Using Formative Research to Inform a BCC Strategy to Increase Food Security and Dietary Diversity Through a Kitchen Gardens Project in Low-Income Rural Communities

Adequate nutrition is a public health priority, particularly in low-income rural communities where there is a high prevalence of malnutrition and stunting. Baluchistan is an arid desert and mountainous Province with the worst health indicators in Pakistan. The objective of this study was to identify current knowledge, attitudes and practices of vulnerable women with young children residing in remote areas of Balochistan and assess their information needs to guide the development of a BCC nutrition strategy materials and activities to support a Kitchen Gardens Project. The needs assessment incorporated formative research via eight focus group discussions with demand and supply-side program beneciaries (n = 124) to assess current nutrition knowledge, attitudes and beliefs toward kitchen gardens and proposed nutrition resources. Semi-structured interviews (n = 16) were also conducted with key stakeholders. A literature review supported the development of discussion agenda based on predominant behavioural theories. Data analysis was conducted with NVivo qualitative software coupled with grounded theory with triangulation of ndings


Introduction
WHO estimates that globally, 2 million children under 5 years of age are wasted, 17 million are severely wasted, and 155 million are stunted, with around 45% of deaths among children under 5 years of age linked to undernutrition, which mostly occurs in low-and middle-income (LMICs) countries 2 . Child stunting and malnutrition is also associated with lower dietary diversity with food insecurity and poor household food choices contributing to the lack in meeting children's nutrient requirements 1,2 .
Additionally, malnutrition is seen to predominantly occur in regions where there is limited access to high protein foods, poor vegetation cover, and a proxy of rainfall or drought 3 . Socioeconomic determinants also play a signi cant role in food insecurity and diversity of food choices, which lead to childhood and maternal malnutrition 4 .
Baluchistan is an arid and remote Province situated in the South West of Pakistan. The Province covers around 44% of the total land mass of Pakistan, making it the largest of the four Provinces in the region.
Health and nutrition are a major problem in Baluchistan which has the worst health indicators of all other Provinces in Pakistan. Many districts in Baluchistan are food insecure with women and children being most affected by food insecurity and malnutrition. As a result, the Province has the highest prevalence of malnutrition, with a globally critical level of stunting of 47% in children under ve years of age 5 .
There are multiple determinants of stunting and malnutrition that pose challenges toward improved nutrition in Pakistan. Recommendations on strategies to improve nutrition outcomes include a greater focus on poverty alleviation, improvement of mother's health literacy, including community-based education and targeted nutritional interventions, and accessibility to health care facilities. Furthermore, behavioural interventions are seen as being much needed in the Baluchistan context. This includes appropriate and culturally sensitive nutrition education materials which address the currently limited access to nutrition information in geographically isolated areas, particularly with most vulnerable groups 6 .
Given these factors, there is a considerable need for formulating best-practice, behaviour change communication (BCC) programs, tools and resources to improve nutritional outcomes and health seeking behaviours in the Province 7 . However, the nature and extent of BCC activities is uncertain due to the current lack of understanding of the cultural dynamics that may affect dietary attitudes, behaviours and practices in these areas. To support greater integration and institutionalising of more strategic, evidencebased BCC approaches a Kitchen Garden Project (KGP) was established to add value to the efforts of inter-sectoral integration of donor programs aiming to achieve improved nutrition outcomes. The Baluchistan Nutrition Project for Mothers and Children is funded through the Australian Government under the World Bank's Multi Donor Trust Fund for Nutrition. Project components are designed to address general malnutrition in women and children, micronutrient malnutrition, BCC and strengthening of institutional capacity. The BCC component was involved with assessing needs for the development and implementation of best practice KGP messages and a set of basic but innovative nutrition communication materials and resources to improve dietary outcomes. Resources for consideration included a training package to improve family dietary and child feeding practices of mothers with infants and a review of other BCC materials which could support greater participation in the KGP for improved nutrition, to be implemented through a Pilot Project.

