Advancing care and support for women and families after stillbirth or neonatal death in Kenya and Uganda: a feasibility study

Article history

The contractual start date for this research was in August 2017. This article began editorial review in July 2023 and was accepted for publication in June 2024. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The Global Health Research editors and publisher have tried to ensure the accuracy of the authors’ article and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this article.

Copyright statement

Copyright © 2024 Mills et al. This work was produced by Mills et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaption in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/. For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited.

2024 Mills et al.

Design and study setting

This work was part of the NIHR Global Health Group programme of improving care and support after stillbirth and neonatal death in sub-Saharan Africa (NIHR 16/137/53). A prospective, observational feasibility study was conducted in urban tertiary maternity facilities in Kenya (1) and Uganda (1) from November 2019 to December 2020, using a ‘pre’ and ‘post’ design. Research governance approvals were obtained from Kenyatta National Hospital/University of Nairobi (P828/09/2019), Makerere University/Uganda National Council for Science and Technology (UNCST; 2019-059) and University of Manchester (2019-7322-12550) and the study was prospectively registered (protocol available, ISRCTN68506895).

Community engagement and involvement

The NIHR Global Health Group community engagement and involvement (CEI) groups of parents with lived experience of the death of a baby, and stakeholder group of health workers, researchers and policy-makers in Kenya and Uganda contributed to all stages of the research, including planning, conduct and final interpretation of the findings.

Study interventions

The study intervention, informed by evidence synthesis and previous exploratory research conducted in both countries,9,10,17,21 was complemented with use of the behaviour change wheel, a structured approach to designing behaviour change interventions. 22 Two intervention components, addressing CEI group and stakeholder priorities of improving facility care and postnatal support for women after discharge identified, were co-produced with CEI members, health workers and service managers at a series of meetings.

Sample and recruitment

Informed by guidance for feasibility studies,23 the planned sample was 120 women (60 per country) ≥ 18 years who had experienced the death of a baby after 28 weeks’ gestation and before, during or 0–6 days after birth (neonatal death in the health facility) following the current pregnancy. This was considered sufficient to permit implementation of the study intervention in two sites and assess the feasibility of a large-scale trial. In the pre-intervention phase (phase 1; scheduled in month 1–6), recruitment of 60 women (30 per country) offered the usual care and support after the death of their baby was planned; in the post-intervention phase (phase 2; scheduled in month 7–12), a further 60 women would be recruited. Recruitment was planned to be completed in the first 3–4 months of each phase, to allow follow-up within 8 weeks after birth. Staff workshops were held in the included facilities, prior to commencement, to inform all staff about the study. Eligible women were identified and approached, prior to discharge from hospital, by health workers providing postnatal care or community health workers. Women were given a brief introduction to the research and those who were interested provided written consent to be contacted by the research team, not sooner than 2 weeks after the birth. At this contact, the research assistant (midwife or nurse) provided verbal and written information, gave an opportunity to ask further questions and allowed the potential participant time to consider. Informed consent was then confirmed in writing. During phase 2, the bereavement champion intervention was implemented on the facility level; therefore, individual consent was not sought for this component. Peer support was offered to all women participating in the study during phase 2; however, it was emphasised during recruitment that women could decline peer support and still participate in the research. Partners (or family members) of women participating in phase 2 of the study were also approached, via woman participants, to explore experiences of care and the research via qualitative interviews. A partner or family member’s unavailability or unwillingness to participate did not impact the woman’s eligibility. Health workers and others directly involved in the research as bereavement champions or peer supporters were also invited to participate in qualitative interviews at the end of phase 2.

Outcome measures

Recruitment and retention of women in the study were the primary feasibility outcomes. The criteria for a trial to be considered feasible was achievement of recruitment targets (n = 120 women) and ≥ 70% of women being retained until study completion. Other key outcomes included the acceptability of the intervention, uptake of peer support and experiences of the research. The feasibility and acceptability of data collection, characteristics of the proposed psychological and clinical outcome measures and process outcomes, including quality of implementation, were also assessed.

