Factors associated with exclusive breastfeeding at hospital discharge: a study using data from the Georgian Birth Registry

The World Health Organization (WHO) recommends exclusive breastfeeding for the first six months of a child’s life, defined as no other solids or liquids besides breast milk and essential vitamins or medicines, and continued breastfeeding up to the age of two years or beyond [1]. To facilitate exclusive breastfeeding, the WHO promotes immediate skin-to-skin contact between mothers and newborns, and initiation of breastfeeding within one hour of delivery [2]. In low- and middle-income countries, only 37% of children under the age of six months are exclusively breastfed [3], despite the fact that breastfeeding is associated with numerous short- and long-term health advantages for both the mother and child. For mothers, long-term benefits include lower risk of breast cancer, ovarian cancer, and type II diabetes mellitus [4]. In children, breastfeeding has a protective effect against gastrointestinal and respiratory infections below five years of age [5], with a clear dose-response relationship [6] of long-term effects include lower odds of obesity and overweight in childhood and later life [7, 8].

Many factors affect breastfeeding; they can be maternal, newborn, or obstetric in nature and are often interlinked. Older age of the mother has been positively associated with exclusive breastfeeding (EBF) at discharge [9]. The relation between maternal age and EBF might be partly explained by parity, but this link is uncertain. Overweight and obesity has been associated with lower odds of exclusive breastfeeding [10], which may be related to both physical and psychosocial factors. Higher education and high socio-economic status of the mother are associated with increased rates of breastfeeding initiation and duration in high-income countries, for both exclusive and any breastfeeding [10‐14], whereas in low- and middle-income countries this association is inversed [15‐19]. Studies from all over the world indicate a negative association between caesarean delivery and breastfeeding [9, 14, 15, 20‐25]. In addition to maternal and/or newborn distress, the effect of caesarean delivery on early breastfeeding might be related to delayed onset of lactation, problems with newborn suckling, disrupted early skin-to-skin contact and mother-newborn interaction, and postoperative hospital practices [23].

Located in the Caucasus region, Georgia is categorized as an upper-middle income developing country, ranked 70 of 188 countries in the 2017 Human Development Index [26]. Georgia has a population of 3.7 million; 87% consider themselves as ethnic Georgians and almost 11% are from the neighboring countries Armenia and Azerbaijan [27]. The infant mortality rate decreased from 22.5 in 2009 to 9.6 per 1000 livebirths in 2017, with a fertility rate of 2.1 in 2017 [27]. All medical facilities in Georgia are private. The beneficiaries of private insurance have decreased since Georgia introduced Universal Health Care in 2013. Basic obstetric care is free, which currently includes eight antenatal visits and the delivery. All Georgian mothers deliver at maternity wards with qualified healthcare personnel (99.9%) [28], and the average length of stay for vaginal delivery is 3–4 days and for caesarean delivery 5–6 days [29].

During the early 2000s, awareness increased about the advantages of breastfeeding in Georgia. In 2004, 14 out of 78 maternity wards in the country were designated as baby-friendly through the Baby-Friendly Hospital Initiative (BFHI), a global initiative to promote and support breastfeeding by WHO and the United Nations Children’s Fund [30]. However, today there are no baby-friendly hospitals left in Georgia due to lack of follow-up on the initiative and breastfeeding support after discharge is lacking at primary healthcare level. Data on breastfeeding practice in Georgia are limited. The information that does exist are provided by national surveys, which have presented inconsistent numbers. One survey from 2010 showed that 20% of mothers initiated breastfeeding within one hour of delivery [29], while another survey from the same year reported a proportion of 66% [31]. According to a survey from 2005, only 11% of children under six months were exclusively breastfed [32], whereas in 2010 this proportion was reported to be as much as 55% [31].

Georgia introduced a national birth registry in 2016, the Georgian Birth Registry (GBR) [33], which collects medical data from antenatal visits, as well as data from the delivery and postpartum period until discharge from the maternity ward/hospital. Reporting to the GBR is mandatory. Designated health professionals receive training in how to manage the register, and are then responsible for teaching the staff at their own facility. The GBR includes information on early postpartum breastfeeding. The objective of this study was to identify factors associated with exclusive breastfeeding of term newborns at hospital discharge in Georgia, using national registry data.

