Factors associated with adoption of beneficial newborn care practices in rural Eastern Uganda: a cross-sectional study

In 2005, Lawn et al. estimated that about 4 million neonatal deaths occur every year in the first 4 weeks of life, contributing 38 % of the estimated 10.5 million deaths that occur among children under-5 years of age worldwide. The highest numbers of neonatal deaths occur in sub-Saharan Africa. Few countries (with some exceptions) have made progress in reducing such deaths in the past 10–15 years [1]. In 2013, neonatal dates still accounted for 41.6 % of under-5 mortality. The rising numbers of births especially in sub-Saharan Africa have lead to more 1.4 million more child deaths [2]. The precursors of these neonatal deaths have partly been traced back to lack of practice of beneficial newborn care practices and attributable to largely infections, intrapartum birth conditions, and preterm births [3‐5]. According to WHO, Save The Children and UNICEF [6‐8], beneficial newborn care practices are described as interventions such as immunization of baby (with birth dose of oral polio and Hepatitis B vaccines) after birth [9], assessment for birth weight, gestational age, congenital defects and signs of newborn illness [6], good cord care practices such as use of antiseptics to dry the umbilical care [10, 11], skin to skin care and maintaining warmth [11‐13], as well as early and exclusive breastfeeding [14‐16]. Studies have shown that beneficial newborn care practices at population level can save many newborn lives [11, 17, 18]. In addition, these beneficial newborn care interventions are simple, cost effective, and acceptable: a single skilled birth attendant caring can effectively provide many of them for the mother and the newborn and sometimes by the mother herself if taught [1, 16, 19‐22].

In Uganda, studies have demonstrated low levels of newborn care knowledge among certain cadres of health workers-particularly midwives: a gap in the promotion of beneficial newborn care practices among newborn mothers and caregivers [23]. In addition, disparities in equitable access to health care are likely to hinder promotion of beneficial newborn care practices among mothers and caregivers since 52 % of deliveries do not take place in health facilities [24, 25]. This creates a gap in delivery of these interventions at health system level in addition to the existing social inequity to healthcare access among the rural poor [3, 16, 26, 27]. The government of Uganda, like other high burden states, has undertaken several steps including policy action to prioritize newborn health, improving midwife coverage and health system strengthening to help improve equitable access to healthcare [28‐30]. Additionally, she has defined a minimum package of health services for all levels of healthcare, abolished user fees, and ensured the existence of a functional health facility within reach (within 5 km) of the majority of households [6]. Furthermore, the Ugandan Ministry of Health has strengthened community support systems through use of community health workers such as village health teams (VHTs) to create demand for services and improve adoption of beneficial newborn care practices among caregivers and newborn mothers, especially in the early newborn period [31, 32].

Despite these efforts, access to newborn health care services remains inadequate in most parts of the country, and adoption of beneficial newborn care practices remains largely sub-optimal [3, 33]. Few studies, if any, have explored the factors associated with adoption of beneficial newborn care practices among mothers [3, 23], thereby presenting a missed opportunity for improving neonatal care in Uganda. We assessed prevailing beneficial newborn care practices among mothers resident in Buyende and Luuka health sub-district in eastern Uganda; and assessed the factors associated with adoption of these practices in the two health sub-districts. Our study findings will help to inform policy on implementation and rollout of the newborn care aspects of the child survival strategy in the light of high neonatal mortality in Uganda and to track progress of newborn care interventions [24, 34‐37].

A total of 1,616 mothers were enrolled into this study. Of these, 797 (49.3) were enrolled from Luuka while 819 (50.7) were enrolled from Buyende health sub-district (Table 1). Of the 1,616 mothers enrolled, 622 (38.4 %) were aged 25–34; 1,096 (67.8 %) had primary education, 1,472 (91.1 %) were married; while 1,357 (84 %) were laborers or peasants. Forty four per cent (726) of the mothers were of parity 5+; 1,583 (98 %) reported antenatal care attendance; 979 (61.8 %) attended ANC for 3-4 times while 889 (56.2 %) had their first ANC visit in trimester 2. Skilled delivery was reported among 1,183 (73.2 %) of mothers while 1,569 (97.1 %) of mothers had a normal delivery. Distance to place of delivery was not known among majority of mothers 1,154 (71.4 %).

Table 2 shows the distribution of selected newborn care practices in Luuka and Buyende health sub-districts. Overall, adoption of all beneficial newborn care practices was 11.7 % (189 of 1,616); significantly lower in Luuka (9.4 %, 75 of 797) than in Buyende health sub-district (13.9 %, 114 of 819; Odds Ratio [OR] = 0.64, 95 % Confidence Interval (95 % CI): 0.46, 0.87). Good cord care (83.6 % in Luuka; 95 % in Buyende health sub-district) and immunization of newborn (80.7 % in Luuka health sub-district; 82.5 % in Buyende health sub-district) were the most prevalent newborn care practices reported among mothers. Weighing the baby after birth was fairly practiced in both Luuka (68 %) and Buyende health sub-district (59.8 %). The proportion of mothers reporting good optimal thermal care was lower in both Luuka (40.3 %) and Buyende health sub-district (38.6 %) as was the proportion of mothers reporting good feeding practice (47.2 % in Luuka; 50.6 % in Buyende health sub-district).