Data collection

A screening and recruitment log included data for women providing consent to contact, those recruited and participants leaving the study before completion. Women who declined to participate or left the study early were requested to provide reasons, if they were willing. Data were collected for each participant at two visits with the study research assistant, first at recruitment (around 2 weeks after the birth) and the second follow-up visit (approximately 6–8 weeks after birth) at study completion. Investigator-designed case report forms (demographic, pregnancy and birth, clinical data including healthcare and external support use) were completed at both visits and questionnaires (researcher-administered psychological measures) at follow up. Psychological measures included the perinatal grief scale (PGS) a 33-item measure assessing perinatal grief24 and the Edinburgh Postnatal Depression Scale (EPDS), a 10-item tool, originally designed to assess risk of postnatal depression but subsequently validated to assess anxiety and depressive symptoms in pregnancy. 25,26 Semistructured qualitative interviews with women (n = 37) and partners/family members (n = 13), interviewed separately, were conducted at or shortly after the follow-up visit (in phase 2 only) and with health workers champions (n = 16) and peer supporters (n = 10). Depending on COVID-19 restrictions in place locally, interviews were conducted at a place of the participant’s choice, for example, home or place of work following a risk assessment confirming physical distancing was possible, or by telephone. Interviews were audio-recorded and transcribed verbatim (Topic guides, Report Supplementary Materials S1–S5). To capture wider staff experiences of the intervention and research, a short investigator-designed questionnaire was distributed to health workers in maternity and neonatal wards in the facility at the end of phase 2 (see Report Supplementary Material S6). An intervention log, completed by the research assistants, captured training and meeting records for the bereavement champions and peer supporters. Peer supporters kept a log of the number and length of calls for each woman, space for brief comments was included, but it was not expected that the content of calls be documented.

Data analysis

Case report data, collected on paper, or directly inputted into a custom-designed REDCap database, was verified by a second researcher and exported to SPSS (SPSS Inc., Chicago, IL, USA) for analysis. Demographic, clinical and other outcome data were compared descriptively, reporting frequencies and percentages for categorical variables and descriptive statistics including means, standard deviations (SDs), medians and ranges for numerical variables. Scores for psychological tools were compared to determine whether characteristics were comparable across different measures, including a comparison between rates of missing data for different tools and their component items. For the psychological outcomes, adjusted (for country) 95% confidence intervals (CIs) for differences between the groups (estimated using analysis of covariance) and within-group SDs were calculated to inform the design of a subsequent evaluation.

Qualitative interview data were analysed using an inductive approach, following the five interlinked phases of framework analysis, moving from descriptive accounts to conceptualisation of meaning. 27 Initial analyses were conducted by two researchers and focused on experiences of research participation, recruitment processes, the acceptability and implementation of the intervention, impacts on care and burdens of data collection. The final interpretation of the findings was discussed and agreed by the wider research team.

Participants

Demographic, pregnancy and birth data for participants are presented in Table 2.

The mean age of women was 28 years (SD 5.8), similar across study sites and phases, all were born in Kenya or Uganda and most resided in urban locations (88%). A majority were married or living with a partner (83%), 90% described themselves as ‘Christian’, with 9% as ‘Muslim’. Secondary education was received by 75% of the sample. Over half women were employed, either full time or part time (55%), more women in Uganda were employed (61% vs. 47% Kenya). The proportion of women in their first pregnancy was similar across phases (phase 1, 21% vs. phase 2, 29%). Eight women reported having a previous stillbirth (four women in each phase) and six had experienced a previous neonatal death (three in each phase). Index pregnancy, birth and postnatal data are presented in Table 3.

Most women were booked for antenatal care (93%); 64 babies were stillborn, 21 were intrapartum stillbirths, 44 babies died in the early neonatal period (including one set of twins), surviving for median 1 day. Gestation and birthweight were lower in phase 1 (32.0 weeks vs. 36.5 weeks; 2086 g vs. 2494 g, phase 2). Among reasons given by women for the baby’s death, fetal compromise (13%), preterm birth (16%) and maternal hypertension (19%) were the most common; however, 27 (25%) stated they did not know the cause. Only one baby was reported to have had a post-mortem examination. Participant flow is presented in Figure 1; of 107 women recruited, 91 (85%) completed all data collection. In Kenya, 86% of participants completed all data collection, during phase 2 recruitment was 70% of target (vs. 100% phase 1), six women withdrew from the study, and one was lost to follow-up (vs. 0 phase 1). In Uganda, 86% of participants completed all data collection, four women withdrew from the study before the follow-up visit (all during phase 2). Additionally, five women (four during phase 1, and one in phase 2) were not contactable for visit two and lost to follow-up. Among reasons provided for withdrawal, three women moved away from the research site, two returned to full-time work and two cited partner objections. For comparison, data for all stillbirths and neonatal deaths occurring at the included facilities during the recruitment period, was collated from the hospital birth registers and is presented in Appendix 1, Table 8.