The study population consisted of all births registered in the GBR from November 1st to December 31st 2017 (n = 8159 newborns). Because this study looks at term newborns only (born between gestational age 37^+ 0 and 41^+ 6 weeks) [34], all preterm and post-term births were excluded (n = 721). Gestational age was estimated by the first day of the last menstrual period in 70.1% of the sample, with the rest being estimated by ultrasound. Stillbirths (n = 76) and neonatal deaths (n = 38) were also excluded, as were newborns with a HIV-positive mother (n = 5) or a surrogate mother (n = 52). Multiple births (n = 246) were excluded because these mothers face greater challenges to initiate and to continue breastfeeding [9, 35], as were newborns with missing information about breastfeeding status or unknown breastfeeding status at hospital discharge (n = 319). The final study sample consisted of 7134 live, singleton, term births, with information about breastfeeding at hospital discharge (Fig. 1).

Of the 7134 newborns in the study sample, 6583 (92.3%) were exclusively breastfed at hospital discharge and 551 were not. Maternal, delivery, and newborn characteristics differed by breastfeeding status at hospital discharge (Table 1).

In adjusted analyses, education level, mode of delivery, birthweight, and admission to NICU were identified as factors associated with nonexclusive breastfeeding at hospital discharge. Mothers with higher education were 25% less likely to exclusively breastfeed at hospital discharge compared to mothers with secondary education or less (AOR 0.75; 95% CI 0.59, 0.97). Newborns born by caesarean delivery were 53% less likely to be exclusively breastfed at hospital discharge compared to those born by vaginal delivery (AOR 0.47; 95% CI 0.37, 0.60). The newborns with the lowest birth weight (< 2500 g) had 49% lower odds of receiving exclusive breastfeeding at hospital discharge compared to those weighing 3000–3499 g (AOR 0.51; 95% CI 0.27, 0.97). In addition, admission to NICU after delivery was strongly associated with nonexclusive breastfeeding at hospital discharge. Newborns admitted to NICU were 98% less likely to be exclusively breastfed at hospital discharge compared to newborns who were not admitted (AOR 0.02; 95% CI 0.02, 0.03). None of the tested interaction terms were significant (Table 2).

The study identified several factors associated with nonexclusive breastfeeding at hospital discharge: maternal higher education compared to secondary education or less, caesarean delivery compared to vaginal delivery, birth weight < 2500 g compared to 3000–3499 g, and admission to NICU.

In line with research from other low- and middle-income countries [15‐19], the current study showed that mothers with higher education were less likely to exclusively breastfeed their newborns at hospital discharge. This finding is the opposite of what is seen in most high-income countries [10‐14]. The association between education and exclusive breastfeeding in Georgia might be related to the work situation of mothers and the prospect of paid maternity leave. The unemployment rate among women in Georgia is low (11.2% in 2018, which is 2.7% lower than men) [40]. According to the Labour code of Georgia, a mother is entitled to 183 days (~ 26 weeks) of paid maternity leave financed by the state with a maximum of 1000 Georgian Lari [41], or around 305 Euro for the whole maternity leave. The legislation also states that employers and employees can agree on additional benefits [41]. However, for higher educated mothers with a well-paid job and for those without extra benefits, the compensation may be insufficient. If the mother returns to work early for economical or other reasons, exclusive breastfeeding in the postpartum period would probably not be a top priority.

Newborns born by caesarean delivery were less likely to be exclusively breastfed at hospital discharge compared to newborns born by vaginal or assisted vaginal delivery. This finding is in accordance with other studies [9, 21‐23]. The proportion of caesarean deliveries is noticeably higher in Georgia (43.5% in 2016) [33] compared to the mean in European countries (25.2% in 2010) [42]. Considering this high rate, the negative association with early postpartum exclusive breastfeeding is of particular interest. Indeed, excessive rates of caesarean delivery is associated with several short- and long-term health risks for both the mother and newborn [43], and lower odds of exclusive breastfeeding at hospital discharge adds to the list of health risks. One of the Ten steps to Successful Breastfeeding in the BFHI is immediate skin-to-skin contact and early initiation of breastfeeding [44]. Without the designation of baby-friendly hospitals, Georgian hospitals and maternity wards may not pay enough attention to the early initiation of breastfeeding after a caesarean delivery. One systematic review suggested that adequate breastfeeding support after a caesarean delivery reduces the negative association between caesarean delivery and early initiation of breastfeeding entirely [45]. Another review found that, among mothers that successfully initiated breastfeeding, there was no difference in exclusive breastfeeding between babies delivered by caesarean and those born by vaginal delivery at six months [23]. These findings indicate that a supportive breastfeeding environment after a caesarean delivery, with proper postoperative pain management, could substantially improve the rates of exclusive breastfeeding at hospital discharge and later in the postpartum period. The Georgian Ministry of Health is addressing the high caesarean rates, and hospitals are under scrutiny to reduce their rates. However, it is too soon to evaluate the results of this intervention.