Table 3 shows bivariate analysis of maternal characteristics and adoption of beneficial newborn care practices. At the bivariate analysis, having primary education for her husband (p = 0.002), being a laborer or peasant (p = 0.028) and having a first ANC visit in the second trimester (p = 0.047) and residing in Buyende health sub-district district (p = 0.005), were negatively associated with adoption of all beneficial newborn care practices. On the other hand, Secondary or higher education for the mother (p = 0.03), making 3-4 (p < 0.001^*) or 5+ ANC visits’ (p = 0.003), delivering under skilled birth attendants (p < 0.001^**), delivery mode (p = 0.036), belonging to middle (p = 0.002) or higher (p = 0.002) socio-economic status, were significantly positively associated with reporting adoption of all beneficial newborn care practices. Mother’s age category, marital status, having first ANC visit at trimester 3, parity, delivery mode, and distance to the health facility where the mother delivered from were not significantly associated with reporting adoption of all beneficial newborn care practices.

Additionally, Table 3 shows unadjusted and adjusted OR and 95 % CI associated with adoption of all beneficial newborn care practices among mothers who reported at least one ANC visit at their last pregnancy. Results show that higher ANC visits (3-4 vs. 1-2: Adj. OR = 1.69, 95%CI = 1.13, 2.52); skilled delivery (Adj. OR = 2.66, 95%CI = 1.92, 3.69); middle-level socio-economic status (middle vs. low: Adj. OR = 1.57, 95 % CI = 1.09, 2.26) were positively significantly associated with adoption of all beneficial newborn care practices. Residence in Buyende health sub-district (Adj. OR = 0.71, 95 % CI: 0.51, 0.99) was significantly associated with poor adoption of all beneficial newborn care practices.

Our study of adoption of beneficial newborn care practices among mothers in Luuka and Buyende health sub-districts found that only 12 % of mothers in both health sub-districts adopted all beneficial newborn care practices. The low adoption of all beneficial newborn care practices may be related to poor quality of services especially at ANC [41‐44] and deep-rooted cultural practices and rituals for newborns [3, 45, 46]. These findings suggest a need for improved sensitization of mothers about the benefits of adopting all beneficial newborn care practices when they come for ANC and the need for interventions that address deep-rooted cultural practices and rituals that still inhibit adoption of newborn care practices [17, 47].

We found that adoption of beneficial newborn practices was associated with skilled delivery and number of ANC attendances, with those attending ANC for 3-4 times more likely to report adoption of all beneficial newborn care practices than those who attended once or twice. These findings suggest that attending ANC for three or four times coupled with increased use of skilled delivery can improve adoption of beneficial newborn care practices [6, 18]. However, these findings should be interpreted in light of the fact that despite nearly three-quarters of mothers delivering under skilled attendance and 62 % of mothers attending ANC services for 3-4 times, only a small proportion of mothers reported adoption of all beneficial newborn care practices. This could partly be due to the fact that the health workers themselves who should have educated mothers about the importance of adoption of beneficial newborn care practices have limited knowledge of these practices [20, 23, 48] and partly due to the continued existence of harmful newborn care practices such as bathing the baby immediately after birth [49], delayed initiation of breastfeeding [50] and putting powder, salty water or lizard droppings on the umbilical cord [3]. Evidence suggests that these practices continue to be practiced even among mothers who have attended ANC and/or delivered under skilled attendance. Note: We acknowledge that some beneficial practices such as optimal thermal care, weighing the baby, and immunization are likely to be done by health facility staff and are dependant of a facility to provide them.

Prior studies in rural Uganda have found that delayed bathing of the baby and putting nothing on the umbilical cord are not acceptable practices among mothers and health care providers [20, 48]. In India, Shah & Dwivedi [49] found that uptake of beneficial newborn care practices is largely hampered by cultural practices, including one known as “Chatti Puja” – that restricts wrapping of the baby in clothes or exposing them to sunlight until the seventh day after birth [49]. In Afghanistan, Newbrander et al. [50] found that initiation of breastfeeding may be delayed until up to 3 days after birth – or until a woman has had her first bath after delivery [50]. These findings suggest a need for innovative interventions (e.g. use of mentor-mothers) to address barriers that inhibit adoption of beneficial newborn practices among mothers on the one hand and interventions that target health workers (e.g. through on-the-job, hands-on training) on the other hand to increase their knowledge about the importance of beneficial newborn care practices in improving child survival.

