Study design

Details about the applied methodology have been described elsewhere [20] but briefly; The study adopted an exploratory qualitative design conducted in Uganda among national-level key informants drawn from the existing maternal and child health policy networks. This manuscript aimed to capture national-level perspectives regarding the implementation processes of policies and interventions to address the national stillbirth burden. It took a qualitative descriptive approach to gather a deeper understanding of how these interventions were unfolding and the emerging challenges that were being confronted from the national-level stakeholders’ perspectives.

Study setting

The study was conducted in Uganda at the national level. The country enjoys substantial external support through development aid with the health sector as one of the leading beneficiaries. Most recently, donor support for health accounted for up to 43% of the total health budget with much of it off-budget support [15]. RMNCH is one of the key programs’ beneficiaries of this donor support. The country adopted drastic measures as reflected in policy alternatives to address the national stillbirth burden as advocated for during the global campaigns from the early 2010s and have been integrated into national RMNCH policies [15]. Conceptually these policy recommendations originated from; 1) World Health Organisation (WHO) recommendations, 2) local evidence generated from donor-supported projects, and 3) insights from routinely collected data and implementation experiences by the Ministry of Health (MoH). The overarching policy expectation is to promote deliveries and antenatal and postnatal care within health facilities. These have been reflected as strategic interventions for health under National Development Plan (NDP) III [15]. Broadly they include; improving maternal neonatal, child and adolescent health at all levels of service provision, improving the functionality of the health system to deliver quality and affordable health care, and improving social determinants of health, Universal Health Coverage (UHC) as some of the key result areas for monitoring. The country has registered substantial progress in reduction of facility-based fresh stillbirths currently standing at 7.7/1000 and is on course to achieve the national target of 4/1000 by 2025 as reflected in Fig 1. There are regional variations in Institutional Perinatal Mortality Rate (IPMR) with Kampala disproportionately performing poorly at 40.2/1000 while Teso region has the lowest at 14.5% compared to the national target of at least 15/1000 [15]. Overall, there have been progressive improvements in performance regarding perinatal death notification and reviews over the past five years (2016–2021). Of the total perinatal deaths reported, up to half (50.8%) were notified within the recommended 24 hours and only 41.2% were reviewed and entered into the DHIS event report [15]. Regional variations exist where the Kampala region registered the highest notification rates at 100% while Tooro only reported 14.2% of the total cases within the recommended time. These variations are also visible when it comes to perinatal death reviews where Ankole region had the highest review rates at 75.9% while Bunyoro had the lowest at 8.8% of the national target of at least 14% [15]. It is within this context that this study was conducted.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. National stillbirth rates and targets over the years.

Source: Ministry of Health, Uganda.

https://doi.org/10.1371/journal.pone.0285172.g001

Study population and sample

The study population comprised purposively selected respondents drawn from the national-level MCH policy networks. Potential respondents were pre-identified as eligible with guidance from individuals familiar with national-level MCH programming and policy processes. A list was generated with the contacts from which they were approached. Additional respondents were selected based on leads from the Ministry of Health, fellow respondents and information from reports detailing their contribution to national-level policy processes. Overall 20 participants were interviewed. We employed a process tracing methodology together with purposive sampling to identify key documents for review.

Data collection procedures

Data collection was conducted between March and June 2019. It primarily consisted of two methods which included key informant interviews and document reviews. The exercise was conducted by the first author (ES) with support from two graduate-level research assistants who were experienced in qualitative methodologies. At the design stage, the inclusion criteria were set to include only those respondents who were familiar with national-level MCH policy processes and had at least spent two years in their current positions. An initial list of potential respondents was pre-generated from where the final sample would be drawn. They were approached through telephone calls where the aims and objectives of the study were explained. For those that agreed to participate, a day and time of the interview would be agreed upon. Two callbacks were made in case the respondent was not available during the first attempt after which a replacement would be identified. Face-to-face interviews lasting from 45 minutes to one hour depending on the point of saturation were conducted at the respondents’ places of work. During the interviews, they would be asked to recommend any other potential respondent whom we could also include in our sample. These were snowballed and contacted for possible participation after the commencement of the study. An interview guide consisting of open-ended questions developed for this study was used. On the day of the interview, a safe and secure place would be identified from where the interview would be conducted. Field notes were taken at the end of each field day during the daily debrief meeting attended by the data collection team. Overall 20 out of the 23 respondents who had earlier on been contacted took part in this study the point at which we attained saturation. The three who had earlier expressed willingness to participate could not be accessed after the third callback. 17/20 were female and drawn from Professional associations (6), the Ministry of Health (5), Non-Governmental Organisations (NGOs) (4), Private not for Profit (2), Academia/Researchers (2) and the Private-for-Profit sector (1).

