Acceptability of evidence-based neonatal care practices in rural Uganda – implications for programming

Few women reported attending ANC four times during pregnancy. The reasons given for this included knowledge barriers and service delivery gaps; cultural, traditional beliefs and practices; and financial constraints. These are summarized in table 1 according to the three delays model

In Busoga, ANC attendance is called "okunhwa obulezi", literally translated "drinking medicine". Thus, the community associates ANC with being "ill". Consequently, only women who feel "ill" do attend ANC and when they do so they expect to get a lot of medication. For instance, intermittent presumptive treatment (IPT) of malaria with fansidar is one of the key interventions to reduce malaria in pregnancy and improve maternal and fetal outcomes. However, when women get only three tablets of fansidar (the appropriate treatment), they get frustrated, as was reported in one FGD:

TBAs are perceived as effective care-givers since they provide herbal medicine, and as more mature providers with 'better' personalized care compared to health workers. Furthermore, TBAs are easily accessible to the community day and night. In addition, it was believed that herbs from TBAs give magical effects which modern medicine may not be able to provide.

Herbs are believed "to make pelvic bones flexible at the time of delivery" so as to help women avoid the risk of caesarean birth, which they associate with delivery from hospitals. Other cases where herbs are used included malaria and syphilis treatment.

Financial constraints are a critical barrier for women who know the importance of four ANC visits, and some women attend ANC only once to get a card. As one young woman in a focus group put it:

Despite these challenges, respondents said that attending ANC four times during pregnancy is acceptable if the barriers are removed.

Birth preparedness (BP) promotes active preparation and decision-making for births, including pregnancy/postpartum periods, by pregnant women and their families. Couples are sensitized about BP in ANC but implementation is poor due to limited knowledge, financial ability, and strong traditional beliefs. Participants revealed to us that men are a key barrier to BP since women rely on them for decisions and support including finances for pregnancy related care.

In an effort to improve BP, the Uganda Ministry of Health tries to provide maama kits for pregnant women. However, these are often out of stock in health units, thus requiring couples to buy them. Figure 1 shows a picture of the maama kit – a birth-preparedness package promoted in Uganda.

We group current birth preparedness practices in this region into four categories as shown in Figure 2:

Participants were able to identify some maternal danger signs including high fever (omusudha), severe pre-eclampsia (amakiro), hemorrhage (okuva omusayi), abdominal pain (okulumwa munda), and swelling of feet (ebigere okuzimba). Amakiro or severe pre-eclampsia was attributed to a partner being promiscuous. Other problems which respondents identified as danger signs but not necessarily medically recognized as such included syphilis (kabotongo) and vomiting (okusesema). We found that in the community, danger signs are usually first treated with herbs, and they only seek medical care when the condition becomes worse.

Given these barriers in accessing the formal health system, the possibility of getting services on credit from TBAs to offset delivery expenses was considered an easier option.

Respondents were not aware of the need for postnatal health care attendance except for immunization of children. In one of the FGDs, postnatal care as part of maternal and newborn care was considered something new:

The newborn care practices which were deemed acceptable to the community included: maintenance of warmth and cleanliness, exclusive breast feeding and skin-to-skin contact at birth. However, the recommended evidence based practices of delayed bathing and putting nothing on the umbilical cord met strong objections from both the community and health workers, as babies are believed to be born 'dirty', with a 'bad smell', and as such needed to bathe so as not to feel "uncomfortable". Early bathing of newborns is a common practice among health workers, TBAs and the community.

However, women were agreeable to the suggestion to wipe the baby with a wet warm clean cloth instead of early bathing.

For both facility and home births, the umbilical cord is usually cut with some form of 'sterile' equipment, usually a new razorblade. Cord-care practices included application of substances such as baby powder, spirit, herbs, soapy water, and salty water. These substances are believed to help the cord to dry/heal faster to enable women get back to their routine chores early. The fast healing of the cord is also believed to stop the pains that women feel following birth of a newborn baby. We also found that a baby whose cord has not fallen off is traditionally kept inside the house.

Interestingly, most newborn danger signs were well known to all participants including child-minders. When they were asked which signs show that a baby's health may be in danger, participants mentioned inability to suckle, weak or abnormal cry, lethargy, vomiting, fever, difficult breathing and convulsions. Only jaundice and umbilical redness were not recognized as newborn danger signs. However, due to the already outlined barriers, care seeking for newborn danger signs was reported to be poor. Table 2 outlines the perceived acceptability of recommended newborn care practices and their associated barriers.

All participants welcomed the proposed intervention to improve care of mothers and newborn babies through home visits to households by community volunteers during pregnancy and in the first week after delivery.

Respondents mentioned the following as the desired qualities of community volunteers who would visit homes: being a woman who is a permanent resident in the area; literate and well behaved and respected in the community; ability to walk long distances; good tempered; knowledgeable; punctual; a good listener; confident; experienced in parenting; properly married, and a kind hearted person.