FIGURE 1.

Participant flow.

Intervention implementation

The study intervention was implemented, mostly as planned in both countries. To minimise potential contamination between the study phases, bereavement champions and peer supporters were recruited and trained shortly before the start of phase 2. Champions were recruited from clinical staff [n = 19; (n = 7 Kenya, n = 12 Uganda)], following information workshops, and with the agreement of line managers. Midwives and nurses from maternity (antenatal, labour and postnatal wards), neonatal and paediatric departments were included in both countries; in Uganda a hospital social worker was also involved. Peer supporters (n = 12, all bereaved mothers, n = 5 Kenya, n = 7 Uganda) were identified via following an information session in December 2019. Three-day interactive training workshops were held in both sites, facilitated by the UK and country research teams, including a mental health specialist, in January 2020. Both bereavement champions and peer supporters attended an introductory day; this included the experiences and needs of bereaved parents, an introduction to the study intervention and opportunities for both to understand their complementary roles and develop supportive networks. Bereavement champions, alone, attended Day 2, which included strategies for set up and sustaining groups, and workshop on communication skills, including an introduction to behaviour change. Day 3 included peer supporters alone and focused on listening skills, managing relationships and resolving difficult situations, boundaries, and self-care. Copies of all training materials were provided to all participants. Peer supporters also received a study mobile phone, pre-paid airtime and signed a ‘peer supporter’ agreement outlining expectations, limitations of confidentiality and support available during delivery of the intervention. WhatsApp groups were set up in each country for peer supporters and research teams to enable communication of any issues.

Bereavement champions

In-person champion group meetings were initiated in February/March 2020, but paused after the first meeting in both countries due to COVID-19 restrictions. It was not possible to resume group meetings in Kenya due to restrictions on facility gathering, staffing pressures, redeployment and lack of devices to facilitate remote meetings. However, the research assistants met individual champions one-to-one, providing support and feedback in developing planned actions. In Uganda, champion meetings restarted in late July and continued monthly until the end of the intervention period in December 2020 (5 meetings), attended by 8–12 champions and research team facilitators. Potential activities to develop care for bereaved parents were discussed during the initial training workshop and activities were agreed at the first meetings. In Kenya, the priority was to improve the environment for bereaved women through identifying private spaces for postnatal care and facilitating mothers and families contact with the baby after birth if they wished. In Uganda, the group targeted improved communication between staff and bereaved women, including introductions, information giving and providing opportunities to discuss experiences and emotions before discharge home.

Peer support

Peer support was offered to all women accepting participation in phase 2 at recruitment. All participants in Kenya, and 27 of 30 (90%) in Uganda agreed to be linked with a peer supporter. Following this, verbal and written information to explain the service was provided. Some women chose to call the peer at a pre-arranged time, while others preferred to use WhatsApp message. In Uganda, women usually initiated the first contact with call or text, while in Kenya some women preferred that the peer supporter called or messaged them. One peer supporter in Uganda declined to continue involvement after the first woman was matched, but before contact was initiated; therefore, the woman was reallocated to another peer. Peers could not establish contact with one woman in Kenya and six women in Uganda. During the first call, the peer supporter explained the role, her volunteer status and provided availability for next call. According to the preferences of the woman receiving support, further contacts, either by phone or WhatsApp were arranged. All peer supporters returned the logs at the end of the invention period. Table 4 summarises peer support contacts, peers provided support for 1–6 women (median 4) during the intervention period. Women received median of 3 contacts (range 1–12), mostly voice calls. Five women in Kenya and three women in Uganda received only one contact, with no response to subsequent call or messages from the peer. Some women and peers exchanged messages by WhatsApp, mostly once regular contact was established. The length of calls varied considerably; the median call duration was 6 minutes (range 1–80 minutes), call duration was longer in Uganda, at 9 minutes (range 1–80 minutes), than in Kenya 3.5 minutes (range 1–27 minutes). However, a similar pattern of longer initial calls, with shorter follow-up contacts was observed in both. Research assistants, who were experienced midwives, provided individual and group support to peers throughout intervention delivery, including facilitating a peer WhatsApp group. One woman requested a change of peer for personal preference, but no other significant issues were reported.