The document review exercise was conducted by the first author with key documents reviewed including Ministry of Health reports and Strategic Plans, and project reports from donor-funded projects. A grey literature search was conducted on the internet and it prioritised websites of leading organisations, projects and donors involved in national-level programs to address the stillbirth burden.

Analytical framework

Translation of policy and delivery of interventions aimed at responding to the national level stillbirth burden was conceptualised from the perspective of their implementation as intended. This would therefore be achieved through the provision of quality RMNCH services guided by evidence-based interventions as recommended by the relevant policy prescriptions. We, therefore, adopted the Donabedian model of quality healthcare service delivery as a guiding conceptual framework [18]. It approaches quality in terms of inputs, processes and outcome expectations. We analysed stakeholders’ views regarding the inputs that went into the implementation of these interventions in terms of the health systems building blocks such as trained health workers, financing and availability of medical products and supplies. The processes as viewed from technical quality by the service providers (guideline adherence) and perceived social quality from the service users’ perspectives for example user experiences during their interface with the health systems. The outcome expectations considered the perceived effectiveness of policy alternatives in addressing the stillbirth burden and effective coverage in terms of universal access for all as reflected in the overall effect of the stillbirth burden so far. The three aspects accounted for the overall health systems readiness to address implementation challenges to deliver interventions that would respond to the stillbirth burden in Uganda.

Data management and analysis

All interviews were conducted in English and recorded verbatim using a digital audio recorder. They were then transcribed using Microsoft office word by two research assistants who took part in the data collection exercise and stored in a secure folder on a password-protected laptop computer only accessed by the study team. All transcripts were read through by the first author before exporting them to Atlas. ti a qualitative data management software for coding [25]. Using the guiding framework, each of the aspects of the Donabedian model was entered a priori as a code with its sub-codes to aid the coding process following a deductive approach. Text relating to a particular code was identified from the transcripts, highlighted, dragged and dropped in the respective codes in the process of assigning textual data to codes. Thereafter query reports were run for each of the codes using a query tool function within the software. a manual pile sorting exercise was done for each of the queries by reading through each of the texts to identify underlying meanings. Text with similar or related meanings was then grouped into separate piles. Each pile was then read through once more to identify underlying themes and sub-themes. A similar exercise was followed while analysing textual data from the documents that were reviewed. We then used the Consolidated Criteria for Reporting Qualitative Results (COREQ) checklist to guide the reporting of results with typical quotes used to bring out the respondents’ voices. Triangulation was done through a comparison of results from the different respondents’ categories and data collection methods.

Perceptions

Inputs

Processes

As technical aspects were strengthened, attention was being directed more to the quality and skills of the health workers and managers especially as it related to their ability to steer the subnational health systems into a functional one. Integration of proposed interventions within the available resources at the sub-national level was still a challenge in some areas. The planning and budgeting skills beyond the clinical/technical competencies were reported as requiring special attention. Many of the policy aspects required managerial competencies and decision-making. Respondents mentioned how this was reflected in managers’ ability to work within limited resources to ensure the interventions were implemented smoothly.

There were observed challenges to the effective coverage of available RMNCH interventions. Respondents noted that whereas some were supply-side systemic factors, others stemmed from the demand side. Despite the increased average uptake of antenatal care services, many users did not complete the recommended visits. There were still instances of mothers attending the recommended antenatal care visits but still would not deliver from the health facilities. Even for those that delivered from health facilities, negative outcomes were still being reported which meant that some risk factors were still sipping through the system.

A lot of effort has been done to encourage mothers to come and deliver in health facilities but when they come they are losing their children. Antenatal coverage we are saying has gone up but when you look at antenatal 4 as to 1, the catchment is still below standard. #2 MOH

Some of the interventions, although implemented within the same health systems were not well aligned to provide a forward and backward linkage hence averting any possible stillbirth risks. This posed further challenges for the likelihood of having many interventions within the health systems that did not align well with each other and operated in parallel. A case in point was the nutrition interventions among adolescent girls and women.