We found that mothers of middle socio-economic status were more likely to practice beneficial newborn care practices compared to those of low socio-economic status. This is in contrast with an earlier study in the same region that reported no significant difference in the adoption of newborn care practices by women’s socio-economic status [1]. Women of higher socio-economic status may be more likely to practice beneficial newborn cares practices because: i) they can easily access health interventions at fixed health facilities which tend to be inequitably distributed, ii) they are able to access skilled delivery, and iii) they can make three or four ANC visits compared to women of low socio-economic status [51, 52]. The mixed findings we see among women of middle socio-economic status compared to low socio-economic status and high compared to low socio-economic status could be because women of high socio-economic status are prone to work pressures, unfavorable work conditions and socio-cultural factors such as maternal and significant other’s beliefs about uptake of recommended newborn care practices such as exclusive breastfeeding [53‐56], thus making them less likely to adopt beneficial newborn care practices. Furthermore, the effects of urbanization, maternal education, and socioeconomic status act through the intervening variables of sociocultural factors, health services, employment status of women, availability of breast milk substitutes and use of powder with a resultant effect of failure to adopt beneficial newborn care practices [57]. Waiswa et al. [3] have described this phenomenon among women of high socio-economic status as “modernistic”. On the other hand, the deep-rooted cultural beliefs and practices among women of low socio-economic status are because of lack of relevant health information and access to quality health care. These cultural practices may not be beneficial to newborns [58, 59]. These findings suggest a need for targeting women of low socio-economic class with correct information on the importance of beneficial newborn care practices in influencing a baby’s growth and survival through outreaches or intensified use of community health workers on one hand, and promoting newborn care practices among women of high socio-economic status through quality ANC services at fixed health facilities (both private and public) they attend, improved conditions at work for breastfeeding and supportive policies at national level for uptake of beneficial newborn care practices.

Our study has several limitations. For instance, there is a likelihood that women interviewed for our study did not remember all the beneficial newborn care practices that they practiced when they gave birth to their babies – given that we interviewed mothers who delivered in the past 12 months. It is likely that mothers who delivered exactly 12 months from the time of interview might be less likely to recall all the newborn care practices that they used compared to those who had delivered in a month or so to the time of interview. However, since we did not collect data on the duration between delivery and time of interview, we are unable to assess the extent to which recall bias varied over time. Nevertheless, since women tend to remember what happens to their babies [60‐62], it is unlikely that recall bias affected the results reported in a substantial way.

We found that distance to the facility was not significantly associated with adoption of beneficial newborn care practices at the bivariate analysis level. This finding should be interpreted with caution given that 71.4 % of the mothers did not know the distance to the health facility where they delivered from. Indeed, the association between distance to the facility and adoption of beneficial newborn care practices was assessed in only those mothers who knew the distance to the facility. Studies have shown that women reporting long distances to the health facility are less likely to complete the minimum recommended number of antenatal care visits [63‐65]. These findings suggest that distance to the facility is an important covariate of the adoption of beneficial newborn care practices among mothers. However, since only 28.6 % of mothers knew the distance to the facility, our study is not able to draw any conclusions about the effect of distance on adoption of beneficial newborn care practices; warranting further inquiry on this aspect.

It is also important to note that the proportions of beneficial newborn care practices reported in this paper are based on self-reports of what women did when they gave birth to their babies. Since we could not verify individual newborn care practices, there is a possibility that some practices may not have been practiced as reported. However, given that almost similar proportions of use of newborn care practices were reported in two separate districts, it is very likely that women’s reports of beneficial newborn care practices might reflect the actual use of these practices in the communities. Additionally, there could be residual confounding, which we were unable to assess, explaining the associations we see in the data.

Despite these limitations, our findings are crucial in the implementation of interventions aimed at improving child growth and survival. The Ministry of Health recommends that pregnant women attend ANC up to four times, and that all mothers deliver under the hands of skilled attendants. At the moment, only 57 % of mothers deliver at the hands of skilled attendants, and ANC attendance declines substantially from 95 % at the first visit to 48 % at the 4^th visit [38]. Our findings suggest a need for provision of quality antenatal care, promotion of ANC attendance up to the 4^th visit, quality skilled delivery services and skilled delivery attendance in order to increase the proportion of mothers who can and are able to adopt beneficial newborn care practices for the benefit of their babies’ growth and survival.

Our study shows that adoption of all beneficial newborn care practices was sub-optimal. The independent factors associated with adoption of beneficial newborn care practices were skilled delivery attendance, middle-level socio-economic status and attending ANC for 3-4 visits. These findings suggest a need for improvement of quality of ANC, skilled delivery attendance and the need to target women of both low and high socio-economic status with newborn care health educational messages.