Municipality recruitment.

Purposive sampling was used to select information-rich municipality cases for the most effective use of limited evaluation resources [28]. Inclusion criteria to select five Brazilian municipalities were (1) population size, (2) implementation length, (3) implementation model, and (4) geographical region. Four large urban municipalities of Campo Grande-MS, Brasília-DF, Fortaleza-CE, and São Paulo-SP were selected predominantly due to the challenges of implementing PCF in large urban centers [18], and one small predominantly rural municipality (Cuité-PB) was selected for analytical comparison. These municipalities are situated in the North/Northeast (n = 2), Central-west (n = 2), and South/Southeast (n = 1), covering all Brazilian geographical regions.

Implementation team’s recruitment.

In each municipality, a purposive sampling approach was utilized to identify participants were teams implementing PCF (i.e., home visitors, supervisors, municipal managers, and managers from different sectors interacting with PCF) and families participating in the PCF, both for at least 6 months. The sample size was designed to reach thematic saturation per municipality; therefore, in each municipality, families with different characteristics (i.e., pregnant, caregivers of children under 36 months, child beneficiaries of Cash Transfer Programs) were intentionally interviewed. In-depth interviews were conducted virtually (with or without video according to the participant preference) or via phone with a sample of (a) municipal PCF teams, including municipal coordinators, supervisors, and home visitors (N = 132); (b) municipal managers from other sectors working on the implementation of the PCF, such as health, education, and social assistance (N = 17); and (c) families participating in the PCF (N = 95) (participant characteristics detailed in S3 and S4 Tables). No participants refused to participate. The data were collected from June 2021 to May 2022.

Method of approach.

The research team hosted a virtual meeting with each municipality to introduce the study and establish the protocols for scheduling virtual or telephone interviews. For selected families, the supervisor and home visitor assigned to each family were contacted and granted authorization from the family to share their contact information with the research team. A research assistant contacted the families by phone to explain the study’s purpose and interview procedures, as well as share a virtual copy of the consent form. If the participant agreed to be interviewed, the research assistant scheduled a convenient time for the interview. A similar approach was followed with PCF teams and municipal managers from other sectors working with PCF.

Interview guides.

Interview guides were developed to capture factors influencing PCF implementation contexts. In the interview guide for PCF team and managers, questions followed the RE-AIM Framework dimensions, i.e., Reach, Effectiveness, Adoption, Implementation, Maintenance [29,30]. Interview guides were customized to address the specific needs of different program audiences, resulting in two distinct versions pre-tested with a member of the PCF municipal team and a family member. Pertinent to this study, specific questions about programming approaches to address food insecurity were taken from the implementation dimension. In the interview guide for families, questions followed the Nurturing Care Framework dimensions [31]. Relevant to this study, specific questions about food insecurity and participation in social protection programs in addition to the PCF were taken from the adequate nutrition dimension (S5 Table). Both interview guides had a section on sociodemographic characteristics of participants, which included the validated two-item screening tool (TRIA) to identify Brazilian households at risk for food insecurity (sensitivity of 79.31%, specificity of 92.95% [32]. Specifically, TRIA uses questions 2 and 4 of the Brazilian Food Insecurity Scale (EBIA): question number 2 in the EBIA (“Nos últimos três meses a comida acabou antes que você tivesse mais dinheiro para comprar mais?”) corresponds to HFSSM question Q2: “In the past 12 months, the food that (I/we) bought just didn’t last, and (I/we) didn’t have money to get more?”); and question number 4 in the EBIA (“Nos últimos três meses, você teve que se arranjar com apenas alguns alimentos para alimentar os moradores com menos de 18 anos, porque o dinheiro acabou?”) corresponds to HFSSM question Q4: “In the past 12 months, did (you or other adults in your household) ever cut the size of your meals or skip meals because there wasn’t enough money for food?”.When a participant caregiver responded affirmatively to both items of the instrument, their household was considered at risk for food insecurity [32].

Research team’s positionality.

Interviews were conducted in Portuguese by trained native Portuguese speaker female researchers with PhD training (GB, MBG, PPS). Data analysis of the food insecurity component of interviews was conducted by female faculty members and co-authors trained in public health nutrition and with experience in research on food insecurity and early childhood in Brazil and the U.S. They all had prior experience conducting qualitative interviews and had no prior relationship with the participants. Participants were told the goals of the study prior to the interview.

Data management.