Newborn admission to a NICU had a large negative impact on exclusive breastfeeding at hospital discharge, a finding that is in line with other studies [46, 47]. This is an expected finding, as exclusive breastfeeding often cannot be prioritized in an intensive care setting. However, although the present study excluded premature newborns, the association between NICU admission and exclusive breastfeeding at hospital discharge was very strong (AOR 0.02; 95% CI 0.02, 0.03). Additionally, newborns with a birthweight < 2500 g were less likely to be exclusively breastfed at hospital discharge compared to newborns weighing 3000–3499 g. Low birthweight has been associated with lower odds of exclusive breastfeeding in previous studies as well [48]. An expansion of the BFHI has been developed for use in NICUs, where the Ten Steps to Successful Breastfeeding are adapted for preterm and sick newborns. The steps include early and prolonged mother-newborn skin-to-skin contact, also known as Kangaroo Mother Care, and support of early breastfeeding, with newborn physiological stability as the only criterion,- not newborn age, weight, or other criteria [49]. This is important, as early initiation of breastfeeding reduces the risk of neonatal mortality, also among low-birthweight newborns [50]. The BFHI expansion to NICUs recommends rooming-in, where mothers and newborns stay together in the NICU, as a step to facilitate continuous breastfeeding [49]. Rooming-in at NICUs may not be feasible in all settings, but the mother should then get the opportunity to stay close to the NICU.

The absolute majority of the term newborns in the study were exclusively breastfed at discharge (92.3%). The high proportion raises concern of reporting bias. Nurses and midwives should record newborn feeding twice a day, but we cannot be sure that all healthcare personnel follow the same procedures, with the risk of misclassification by type of newborn feeding at discharge. However, before applying the exclusion criteria, 85.0% of all newborns were exclusively breastfed at hospital discharge. Prevalence figures of exclusive breastfeeding at hospital discharge in other countries vary substantially, from 61.6% in Canada (only term newborns) [9], 82.7% in rural Western Australia [47], 86.9 -93.1% in the Czech Republic [51], to 93.5% in rural China (only healthy singletons) [52]. The prevalence figures from other countries support the plausibility of the high proportion of exclusive breastfeeding we observed in Georgia. Although it is previously reported a lower prevalence of breastfeeding in Georgia [29, 31, 32], the data are 10–15 years old and based on surveys. Furthermore, the data are not directly comparable as the surveys display the initiation and duration of exclusive breastfeeding. The rate of exclusive breastfeeding is likely much lower at 4–6 months, because the majority of mothers have returned to work or studies. A pilot follow-up program (up to 6 months) in the GBR was launched in 2019 in the region of Ajara in West Georgia. When this program is expanded, we will be able to conduct a follow-up study on the duration and potential obstacles of breastfeeding after hospital discharge.

The GBR was launched in 2016, and it takes time to establish good practices for proper reporting [33]. This is evident for some variables, for instance maternal intrapartum complications. Even though the variable merged several complications the mother may experience during delivery, the prevalence of these complications was low (total 4.9%), indicating a considerable underreporting of complications in the GBR. This underreporting can mask a potential effect of maternal intrapartum complications on exclusive breastfeeding at hospital discharge. Additionally, data are entered into the GBR by maternity wards all over the country, indicating that the healthcare personnel have different training in how to assess certain maternal and newborn conditions. The variable maternal intrapartum complications are particularly exposed, because many of the complications included in the variable are based on the judgement of the healthcare personnel attending the birth. Registry data is also vulnerable to data entry errors. Both of these problems could potentially introduce bias, but we do not know to what extent. Data on the validity of the variables in the GBR are still not available, hence the findings should be interpreted with some caution.

To the authors’ knowledge, this is the first time exclusive breastfeeding at hospital discharge have been studied in Georgia, and certainly for the first time using national birth registry data. The study identified several factors associated with nonexclusive breastfeeding at hospital discharge in Georgia in term newborns: maternal higher education, caesarean delivery, low birthweight and admission to NICU after delivery. Potential steps to increase rates of exclusive breastfeeding are a re-introduction of the Baby-Friendly Hospital Initiative, reducing the high caesarean delivery rates in Georgia, as well as strengthening the length and economical support of the maternity leave. These findings will be valuable for national health authorities when setting new priorities in maternal and child health, as well as for non-governmental organizations working with breastfeeding dyads. Hopefully, the findings will increase awareness about breastfeeding in maternity wards and hospitals all over Georgia. The results could have validity in countries with high rates of caesarean delivery and similar health system structures as in Georgia.