Background
Reduction of neonatal mortality is one of the priority areas in the sustainable development agenda [
1,
2]. To attain the sustainable development newborn target, Uganda has the challenge of reducing neonatal mortality from 27/1000 live births [
3] to 12/1000 live births [
1] by the year 2030. In order to achieve this target, key causes of neonatal death have to be addressed. Newborn infections account for about a quarter of all newborn deaths [
4]. Infections also contribute significantly to deaths from other causes such as prematurity [
5]. The umbilical cord stump, (hereinafter umbilical cord) is one of the major routes of infection in the neonatal period and a significant number of systemic infections are believed to progress from umbilical cord infections [
6].
Proper hygienic care of the umbilical cord is recommended by the World Health Organization to reduce umbilical and systemic newborn infections [
7]. Currently, the Ugandan Ministry of Health recommends dry umbilical cord care to discourage mothers from putting unclean substances on the umbilical cords, and this is in agreement with the World Health Organization (WHO) recommendations [
7]. However, studies done in Uganda show low levels of adoption of the recommended dry umbilical cord care practice, with most authors quoting an uptake less than 50% [
8‐
10]. Dry umbilical cord care seems to conflict with popular beliefs and cultural practices [
11‐
13]. A common practice is to apply various substances on the umbilical cord to hasten umbilical cord separation [
14].
Chlorhexidine, a topical antiseptic [
15], is advocated for use in health facilities as an alternative to dry cord care in areas with high neonatal mortality and where dry cord care is unacceptable by the community [
12]. Application of chlorhexidine to the umbilical cord has been found to reduce the incidence of both newborn infections and newborn deaths in Asian countries [
6,
16‐
18] and is listed as a prioritized intervention in newborn health [
5,
19]. A trial assessing the effect of a single application of chlorhexidine on the risk of infection of the umbilical cord stump (omphalitis) and newborn severe illness is currently underway in Uganda [
20], and it is within this trial that we conducted this acceptability study.
Sekhon et al.
(2017) have defined acceptability of healthcare interventions as a multifaceted construct that reflects the extent to which people delivering or receiving a healthcare intervention consider it appropriate, based on anticipated or experienced cognitive and emotional responses to the intervention [
21]. The authors propose seven aspects to evaluate acceptability; individual’s affective attitude, burden, intervention coherence, ethicality, opportunity costs, perceived effectiveness and self-efficacy [
21].
In this paper we draw upon Sekhon’s framework to explore the acceptability of single dose chlorhexidine solution for umbilical cord care among infant care providers and health workers in the districts of Kampala and Mukono in Central Uganda. With the overall aim to clarify conditions for scale-up we bring attention to social and cultural factors and argue their centrality for the emotional and cognitive responses addressed by Sekhon et al.
Methods
This qualitative study was embedded within the chlorhexidine randomized controlled trial (RCT) designed to assess the efficacy of a one time application of chlorhexidine (4%) on the umbilical cord in reducing the incidence of neonatal severe illness [
20]. The RCT is undertaken at three health centers in or close to Kampala, the capital of Uganda, and recruits newborns on the first day of life. After obtaining informed consent from the mother, research nurses/midwives apply the chlorhexidine solution on the umbilical stump of the newborns that are randomized to receive such care. Mothers are instructed not to put any substance onto the umbilical cord after the chlorhexidine has been applied, and only to wash it with plain water to remove dust and dirt if necessary. The RCT aims to recruit approximately 4700 newborns over a 3-year period. The newborns are followed up for 28 days and are examined on days 1, 3, 7, 14 and 28. The newborns in the control arm receive the standard of care, which is dry cord care with the same instruction not to apply anything on the umbilical cord, and only use plain water when bathing the baby. To the best of our knowledge, the RCT is the first project to introduce chlorhexidine for umbilical cord care in the study area and the participants did not have prior experiences with chlorhexidine.
The qualitative study was conducted between June 2016 and January 2017 in Kampala and Mukono district and had both a health facility component and a community component. The sites were chosen because they were the sites of the CHX randomized controlled trial. The Baganda, who are Bantu people, inhabit the central region of Uganda, and constitute the largest ethnic group in the study area as well as in the country at large [
22]. Kampala district is mainly urban, whereas Mukono district is partly peri-urban but mainly rural. About 30% of people in Mukono district live 5 km or more from the nearest public health facility. A dominant role of husbands in accessing maternal and child health care services has been observed in this region [
13]. Approximately 20% of women over 18 years in Mukono are illiterate [
23] and there a high proportion of girls aged 12–19 years have given birth [
23]. In Kampala, 98% are reported to attend skilled antenatal care with skilled, 94% deliver at a health facility, and 78% have a postnatal check within the first 2 days after birth [
3,
23]. Traditional birth attendants (TBAs) are a popular source of knowledge about maternal and newborn care, but after the Ministry of Health terminated their previous working relationship with traditional birth attendants [
24], their practice is considered illegal and liable to prosecution by local authorities [
25]. Nevertheless TBAs remain popular and the communities protect them and continue to consult them for advice on pregnancy and newborn care [
24]. As a result of their popularity, TBAs in the community where we conducted our study had been trained and incorporated into the village health team. Their main task was health promotion and referring women to the nearby government health center. The main groups of study participants are summarized in Table
1 below.
