The present study highlights that home deliveries are not only common in rural areas but also in urban areas where maternity services are relatively easily accessible. An earlier study from rural Nepal reported that 93% of the deliveries took place at home [
9]. The proportion of home deliveries in the present study is similar to that reported from earlier studies in Kathmandu and its surrounding areas. These studies reported that the proportion of home deliveries increased, the farther one gets from urban areas [
6,
7].
Interestingly, some findings of the present study are similar to the previous study from rural Nepal [
9]. It was surprising that skilled attendance of government health workers or traditional birth attendants, use of CHDK and hygiene practices during delivery was low in urban areas also. Practices like heating the birth place, applying mustard oil to the stump of umbilical cord, and bathing the baby soon after birth were common in urban areas. Early initiation of breast-feeding, use of prelacteal feeds and breast-feeding from another woman are also common practices prevalent in urban areas.
Attendance during delivery
Most deliveries took place either in a separate room or some place inside the house which is similar to the report from an earlier study [
9]. An earlier study has highlighted that cattle-shed deliveries were contributing to higher rates of infant mortality in the remote rural areas of Nepal [
18]. Such practices were not reported in our study. Studies from rural areas have underscored the role of mother-in-law for assistance during the delivery and care of newborn [
9,
10]. But in our study, mother-in-law was present during delivery in only a small proportion (5%) of home deliveries. More than half of the deliveries were attended by neighbors. Such a difference may be due to demographic structure of the urban population in which many families may be economic migrants and nuclear families. Earlier studies have confirmed the extremely low presence of skilled government health staff or traditional birth attendants during delivery in rural areas of Nepal [
9,
10,
19,
20]. Maternal and child health workers who are identified as key birth attendants by the policy makers were not present at delivery in our study either. This study highlights that skilled attendance at home deliveries is very low in urban areas also. Previous studies found that about 15% of the mothers had delivered alone at home [
9,
19,
20]. This may emphasise the low status of women in the society and the gender inequities in health. For many of these urban families pregnancy and the process of childbirth may not be a concern or priority. It takes huge efforts to change this tradition of home deliveries and lack of skilled attendance during delivery in home setting. There is an ongoing debate about reinforcing home-based birthing strategies with skilled attendants in developing countries [
21]. A recent study from Nepal suggested that such a strategy might cost a substantial amount [
22]. Hence there is a need for research comparing the feasibility, cost-effectiveness, acceptability, and equity implications of skilled home-based and facility-based obstetric care [
21].
Hygiene and thermal control
A study from India has reported that Infection accounts for up to 40% of neonatal deaths [
24]. Therefore the WHO emphasizes on five cleans during the delivery. The 'five cleans' are: a clean place; a clean surface; clean hands; clean cord and dressing; and a clean tie. In our study only one-third of the attendants had washed their hands before delivery which is less than previous reports from rural area [
9]. CHDK was used only in 19% of the deliveries which is higher than earlier reports [
9,
19,
23]. A qualitative study from rural areas of Nepal reported that despite perceived usefulness and awareness, the use of CHDK was low, and the common reasons cited for their non-use were lack of awareness about the kit or difficulty in procuring a kit locally. The study also reported that CHDK had a limited influence on general hygiene practices [
23]. In our study, despite the low usage of CHDK, new/boiled razor blade was used to cut the cord in a majority (90.4%) of deliveries. This practice is encouraging as compared to practice in rural areas where sickle or wooden knife was used in nearly one-third of deliveries and old/unboiled blade in 23% [
9]. This practice was complemented by leaving umbilical stump undressed which is similar to the practice in rural areas [
9]. The practice of applying unsterile substances like oil or ghee is more important risk factor than the means of cutting the cord as reported in earlier studies [
25,
26]. The common substances applied to the cord were mustard oil, turmeric and disinfectants. This practice was similar to the reports from earlier studies from rural Nepal and urban settlements of Karachi, Pakistan [
9,
11,
26].
The WHO has focused on thermal control of newborn in the essential newborn care [
27]. Previous studies from Nepal, India and Bangladesh have reported on health beliefs about pregnancy and childbirth. The common view is that pregnancy is a 'hot' state and postpartum is a 'cold' one [
28‐
30]. Neonatal hypothermia in Nepal has been described earlier [
31]. In our study, we found that in 46% of the instances birth place was heated during or throughout the delivery. This proportion is less than that observed in a rural study [
9]. The practice of waiting for the placenta to deliver before cutting the umbilical cord was observed in 64% of the deliveries, a rate similar to that reported in a qualitative study. [
28]. This practice delays immediate wrapping of the baby. This was further compounded by bathing the baby soon after birth which seems to be a universal practice. In our study, 96% of the newborns were given a bath within one hour after delivery. Similar practices have been reported earlier [
9,
11].
Application/massage of mustard oil to the newborn is a well established practice in Nepal [
26]. In our study, 60% of the newborns received a mustard oil massage soon after birth as in an urban population of Pakistan [
11]. Recently there have been increasing concerns about this traditional practice of massage with mustard oil as it is thought to predispose the newborn to the risk of hypothermia. Earlier studies have raised concerns about the traditional practices of applying mustard oil to the umbilical cord stump and mustard oil massage after delivery [
26,
32].
