Background
The leading cause of neonatal mortality is complications from preterm birth (WHO, 2016a). It is estimated that globally, 15 million babies are born preterm every year, with two-thirds of these born in low- and middle-income countries (LMICs) [
1]. Although previously it has been estimated that only 50% of babies born at 32 weeks of gestation in LMICs survived [
1,
2], more preterm babies than ever before are surviving to discharge [
3‐
5]. When discharged home, parents of preterm infants are still required to provide additional care in the community in order to prevent infection and hypothermia, and to support growth and development,. There are a number of proven interventions that neonatal healthcare workers expect parents of preterm infants to provide after discharge. These include; kangaroo care, exclusive breastfeeding, and/or expression and administration of breast milk to their infant and infection prevention [
5]. Kangaroo care significantly improves neonatal survival rates through prevention of hypothermia, facilitation of breastfeeding, and prevention of infection [
6‐
8]. In preterm infants, breast milk improves growth as well as reducing the risk of developing sepsis, necrotizing enterocolitis, retinopathy, neurodevelopmental impairment [
9‐
15]. Furthermore, preventing the transmission of infections through good hand hygiene is also an effective, low-cost means of infection control [
16,
17].
To support these vulnerable preterm infants after discharge, the ability of parents to continue the provision of these key, life-saving interventions at home is vital. Despite this, it is unclear how well these high-risk infants are managed after they are discharged home. Data from Uganda show that almost one-third of very low birth weight infants die in the first 8 months after discharge from a neonatal intensive care unit, with two-thirds of these deaths occurring in the first month after discharge [
18]. Few studies, particularly from LICs, have investigated how caregivers of preterm infants think and feel once they are finally discharged home and how capable they feel to continue with the recommendations and care they were taught in the hospital [
4,
19,
20]. Some studies from HICs have demonstrated that many parents have feelings of insecurity, fear and anxiety after discharge, with these feelings continuing for a few months [
21‐
24]. The lack of preparation for discharge has been identified as a key source of this fear. These feelings are often mixed with relief that their preterm infants are considered healthy enough to be discharged and the relief to be leaving the stressful environment of the neonatal unit [
22,
23]. Caregivers also report feeling ‘isolated’ following discharge, often due to feeling an overwhelming need to provide constant surveillance of their infant and a fear of contracting infections, both of which inhibit their chances of socialisation with other community members [
21,
22]. Some fathers are reluctant to be involved in the care of the preterm infant, leaving the mother alone at home and adding to mother’s feelings of isolation [
22].
Some studies from both HICs and LMICs have demonstrated how responsibilities in the home, including housework, cleaning and care of siblings provide a considerable challenge for caregivers [
19,
20,
25‐
27]. There are often competing demands for a mother’s time including having to spend much of the day working in the fields [
28]. This results in a reduced amount of time available for the care of their preterm infant and can contribute to feelings of exhaustion [
24,
26].
It is possible for these preterm interventions to be taught to parents during the inpatient admission and for them to continue them after discharge. Some studies have highlighted lower levels of parental education to have a negative impact on the continuation of preterm care in the home [
23]. Other studies have reported inadequate training of the caregivers in appropriate feeding and kangaroo care before discharge, as well as a lack of knowledge of the caregivers of the benefits of continuing these practices at home to be key factors [
23,
25,
26].
Most research on the barriers and facilitators to providing care to preterm infants post-discharge has been confined to HICs [
22,
29‐
33]. In LMICs, research has focused on the initiation of care in the community for moderate to late preterm infants [
20,
26,
34‐
36]. There is limited qualitative research in sub-Saharan Africa (SSA) on the experiences and journeys of parents once their preterm infant is discharged from a neonatal unit. Given that in-patient survival rates of increasingly preterm infants in LMICs are improving, particularly in SSA, understanding how best to improve the care of these high-risk infants post-discharge is vital. It is evident that, caregivers need to have the knowledge and the ability to meet the basic needs of their infants to prevent them from becoming unwell, and to promote good nutrition and development [
32]. It is also clear that many caregivers in LMICs are unable to continue an adequate level of care for preterm babies in the community. This qualitative study examines the perceptions and experiences of mothers, caregivers, and healthcare workers on the journey following discharge of preterm infants from a neonatal unit in eastern Uganda. This study has been conducted to identify perceived barriers and potential facilitators to improving post-discharge care of preterm infants in the community.