Methods
The formative research method used for the development of the BCC Pilot Project Plan comprised of Rapid Assessment and Response (RAR) approach. RAR has been found to be a cost-effective, pragmatic method of public health and social issues research allowing for rapid collection of program intelligence to support the development of BCC plans 8, 9,10 . The application of RAR was of particular importance in the resource constrained and insecure settings of Baluchistan with the eldwork needing to be conducted in a short time period, while ensuring adequate levels of security for eld teams. Despite the challenges, a best practice approach to designing effective behavioural change interventions was embarked upon, including an assessment of participants' health behaviours and examination of the type of BCC resources and actions needed, with the objective of tailoring the interventions to best meet participant needs 11 .
The RAR method included academically robust, qualitative approaches, including: semi-structured interviews (SSIs), Focus Group Discussions (FGDs), observational research, and case study approaches and narratives, which are seen to have generally been underutilised in public health program research. 12 These qualitative methodological approaches to data collection were seen as potentially powerful tools for understanding the culturally-speci c in uences regarding food, dietary practices and nutrition education challenges in Baluchistan 13 .
The rst step of the study involved an audit of existing nutrition resources in the Province with the intention of pretesting key resources as a component of the formative research. Next, was completion of a literature review to provide context for the program. International and national literature was sourced using a rapid review procedure in line with WHO recommendations for swift knowledge generation for priority health issues in resource limited settings 14 . A broad number of online databases were utilised as well as grey literature and internal reports sourced from Provincial stakeholders. Predominant behavioural theories applicable to the study were also identi ed with key variables incorporated into the program logic framework objectives, SSI and FGD agenda. Behavioural theories most relevant to the nutrition project  19 . BCC program objectives emanating from the behavioural theories included increasing target audience awareness and knowledge about the risks of malnutrition; changing attitudes, beliefs and self-e cacy perceptions; shaping community social norms through in uencers and other role models, increasing motivation to engage with the program through incentives; and building practical skills and behavioural intentions toward establishing KGs to grow nutritious, protein dense, produce.
Fieldwork approaches included the development of tailored discussion agenda for the SSIs. Given the ongoing insecurity in the Province a small eld team of three staff was deployed to travel to the capital city of Quetta to conduct the SSIs, and travel onward to project intervention Districts of Kharan and Nushki. The eldwork was conducted over 12 consecutive days.
SSIs were facilitated with 16 key informants as the approach has been found to offer a exible and powerful tool to capture local people's voices and how they create meaning from their experience 20 . Key informants interviewed included Program Directors, Deputy Directors and Program Managers, Provincial BCC Consultants, District Nutrition O cer's, the Head of the Lady Health Workers, and Provincial Nutrition Directorate staff working with at-risk groups. Additionally, Technical Advisor's and Team Leaders from the Women's Economic Empowerment and Agri-Business Units were interviewed as well as Trainers from the Food and Agriculture donor agency, as well as representatives from UNICEF, World Food Program, and the World Health Organisation.
Insights from the SSIs and the literature review informed the development of discussion agenda for FGDs. Eight FGDs were conducted with program bene ciaries as the approach has been found to be a pragmatic method of formative research with community participants who, on their own, may be reluctant to be interviewed or who may feel they have nothing to say 21 . Convenience samples of participants for the FGDs were selected following training workshops and during the eld visits. Group segmentation included demand and supply side participants segmented by gender, age, socioeconomic status and location. A structured screening instrument was used to screen participants prior to recruitment (see Table 1). All study methods were carried out in accordance with relevant guidelines and regulations with reporting experiments on humans. Participants under 18 yrs of age were excluded from the discussion groups, while informed consent was obtained from all other participants through the FGD screening instrument, prior to taking part in the study.
A total of 124 participants took part in FGDs with 8-19 participants attending each of the groups. Male and female group moderators were trained to facilitate like-minded gender groups in local language, with each FGD taking approximately 60-75 minutes to complete. Data collection involved recording of all discussions and then compiling of the recordings into Word format. Recordings were supplemented through dialogue between team members, immediately following each group discussion. Notes were compiled in a question-by-question format to capture what the individuals had to say in regard to each topic with the notes from the SSIs and FGDs totalling 48 single-spaced, A4 pages.
Data analysis was conducted using NVivo software to uncover the most signi cant themes and issues related to the behavioural theories. Software analysis was supplemented with grounded theory approaches which use open, axial and selective coding to identify and categorizing the relationships of various categories as well using selective coding to identify other categories which may be part of the core categories 22 . Data from the literature review, the SSIs and FGDs was triangulated to provide a rich data-source for comparative analysis 23 .

Results
A number of biopsychosocial and cultural determinants emerged from the analysis of the data-sets with the themes highlighted within core categories of project challenges, strengths and opportunities to support the behavioral interventions. Participant responses were also categorised according to key determinants from the predominant behavioural theories. Key challenges for the KGP included: Water scarcity/quality, human resource capacity, and lack of basic infrastructure, environmental and border challenges, gender inequality, and other cultural factors as well as behavioural issues related to poor knowledge and self-e cacy, particularly of women farmers, towards improved nutrition (see Table 2).
A number of strengths were also identi ed to provide insights for improved BCC nutrition interventions in Baluchistan. These included closed but also strong and motivated community networks, opportunities for greater program integration and ownership, cultural factors to facilitate message dissemination and behavioral change, and insights on potential in uencers and predominant communication channels in the remote rural areas (see Table 3).
More speci cally, a number of opportunities for the BCC project emerged from stakeholder and bene ciary discussions. Opportunity categories related to program messages, communication channels, human resources and what incentives could stimulate engagement by community members in nutrition initiatives (see Table 4).

