Using scenario‐based assessments to examine the feasibility of integrating preventive nutrition services through the primary health care system in Bangladesh

Abstract The National Nutrition Services of Bangladesh aims to deliver nutrition services through the primary health care system. Little is known about the feasibility of reshaping service delivery to close gaps in nutrition intervention coverage and utilization. We used a scenario‐based feasibility testing approach to assess potential implementation improvements to strengthen service delivery. We conducted in‐depth interviews with 31 service providers and 12 policymakers, and 5 focus group discussions with potential beneficiaries. We asked about the feasibility of four hypothetical scenarios for preventive and promotive nutrition service delivery: community‐based events (CBE) for pregnant women, well‐child services integrated into immunization contacts; CBE for well‐children, and well‐child visits at facilities. Opinions on service delivery platforms were mixed; some recommended new platforms, but others suggested strengthening existing delivery points. CBE for pregnant women was perceived as feasible, but workforce shortages emerged as a key barrier. Challenges such as equipment portability, upset children and a fast‐moving service environment suggested low feasibility of integrating nutrition into outreach immunization contacts. In contrast, CBE and facility‐based well‐child visits emerged as feasible options, conditional on having the necessary workforce, structural readiness and budget support. On the demand side, enabling factors include using interpersonal communication and involving community leaders to increase awareness, organizing events at a convenient time and place for both providers and beneficiaries, and incentives for beneficiaries to encourage participation. In conclusion, integrating preventive and promotive nutrition services require addressing current challenges in the health system, including human resource and logistic gaps, and investing in creating demand for preventive services.


| INTRODUCTION
Nutrition has become central to the development agenda with 12 of the 17 Sustainable Development Goals being directly or indirectly linked to improving it (Grosso et al., 2020). Globally, large strides have been made to address undernutrition in the past decades, yet maternal and child nutrition remains a significant public health concern, particularly in low-and middle countries (Black et al., 2013;Victora et al., 2021). In Bangladesh, stunting among children under 5 years declined from 60% to 31% between 1997 and 2017, wasting from 21% to 8% and underweight from 52% to 22% (NIPORT, 1997(NIPORT, , 2020. Much of these changes in nutrition are explained by nutrition-sensitive improvements such as increases in income, education and access to family planning, but coverage of nutrition-specific interventions remained low (Nisbett et al., 2017). Less than half of women (47%) attended at least four antenatal care (ANC) visits, 49% had institutional delivery (NIPORT, 2020) and 35% received child growth monitoring (Nguyen, Khuong, et al., 2021).
Community-based nutrition interventions have been found to improve maternal and child nutrition status in low-and middle countries (Majamanda et al., 2014) and have long been a policy focus for the Government of Bangladesh. The first communitybased nutrition interventions were implemented between 1996 and 2011 (Saha et al., 2015) called the Bangladesh Integrated Nutrition Project and later the National Nutrition Program. These programmes covered only 110 Upazilas (subdistricts) with negligible involvement of primary health care frontline workers in service delivery. A large-scale nutrition initiative, the National Nutrition Services, began in 2011 (NNS OP, 2011, enabling the provision of mainstreamed nutrition interventions through the existing health system structure, the Bangladesh Essential Services Package. The service delivery platforms spanning health facility and community levels are mainly focused on curative services, outreach at satellite clinics, Expanded Program on Immunization (EPI) and ANC (Government of Bangladesh, 2016).
Delivery platforms such as primary health curative care facilities are less likely to invest in preventive outreach programmes (Saha et al., 2015). For example, in rural Bangladesh, growth monitoring and promotion (GMP) is integrated into facility-based curative care, bearing several challenges related to inadequate coordination, training, supervision, logistics and supplies, hindering the implementation of GMP (Billah et al., 2017). In addition, only a subset of children is reached by services provided during sick child visits, for whom preventive services are a lower priority than curative services. Although nutrition assessment and counselling are a key component of Integrated Management of Childhood Illness, these are not prioritized (Saha et al., 2015) as there are no dedicated frontline nutrition workers (NNS OP, 2011). For pregnant women, nutrition interventions are mainly delivered during facilitybased ANC, which is a preventive platform. However, other challenges are present including low coverage of at least four ANC (NIPORT, 2020), suboptimal ANC quality (Nguyen, Khuong, et al., 2021) and persistent inequalities in the accessibility of quality care (Anwar et al., 2015;Hajizadeh et al., 2014). The frontline workers in the public health system are designated to provide nutrition services, but there are missed opportunities to prioritize and raise awareness of the importance of nutrition with families and communities since the community-based services are primarily focused on family planning and routine immunization (Bangladesh, 2011).
To improve service coverage and quality, there is a need to strengthen nutrition services by optimally utilizing contact points, while integrating key nutrition interventions through new service delivery points. Previous assessments documented nutrition service delivery gaps (Billah et al., 2017;Saha et al., 2015), yet limited attention has been paid to the feasibility of reshaping service delivery to close gaps. Our study assesses the feasibility of strengthening and reshaping existing service delivery platforms at the community level to provide preventive nutrition services to pregnant women and young children, focusing on two key research questions: (1) What is the perceived feasibility of reshaping existing platforms or introducing new platforms to deliver preventive nutrition services? and (2) What are the barriers and facilitators to implementing these interventions? 2 | METHODS

| Study setting
This qualitative study took place in two divisions, Chattogram and Sylhet. These divisions have been prioritized by the government to strengthen the core management systems and delivery of essential health, nutrition and population services. Two districts namely Feni (Chattogram division, South-East of Bangladesh) and