Methods
Three different methods of data collection, IDIs, FGDs and case-studies, were used in this study to collect data from carers and key-informants. These different methods of data collection and sources of data are necessary for triangulation and were chosen to ensure a thorough exploration of key themes [
37]. We used purposive, non-random sampling to identify the most suitable participants for the IDIs, FGDs and case-studies. Although this may have introduced bias, it was necessary to select health workers and family members that had experience of caring for a preterm infant. Individual characteristics also provided the basis of selection, ensuring that the research sample reflected the diversity of people living in the community including participants from different tribes, different roles and different ages. For example, mothers, grandmothers and fathers of different ages were selected. This ensured that there was appropriate diversity between participants. The suitability of key-Informants was determined, and selection made, following the advice of the neonatal lead at the MRRH. This ensured that healthcare workers who had a broader and more extensive experience of caring from preterm infants were included, as well as a range of different cadres of healthcare workers including nurses and clinicians.
Data collection took place from May to July 2017. We recruited 39 participants in the overall study (a total of 35 separate sessions including 18 IDIs (12 caregivers and six HCWs), 3 FGDs (17 caregivers (mothers, grandmothers, and fathers)), and a further four case studies (14 separate IDIs) were carried out with caregivers (three mothers and one grandmother), followed up for 4–5 weeks after discharge. IDIs and FGDs took place at MRRH or in participants’ homes (Table
1). Further demographic details are provided in supplementary material (Supplement
2).
Table 1
Numbers and roles of the participants in the IDIs, FGDs and case studies
IDIs with Caregivers | 8 | 3 | 1 | - | - |
12
|
IDIs with HCWs | - | - | - | - | 6 |
6
|
FGD1 | 3 | - | - | 1 | - |
4
|
FGD2 | 4 | 1 | 1 | - | - |
7
|
FGD3 | 3 | 2 | - | 1 | - |
6
|
Case-Studies | 3 (11 IDIs) | 1 (3 IDIs) | - | - | - |
4
|
Total
|
24
|
5
|
2
|
2
| 6 |
39
|
Setting
This study was conducted in Mbale Regional Referral Hospital (MRRH), a public hospital that serves a population of 4.5 million people in eastern Uganda. MRRH has nearly 10,000 deliveries a year and also receives neonatal referrals from surrounding health facilities [
38]. MRRH has a dedicated level-2 neonatal unit (NNU) that can provide intravenous fluids and medications, nasal oxygen, and bubble continuous positive airway pressure with up to 2,500 neonatal admissions per year, including almost 800 preterm neonates. In addition to one-on-one explanations to mothers, education on cord care, kangaroo care, hand hygiene, and temperature monitoring is delivered daily on the NNU by a healthcare assistant.
MRRH is a regional centre of excellence for neonatal care in Uganda and has achieved a 57% reduction in preterm in-patient mortality over the previous seven years [
3,
39]. The neonatal lead and co-author (KB) had observed poor follow-up attendance and poor compliance with preterm care practices after discharge from the NNU. The need to improve understanding of the challenges that existed in the community after discharge as a priority in improving preterm outcomes had been identified.