Concerns emerged regarding interventions that focused on a single health worker. This was deemed to create work overload for the midwives within the maternity ward. This was feared to negatively affect the quality of maternal and child health services. It was observed that as health workers get overwhelmed policy implementation would be suboptimal due to fatigue.

there is a lot of health worker fatigue because there are many programs that come in and they are focusing on the same health workers. So sometimes health workers have tended to place more emphasis on areas where maybe they have been motivated; #1Professional association

The continued implementation of pilots alongside mainstream policy translation of scaled-up interventions was observed as another likely barrier. Respondents noted that it was creating patches of “islands of success” within the same health systems which was producing different results. Different donors and implementing partners in some aspects continued to promote different priorities and interventions. In some frontline contexts, health workers would implement policy recommendations alongside project-supported pilots which came with varying incentives compared to standard routines on the sidelines.

we also realized that they are overwhelmed by the interventions, everybody is doing something. So, I think it is also a problem that there are so many cooks in this kitchen that the food ends up not being okay. #6 Professional association

The inability to respond to other resource constraints at the frontline such as medical records was feared to create redundancies in the promoted policies. Respondents expressed concern that there was a likelihood for the guidelines distributed at the frontline to be shelved and not serve the intended purpose;

For example, emphasis is again laid on this issue of the partograph but the use is not very okay, it is ill-used simply because some of these labour wards and hospitals get so many mothers and the number of health workers or midwives to carry out systematic WHO layout when you are conducting delivery…..they can’t follow it because of the number of the mothers. #2 Professional association

Outputs

Inputs

Health systems inputs are key to the successful implementation of interventions responding to a public health issue. The absence of essential inputs will impact the level of the health system’s effectiveness [26]. Besides, they demonstrate the level of the health system’s readiness to implement policy recommendations [27]. Our results revealed that although efforts were made to prepare the health systems with many essential inputs to implement evidence-based interventions, some challenges regarding the interventions were persisting.

One of the key findings revealed was the slow recruitment of the required human resource to implement evidence-based interventions to address stillbirth. This observation may have been due to the mismatch or uncoordinated planning within the different government ministries responsible for resource allocation, recruitment and deployment of health workers. Whereas the Ministry of Health (MoH) may have played its part to highlight and address the human resource challenges for preparation of the health systems to implement interventions to respond to the stillbirth burden, recruitment and allocation of finances to facilitate recruitment and subsequent remuneration of human resources was vested outside the ministry of health mandate. Whereas recruitment falls within the health service commission mandate, appointment into public service is by the Ministry of Public Service and their remuneration pretty much rests with the Ministry of Finance Planning and Economic Development mandate. The country has witnessed on-and-off freezes on staff appointments into public service with the current health service staffing structure a mandate under MoH said to be limited about the service delivery mandates. This has coincided with the shrinking budgetary allocation for health from the government which continues to reduce concerning the Abuja declaration recommendation of at least 15% of the national budget. This points to the need for well-coordinated planning and management of intersectoral collaboration if such interventions are to succeed when translated into routine services [28]. It also highlights an important revelation that addressing the stillbirth burden thrives well in a multisectoral approach just like any other maternal and child health intervention [29]. Elsewhere, the value of a multisectoral approach towards addressing public health challenges has been emphasised [28]. within maternal and child health programming, it has been highlighted as an essential ingredient in responding to service delivery bottlenecks [30].

Another key finding was the health workers’ motivation and attitudes. It emerged that both motivating the health workers and their attitudes towards implementation of the promoted interventions were in some ways a challenge. It impacted the effective delivery of evidence-based interventions to address stillbirth. Overall, the factors which affected health worker motivation were the low wages relative to the workload expectations and the inadequate staff accommodation onsite which affected their full-time delivery of maternal and child health services including interventions responding to stillbirth risks. This may have been the case because whereas there are efforts to improve service delivery to address stillbirth through the implementation of evidence-based interventions, resource allocation and focus were not evenly dedicated between policy compliance and other workplace environmental bottlenecks which were likely to affect staff motivation such as accommodation and commensurate salary enhancements. Intrinsic motivation of health workers is a key factor for the effective delivery of health services when considering the realisation of policy objectives [31]. More so, it is even paramount when previously piloted interventions get rolled out within routine health services [32]. Within the Ugandan context, a study conducted by Källander et al shared similar findings where the value of intrinsic motivation of health workers through supportive supervision meetings was highlighted as important for their productivity [33]. Factors like staff accommodation, recognition and appreciation of tasks performed, being availed with adequate resources to provide the required care and fair remuneration were particularly singled out as important for better health service delivery [34]. Indeed, for years now, the MoH has prioritised addressing staff accommodation in efforts to improve health service delivery [35–37]. It is therefore important that while focusing on the rollout of evidence-based interventions to respond to the stillbirth burden it was equally important to pay equal attention to other health systems factors which were likely to affect the productivity of health workers such as salaries and accommodation [35, 36].