Key messages
• This study uses a scenario-based feasibility testing approach to explore potential interventions to strengthen preventive and promotive nutrition service delivery through the primary health care system.
• Our findings highlight three highly feasible potential platforms (community-based events [CBE] for pregnant women, CBE for well-children and well-child visits at facilities) to expand preventive services. Scaling these community-based services requires addressing current challenges in the health system (including human resource and logistic gaps) and investment in demand creation for these services. Union facilities, and (4) Community clinics. In each district, we selected two Upazilas (subdistricts), yielding a total of four Upazilas. From each Upazila, we randomly chose one union (a total of four unions), and from that union, we randomly selected one community clinic (a total of four community clinics) through a manual lottery. Details of the sampling frame are presented in

| Participants
We conducted in-depth interviews with service providers, supervisors (i.e., Health Inspectors and Family Planning Inspectors) and managers (n = 31, including 7 at the Upazila level, 8 at the union level and 16 at the community level) (Figure 1). We also conducted key informant interviews with policymakers at the national level (n = 7, including representatives of the public sector, civil society and development partners) and district level (n = 5, including one Deputy Director of Family Planning, two civil surgeons, and two nongovernment partners). We conducted five focus group discussions with pregnant women and mothers of children <2 years (each with six to eight women).

| Data collection
We developed scenario-based in-depth interview guidelines for each participant category based on four potential platforms to reach beneficiaries, focusing on preventive rather than curative care (Table S1). These included one platform to reach pregnant women, and three to reach well-children under the age of 5 years ( Figure 2). The scenarios were developed by the researchers from evaluation and programme teams, in consultation with government stakeholders, based on reviewing the literature on current policy and programmes to deliver nutrition-related services for pregnant, F I G U R E 1 Sampling frame and sample size NGUYEN ET AL. | 3 of 14 lactating women and children. Team discussions were used to conceptualize how each of these platforms would be defined and presented to respondents, and to frame feasibility-related questions on where these contacts could be, what services could be provided there, who would provide them and how, and anticipated supply-and demand-side challenges.

| Community event for pregnant women
This was defined as a counselling-focused event held in the community for pregnant women, and open to family members. The discussion could centre around topics such as nutrition and care during pregnancy. Health care providers could be present to answer client questions, discuss challenges and provide counselling messages. Counselling would be group-based and not one-on-one.

| Well-child services for children
Well-child services were presented to respondents as preventive nutrition services delivered to children either at an EPI session, a separate community event or facilities, where they would be brought for routine growth monitoring and counselling, and not when they are sick. These services would enable caregivers to receive agespecific advice and counselling messages on feeding and caring for their children. Messages or referrals would also be tailored to the child's nutritional status, which would be assessed through services such as length/height, weight and/or mid-upper arm circumference measurement. Potential platforms for young children could either be fixed-day/fixed-service (such as integrating services into EPI or holding a separate community event) or a routine service available at facilities (such as a well-child visit protocol at facilities).
Providers, managers and policymakers were asked what kind of nutrition services could be provided through these platforms, how feasible these are, which providers would be responsible for delivering services, how to implement the events and key considerations. Beneficiaries specifically were asked whether these services would be useful to them, why they would or would not utilize them and what kind of challenges may arise.
All interview guidelines were pretested in the Bhaluka district of the Mymensingh division by experienced Research Investigators and Research Assistants to check on the flow, contents and consistency of the questions, and to contextualize the guidelines in real settings. All feedback from pretesting interviews was discussed and incorporated into guidelines. Data were collected by a team of well-trained qualitative researchers from icddr,b. Interviews were conducted in Bengali and recorded in their entirety with consent from the respondents. All focus group discussions were conducted by at least two researchers, including one facilitator and one notetaker.

| Data analysis
Data analysis took place using the framework approach (Smith & Firth, 2011), providing a systematic structure to manage, analyse and identify themes (Ritchie & Spencer, 1994). Recorded interviews were transcribed verbatim in Bengali. Field notes and interviewers' observations were incorporated into the transcripts. Transcribed data from the early interviews were compared to assess how similar issues were discussed by different types of interviews and to identify gaps in data exploration which could be investigated further during subsequent interviews. Before working with raw data, a set of a priori codes were identified based on interview guidelines and study objectives, allowing for emergent codes during the analysis F I G U R E 2 Data collection framework (Table S2). The final code list was developed when all interviews were coded and condensed. Transcripts were analysed by identifying emerging themes and subthemes and highlighting common ideas and recurrent themes. Key issues, concepts and themes were based on the objectives of the study. Finally, data were systematically indexed and coded, synthesized, and interpreted. Results on the same issues from different types of respondents and areas were compared to strengthen the validity of the findings. To ensure quality, multiple researchers coded the same transcripts and at least two researchers coded each cadre of interviews, with regular discussions to resolve divergent results.