Individual interviews
Caregivers were identified from MRRH-NNU or the weekly neonatal follow-up clinic of recently discharged patients. All caregivers (mother, grandmother (“jaja”) or father) of any preterm infants who had recently been admitted to NNU were eligible to take part. There were no exclusion criteria. The suitability of key informants was determined, and selection was made following the advice of the neonatal lead and based on those with the most experience on MRRH-NNU. Potential participants were provided with verbal information as well as an information leaflet either in English or Luganda. Informed written consent was sought from all participants, those who were illiterate gave a thumbprint to sign their consent form. Case-study participants were re-consented at each interview. There was no remuneration for involvement. Unique letters were assigned as identifiers to each participant and used throughout in place of the participant’s name to ensure that the identity of each participant would not be recognised by anyone outside the research team. All data used was kept securely on a password-protected laptop.
Focus groups
Mixed groups were used for the FGDs to ensure diversity as it was difficult to recruit enough men to have a group on their own. The FGDs were moderated by the research assistant to ensure that all the participants gave responses. We conducted IDIs with HCWs who were currently working, or who had recently worked, in MRRH-NNU.
Case studies
We conducted four case studies with caregivers to allow for the exploration of individuals’ experiences within their real-life context and a holistic analysis of the case. We followed these caregivers for a period of four to five weeks after discharge every one or two weeks.
Data collection
All IDIs with the HCWs were conducted in English by the primary researcher (HL), a white British young female Masters student. A research assistant, who was a young Ugandan female midwife (Diploma) with three years of experience in qualitative research and who spoke both English and Luganda, conducted all other IDIs and FGDs in either English or Luganda, as appropriate. Neither the primary researcher nor the research assistant were known to the participants prior to the IDIs/FGDs. IDIs lasted between 30 and 60 min and FGDs lasted between one and two hours.
IDIs were conducted either in a quiet room at MRRH or at the participant’s home enabling good rapport between interviewer and participant [
40]. Interviews were structured using topic guides (Supplement
1), and depending on participants’ answers, the interviewer questioned further to gain a complete understanding of the participant’s perspective [
40]. FGDs involving mothers and caregivers were structured using topic guides (Supplement
1) based on key themes that had emerged during the IDIs [
40]. Topic guides were reviewed by the research team, piloted before interviewing began, and were iteratively adapted throughout.
All IDIs and FGDs were recorded on a digital voice recorder. To ensure consistency, interviews not undertaken in English were carried out by the same research assistant, who was familiar with the topic guides. Audio files were then transcribed and translated. All transcription of non-English interviews was carried out by the same transcriber. A reflective diary and field observation notes were taken by HL as part of the research work to support the analysis of themes and ideas.
Analysis
Analysis was iterative, taking place during data collection and continuing afterwards.
Data analysis was carried out using the framework approach as this method is systematic and ensures rigour and transparency [
41]. All interviews were transcribed and discussed with the research team to also enable for adaptation of the topic guides. Analysis and coding began deductively with reference to the topic guide, identifying emerging themes in the transcripts. Analysis became more inductive as transcripts were then coded to finally establish a coding framework. To ensure a clear process of coding, initial transcripts were independently coded by three members of the research team. This meant that different viewpoints were considered, and no one point of view dominated the overall analysis [
41]. After this, coding was completed by HL. Data analysis then continued using the framework approach as this method is systematic and ensures rigour and transparency [
41]. NVIVO software (version 11) was used to chart the interview data into the framework. Throughout this process, the framework was continuously revised, and further themes were identified. The data from the FGDs and the IDIs with both the HCWs and caregivers were combined thematically. Data from the case studies was reported separately as these data focused on the challenges that emerged over a period of time after discharge. Data was then interpreted and explanations and reasons for the emergence of phenomena were sought. To begin with, the sample size was flexible and was guided by the concept of ‘saturation’. When additional data from further participants resulted in repetition and no new additional ideas being gathered, ‘saturation’ was assumed and no further IDIs or FGDs were held.