Key resources

Data for attendance at training for the bereavement champions and peer supporter were collected following the events and reported above. In the intervention phase of the study, the restrictions associated with COVID-19 interrupted access to facilities and participants, intermittently. As a consequence, research assistant workload increased and availability was reduced. These constraints prevented other resource data being collected as planned.

Psychological outcome measures

Psychological outcome data for the EPDS and the PGS are presented in Table 6.

Questionnaires were completed by all women remaining in the study at follow-up and no required question on either scale was repeatedly omitted. During interviews with participants, no issues were raised with wording or content of questions, or burden or completion for either instrument. These findings support acceptability for use in a subsequent evaluation. The mean EPDS score in phase 1 was 14 and scores were higher in phase 1 than phase 2 (14 vs. 9.1; adjusted difference 4.6, 95% CI 2.0 to 7.2). The mean PGS score in phase 1 was higher than phase 2 (106.2 vs. 84.7; adjusted difference 22.6, 95% CI 13.3 to 31.9). Where EDPS scores were above 10 or above 1 on item 10 (thoughts of self-harm) at follow-up, further assessment, referral to secondary care or psychological counselling services was offered. In Kenya, 5 women recruited in phase 1 scored > 1 on question 10 indicating potential thoughts of self-harm (all in phase 1), in Uganda 2 women scored > 1 (1 in phase 1, 1 in phase 2), all were offered immediate support and referral to secondary/psychological services. These, along with unexpected adverse events affecting participants, were reviewed by principal investigator and presented to the NIHR Stillbirth Group Advisory Board, and none were considered related to the research or the study intervention.

Lessons learned

Alongside prevention of stillbirth and neonatal death, the World Health Organization/UNICEF ‘Renewed Call for Collective Action’ (2020) stressed the urgency of provision of appropriate bereavement support for all women and families. To date, few studies have addressed improving facility care or community support in LMICs including Kenya and Uganda. While this study did not evaluate the impact of the intervention on outcomes, EPDS scores were above suggested cut offs for depressive symptoms in this context29,30 and PGS scores suggested intense grief. 31 This reflects qualitative evidence suggesting commonality of psychological responses across settings and supports assessment of grief and depressive symptoms as outcome measures for studies exploring experiences and improvement of care. The intervention components were designed to address prominent factors identified in negative experiences in Kenya and Uganda including poor communication, lack of supportive care after birth and social support at home.

Introducing trained ‘bereavement champions’ in wards and departments where bereaved women and families were cared for in included facilities was designed to address key ‘opportunity’(physical and social resources) and ‘motivation’ (conscious intention and subconscious experience) barriers, in addition to traditional strategies targeting knowledge and skills. 32 Activities revolved mainly around promoting good practice and role modelling, particularly in communication, although environmental changes including privacy and separation from other mothers and babies for bereaved women were also achieved. Group activities were curtailed in Kenya due to COVID-19 restrictions within facilities, but regular meetings were facilitated in Uganda via video conference. Facility surveys confirmed that champions were visible to other maternity staff, who also recognised increased attention to bereavement care and changes made during the study period. However, this was less evident in Kenya, also likely due to COVID-19 restrictions. The champions appeared to possess many characteristics, identified by Bonawitz et al. (2020), as key in leveraging change including presence, influence, investment in improvement and tenacity. 33 However, while the group included senior ward-based staff (‘in-charges’), service managers were not directly involved. Including managers as bereavement champions could increase success in overcoming organisational barriers encountered, through strategic knowledge and relationships within the institution. Another issue raised by champions was the lack of knowledge of bereavement care among maternity and neonatal staff in facilities. Additional formal training was perceived as important and could also lead to further champions ‘emerging’ through self-selection. This has been identified as a mechanism of sustaining effective champion networks in other health systems with benefits over reliance on nominations alone. 34