Another key finding from our study was that inadequacies in medical supplies impeded the realisation of the policy goals. Respondents mentioned that there were instances when they would be faced with stockouts of supplies which impeded policy implementation. The UN commission designated a basket of essential RMNCH commodities which need to be prioritised to facilitate the implementation of evidence-based interventions [38]. Indeed, the MoH went ahead to localise these recommendations into the routine standard of care where the specific commodities were included in the list of essential medicines to be supplied [39]. Perhaps the observed mismatch could have been brought about by the logistical bottlenecks to the last-mile delivery of drugs and medical supplies in addition to ordering and procurement bureaucracies within the existing practices [26]. This has for some time persisted as a major challenge causing drug stockouts at the final points of health service delivery [40]. From this observation, the key lesson revealed is that for interventions to be effective, there is a need to pay attention to both the inputs such as medicines as well as the promoted guidelines. The requirement to strike a balance is key. The volumes and availability of drugs and medical supplies are essential for the effective implementation of policy recommendations.

Processes

Health managers often faced bottlenecks in integrating new interventions within routine care for the effective delivery of quality healthcare. The promoted evidence-based interventions not only required guideline adherence not only from the service providers but also from the service users. The inability to complete the recommended antenatal care visits and delivery within the health facilities compromised the effectiveness of the promoted interventions. This could have been due to variations in resource availability and allocation to support their implementation between effective pilots and when they were integrated into routine health services. It is especially the case since small pilots tend to be heavily resourced and post impressive results while routine services suffer from chronic resource constraints to support their implementation [32]. Elsewhere, many programs which were found to be effective during pilots tended to fail when taken to scale [41]. This poses a challenge hence leading health systems into a vicious cycle of “pilotitis” (an expression of frustration due to the inability to scale up successful pilots) in the quest to prove that certain interventions work [42, 43]. It remains a major challenge to the sustainability of many interventions which start as pilots in low- and middle-income countries. The implementation context is one attribute that tends to lead to such results [41] as they vary from pilots to real-life implementation. Integration of interventions into routine care requires careful attention to facilitate a seamless transition process. Service users’ feedback and cooperation for the effective implementation of the promoted interventions are as well important. While designing interventions, focusing on the demand side to incorporate user feedback and interests would address some of the barriers while securing their commitment to collaborate on the promoted interventions [40].

Outputs

Despite the translation of interventions into routine care, incidences of stillbirth continued to occur within the health facility settings. It negatively impacted the expected outcomes of these interventions and policy provisions. This was likely due to causes and risk factors which rested outside the critical periods of antenatal care, labour and delivery. Broadly these speak to the social determinants of health. This may have been the case because most of the prioritised interventions save for data and registration systems focussed on the critical periods when most risks are heightened. Due to resource constraints, minimal attention was paid to responding to risk factors that were outside the critical periods such as risk factors along the reproductive maternal life course. This may be the case because of the complexity and the many intervening factors contributing to the eventual risks that would need to be attended to when such interventions are prioritised. Besides, such an intervention would require more investment targeted and “whole of health” and yet results would only be visible in the long term after consistent implementation processes [44, 45]. Such interventions are difficult to secure funding commitment for the duration when results would be realised. Given the current donor and the funding terrain, there tends to be a preference for programs that can demonstrate feasibility within the short and medium term. The complexity of social behavioural interventions has been emphasised before [46, 47]. Whereas social behavioural interventions do not easily secure short-term gains, their benefits in addressing public health challenges in the long term ought not to be compromised. It is therefore imperative that governments and donors equally give priority to the social determinants of health when responding to the stillbirth burden in the long term.

Limitation

This study is not without limitations; first, the purposive sampling and the limited duration of the study may have left out perspectives of other key actors, and therefore the results reported here may have represented a one-sided story more so from health workers directly involved in the implementation of MCH. Secondly, the study relied on information provided by the respondents, and in the context where a lot goes on in the implementation of maternal and child health programs, views may be subjected to a recall bias. Therefore, in the absence of documented experiences, we were unable to triangulate the respondents’ information. Third, there was a risk of social desirability associated with research related to program implementation experiences where respondents may systematically edit out views that may appear to reflect the poor performance of entities they represent. A key strength of this study is that the analysis and discussion draw on perspectives from multiple sources of data including both national-level respondents and key strategic documents and reports related to the implementation of maternal and child health policy in the country. This enabled the triangulation of information provided from more than one source which ensured the reliability of the information collected. Besides, the analysis was conducted deductively drawing on the main study themes, where particular text fitting into a specific theme was applied.