Quality control
During this study, every attempt was made to ensure the credibility, dependability and confirmability of the research [
37]. Different methods of data collection and sources of information ensured that results are comprehensive to allow patterns of convergence between sources to develop [
37]. All members of the research team had experience in this particular medical setting and were known to participants. They were therefore able to build trust and rapport with the participants and this minimised the Hawthorne Effect [
42]. The primary researcher received formal training from Liverpool School of Tropical Medicine (LSTM) in qualitative data collection and methods in order to ensure the quality of data collected [
43] All research assistants also received local training in these methods and had previous experience. Participant checks throughout the data collection period were undertaken. If ambiguous answers were given, interviewers questioned participants to clarify the participant’s views. Following interviews, data was analysed and discussed within the research team. Any ambiguity in participant responses were then clarified, either by questioning the same participant, or where that was not possible, by conducting further interviews with different participants [
44].
Discussion
Our study presents data regarding the care of preterm babies at home following discharge from a neonatal unit in Uganda and looks at how caregivers were prepared for discharge, how they coped in the home setting after discharge and the challenges they faced.
The first key theme identified was preparation for continuing preterm care in the home. Overall caregivers felt well-prepared for discharge and many caregivers felt confident about continuing care post-discharge. They reported that they had received health education and opportunities to carry out different aspects of preterm care under the supervision of staff. Learning best practices in caring for preterm infants before discharge was seen as the key factor in enabling caregivers to provide high standards of care in the home. At this healthcare facility, the importance of education was recognised, and emphasis was placed on group teaching on the ward by a nursing assistant and supervised practice of different aspects of care before discharge. This proved effective, and compared to other settings, caregivers in this study appeared to have better knowledge about care on discharge from NNU [
23,
25,
26]. However, as in Papua New Guinea and India, the education received by caregivers was limited by the fact that that staff were busy and did not always have sufficient time to provide education [
45,
46] Limited human-resource in low-resource settings is a significant barrier to improving this and advocacy for additional roles in training and education is needed.
One key gap in the caregivers’ knowledge was highlighted: neonatal resuscitation. Many caregivers felt concerned that they would not know what to do if their baby stopped breathing. Research into caregivers’ awareness of neonatal resuscitation is limited. A qualitative study undertaken in a HIC similarly found that caregivers had little knowledge of preterm apnoea or bradycardia and were not able to say how they would respond to these situations. One literature review concluded that teaching neonatal resuscitation was beneficial in improving preterm mortality rates [
47]. This, however, was based on limited evidence.
The second key theme identified in our study was the psychosocial challenges to providing preterm care in the home. While on NNU, caregivers benefit from 24-hour surveillance and have immediate access to support, advice and treatment when needed. Following discharge, this support is withdrawn as caregivers continue the care of their preterm baby alone. A qualitative study in Iran demonstrated that once in the home setting, caregivers’ confidence declined, as did their competence, due to the fact they had been dependent on staff at the unit [
31]. Our research found that caregivers struggled at home where the NNU staff were no longer accessible to them. Similar findings have been reported in other settings [
21,
22,
27]. Many caregivers stated that home visits during the initial weeks following discharge would have been desirable.
An important finding in our study was that caregivers felt isolated following discharge. In case studies, it was reported that only five weeks after discharge did caregivers feel comfortable leaving the home. An underlying reason for this was the carer’s belief that it was necessary to keep the baby away from other people to reduce the risk of infection. Similarly, in Tanzania, caregivers spent an average of 40 days isolated indoors with their preterm babies [
35]. On NNU, the importance of restricting visitors to minimise transmission of infection is stressed. This may well contribute to caregivers’ isolation in the home as caregivers continue to adopt these measures in the home setting for several weeks post-discharge. Advice as to when restrictions can be eased may well contribute to caregivers accepting more support and feeling less isolated once back in the community.