In-depth interviews were conducted throughout the months of June 2021 and May 2022. Debriefing meetings were held at the end of each day of interviews to expand field notes and conduct preliminary data coding and analysis. Through this approach, when thematic saturation was achieved in each municipality, the sample size was deemed sufficient for that municipality [33]. Interviews lasted 40–70 min each, were audio recorded with permission after verbal consent was granted, and transcribed verbatim by a professional Portuguese-speaking service. Original audio recordings were compared to transcripts to ensure accuracy.

Analytical framework.

The GTE framework was used in this study to integrate qualitative data analysis findings. The GTE framework was first published in 2018 [23], and applied specifically to food insecurity in 2020 [24]. It provides solution-oriented guidance aimed at reducing health disparities, and to eliminate structural factors that inhibit food and nutrition security for all, especially low-income children and families [24]. The GTE has four categories (1) increase healthy options, (2) reduce deterrents, (3) improve social and economic resources, and (4) build on community capacity that help translate the intention to achieve equity into action.

In this study, the premise for using GTE is that disparities in early childhood development cannot be overcome without addressing underlying inequities that cause food insecurity [7–9]. Therefore, in our study, food insecurity is not considered the root cause of existing inequities rather is a metric used to understand who was most affected by the root causes (e.g., poverty). Thus, inequities that GTE informed-recommendations are addressing contribute to food security as a consequential issue.

Thematic analysis approach.

Rapid qualitative methods, defined as research designed to be responsive and adaptable to changes in context while the study was ongoing [34]. This methodology has been widely used in the field of implementation science to expedite knowledge translation and inform policy decisions [34,36]. A three stage thematic analysis approach was used to analyze data (Fig 1) [35].

Download:

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

Fig 1. Stages of data analysis.

https://doi.org/10.1371/journal.pone.0329310.g001

In the first stage of analysis, two researchers listened to interviews’ audios and recorded information in a predefined matrix to capture themes related to food insecurity throughout PCF implementation. A third researcher compared and validated the information recorded. Questions were discussed and confirmed throughout revisiting original audios. Then, an open-coding process [35] was used to inductively label strategies to addressing food insecurity. During this stage, frequency, complementarity, expansion, and divergence of coded information were considered in order to identify central codes related to food insecurity at family, organizational, and system levels [35,36]. Several central codes related to food insecurity emerged as challenges to the implementation quality of the PCF [18] (Fig 1). The preliminary findings of the first stage of thematic analysis were presented to each municipality for feedback, fostering reflection between implementers and the research team. Insights from this interaction generated an important research question of how the high prevalence of food insecurity among participants influence PCF implementation at family, organizational, and system levels. This question guided the second stage of thematic analysis as well as the integration of data included in this manuscript.

In the second of analysis, in-depth analyses were conducted to understand how food insecurity impacts the internal and external contexts of PCF operations. Thus, the central codes related to food insecurity identified in phase 1 of data analysis were deductively grouped into two major themes by similarity, complementarity, and meaning as follows: (1) internal context in which PCF operates (i.e., organizational-level factors such as program curriculum) and (2) external context in which PCF operates (i.e., family-level and system-level factors). Quotes from interviews that best represented themes were selected to illustrate the results and were translated from Portuguese (Fig 1).

In the third stage of analysis, based on the preliminary findings of the second stage of analysis and literature review [23,24], our team operationalized definitions for the four categories of the GTE. These definitions were used by co-authors to integrate findings from the second stage of thematic analysis into a cohesive and pragmatic set of illustrative examples of equity-focused recommendations to navigate food insecurity in the context of early childhood programming (Fig 1).

Theme 1. The internal context in which PCF operates.

1.1. Program implementation lacks guidance on how to engage families navigating food insecurity. All interviewees reported that the COVID-19 pandemic exacerbated existing economic and social vulnerabilities leading to high prevalence of food insecurity risk among households participating in PCF. Several families participating in PCF described challenges in accessing food due to limited income and high food prices. Specifically, caregivers shared that the uncertainty of having enough food was a major household stressor, which strained the relationships in the household, as described by a mother participating in the PCF:

Here our financial conditions to buy everything to eat is difficult, it’s a little tight, isn’t it? (...) To be honest, I don’t remember (...) when I last went to the store to buy lettuce or squash, or okra, a maxixe, I don’t remember when I went (…) It’s pretty bad [dealing with food insecurity]. Sometimes I get stressed. I take it out on my husband, but it’s not his fault, is it? Thank God he works. But... things these days are so expensive. Everything is difficult (caregiver, Campo Grande).