Peer support for bereaved parents evolved in HIC, as a response to unmet needs for psychological support and is often provided by third-sector organisations. In this study, telephone peer support was offered. Peer supporters were past or present members of the CEI groups supporting the research programme in Kenya and Uganda, and highly committed to improving care and support for women. Most women offered peer support accepted it, and established contact with the peer. In Uganda, the requirement for the woman to initiate the first call was identified as a potential issue and several peer supporters reflected that giving women a choice to be contacted, as was done in Kenya, might have increased reach. Considerable variation in contacts was observed; it was notable the Uganda participants had greater frequency and length of contacts, which could be related to less routine postnatal follow-up with health workers than observed in Kenya. Overall, women were positive, citing shared experiences, hope and validation of loss, as distinctive benefits of peer support in perinatal bereavement similar to those reported in HIC. 35,36 A review of the peer support used across multiple maternal and newborn health contexts in sub-Saharan Africa also demonstrated a range of positive impacts including empowerment through knowledge and agency and reduced perception of stigma. 37

Peer supporters generally understood their role, felt prepared and supported during the research and gained considerable satisfaction, reflecting experiences reported in HIC. 21 Peer support was a significant commitment of time, in addition to emotional resources. Co-developed guidance shared with participating women, provision of study specific phones, limiting the number of women allocated to each peer and the fixed duration of support offered in the study, helped most peer supporters balance their commitments. Several peers expressed a desire for further or extended training to build confidence, and provision of ongoing development activities have also been highlighted as important for the long-term sustainability of similar services in Australia and Finland. 21,38 Providing refresher training sessions at regular intervals or during meetings with the research team would potentially increase peer supporters’ confidence in the quality of their interactions with women.

Strengths and limitations

This feasibility study was conducted in two urban public maternity facilities in Kenya and Uganda. These are lower middle- and low-income settings with stillbirth and neonatal death rates significantly above international targets. The study was impacted by the COVID-19 pandemic, during phase 2, restrictions on contact and access to health facilities impacted recruitment and some aspects of intervention implementation and data collection, in both sites. In both countries, participant retention was also decreased in phase 2, which might also be related to impacts of COVID-19. The focus on urban sites might also reduce the transferability of findings to rural communities in sub-Saharan Africa, where birth outside facilities, geographical distance and sociocultural differences might impact.

CRediT contribution statement

Tracey A Mills (https://orcid.org/0000-0002-2183-7999): Conceptualisation (equal), Data curation (lead), Formal analysis (equal), Funding acquisition (supporting), Methodology (lead), Project administration (lead), Supervision (equal), Visualisation, Writing – original draft (lead), Writing – reviewing and editing (lead).

Valentina Actis Danna (https://orcid.org/0000-0003-2476-1659): Data curation (supporting), Formal analysis (supporting), Writing – reviewing and editing (supporting).

Elizabeth Ayebare (https://orcid.org/0000-0002-5636-5258): Funding acquisition (supporting), Project administration (supporting), Investigation (lead) Writing – reviewing and editing (supporting).

Carol Bedwell (https://orcid.org/0000-0001-8031-7793): Funding acquisition (supporting), Writing – reviewing and editing (supporting).

Lucie Byrne Davis (https://orcid.org/0000-0002-9658-5394): Methodology (supporting), Writing – reviewing and editing (supporting).

Karina Lovell (https://orcid.org/0000-0001-8821-895X): Methodology (supporting), Writing – reviewing and editing (supporting).

Raheli Mukwhana (https://orcid.org/0000-0002-0345-6316): Investigation (supporting), Writing – reviewing and editing (supporting).

Allen Nabisere (https://orcid.org/0009-0003-9397-6351): Formal analysis (supporting), Investigation (supporting), Writing – reviewing and editing (supporting).

Marion Okello (https://orcid.org/0009-0008-8166-6811): Writing – reviewing and editing (supporting).

Grace Omoni (https://orcid.org/0000-0002-0345-6316): Funding acquisition (supporting), Project administration (supporting), Investigation (lead), Writing – reviewing and editing (supporting).

Chris J Sutton (https://orcid.org/0000-0002-6406-1318): Methodology (supporting), Formal analysis (supporting), Writing – reviewing and editing (supporting).

Vicky P Taxiarchi (https://orcid.org/0000-0003-0737-098X): Formal analysis (supporting), Writing – reviewing and editing (supporting).

Sabina Wakasiaka (https://orcid.org/0000-0002-7925-506X): Project administration (supporting), Investigation (lead), Writing – reviewing and editing (supporting).