Barriers to providing preterm care at home was the third and largest key theme we identified. Our study highlighted the fact that some caregivers commenced inappropriate practices on their own initiative, including supplementing breast milk with alternative feeds, for example, cows’ milk, and using charcoal stoves to keep their babies warm. Other research in similar African settings in Tanzania, Ghana and Uganda had similar findings [
4,
35,
48]. In Tanzania and Ghana, caregivers often fed their babies sweetened water in place of breast milk [
35,
48]. In Uganda, caregivers reportedly use charcoal stoves, or jerrycans of hot water, close to the baby’s bed to keep the baby warm [
4,
49]. In India caregivers often wrapped their babies in blankets to maintain temperature, rather than practising kangaroo care [
49]. Similarly, in this Ugandan setting caregivers wrapped their babies in blankets in an attempt to maintain their temperature, although kangaroo care had been taught as the most effective means of thermal care. Another, earlier finding in Tanzania indicates that caregivers would begin to bathe preterm babies too [
35]. In our study, caregivers were unclear as to the appropriate time to start bathing at home. These are areas of care that could be clarified and discussed before discharge.
Another key finding from this study was that very often the mother was not the primary caregiver during the baby’s admission to NNU. Other family members took on this responsibility while mothers recovered. Mothers then took over the role of the primary carer in the home, having not received appropriate levels of training as information was not passed on. This finding has not been reported in previous studies and needs further exploration and intervention to minimise the negative impact.
Caregivers stressed the importance of support from family and friends following discharge, which was also the number one enabler in caregivers being able to provide good care. This was reported in a comprehensive, systematic review on the barriers to and enablers of care for preterm infants [
50]. It has been found in other LICs, that support from partners was frequently non-existent due to work commitments and gender role stereotypes [
51‐
53]. Similarly, in this study, it was frequently reported that care from family members decreased following discharge. It was only possible to interview two fathers as the majority were not available for interview. Interviewing a larger number of fathers would have provided more insight into how they regarded their role in the home. Research into the care given by fathers in the days following discharge is limited, however encouraging male involvement could improve the care of neonates in the community [
54]. Research undertaken in HICs found that fathers’ involvement in care post-discharge was beneficial. Tackling perceptions concerning gender roles could therefore contribute significantly to improved support mechanisms for caregivers and involvement by fathers could enhance the quality of care.
Another finding in our study was that mothers were unwilling to accept help with the direct care of babies, their main reason being that they could trust no one else. This has been reported in other settings [
25,
27]. Whereas caregivers in this study were not practising continuous kangaroo care, 99.5% of caregivers in Ghana, who accepted help from spouses and other family members, were still practising continuous kangaroo one week after discharge [
55]. This stresses the importance of family members being involved in the care of preterm infants. It also highlights the need to explain to mothers that managing alone as sole carer for a preterm baby is a huge challenge and help, whenever offered, should be accepted in the best interests of their baby. Religion is also known to play a significant role in enabling caregivers in particular areas of the world. In Iran, the majority of caregivers put complete faith in religion, asking guidance from God as they continue to care for their preterm infants [
27]. Partner involvement is also considered crucial to the successful continuation of kangaroo care in the community [
34].
Although unwilling to accept direct support in the care of babies, caregivers mentioned requiring support regarding other household tasks. The need to divide time between caring for their preterm baby and other responsibilities such as cleaning, cooking and gardening. This was found to be a key challenge both in this research and across many other settings [
4,
26,
53,
56]. In Papua New Guinea, where family members took over all other responsibilities in the house following discharge, caregivers reported practising kangaroo care for longer periods than in our study [
45]. This provides evidence that continued support from others in sharing the workload is vital if preterm babies are to receive optimum care.
Caregivers experienced difficulties commencing breastfeeding or lacked the incentive to persevere with it. Preterm infants start breastfeeding at a later age than babies born at term, and in MRRH-NNU they were encouraged to express their milk and give it by nasogastric feeding tube or spoon. Although they are taught to hand express, most mothers did not get support with breastfeeding in NNU and were discharged into the community without having had to practice in this area. Most mothers began breastfeeding at home, without advice and support, and reported challenges such as experiencing pain and lacking sufficient milk. This contributed to under-confidence in their ability to breastfeed. This was also found to be the case in other settings [
31,
57] Research indicates that nurses play an important role in providing support and reassurance to new mothers as they commence breastfeeding [
57‐
59]. This demonstrates a need for continued close contact between mothers and HCWs following discharge.
