NJ in the first week after birth. Breastfeeding was the most commonly known risk factor of NJ, followed by preterm birth, a G6PD deficiency, and blood group incompatibility. Furthermore, over a half of the parents did not know any complications of severe NJ. Liver damage was the most commonly known complication of NJ. As for assessing NJ, more than half of the parents knew of the following three methods to assess the severity of NJ—a dark-yellow skin color, yellowish skin color on the entire body, and the TcB level. In addition, 408 (91.5%) parents rated PT as management for NJ. Nevertheless, 107 (24.0%) parents rated cessation of breastfeeding as a treatment choice for NJ. Overall, 282 (63.3%) parents were worried about NJ occurring in their child.
We further compared parents' conceptions toward NJ between infants with and those without PT at the day of questionnaire completion. Specifically, parents whose children had received PT more often rated breastfeeding as a risk factor for NJ than those without PT (p = 0.008). In addition, they were more aware of poor appetite being a clinical presentation of severe NJ (p = 0.001). They less often rated oral drugs as being helpful for managing NJ (p = 0.024). There was no significant difference in knowledge sources between the two groups.
Characteristics of parents rating breastfeeding as a risk factor for NJ
Breastfeeding was regarded as the most important risk factor for NJ during the first week after birth. Thus, we compared the belief that “breastfeeding is a risk factor for NJ in the first week after birth” with parents' knowledge and demographics (Table 4).
Parents' Belief in Breastfeeding as a Risk Factor for Neonatal Jaundice
Parents who rated breastfeeding as a risk factor more often knew of the complications, assessments, and management of NJ. Furthermore, they more often had obtained knowledge of NJ from medical doctors, the internet, and their previous experience. In contrast, they less often had obtained knowledge from relatives and friends. In addition, parents with the following characteristics—being a father, aged >35 years, the child being primiparous, and with an infant undergoing PT—less often rated breastfeeding as a risk factor. Other information—educational level, occupation, and feeding type—revealed no significant difference between the two groups.
Characteristics of parents rating cessation of breastfeeding as management for NJ
We further compared the belief that “cessation of breastfeeding is a management option for NJ” with parents' knowledge and demographics (Table 5). Parents who believed that cessation of breastfeeding was a management option for NJ were more knowledgable about the risk factors, complications, and assessments of NJ. In addition, they more often had obtained their knowledge from medical doctors and their previous experience, and less often had obtained their knowledge from relatives and friends. Furthermore, parents aged >35 years or who had a child that was primiparous more often rated cessation of breastfeeding as a management option for NJ. Other characteristics—educational level, occupational status, having an infant that had undergone PT, and feeding type—revealed no significant difference between the two groups.
Parents' Belief in Cessation of Breastfeeding as a Management Option for Neonatal Jaundice
Discussion
This study revealed a knowledge gap among Taiwanese parents as to risk factors, complications, assessments, and management of NJ. To our knowledge, this is the first quantitative survey to compare parents' perceptions toward NJ with exclusive breastfeeding. In this prospective study, we compared clinical manifestations between infants who were and those who were not exclusively breastfed during their first few days after birth. The results illustrate a couple of important issues for clinical implications of breastfeeding. First, there was no significant association between exclusive breastfeeding and hyperbilirubinemia in infants in the first few days after birth. This finding is in accordance with previous reports, showing that breastfeeding is not a risk factor for severe NJ during the first week of life.8,18 Second, weight loss of >7% was more common in infants who were exclusively breastfed. Body weight change is a parameter indicating whether an infant is dehydrated. Although it is recommended that weight loss for healthy breastfed newborns should not exceed 7%,8,19 our study demonstrated that weight loss of >7% was common among breastfed infants. These results are consistent with previous studies, which showed that more than half of healthy breastfed newborns lost 7% of their birth weight.12,20,21 Nevertheless, our analysis did not find a significant correlation of hyperbilirubinemia with weight loss >7% (data not shown). Thus, our results suggest that weight loss of >7% is a normal phenomenon among breastfed infants. In addition to weight loss, exclusive breastfeeding was less common in primiparous mothers and mothers who had undergone a cesarean section. We speculated that wound pain might be an important reason not to initiate breastfeeding immediately after birth.
In this study, the rate of cesarean section was relatively high. A probable explanation is the high average age of enrolled mothers (33.9 years). Our previous report analyzing >2 million infants in Taiwan showed that rate of cesarean section was significantly higher for women of an older age. 22 Since our hospital is a tertiary hospital, we may enroll more mothers with a high-risk pregnancy, including a high age.
Our study used a questionnaire survey to determine parents' knowledge of NJ in relation to breastfeeding. Unlike other questionnaire studies that selected mothers with hyperbilirubinemic infants23,24 or mothers with older children, 16 we randomly selected parents within a few days of having a new baby. In particular, we investigated perceptions of both mothers and fathers. Fathers increasingly supported breastfeeding. 25 Their support can influence breastfeeding decisions and behaviors. 26 In our study, fewer fathers rated breastfeeding as a risk of NJ. Their knowledge of NJ greatly differed from that of mothers. With support from fathers, breastfeeding outcomes can improve. 27 In addition, our analysis differed from other questionnaire studies because we focused on NJ in relation to breastfeeding. Parents may suffer negative emotions when they need to monitor the bilirubin levels of their infant or when their infant must undergo PT. Thus, our study showed that a majority of parents were worried about the impacts of NJ on the health of their child.