Discussion
The formative research to inform the design of BCC messages and interventions for the food security and KGP related nutrition programs have highlighted a number of challenges. The barrier and bene t analysis has also highlighted program strengths and identi ed opportunities to engage communities more fully in the behavioural change process. Given the dearth of prior BCC program intelligence on what may work in challenging cultural contexts, this formative research was critical in starting to build this evidence base to more fully explore the cultural, and social norms and other environmental and structural factors which need to be addressed through BCC to ensure the success of the program. This was particularly important as the review of behavioural theories highlighted that a number of maladaptive responses may be grounded in long held community customs and deeply entrenched cultural habits 24 . As such, this study has provided program planners with a better understanding of the cultural, socioeconomic, psychological, gender and other barriers to behaviour change in Baloch rural communities as well as providing opportunities on how best to leverage the scarce resources available for improved BCC interventions in the Province.
Cultural factors including deep-rooted dietary habits, gender inequality which disempowers women from decision making, and limited access to health and agricultural information was found to further limit selfe cacy and perceived behavioural control toward desired nutritional behaviour changes 25 . However, other researchers have suggested that greater nutrition-focused approaches to agriculture and the design and training of agricultural inputs may also empower caregivers to healthy dietary practices and thereby improved nutrition to their children 26 . Barriers categories also identi ed that the KGP and associated BCC activities will not be successful if issues of water scarcity and infrastructure are not adequately addressed through a combination of interventions. This may require the development of more speci c programs to ameliorate water insecurity, alongside the nutrition education and counselling interventions 27 . The triangulated ndings highlight the need to address a range of behavioural determinants with BCC messages and interventions, including the provision of ongoing practical advice on planting, growing, maintaining and preparing produce from KGs, while being respectful of cultural issues when developing resource materials.
"We need to be considerate to the local context and culture. Pictorial messages should include both men and women, but women should be properly dressed and covered, and men should be dressed in shalwar and kameez and local hat".

FAO Stakeholder -Quetta
Additionally, opportunities identi ed from the formative research include the willingness of opinion leaders and program bene ciaries to more fully engage in the program if training and educational items are provided to incentivise participants toward improved nutrition outcomes. This con rms ndings from similar programs in other regional countries designed to empower communities toward improved agricultural practices 28 . Other opportunities to support empowerment identi ed through the ndings include the potential for KGs to improve food security while facilitating easier access to food, reducing household expenditure, and improving household diets 29 , including food sovereignty and the localization of food systems 30 through more integrated, holistic approaches. 31 This study found limited diversity in food products in the Province. This may be addressed by applying 'Diffusion of Innovation' approaches and identifying opinion leaders in the agricultural communities to grow and advocate for previously not considered protein dense, farming produce. This could include the provision of BCC materials and training supports for opinion leaders to facilitate the diffusion of greater range of innovations in the form of new seasonal food options to grow in KGs. The approach supports other studies which point to the need to promote dietary diversity as a critical nutritional feature for rural communities located in semi-and arid areas as well as promoting the value of grain legumes, rich in proteins and micronutrients, which also possess good adaptability to marginal rural environmental conditions. 32 Pre-testing of a range of BCC resources also provided programmers with greater insights on the types of materials preferred by participants, including planting calendars, food mats promoting dietary diversity through locally sourced foods, posters, ip-charts and wall branding opportunities. Feedback from women on their motivation to engage in the nutrition related activities and the types of BCC materials preferred to support trainings, highlights the important role of women in achieving family food security and subsequently the need for greater women's empowerment in the conservative rural settings. This emphasis on supporting greater gender equality by building women's self-con dence and control over their food supplies and income is also evident in other country programs1 9, 33 . Additional bene ts not yet realised through the program could include climate change adaptation and the preservation of biodiversity. 34 Last, is the potential to build on the currently limited access to nutrition information cited by bene ciaries through speci cally tailored messages to at-risk groups, and the dissemination of messages through more integrated and culturally appropriate communication platforms. The needs assessment identi ed that women in particular had limited opportunities to build knowledge about KGs, dietary diversity, economical food choices and nutritional bene ts, given literacy challenges and the relatively low access to mass media communication channels. However, respondents did identify various opportunities for BCC program messages to cut through at community levels and through interpersonal communication.
This included communication from trusted lady health workers, already living within these communities, coupled with training and materials incentives, with these activities having the greatest potential to impact on women's knowledge, attitudes and beliefs about KGs and nutrition. Given the critical importance of more fully engaging vulnerable women in remote Districts of Baluchistan, consideration should also be given to including their husbands in BCC activities as a result of their gatekeeping role, highlighted from the study and also identi ed in neighbouring communities 35 .