An important finding in this study is that caregivers believed that it was impossible to undertake other tasks while their baby was in kangaroo. Caregivers strongly believed that any time spent practising kangaroo care would need to be spent sitting still and they saw this as time-wasting given the many other responsibilities they had. Being encouraged to move around more whilst practicing kangaroo care in NNU might encourage caregivers to do so at home and therefore increase the time spent in kangaroo.
Interviews with caregivers in Uganda and Ghana have identified that some mothers fear that kangaroo care might harm their baby as traditionally, mothers carry babies on their back [
4,
26]. This results in caregivers practicing kangaroo care for shorter lengths of time at home than in neonatal units [
23]. Furthermore, a Ugandan study identified low income, a heavy workload and a lack of decision-making power as key barriers to the practice of kangaroo care at home [
34].
Caregivers in this study believed kangaroo care might harm their baby. This was in line with findings in other developing countries [
26,
50] Caregivers feared that it was uncomfortable and detrimental to the baby’s breathing or that the baby might fall if carried around. Addressing these misconceptions could therefore increase caregivers’ willingness to continue practising at home. However, caregivers in this country experienced less pain and fatigue related to kangaroo care, and fewer cultural barriers than in many other settings [
36,
55,
60,
61].
Financial considerations were responsible for caregivers’ failure to return to the hospital for follow-up appointments, or to NNU if their baby fell sick. This has also been reported in Malawi [
62]. Transport problems were given as the main barrier to seeking healthcare advice following discharge. Caregivers and HCWs expressed a strong preference for returning to NNU at the referral hospital rather than attending community hospitals or healthcare facilities in the neighbourhood.
The last key theme identified was suggestions for improvement of preterm care at home. In HICs, good communication between caregivers and healthcare workers in neonatal units, as well as the need for continuing education and support in the community, have been highlighted as key enablers to comprehensive preterm care following discharge [
21,
24,
25,
27,
63]. A Swedish study assessing the benefits of real-time video conferencing between a neonatal unit and preterm caregivers at home, together with a contact by phone, was shown to reduce anxiety amongst caregivers and to improve their ability to provide appropriate care [
29]. A Neonatal Transitional Care Program in the United States that involved 10 to 20 follow-up home visits by HCWs over four months found reduced stress levels for caregivers, more successful breastfeeding routines and a reduction in the number of visits to hospital emergency departments [
64]. Education and support groups commenced before discharge have also proved beneficial in aiding the transition from hospital to home, providing information to parents and opportunities for discussions with peers sharing similar experiences [
30,
63,
65].
Strengths and limitations
Purposive sampling methods used to achieve the study population contribute to the validity of the research [
66]. However, in sampling, selection bias may have been introduced. We were unable to access many primary caregivers who did not attend follow-up appointments and who may have been facing greater challenges at home. Therefore, it would have been valuable to have followed up more caregivers by travelling to their homes. To mitigate this, caregivers were also identified and recruited through patient logbooks and offered transport costs to be able to attend interviews. Many fathers were unable or unwilling to take part, thus limiting insights into perceived gender roles.
Caregivers may have reported practising the care that they knew they should be providing, rather than the care that they were providing. An attempt was made to mitigate this by visiting some caregivers at home where they felt more at ease. This also made it possible to get a sense of the home environment and establish what care was occurring.
Due to the number of languages spoken in the area, interviews were often not conducted in the participants’ first language. This may also have had an impact on the quality of the results obtained.
Although the results of this research might not be generalisable to other settings due to differences in culture, beliefs and resources, the findings related to the care of preterm babies in the community in eastern Uganda have uncovered several important issues. These findings could, therefore, be used to generate hypotheses as a basis for future research in other low-resource settings.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.