Home visitors found it challenging to engage caregivers struggling with food insecurity in early learning activities such as play, games, and fun activities. They questioned whether teaching these activities should be a priority when people may be experiencing hunger. From their perspective, these activities designed to primarily strengthen the bond between caregiver-child to promote early learning are neither feasible nor appropriate in the context of many competing social needs. In this context, from the perspective of home visitors and families, the effectiveness of learning activities available from PCF was undermined by the food insecurity experience, as emphasized by a home visitor:

Many [of the caregivers] do not see this positive side [in a child’s early learning provided by the PCF]. It’s very difficult to make a family that is often hungry understand the objective [of the PCF], what’s gained in learning, [what is the point] of playing with the child (home visitor, São Paulo).

1.2. Program curriculum lacks protocols to guide the implementation team in addressing food insecurity. When asked about protocols to assist households at risk for food insecurity to access food, home visitors and supervisors reported the absence of a government official guidance on how to navigate food insecurity within PCF. The most immediate response from PCF teams to address food insecurity was finding ways of accessing food baskets to donate to families in need. However, PCF teams reported that the supply of food baskets was insufficient to meet the demand from families in need, as reported by a home visitor:

The difficulty we have at the moment is to meet people’s demands because hunger is constant and food is a daily necessity. (…) We have a low supply [of basic food baskets] and a huge demand. There’s never enough money (…). And that’s what makes everyone anxious (…), but often our hands are tied, because we can’t help all the families. (…) It’s very complicated. (home visitor, Cuité).

Despite the lack of specific official guidance to manage screening and referral of families reporting having difficulties accessing food, home visitors used available resources to mitigate the issue. For example, they mentioned getting food baskets to households, which involves connecting families with official social protection services. The process of connecting with social protection services was described as bureaucratic, hierarchical, exhausting, and many times did not result in solutions for families, as highlighted by a home visitor:

The person needs a food basket (…) You have to make a request, and you go to the supervisor, and the supervisor goes to I don’t know where and nothing happens. When you go to see the person, they haven’t even received anything (home visitor, Campo Grande).

The lack of resources and guidance generated a lot of frustration among home visitors when working with food-insecure households. Indeed, some home visitors shared that when no formal resolution for cases was found through the official social protection services, they used their personal or collective financial resources to purchase food baskets for families, as stated by a home-visitor:

When you try to get a basket and you can’t find it. Then, you have to do it yourself. But sometimes you also do not have the money to buy it, because nowadays everything is very expensive. Then I started asking [colleagues to chip in with money] to buy a food basket (…) (home visitor, Brasilia)

Although reporting informal resolutions, home visitors understood these were not long-term solutions for families. However, most of them shared that in critical situations like hunger, they could not move on with their day without providing any type of support to the families.

Theme 2. The external context in which PCF operates.

2.1 Family expectations concerning the program. Families with young children assisted by the PCF reported the COVID-19 pandemic as a major event exacerbating economic strains (e.g., lack of money for food and other competing basic needs) and social vulnerabilities (e.g., unemployment, domestic violence, household stress). Coping strategies were also reported to manage strains and vulnerabilities. Specifically, families struggling with food were prone to ask for help from other people to get food, skip meals or not eat meals to save food for their children’s next meal. When families were asked whether PCF could help them navigate food insecurity, caregivers expressed the expectation that PCF would provide stipends or food baskets tied to program participation, as illustrated by the following quote:

How difficult is that? Because we don’t have any benefits [participating in PCF], and we eat what we have, [with the] help of other people, so we have to save as much as possible, especially for our children, we stop eating to give to our children (…) [The PCF] could help me get something. Especially with food stamps, right? With gas, everything helps… (mother, Fortaleza).

2.2. Family involvement in food and income assistance programs. It is important to acknowledge that the families’ expectations were unrealistic since there is no financial assistance tied to PCF participation. However, supervisors and home visitors from one municipality shared a unique experience that proved two opposite points. On the positive side, policies to mitigate food insecurity can be rapidly enacted. On the negative side, rapidly enacted policies can pose a threat to program implementation, i.e., leading to confusion about the inclusion criteria of one-time benefits and the continued home visits offered by PCF During the pandemic, a specific municipal government decided to prioritize families participating in PCF to receive a one-time food stipend (R$250 [~US$50]). PCF teams were responsible for confirming families that were actively participating in program activities and the social assistance secretary was responsible for registering eligible families in the food stipend program. However, this one-time effort to mitigate food insecurity had limited coverage and left many families in need without access, which created confusion among families as to why some received the stipend and others did not, as highlighted in the quote from a home visitor:

As the PCF team, we confirmed the list of people active in the program, including information on those who have already received the food stipend program, [those who] are no longer receiving it, and [those who] will receive again (…) The process that takes the longest to put these people in the food stipend program is the interview with the [social assistance] worker. Thus, the only thing we [as a PCF home visitor] can do for those asking about the food stipend program is to request [the families] to call again [to schedule an interview with the social assistance worker] (home visitor, Brasilia).