Conclusion
The formative research rapid assessment method for application in resource constrained and insecure settings has provided considerable insights on how to increase the impact of nutrition programs through the development of culturally appropriate BCC materials and KG engagement in Districts of Baluchistan. The study contributes to the very limited research conducted in insecure, closed and conservative communities, while outlining an iterative, participatory, user-centred approach to support the development of tailored BCC messages and materials, and the identi cation of predominant message dissemination channels. Given these factors, the methodological and eldwork approaches undertaken, and needs assessment considerations would also be relevant to the development of BCC programs in other arid and remote areas of a number of Middle Eastern and African countries, which also have closed, conservative communities, varying levels of insecurity, and facing similar drought/famine cycles. BCC recommendations emanating from the ndings include greater consideration of the use of empowering and personalised messages from community opinion leaders to highlight the bene ts of establishing KGs, the building of women farmer's skills and con dence to improve self-e cacy perceptions and perceived behavioural control, while addressing cultural and gender sensitivities. Insights on program branding and BCC materials designs were also identi ed (to be reported on in another article) as well as highlighting the bene ts of greater engagement by incentivising participants in the process of behaviour change. The issues identi ed through the needs assessment also point to a number of factors which may not be able to be directly impacted with BCC approaches, on their own. Therefore, greater commitments to the provision of infrastructure and nancial supports can facilitate the behaviour change process while further building con dence in the KGP desired nutrition outcomes.

Limitations
There are a number of limitations in conducting eld research in resource constrained and insecure settings such as those found in remote rural Baluchistan. This includes limitations in the time available to collect the data given the risks to eld staff. External validity may have also limited study ndings with the application of subjective, nonprobability sampling using both purposive and convenience samples with voluntary participation. This included the limitation to access trainers and in uencers of broad age ranges of 16-45 years that were available in the local communities, rather than having the time and human resources to segment a larger number of groups by age and location. This was due to the short time span for the work, the deliberate selection of participants based on what was needed to be known and the availability of knowledgeable or experienced people who were willing to provide the information, coupled with the general lack of access to women in the conservative cultural environments. Cultural constraints created additional limitations in the number of participants attending groups in some areas.
Internal bias was minimised through the selection of items for the discussion agenda emanating from the literature review and advice from key informants.

Declarations
Ethics approval and consent to participate All participants provided informed consent to participate in the study prior to the commencement of interviews and were offered the option to opt out at any time during the enquiry. All study methods were carried out in accordance with relevant guidelines and regulations with reporting experiments on humans. Ethics approval for the study was obtained from Dow University of Health Sciences, Ethics Committee (Approval number: IRB-/399/DUHS/Approval/2019).

Consent for publication
The authors provide their consent to publish this study.

Availability of data and materials
Additional data and materials pertaining to this study can be made available on request to the authors.

Competing interests
The authors declare no competing interests, or other interests that might be perceived to in uence the results and/or discussion reported in this paper..

Funding
This project was funded by the Australian Government as part of the Australia Baluchistan Agri Business programme (AusABBA) and the Baluchistan Nutrition Project for Mothers and Children (BNPMC). Tables   Table 1. Sampling frame, group pro les and locations for focus group discussions conducted in intervention Districts of Baluchistan (n=124) GROUP   Female School Teacher -Nushki Low blood (anaemia) is very common in our community especially in girls and women. They feel weakness all the time".

Lady Health Worker-Kharan
Cultural Factors "Our men like to eat meat, for lunch and dinner. We mostly have vegetables and beans but for dinner. I try to cook at least one meat item. But not all the time; maybe two or three times a week".

Female Farmer -Kharan
Gender Inequality "In our society men are given preference over women. More and better food is served to men compared to women".
Female Farmer -Nushki Environmental Factors -Insecurity "Due to droughts, unrest and trade restrictions at the Afghan border; people have now migrated to towns and cities. This in uenced the culture of sourcing food from the land which has increased poverty and badly in uenced health and food insecurity.
Political Leader -GoB We are from this area, they are from this area, and we are neighbours. We go there; they offer us food and drinks. We are like their family; they listen to each and everything we say and try to follow us".

Lady Health Worker -Nushki
Program Integration "We give awareness on Nutrition to mothers' groups, but they lack decision making authority about their children's health. We think BCC will work better if we also educate husbands" Lady Health Worker -Nushki Cultural Factors "When you speak the message, it should be in local language but if it's a written message it should be in Urdu. Well, for anyone that is literate enough to read, Urdu is the basic language."

Male School Teacher -Kharan
Interpersonal Communication "The best channels would be face to face communication. In that way we can certainly convince the other person. Balochistan's local people respect guests and take their advice. Lady health workers and others may be a very good option".

Ministry of Health Stakeholder -Quetta
Motivation to Engage "We would like to attend more activities such as Kitchen Garden trainings and nutrition education programs if they're arranged within our community.
Female Farmer -Nushki) Table 4. Program opportunities identi ed from bene ciaries and stakeholders