Table 1
Summary of Study participant’s in the chlorhexidine acceptability study in Central Uganda
| Mothers-18 |
Healthworkers-8 | Mothers - 3 |
TBAs-2 | Study nurses- 1 |
Others-2 |
Participants were chosen purposively, aiming for participants with rich experience in newborn care and for variation in experience and perspectives according to education, caretaker role and age. The health care workers recruited were study nurses who were employed specifically to work on the RCT. In addition, influential traditional birth attendants, a father and a grandmother were included. Two TBAs who also acted as members of the village health team, and therefore responsible for postnatal visits in their community, were given chlorhexidine bottles to use on babies in their communities and then interviewed. A father and a grandmother of a baby who had received chlorhexidine were also interviewed. This was done in order to enrich our data by soliciting views from various participants. All the participants had had an experience with chlorhexidine use for umbilical cord care. The TBA interviews were conducted in the community. We conducted 30 in depth interviews (IDIs) with mothers (18), health workers (8), TBAs (2), a father (1) and a grandmother (1) and 4 FGDs, 3 with mothers and 1 with health workers. We also video recorded the preparation of kyogero; a local herbal mixture composed of a variety of herbs, which are mainly in the form of leaves and stem barks. Kyogero is prepared by boiling these herbs to form a solution.
Two researchers; a social anthropologist and a medical doctor trained in qualitative research data collection; conducted the interviews. We used semi-structured interview guides for the different categories of study participants and topic guides for the different FGDs, which were both used flexibly and modified as need arose in the course of the study [
26]. The interviews lasted between 20 to 80 min. We conducted most interviews in
Luganda, the local language and a few in English mainly for health workers. A moderator and one note taker guided the focus group discussions (FGDs). We used video recording as a method of data collection to provide a detailed description and to increase the trustworthiness of the findings.
All the interviews were audiotaped and notes written down during the interview. The collected data was kept confidential and stored on a password-protected computer. A professional transcribed and translated interviews/FGD that were conducted in a language other than English. The principal investigator, who was present during all the interviews, proofread the transcripts comparing them to the audio recording. Data analysis was a continuous and iterative process guided by qualitative content analysis [
27]. Two other independent researchers looked at the transcripts, categories and the subthemes generated. We used Nvivo 11.0.0 (QRS International, Cambridge, MA) to organize the analysis process. We borrowed upon Sekhon’s acceptability model when presenting our findings.
Discussion
The findings suggest that chlorhexidine use for the umbilical cord is generally well accepted but that there are customary practices like ritual washing of the newborn and preferences like quick detachment of the umbilical cord, which may hinder exclusive use of chlorhexidine for umbilical cord care.
Our participants commented favorably on the physical attributes of chlorhexidine: color, smell and liquid formulation. This is in agreement with a study in Zanzibar, which showed preference in use of the liquid chlorhexidine formulation compared to the gel formulation [
28]. Chlorhexidine use on the umbilical cord has been advocated as a means of discouraging unhygienic practices [
12,
14,
29], especially in areas where dry cord care has been perceived as inappropriate [
11,
12,
14]. In our study, we found that acceptance of chlorhexidine for umbilical cord care did not automatically lead to abandonment of the customary practices related to umbilical cord care. Participants were willing to abandon some of the substances applied directly onto the umbilical cord in favor of chlorhexidine, but were unwilling to abandon the washing of newborns in
kyogero for the perceived benefits to the baby now and later in life
. In effect, chlorhexidine was seen as an addition rather than a substitution to
kyogero use for childcare. One of the reasons our participants were unwilling to abandon traditional substances applied to the umbilical cord, was their belief that these substances quickened umbilical cord separation.
One of the major barriers to chlorhexidine acceptance in our study was the suspicion among some participants and health workers that it delayed umbilical cord separation. Delayed umbilical cord separation following daily chlorhexidine for several days has been objectively assessed in a cluster randomized controlled trial conducted in Bangladesh, which revealed a 50% increase in umbilical cord separation time, when chlorhexidine was used [
30]. A number of authors state that appropriate health communication may ameliorate the perceived negative effect of delayed umbilical cord separation [
14,
30]. The perceived negative impact of prolonged umbilical cord separation may be rooted within deeper socio-cultural meanings attached to the cord, and may not solely be caused by a lack of adequate knowledge concerning chlorhexidine side effects. The findings of negative perceptions among study nurses and midwives who had previously been adequately trained in the chlorhexidine side effect gives credence to our assertion. Participants also reported that prolonged umbilical cord separation increased pressure from community members to resort to the herbal formulations previously used to speed up umbilical cord separation. This may obliterate the beneficial effects of chlorhexidine in reducing umbilical cord infections.