| Characteristics of study participants
Among participants involved in key informant and in-depth interviews, 55% were male and 59% were aged 25−44 years. Nearly half had a postgraduate degree while 27% completed higher secondary education. Considering their professional experience, 45% had 1−9 years of work experience and 32% had 20−29 years of experience. All participants in the focus group discussions were female in the young age group (85% aged 20−29 years). Around 80% of them completed primary education and most of them are homemakers.

| Perception of the feasibility of community events for pregnant women
Almost all health service providers (n = 16), a few health managers (n = 2) and national-level policymakers (n = 3) opined that organizing a counselling event for pregnant women on basic ANC components is feasible (Table 1). While some health service providers suggested that this event can be arranged in a certain place (such as a house, school or the community clinic), some other health managers and national-level policymakers further added that such events can be arranged at the EPI sessions and community clinic instead of finding a new place.
We have satellite clinics to reach the pregnant mothers which are organized by Family Welfare Assistant. The Family Welfare Visitors visit the outreach centers twice a week and provide ANC, postnatal care, childcare and various family planning methods. So, there is no need to arrange a separate event as these are available in satellite clinics. UFPO) For the frequency and timing, most respondents suggested that events should be held once a month, but some suggested once every 3 months. Mornings were identified as the convenient time for women as they have a huge workload at noon. A health manager further explained that such events must be arranged on a workday.
It would be good if it is arranged once a month, I wish a doctor could join me to provide services.
(IDI-01, CHCP) One health manager suggested the existing providers will be able to carry out the tasks of a separate event for pregnant women, but others (n = 2) cautioned though, that in such an event the provider

| Well-child services incorporated into EPI
Mixed reactions were found regarding the feasibility of providing preventive nutrition services (i.e., measurement and counselling) in EPI sessions, ranging from enthusiastic positive to strong negative ( • GMP activities should take place before immunization-afterwards, the babies will be upset, and caregivers will not be attentive • Nutrition services must take place before immunization • At least two staff are required-one for taking measurements, and one for counselling • Nutrition services must take place before immunization • Collaborations between the government and NGOs to provide technical support • Any services must be given before

| Separate well-child event
In contrast with integrating well-child service within EPI, the separate well-child event was thought to be feasible and desirable by most respondents ( challenges and offers recommendations to strengthen these services.
Previous government nutrition programmes, that is, the

Bangladesh Integrated Nutrition Project and the National Nutrition
Program could reach only 25% of the entire country (Saha et al., 2015).  (Anwar et al., 2016;Rahman, 2022).
Role clarification for conducting the proposed preventive nutrition service delivery through fixed day fixed service events has been raised as an important concern given the workload and gaps in existing human resources. Though Bangladesh is one of the very few countries that has successfully scaled up and sustained its community-level workforce (El Arifeen et al., 2013), the country has only 3.9 community health workers per 10,000 population (DGHS, 2015 Over the years there have been consistent increases in ANCseeking, facility delivery and care-seeking for childhood illness (Billah et al., 2017). The societal and cultural barriers to careseeking outside the home are diminishing. Increasing the awareness among both parents and health care providers as well as ensuring consistent availability of quality care is likely to improve careseeking preventive nutrition services (Billah et al., 2017). To generate awareness of and demand for these new services, communities should be sensitized through interpersonal counselling and courtyard meetings, and the involvement of community leaders and community groups. Other suggested strategies to improve attendance in CBE include providing free food or snack for beneficiaries, selecting convenient times and places for both providers and beneficiaries, making female health care providers available, even after clinic hours and outside the clinic, and activating and scaling up the involvement of Community Groups, Community Support Groups and Multipurpose Health Volunteers.
This study is influential for informing future implementation research on community-based platforms to strengthen MIYCN services. The scenarios were carefully framed based on prior research and existing knowledge of the health system in Bangladesh. We acknowledge the limitation that responses to the potential platform could be aspirational or overly negative depending on the current workload and circumstances of the participants because they have not fully experienced some of the platforms. To minimize the response bias, we have detailed the hypothetical platform scenarios, given relevant examples from existing practice, and increased probing. The study did not include key informants from the private sector. Investigating their perspective on the nutrition service at health care contacts at for-profit private facilities could complement the public service platforms.

| CONCLUSION
A scenario-based assessment efficiently identified potential platforms to bring MIYCN preventive and promotive services closer to the community in Bangladesh. The approach successfully identified the bottlenecks and pertinent system strengthening areas and enabling factors for such services to be successful. Integrating and scaling up preventive and promotive MIYCN services would require addressing current challenges in the health system, including human resource and logistic gaps, and investing in creating demand for preventive services.

AUTHOR CONTRIBUTIONS
Phuong supported data interpretation; reviewed and edited the manuscript.
Deborah Ash: Reviewed study design and tools; draft discussion; data interpretation and its implications; reviewed and edited the manuscript. All authors read and approved the final submitted manuscript.