In a way or another, PCF families often found themselves without adequate support when experiencing food insecurity. Increasing family income through participation in the national conditional cash transfer program (i.e., Bolsa Familia Program) emerged as a facilitator to reduce economic vulnerabilities, which in turn would directly affect food insecurity. Despite that, several home visitors expressed concerns about the COVID-19 pandemic-era government that changed some of the eligibility criteria of Bolsa Familia, reducing the number of eligible families. From the perspective of those interviewed, this was worrying because it could increase food insecurity among children. Specifically, families stated that the monthly cash transfer was used to buy food for their children, which in turn shielded them from food insecurity, as reported by a caregiver:

My husband is unemployed, we are surviving, getting hourly day work here and there; then he buys a little something, however, our child’s food is specifically [bought with money] from Bolsa Familia. I just buy her food, and the rest of us make do (caregiver, Cuité).

2.3 Formal networks to enhance food security. PCF managers recognized challenges in promoting sustainable solutions to mitigate food insecurity due to bureaucratic processes in accessing government safety net (i.e., social protection services) as well as limited non-government safety net (i.e., community resources). At the time of data collection, when families shared with home visitors, they are experiencing hunger, particularly not having food for the day; then, home visitors would notify supervisors, who in turn would notify PCF managers. Managers were able to refer these families to municipal social assistance centers. These municipal social assistance centers are responsible for assessing the needs of each family and enrolling them in as many social programs as they are eligible for such as Bolsa Familia, as seen in a home visitor’s statement:

“So, we bring to municipal social assistance centers [the needs of each food insecurity family]. Then municipal social assistance centers within the social assistance sector make the move to call families and schedule a time for them to attend an in-person meeting at the municipal social assistance center. When families attend this meeting, they [families] receive what they need for example a food basket, rental assistance, and sometimes, even financial assistance with medication, when the family is very vulnerable, because many are. So, everything comes through the municipal social assistance centers” (home visitor, Cuité).

According to the families and home visitors interviewed, there was a huge waitlist with waiting times varying from one to seven months or more just to schedule an initial interview, which is required to access the services offered by the municipal social assistance centers. These waiting times were reported to be increased due to the high demand during COVID-19. The slow and bureaucratic process to get enrolled into the municipal social assistance centers resulted in families not having prompt access to social assistance programs to help them with economic vulnerabilities. The bureaucratic process was described by the interviewees as a lack of system preparedness during the pandemic. All interviewees agreed that the municipal social assistance sector had limited resolutions to address timely the complexity of food insecurity, as one manager noted:

“It’s noteworthy. The problem of food insecurity exceeds what the social assistance sector can offer at all levels (…) both the municipal social assistance centers and our [social assistance] sector are not prepared and cannot address this complex issue of food insecurity” (PCF manager, São Paulo).

In this context, other community resources such as daycare emerged as protective factors against hunger among young children participating in the PCF. Many families stated that their children depended on the food provided by daycare centers. Thus, the shutdown of daycares during the COVID-19 pandemic meant that their children consumed less food. Therefore, the return of daycare operations was considered by all interviewees a positive measure to mitigate food insecurity among young children participating in the PCF, as highlighted by a caregiver:

“(…) now that school [daycare] is back on, children aren’t so bad off. Because they have food at school. But it’s lunch and dinner. Sometimes there’s no breakfast, no snack. Understand?” (caregiver, Brasilia)

2.4 Informal networks to enhance food security. In search for additional resources to support food-insecure households, home visitors reported engaging informally with community organizations, including non-profit neighborhood organizations, churches or religious places, and local grocery shop owners. Home visitors shared that building this informal/non-governmental network was strategic particularly during the COVID-19 pandemic as they used it to obtain donations of food baskets to families in urgent need of food, as highlighted in a home visitor quote:

“(…) We do it like this: in this case, I seek out [to build a relationship with a] neighborhood association (…) When I notice that a family is in a very vulnerable situation, then I ask [to the neighborhood association] if they have a food basket to give [to that family in need], understand? (home visitor, Fortaleza)

Home visitors acknowledged that many times they were not able to obtain food baskets through these informal networks, which also depend on outside donations. On one hand, building this informal network was valuable for their work. On the other hand, the informality of these relationships made it hard to have it as a sustainable option for food-insecure families.