Sekhon et al. have stated that acceptability of health care interventions depends on the the appropriateness of the intervention in terms of anticipated or experienced cognitive and emotional responses [
21]. They state 7 constructs, which include: affective attitude (how an individual feels about the intervention), burden (perceived effort required to participate in the intervention), ethicality (intervention’s goodness of fit with individual’s value system), intervention coherence (participant’s understanding of how intervention works), opportunity costs (benefits, profits and values to be given up to engage in the intervention), perceived effectiveness (perception that intervention has achieved purpose), and self efficacy (participant’s confidence that they can perform the intervention) [
21]. Our participants generally felt good, clear, and unburdened by the application of chlorhexidine. They were also confident of the effectiveness of the intervention, and did not find any major value conflicts in using chlorhexidine. Interestingly, we found that mothers perceived chlorhexidine to be effective in prevention of umbilical colic, a finding that we cannot explain biologically. Possible hypothetical explanations include the reduction in umbilical cord colonization with microorganisms resulting in reduced colic, or a placebo effect resulting from chlorhexidine application to the umbilical cord. Despite mothers having serious opportunity cost challenges in having to forfeit herbal formulations, we argue that the intervention was considered largely acceptable on the individual level. However, this was not adequate to ensure use. We came across incidences where despite cognitive and emotional appropriateness of chlorhexidine as experienced by mothers, the product was not approved until accepted by significant others with decision making power in newborn care. These significant others were mainly older female relatives. Hence, widespread chlorhexidine acceptance may not be achieved unless these older relatives who are the practical caretakers and decision makers approve of it. In addition to Sekhon’s concept of acceptability as cognitive and emotional appropriateness, [
23] we suggest to include social and cultural appropriateness as a condition for acceptability. Even if the social is seen as implicit in the concepts of affective attitude, opportunity costs, and self-efficacy, we argue that the social and cultural dimension is undervalued in Sekhon’s model and that it should be given more weight. This is particularly relevant in contexts where decision-making and implementation of a health care intervention is not in the hands of the target group - in this case, the mother.
We enlisted TBAs in the study because we regarded them as opinion leaders in the community. Indeed, it seemed the TBAs were very influential as regards umbilical cord care and they reported favorably on their chlorhexidine experience. On the contrary, health workers, who are another group of opinion leaders, were apprehensive concerning chlorhexidine scale up. They expressed concerns with the prolongation of the umbilical cord separation. The apprehension by health care workers could potentially affect the sustainability of the chlorhexidine scale up. We also found that older women (mother’s mothers and mothers in law and sometimes neighbors) whom the mothers were in daily contact with influenced newborn practices, including cord care, much more than the community opinion leaders such as TBAs. This finding is not new, and has been reported from other African settings [
31]. It has been argued by Kumar and colleagues that interventions to change newborn care behaviors should be targeted to households, and not to individuals [
32]. However, none of the health facilities we visited had newborn health promotional activities targeted to persons other than the mothers.
In this study we interviewed participants in a randomized controlled trial, who were exposed to more information on the issue at stake, and hence probably became more health attentive than others [
33]. In addition, since this randomized controlled trial recruited participants who gave birth in a health centre, the applicability of our findings to mothers who give birth at home may be limited. We mainly focused on individual conditions that could favor chlorhexidine scale up and did not tackle larger aspects like economic factors (for example cost of the product) and political factors (for example umbilical cord care policies) These are some of the limitations of our study. To increase the trustworthiness of our findings, we triangulated the data collection methods by using IDIs, FGDs and video, which broadened our understanding of perspectives about chlorhexidine acceptability [
34].
Conclusion
Our participants generally accepted umbilical cord care with chlorhexidine cleansing due to its convenience, and superiority in the reduction of umbilical cord foul smell and perceived abdominal colic facilitates scale-up. However, the perceived prolongation of umbilical cord separation associated with chlorhexidine was a discouraging factor, and could affect the sustainability of possible scale up efforts in the future. We recommend prior communication about its possible effect of delayed umbilical cord separation. Despite willingness to abandon some of the substances used for umbilical cord care, participant were unwilling to abandon the practice of bathing the newborn in the local herbal solution kyogero, in order to use chlorhexidine exclusively. However, they were willing to combine chlorhexidine with kyogero to harness the more holistic benefits of kyogero, which include future wellness and blessings. Participants’ willingness to defer the bathing of newborns in kyogero to a period after umbilical cord separation, offers an opportunity for chlorhexidine scale up efforts. Special attention should be paid to key decision makers like grandmothers and mothers in-law, who are often ignored in the current childcare promotion activities, but were found to be very important in the care of newborns, especially during the first few days after birth. According to our study participants, it is the people closest to the mother of the newborn who seem to be more influential in terms of cord care decision-making than the TBAs and the health workers that we originally had identified as opinion leaders. The identification of a significant bathing practice in a study designed for umbilical cord care, highlights the importance of an explorative approach to implementation science, as the factors that influence adoption of an intervention may be beyond the boundaries of what we initially set out to explore.
Acknowledgements
We thank the Center for International Health, University of Bergen that provided part of the funds for this study.
We also acknowledge the research assistants working with the CHX trial, and our community gatekeeper Ms. Namyalo Florence. Finally, we acknowledge the participants for their cooperation.