In this study, a number of parents rated breastfeeding as a risk factor for NJ and considered cessation of breastfeeding as a means to manage NJ. Those parents had learned about NJ from a variety of sources, including the web, relatives, friends, health care workers, and their own experience. Health care workers in hospital settings might not have sufficient time or adequate knowledge to provide information about NJ. Thus, parents have to learn to adapt by actively seeking information from other sources. Furthermore, we found that the sources were associated with their knowledge of NJ in relation to breastfeeding. Parents with sources from physicians and their own experience more often rated breastfeeding as a risk factor for NJ and considered cessation of breastfeeding as a way to manage NJ. Interactions with physicians may be an important factor mediating the impact of parents' experience with NJ in terms of breastfeeding. 28 These findings suggest that physicians were not well prepared to counsel parents about breastfeeding. To better promote breastfeeding, training for a deeper understanding of NJ should be incorporated into breastfeeding education for physicians.14,29–31
It is noteworthy that most parents knew that PT is standard management for NJ. However, their knowledge of risk factors, assessments, and complications was inadequate, probably because such information is not much mentioned by health care providers. In addition, our results indicated cessation of breastfeeding as the second most common method of preventing NJ. Although breastfeeding carries a significant risk of NJ, it is not a risk factor for neurological sequelae. 3 Thus, cessation of breastfeeding should not be routinely recommended as an option for treating NJ. 13 However, our results suggest that severe NJ may have a negative impact on the willingness to breastfeed. Our findings are in accordance with a study, which showed that bilirubin levels >12 mg/dL adversely affect breastfeeding. 32
In our study, parents with the following characteristics—with a nonprimiparous child, aged ≤35 years, and being a mother—were prone to believe that breastfeeding was a risk factor for NJ and considered interrupting breastfeeding as a means of management. It was reported that demographic factors were associated with knowledge of, attitudes toward, and behaviors associated with NJ as related to breastfeeding. 17 Education about breastfeeding should be focused on this group of parents. Furthermore, most parents in our study had high educational attainment. Nevertheless, their knowledge of NJ in relation to breastfeeding was insufficient. These findings suggest that different educational levels and cultural practices should be considered when designing educational materials.14,30
There are some limitations to this study. First, our survey was a prospective study, not a randomized controlled trial (RCT). However, it is very difficult to conduct an RCT to compare infants with and those without breastfeeding. Second, inaccuracies may have occurred in the questionnaire survey. Third, we did not approach all eligible parents. Parents of infants who did not stay in the nursery, such as those rooming-in, were not approached. Exclusive breastfeeding, neonatal hyperbilirubinemia, and significant weight loss are more common in infants who room-in. 33 Thus, the clinical manifestations of enrolled infants in this study should differ from those of infants who were rooming-in. Fourth, the National Health Insurance of Taiwan regulates maternal stay of 3 days for vaginal delivery and 5 days for cesarean section. Thus, selection bias may exist because of different durations of hospital stays. Accordingly, our method may have included a relatively low rate of breastfeeding and a higher rate of cesarean section. Nevertheless, the rate of PT in this study was similar to a previous report in Taiwan. 34 Therefore, we believe the impact of selection bias on our study results was not very significant in terms of parents' perceptions toward NJ. Fifth, the various times the questionnaire survey was conducted may have resulted in different perceptions toward NJ. Our data suggest that parents whose children received PT had greater knowledge of NJ. Sixth, the generalizability of this study is limited. As for the strength of this study, we conducted a self-administered survey and also an audit of actual practice. Thus, the results reflect the realities of practice under routine clinical care in Taiwan.
Conclusions
Jaundice is a common condition during the first week of life. Our study demonstrated that breastfeeding is not a significant risk factor for severe NJ during the first few days after birth. In addition, we provide insights into parents' perceptions of ceasing breastfeeding to avoid NJ. The findings have clinical implications for strategic developments to encourage breastfeeding. We revealed some misconceptions about NJ pertaining to breastfeeding, which are worthy of note. A significant association between parents' knowledge of NJ and breastfeeding was observed. The data suggest that knowledge of NJ and breastfeeding was insufficient among most parents.
Knowledge of NJ, in terms of sources, risk factors, assessments, complications, and management, varied considerably among parents. Efforts to increase knowledge are required to reduce parents' concerns about NJ. Thus, cessation of breastfeeding is not encouraged according to findings of this study. Regular training workshops or seminars may be helpful in enhancing correct concepts about NJ in relation to breastfeeding.
Footnotes
Acknowledgments
The authors thank the parents who inspired and participated in this research.
Disclosure Statement
No competing financial interests exist.
Funding Information
This work was supported by research grants from Chang Gung Memorial Hospital (CMRPG1F0182, CMRPG1J0111) and the Ministry of Science and Technology, Taiwan (MOST 106-2314-B-182A-032- and MOST 105-2410-H-038-011-SSS).