Attending ANC four times is one of the major WHO recommended interventions for pregnant women, as it is associated with improved maternal and neonatal outcomes [26, 27]. However, we found this to be rare in practice, but was said to be acceptable to the community under study if access barriers were minimized. We believe that in this setting an intervention will need to discuss at least three issues on ANC: the perception of ANC as being for the ill pregnant women; the role of TBAs; and attendance 'to get an ANC card'. In Uganda, it is a common reason for women to attend ANC to get a card as nurses and midwives usually compliment women who present these cards at delivery as good mothers who have been in contact with the health facility during ANC [28].

The low attendance of ANC four times during pregnancy may be partly explained by the local perception among this Basoga ethnic group of ANC as "okunhwa obulezi", literally translated "drinking medicine". As such, the community perceives attending ANC early or more than once to be for pregnant women who are 'unwell' and only the 'ill' women should attend ANC. In other words, women do not see the need to seek care during pregnancy except when they are unwell or to get a card. This is contrary to the view of some medical professionals who perceive that every pregnancy requires medical care and hence women must attend ANC in order to have safe deliveries. Yet it is generally known that pregnancy and childbirth are biological and physiological events which are embedded in a social and cultural setting [29]. Therefore, for ANC to be effective, it needs to be perceived from both medical and social/cultural perspectives [29]. In addition, there is a need to redefine ANC from "drinking medicine" to caring for a healthy pregnancy, while keeping in mind that comprehension and interpretation of health messages are different and new information interacts with previous ideas into "syncretic models" [30].

Other important barriers to seeking care that need to be addressed include transport and reports of requests for illegal payment of health workers by mothers. In addition, men will need to be targeted as key allies in improving ANC attendance. Current strategies in Uganda tend to focus mainly on women, yet it is men who provide financial support and also make decisions to seek care.

Our findings on ANC are supported by a recent systematic review of factors affecting ANC in low-income countries by Simkhada et.al. which found that the key factors affecting antenatal care uptake include: cultural beliefs and ideas about pregnancy, maternal and husband's education, marital status, availability and cost of ANC, household income, women's employment, media exposure and having a history of obstetric complications [31].

We believe that a community-based approach to dialogue with parents and communities to increase "demand" for ANC [32] coupled with a health system intervention to improve the "supply" of quality, affordable ANC will improve ANC attendance in Uganda, as it did in Nepal where women's participatory groups helped reduce neonatal mortality, increased supervised deliveries and ANC attendance [12].

We found that most women prefer to have a supervised delivery, but in practice often do not manage, mainly because having a supervised delivery is too costly for them to afford. Although user-fees were officially abolished in Uganda since 2001, in practice couples need to incur costs for drugs and other medical supplies. These high costs incurred by couples may be one of the major explanations as to why supervised deliveries in Uganda have remained low for over a decade despite high ANC attendance rates and improved physical access to health facilities. In Uganda, 92% of pregnant women attend ANC at least once, and 72% of the population now lives within a distance of 5 km from the nearest health facility [19].

Based on these findings, we suggest that the issue of BP needs to be revisited with the aim of reducing the financial burden on couples [33]. In addition, mothers told us that most labour pains start at night when there is no money and adequate transport to get to health facilities which provide delivery care, implying the need to consider locally available transport as well as waiting shelters for pregnant women in under-served areas [33‐37] if supervised delivery care is to be increased.

Our findings show that postnatal care was barely present in practice and was not promoted by service providers or through the community outreach activities. However, most recommended care practices for the newborn were said to be acceptable to both health providers and community members. Notable exceptions were delayed bathing and putting nothing on the umbilical cord, which were relatively unacceptable. Early bathing within the first 8–12 hours is common and a range of different substances are put on the cord.

Therefore, in order for effective newborn interventions to take effect, the acceptable recommended practices will need strengthening. However, special efforts including compromise practices will need to be defined and promoted for those recommended practices which are not easily accepted. Thus, in order to keep babies warm at birth, the recommended practice of delayed bathing should be communicated better, or a compromise of wiping the baby with a clean warm piece of cloth will need to be promoted along with skin-to-skin care. Similarly, there is need to define some harmless practices for cord-care. We find that this is one of the key areas where a community based approach has a major contribution since most births occur outside health units and facility attendance is not deemed feasible in the first week after birth due to cultural, transport and financial reasons.

While we selected mainly respondents who had recently given birth, efforts were made to encourage them to give their actual experiences, for some recommended practices respondents probably gave us general attitudes prevalent in their communities. Another possible limitation is that the study focused on neonatal care practices and did not target maternal issues other than when they were thought to impact on the newborn. While this was a study in one region only, we believe that some of our findings are generalizable to other parts of Uganda. However, we believe the methodology used to elicit the findings can and should be used across regions and countries in the formative stages of intervention development.