Infant feeding
In our study the only traditional newborn care practice which seems to be healthy and encouraging is breast-feeding. As reported in the earlier studies, rates of initiation and exclusive breast-feeding are high [
9,
33‐
35]. However, practices like prelacteal feeding and discarding colostrum which still persist in urban areas are a cause of concern. Qualitative studies suggest that the traditional practice is to give a taste of non-breast milk food and usually only once [
28]. In our study, 15% of the newborns received a prelacteal feed which is similar to that reported previously [
9,
33]. In our study, only ten percent of the mothers discarded colostrum which was in contrast to 40% in rural area [
9]. Use of formula feeds was minimal and feeding with bottle and nipple almost non-existent in Nepal [
33]. A recent qualitative study from rural Nepal reported that grand mothers held colostrum in high regard did not use prelacteal feeds and also supported early initiation of breast-feeding [
34]. These findings have positive implications on child nutrition.
Reasons for delivering at home
Studies carried out in Kathmandu and its surrounding areas have reported socioeconomic status and multiparity as strong predictors of the place of delivery. In our study the reasons for planned home deliveries were related to 'preference' for home delivery and perception of home deliveries as 'easy' and 'convenient' and experience of previous home deliveries. For unplanned home deliveries the reasons cited were 'precipitate labor' and 'lack of transport' and 'lack of escort' during labor. Similar findings were reported from an earlier study in Kathmandu [
6]. In urban areas there is a mix of traditional families and recent economic immigrant families. In rural areas women have a strong cultural preference for home deliveries because institutional deliveries are inaccessible. This could be the reason for women indicating 'preference' as the reason for delivering at home. The decision making process in the family about the place of delivery is also an important aspect about the reasons for home deliveries. We could not explore the details of this aspect in our study. Therefore we are planning to undertake an in-depth qualitative study to explore the reasons for delivering at home.
The findings of our study suggest that easy access to maternity services may not be enough to ensure the use of such services. Lack of utilization may be influenced by income, education and cultural beliefs. In Pokhara, institutional delivery facilities are available at the Western Regional Hospital, Manipal Teaching Hospital of MCOMS and a few private hospitals. Since the services at all these facilities have to be paid for, financial constraints may be the main reason for not using these facilities. A large section of this urban population may be recent economic migrants from rural areas. This may be the reason for urban-rural similarities observed in home delivery and newborn care practices in our study.
In our study, 70% of the home deliveries were unplanned. These women would have sought institutional delivery if an ambulance service or local facility for delivery was made available. In this respect it may be worth investing on satellite maternity services run by midwives. Mothers might prefer to utilise such local and user-friendly services than a tertiary care hospital. Also the mothers need to have information about how to access a trained traditional birth attendant or a midwife during the delivery.
The National Safe Motherhood Program of Nepal emphasizes the provision of round-the-clock emergency obstetric services including transport and financial assistance. The program has recently implemented the scheme of financial incentive for the mothers who choose institutional delivery [
4]. In our study, one-third of the mothers had not gone for antenatal checkups and only a third of them received two doses of tetanus toxoid during their previous pregnancy. There may be many cultural constraints for use of maternity services e.g. decision of the husband or mother-in-law which often over-rides that of the mother. The reasons for low uptake of maternity services in the urban population may be due to socioeconomic and cultural factors [
6,
7]. Therefore interventions should address not only the medical problems but also need to deal with wider social problems. Interventions should also focus to improve the status of women in society including increasing female literacy and empowerment to tackle the maternal health problems [
36]. A recent qualitative study carried out in Kathmandu reported that attitudes of pregnant women, husbands and service providers were favorable towards encouraging greater male involvement in maternal health services [
37]. Further studies are needed about quality of the available maternity services and cultural beliefs about pregnancy and childbirth.
The present study may have both selection and information bias. Since our survey was carried out in immunisation clinics, selection bias cannot be ruled out. We might have missed to interview those mothers who delivered at home and did not attend immunisation clinics. However, the sample of mothers interviewed may be representative of the urban population since the immunisation coverage is more than 90% in urban areas [
4,
20]. The high infant mortality rate in study area means that ten percent of the children who were born at home may have not reached their first birthday. Therefore, ten percent of the mothers did not attend the immunisation clinics. We included only those deliveries that took place within one year in order to avoid recall bias over a longer period of time. However, we cannot rule out some amount of recall and reporting bias. Hence the interpretation and generalisability our findings may be limited.
In our study, all the women who had delivered at home agreed to participate in the study. The interviewers identified themselves as independent researchers rather than as a part of health service team present in the immunisation clinics. The respondents were assured of providing health services regardless of their decision to participate in the interview and verbal consent was sought. All the women agreed to participate in the study after receiving such an assurance. The possibility that the women considered interviewers as a part of health service team cannot be ruled out. The mean monthly family income of the respondents was approximately US$90. But the annual per capita income of Nepal is US$ 289 [
12]. The reported monthly family income of this urban population is higher than national average. The families in this urban population may be more affluent than an average Nepali family since Nepal is an agrarian economy. In Nepal, majority of the population is mainly dependent on subsistence farming with seasonal migration to India or other countries.
Despite the above-mentioned limitations, our study has obtained important information about home delivery and newborn care practices and reasons for delivering at home. This information has many policy implications about the ongoing safe motherhood and child survival programmes in Nepal. There need to be more focus on the skilled attendance and hygiene during delivery and the use of CHDK in urban population also. Some high-risk newborn care practices like delayed wrapping, immediate bathing, mustard oil massage, applying mustard oil to the cord; prelacteal feeding and discarding colostrum need more attention. This information will assist in planning public health interventions to change the behaviour. Expanding skilled attendance during delivery is an important issue since these urban women 'prefer' home deliveries and home deliveries are perceived as 'easy' and 'convenient'.