Tina Lavender (https://orcid.org/0000-0003-1473-4956): Conceptualisation (equal), Funding acquisition (lead), Methodology (supporting), Supervision (equal), Writing – original draft(supporting), Writing – reviewing and editing (supporting).

Other contributions

Dr Carol Bedwell co-ordinated community engagement and involvement input for this study, Ms Marion Okello provided input and critical review for the writing up of the paper on behalf of the CEI groups in Kenya and Uganda. Mr Jonan Mweteise, Ms Nerea Ojanga and the late Ms Anne Nendela also supported data collection. Special thanks to the CEI members in both countries for their input into the research at all stages of the design, conduct and interpretation. We would also like to thank Ms Claire Storey (UK) for her advice for peer support training, and the health workers and peer supporters in Kenya and Uganda involved in delivering the study intervention.

Data-sharing statement

The data generated for this study are not publicly available due to arrangements specified in the ethics approvals, but requests for anonymised data will be considered on application to the corresponding author.

Ethics statement

The study was approved by the Kenyatta National Hospital/University of Nairobi (P828/09/2019) 27 November 2019, Makerere University/Uganda National Council for Science and Technology (UNCST; 2019-059) 29 October 2019 and The University of Manchester (2019-7322-12550) 31 July 2019. Written informed consent was obtained from all participants, including for use of anonymised verbatim quotes. All study processes including recruitment, data collection and data processing were carried out in accordance with relevant guidance and regulations.

Information governance statement

The University of Manchester is committed to handling all personal information in line with the UK Data Protection Act, and the General Data Protection Regulations (EU GDPR) 2016/679. Under the Data Protection legislation The University of Manchester is the Data Controller and you can find out more about how we handle personal data, including how to exercise your individual rights and the contact details for our Data Protection Officer by contacting: dataprotection@manchester.ac.uk

Disclosure of interests

Full disclosure of interests: Completed ICMJE forms for all authors, including all related interests, are available in the toolkit on the NIHR Journals Library report publication page at https://doi.org/10.3310/JNWA6983.

Primary conflicts of interest: Dr Tracey A Mills is a member of the NIHR Academy Pre-doctoral Clinical Academic Fellowship (PCAF) Selection Panel and Co-I on NIHR Global Health Group (16/137/53) and Unit (132027). Professor Karina Lovell is a member of the NIHR Advanced Fellowship and Senior Investigator Selection Panels. Dr Chris J Sutton is a member of the NIHR Health Technology Assessment Commissioning Funding Committee and Co-I on a number of NIHR funded grants/projects, including HTA (133518, 133418, 131483) Global Health Groups (16/137/53). Other NIHR funded work EME 132622, RFPB 203507, 201093, 203475, 203468, Programme development grant 202044 HSDR 130581. Professor Dame Tina Lavender was chair of NIHR Global Health Groups Round 4 Selection Panel and is a member of NIHR Global Health Groups Round 5 selection panel, was a member of the HTA Obesity Themed Call Board and Independent Steering Group Chair for NIHR CRIBBS Sierra Leone. She is also principal investigator on NIHR Global Health Group (16/137/53) and Unit (132027). The other authors declare they have no competing interests.

Department of Health and Social Care disclaimer

This publication presents independent research commissioned by the National Institute for Health and Care Research (NIHR). The views and opinions expressed by the interviewees in this publication are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, MRC, NIHR Coordinating Centre, the Global Health Research programme or the Department of Health and Social Care.

This article was published based on current knowledge at the time and date of publication. NIHR is committed to being inclusive and will continually monitor best practice and guidance in relation to terminology and language to ensure that we remain relevant to our stakeholders.

Study registration

This study is registered as ISRCTN68506895.

Funding

This research was funded by the National Institute for Health and Care Research (NIHR) Global Health Research programme (NIHR award ref: 16/137/53) using UK aid from the UK Government to support global health research. The views expressed in this publication are those of the author(s) and not necessarily those of the NIHR or the UK government.

About this article

The contractual start date for this research was in August 2017. This article began editorial review in July 2023 and was accepted for publication in June 2024. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The Global Health Research editors and publisher have tried to ensure the accuracy of the authors’ article and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this article.

Copyright

Copyright © 2024 Mills et al. This work was produced by Mills et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaption in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/. For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited.