Abstract
As a response to the rise in microcephaly cases in 2015, Brazilian health authorities recommended a number of Zika virus (ZIKV) prevention behaviors. This study explores the contrast between the Brazilian health authorities’ health promotion response to the epidemic and the context in which the epidemic unfolded. Rapid Anthropological Assessment was used to assess how women in Fortaleza, Brazil, perceive ZIKV, ZIKV prevention recommendations, and the feasibility of adhering to these recommendations. Semi-structured interviews, informational interviews, and observations were used. The ZIKV recommendations and prevention did not represent best practices in health communication and promotion and failed to achieve their goals. Prevention recommendations were delivered without actionable steps and without considering cultural, socioeconomic, or environmental contexts. It is imperative to take advantage of this interepidemic period to apply best practices in health communication, education, and promotion to ensure populations at risk have adequate awareness, information, and resources to prevent infection.
Zika virus (ZIKV) reemerged as a global threat in 2015 with Brazil at its epicenter. The flavivirus can be transmitted by Aedes aegypti mosquito, through sexual intercourse, and from mother to child during gestation (World Health Organization [WHO], 2018b). ZIKV symptoms are like those of other common vector-borne diseases in the area—mild fever, headache, body ache, possible rash, and general malaise (Centers for Disease Control and Prevention, 2019). ZIKV’s ability to pass through the placenta during pregnancy (Adibi et al., 2016) can lead to negative birth outcomes such as microcephaly and other developmental issues. This group of ZIKV-associated birth defects is termed Congenital Zika Syndrome (CZS) (WHO, 2018a). There have been over 200,000 confirmed cases of ZIKV, approximately 60% of which have been reported in Brazil (over 137,288 cases) (Pan American Health Organization [PAHO], 2018). Furthermore, over 3,720 confirmed cases of CZS have been documented, with approximately 80% of those cases in Brazil (over 2,952 cases) (PAHO, 2018). A Public Health Emergency of International Concern was declared by the World Health Organization on February 1, 2016 (WHO, 2016b) and then lifted on November 18, 2016 (WHO, 2016a). Although reported ZIKV incidence has decreased substantially, the virus persists as an endemic disease with local transmission in many areas of Brazil and a continued threat of viral mutations that might bring new consequences. In June 2020, Brazilian researchers at the Center for the Integration of Data and Knowledge for Health published findings of the first new ZIKV viral lineage since 2015, suggesting a potential for a new wave of global ZIKV epidemics (Kasprzykowski et al., 2020).
Best Practices in Health Communication
There is no evidence that best practices for health communication and promotion were followed in response to the ZIKV epidemic. Best practices in health communication create strategies that (a) are informed by behavior change theories; (b) are tailored to the specific priority population, health issue, context, and behaviors; (c) are actionable (audience is advised which behaviors to uptake to avoid health issues); (d) are instructional (audience is taught how to complete those behaviors); and (e) acknowledge the multilevel contextual factors that influence the barriers and benefits to adopting the healthful behaviors (Rimer et al., 2005; WHO, 2017). We argue that these best practices were not sufficiently utilized when Brazilian authorities developed their ZIKV health communication and promotion.
Contrast Between Health Promotion Campaigns
The World Health Organization has promoted three major forms of ZIKV prevention: mosquito avoidance, condom use, and reproductive health counseling with health care providers (WHO, 2016c). In contrast, the Brazilian Ministry of Health created a set of recommendations for ZIKV prevention that focused mainly on mosquito and pregnancy avoidance, poorly publicized the sexual transmission of ZIKV, and deferred the responsibility for family planning and reproductive health counseling to the clinics and health care providers with no additional training for those health care providers (Ministério da Saúde, 2017). The core of the Brazilian health authorities’ response to the outbreak was the Ministry of Education’s “Zika Zero” program (Ministério do Educação, 2016), a nested subprogram of the Ministry of Health’s national “Combate Aedes” mosquito elimination campaign (Ministério da Saúde, 2019). After confirmation of the virus’ sexual transmission in February 2016 (Counotte et al., 2018), health authorities put some effort into promoting condom use to prevent the spread of ZIKV. Health promotion and education pertaining to sexual and reproductive health at the time of ZIKV was minute compared with the effort placed on mosquito avoidance and breeding site reduction (Brito & Fraser, 2016). It did not appear that best practices for health communication and promotion were being utilized as the overwhelming majority of these recommendations were solely directed to women. During the epidemic, the Ministry of Health’s official ZIKV website stated, “[ZIKV] prevention and control measures are like that of dengue and chikungunya. There are no specific measures of control directed at men, since there is not a vaccine or antiviral drugs” (Ministério da Saúde, 2017). The Brazilian Ministry of Health deviated from the World Health Organization’s ZIKV recommendations by predominately focusing on mosquito mitigation and not adequately acknowledging men’s role in the spread of ZIKV nor the sexual transmission of the disease.
Researchers, evaluators, and responders have created a body of literature pertaining to lessons learned since the start of the 2015 ZIKV epidemic. Among these lessons learned are the points that this article attempts to explore: the need for ZIKV health promotion to consider the role of behavioral determinants (Center for Reproductive Rights, 2018), the importance of tailoring messaging to local contexts (Carey et al., 2020; Toppenberg-Pejcic et al., 2019), the need to include considerations of gender norms in health promotion (Wurth, 2017), and the importance of including clear descriptions of instructional behaviors within health promotion materials (Rodrigues & Grisotti, 2019). Since the start of the ZIKV outbreak, we have continued to battle outbreaks of dengue, chikungunya, cholera, plague, and now a coronavirus pandemic. The time taken to learn from past responses—in this case the ZIKV outbreak in Brazil—is time invested in preparing for the next wave of emerging infectious diseases, risk communication, and situations where rapid health promotion is required.
As the world is wrapped in the COVID-19 pandemic and issues of messaging and behavior change, it is should be obvious that there is a need to learn from previous experience to (a) understand how to improve risk communication and health education at times of outbreak, epidemic, or pandemic; (b) continue to enhance the health promotion and education for areas, such as Brazil, that continue to fight endemic ZIKV; and (c) prepare for the eventual return of epidemic levels of ZIKV. This study used qualitative methods to investigate how women in the city of Fortaleza perceived ZIKV, the Brazilian authorities’ ZIKV prevention recommendations, and the feasibility and utility of these recommendations.
Method
We conducted this study with a convenience sample of women participating in a larger cohort study funded by the Brazilian government: “Zika em Fortaleza: respostas de uma coorte de mulheres entre 15 e 39 anos (ZIF)” (Zika in Fortaleza: responses of women 15–39 years old) (Dr. Ligia Kerr and Dr. Carl Kendall, PIs). The research team includes epidemiologists, biostatisticians, physicians, nurses, anthropologists, lab technicians, and doctoral students at the Federal University of Ceará (UFC) in Fortaleza, Brazil. ZIF includes qualitative, quantitative, and laboratory components. As of April 2020, the ZIF team is following 1,498 women. All participating women in the ZIF cohort completed a quantitative survey and underwent a series of lab tests for the detection of ZIKV, dengue, and chikungunya exposure (IgG and IgM). Inclusion criteria were women between the ages of 15–39 years, who utilize(d) the public health sector, who have engaged in sexual intercourse in the past year, and who have not undergone tubal litigation or who are otherwise not able to become pregnant. This qualitative study serves as a complement to the ZIF cohort study, as its qualitative methodology allows for a deeper assessment of factors influencing understanding and uptake of ZIKV prevention recommendations.
Participant Recruitment
The inclusion criteria for this qualitative study were the same as those of the ZIF study with the exception of class identity. In Brazil, socioeconomic class is calculated based on an individuals’ ownership of household appliances, occupation, years of education, monthly income, and access to utilities. Socioeconomic class ranges from A (most affluent) to E (least affluent) (Associação Brasileira de Empresas de Pesquisa [ABEP], 2016). The C-class is commonly perceived as the “middle class,” encompassing the largest proportion (47.7%) of the Brazilian population (ABEP, 2016). C classification is tied to a monthly income (from BRL 2,705 or US$501.07 1 [C1] to BRL 1,625 or US$301.01 [C2]) (ABEP, 2016).
For this study, the research team approached women who identified as C-class to control for extreme differences in income (A, B vs. D, E classes). The C classification covers almost half of the Brazilian population, yet the C1 and C2 subclasses encompass great economic differences. The differences in household income means disparities in access to goods, level and kinds of employment, quality of goods purchased, and type of housing. Generally, members of the C-class have the financial ability to adhere to many of the recommendations like the purchasing of repellent and long-sleeved clothing, but may lack the funds to be able to afford private health services. On the other side, it is unlikely that women from classes A and B would use public primary care facilities. Furthermore, according to reports from the ZIF study, women in D and E classes had already reported low levels of information and knowledge about the recommendations and little behavior change. In summary, this led to our decision to choose participants from the C-class.
Self-reported socioeconomic class
It became evident in the piloting of the interview guide that it was common practice for participants to differentiate their middle-class status, identifying as either high middle class (high C-class) or low middle class (low C-class). Throughout interviews, women were eager to talk about differences in wealth, feeling that there were major differences in lifestyle, access to goods, and stereotypes of people associated with either end of the class spectrum. Women commonly stated that the differences in class created two different cities in Fortaleza—one for those of higher socioeconomic class (high C-class) and one for those of lower socioeconomic class (low C-class). How women saw their socioeconomic status was dictated by how they perceived their physical, social, and natural environment.
Procedure
During the formative research phase, the ZIF research team shared preliminary findings from interviews and knowledge of the communities, convened to strategize the study’s approach, and piloted the data collection instrument. Pilot testing allowed the research team to improve the interview tool content, including topics, questions and prompts, language, format, length, phrasing, and delivery of questions to best collect data pertinent to the research questions. Data collection occurred over a 2-month period, between December 2018 and January 2019. A convenience sampling method of already enrolled ZIF participants was used. The ZIF research team randomly selected individuals from a roster of already enrolled ZIF participants. Each randomly selected individual was called and asked to participate in this new qualitative study component. All participants agreed. We continued recruitment and information gathering until we achieved saturation in terms of the main research topics (Creswell, 2009).
Data collection
Data collection took place in two public health units (PHUs), Posto de Saúde Graciliano Muniz and Posto Unidade de Saúde Escola Casemiro José de Lima Filho, in Fortaleza, Brazil. These PHUs are part of Brazil’s national health care system, Sistema Único de Saúde (SUS). The PHUs serve as a training site for UFC medical school, which supervises the clinic. These units serve as the primary location for basic health care, vaccinations, access to family planning, and pharmaceuticals for residents of two health districts in the city. The first author conducted interviews in private locations inside the PHUs. All of the participants declined permission to audio record the interviews. This was a foreseen possibility, given the severity of topics especially abortion, which is both common and criminalized in Brazil. The first author took thorough notes during each interview and followed up with a second phone interview within 24 to 72 hr of the initial interview. The purpose of the follow-up was to verify content and ensure all data collected adequately represented the interviewees.
Measures
Qualitative methods were chosen to allow a richer understanding of the respondents’ views of the epidemic and the utility, feasibility, and acceptability of the recommendations. We were especially interested in identifying, in the broadest way, the difference between recommendations and how respondents described and justified their responses. Qualitative methods utilize open-ended questions that often collect deeper, lengthier, more detailed responses to questions that capture reasons why behaviors were done, description of thought processes, and an examination of decision-making (Creswell, 2009; Sangaramoorthy & Kroeger, 2020). This study was designed to give women an opportunity to discuss their understanding of ZIKV, describe their experiences, and provide a firsthand perspective to the prevention efforts made in Brazil. To capture this robust information, a qualitative approach appeared appropriate.
We collected data utilizing a Rapid Anthropological Assessment (RAA) based on semi-structured, open-ended, and in-depth interviews, structured and unstructured observations, free listing exercises, scenario discussions, and informal conversations in the study setting (Sangaramoorthy & Kroeger, 2020). RAAs, also called Rapid Assessment Procedures, Rapid Ethnographic Assessment, Targeted Intervention Research, and Focused Ethnographic Surveys, have been used across a variety of health topics and are frequently utilized in formative research to inform interventions or health communication (Kendall et al., 2008; Sangaramoorthy & Kroeger, 2020; Scrimshaw & Gleason, 1992). We built the RAA on the ZIF qualitative instrument as well as preliminary findings from the cohort. The instrument included demographic and family information including social class and education; existing ZIKV knowledge; exposure, signs, and symptoms of ZIKV; pregnancy histories; reproductive intentions and family planning practices; mosquito avoidance/vector control; attitudes and responses to ZIKV health recommendations; and care-seeking during the ZIKV epidemic. The instrument additionally contained free listing exercises as well as questions asking women to rank the effectiveness of health behaviors, the importance of health issues, and perceived risk of diseases. The final section of the instrument included a series of ZIKV-specific scenarios. These prompts asked women to discuss their perceptions of the situation as well as how they believe their friends would respond in each scenario.
Data analysis
Notes were taken by the first author during interviews and converted into fairnotes at the end of each day (Halcomb & Davidson, 2006; Hill et al., 2003). Fairnotes were then coded line by line in NVivo 12 Pro software then thematically analyzed, which narrowed the codes into categories that highlighted key concepts used by respondents to address the research questions (Creswell, 2009). Emerging findings were discussed among the team throughout the analysis process. Selected quotes are provided to illustrate key concepts in each identified theme. Thematic analysis permitted the team to report patterns of interconnectedness, nuance, and complexity of responses as they related to the research question and overall study. This approach was chosen as opposed to limiting responses to a numeric, linear, response frequency, which would not capture the depth or interconnectedness of response (Creswell, 2009). Institutional Review Board (IRB) approval was provided by the Federal University of Ceará (FWA # IRB00004330) and the Tulane University IRB (IRB# 2018-1606).
Results
The sample consisted of 35 women who completed semi-structured interviews lasting between 30 min and 2 hr. Participant ages ranged from 18 to 39 years with a mean age of 25. Nineteen of the participants had children, 16 participants did not have any children, and six of the participants were pregnant at the time of their interview. All measures in Table 1 are self-reported unless otherwise indicated.
Participant Demographics.
Note. ZIKV = Zika virus; CHIK = Chikungunya; DENV = Dengue.
Disease exposure was recorded as either a blood test result or a self-report. IgG/IgM blood testing was provided by Zika in Fortaleza (ZIF); therefore, some participants had their results at the time of their interview. None of the participants had positive IgM results at time of interview. If a participant did not have a test result, self-report was recorded. bEmployment as used here includes having an official job card (Carteira de Trabalho) as well as part-time employment, informal employment, and self-employment. The category “unemployment” refers to women who self-reported having no form of occupation and/or no regular income.
Perceptions of Zika Virus
All interviewees were aware of ZIKV and characterized it as a milder form of dengue. Only when associated with microcephaly or pregnancy was ZIKV viewed as more dangerous than other vector-borne diseases such as dengue, chikungunya, yellow fever, or malaria. Overall interviewees felt ZIKV was no longer an issue, having mostly disappeared after 2017. Respondents justified this comment because of the media’s (especially TV’s) lack of coverage on the topic, the diminished number of cases reported in their neighborhoods, and the lessened emphasis of ZIKV in the health care setting: There was a lot of information on TV! A lot on the news and a lot of commercials talking about mosquitos and pregnant women . . . they looked like what we always see for dengue and [the news and commercials] even said Zika is dengue so we thought “Ok I know dengue so I know Zika so I will do nothing new” . . . and yes there was a lot about pregnancy and microcephaly, which was very sad but if you are not pregnant you do not need to know, you know? (High C-Class, 18 years old) The television and Facebook had all the information about Zika . . . it was everywhere at the beginning but then it stopped, and we all figured Zika was over. Now you see nothing on the television or Facebook or WhatsApp so maybe it is gone. (Low C-Class, 22 years old)
Participants were then asked about past ZIKV risk. Women who self-identified as a higher socioeconomic class felt they were at a slightly higher risk of exposure at the time of the epidemic, but currently felt removed from any risk. As a result, these women were not motivated to adhere to the ZIKV prevention guidelines. Overall, high C-class women believed that because their risk of ZIKV was low, there was no need to engage in ZIKV prevention, and that there were very little consequences of not adhering to the prevention recommendations: Zika is gone by me so we are not worried and so what if you get it? Maybe you get a fever, but I think it is not a problem and I am not pregnant so why worry? (High C-class, 29 years old)
Women who identified as low C-class initially said ZIKV was no longer a problem, but further discussion found there were still feelings of risk among this group. Women in this group recalled feeling extremely vulnerable, helpless, and afraid during the epidemic. Most women in this group had contracted ZIKV or could identify individuals in their community who had ZIKV. One participant spoke to the inevitability and expectation of illness from one or another mosquito-borne illness: Always people are sick . . . dengue . . . chikungunya . . . and now it’s Zika. You can work hard but the mosquito will win. (Low C-class, 24 years old)
All 35 women interviewed recognized that not adhering to the prevention recommendations was risky, yet argued that completing these behaviors came at a price. Associated costs tied to the ZIKV prevention included repellent use being stigmatized, the financial burden of repellent, moral judgment for accessing condoms, condom use possibly alienating partners, and pregnancy avoidance leading to an unfulfilled life. When asked which was more important—the risk of ZIKV or the socioeconomic cost associated with the behaviors—all women said that the associated cost could more seriously affect a woman’s life. In the following quote, the respondent discusses how ZIKV prevention adds to her list of responsibilities: I know Zika is bad if you’re pregnant, but now it is just another disease. For me . . . I need to buy food for my child first, pay for the car second, and a lot of other things before I think about using my money for repellent. (Low C-class, 39 years old)
ZIKV was viewed across interviews as mild and less impactful than the costs associated with adhering to preventive behaviors. This was found to be true in all situations other than in the context of a pregnancy. All respondents agreed that ZIKV prevention was incredibly important if a woman was pregnant due to the possibility of microcephaly or other negative birth outcomes. It was strongly believed that in this situation, it was the mother’s job to protect the fetus and ensure a healthy pregnancy. Pregnant interviewees felt extreme pressure to prevent ZIKV and subsequent harm to her pregnancy: I understand doing these things is important because Zika is serious when a woman is pregnant. I try to do everything correctly to protect my baby because that is my job. (Low C-class, 19 years old)
Visible in this response is the sense of obligation and responsibility for the pregnancy that certainly predates the ZIKV epidemic and response.
Feasibility of ZIKV recommendations for vector control
Mosquitos
Women reported receiving their health information from news channels, television commercials, WhatsApp, and Facebook during the ZIKV epidemic. All informants were aware of a mosquito’s ability to transmit ZIKV and other arboviruses such as dengue and chikungunya. Participants recalled ZIKV portrayed in the media similarly to dengue and chikungunya. All interviewees recalled not paying much attention to the mosquito recommendations advertised because they had “already seen them all,” feeling it was not new information, but recycled dengue information with emphasis placed on pregnant women. Interviewees recalled ZIKV-specific messaging at the start of the epidemic, with a transition to just dengue information after a few months. Respondents assumed this transition symbolized ZIKV was gone and no longer a threat. Furthermore, women noted that the consistent rainy-season-influx in mosquito media fostered complacency toward mosquito messaging, allowing them to ignore the redundant messages. Content of mosquito prevention media listed by participants included general recommendations for repellent use, cleaning the home, and removing stagnant water. All women interviewed expressed that they were not feeling worried about mosquito prevention as they were cleaning their home daily and felt knowledgeable about mosquitos. Women interviewed spoke of mosquitos as a common occurrence, one which Brazilians expect to live with: Mosquitos are normal? (laughs) Yes mosquitos are our neighbors, they’re always around and come back with the rain. Annoying but we are accustomed to it, you know? (Low C-class, 32 years old)
The number of mosquitos within the community was perceived differently across socioeconomic classes. High C-class interviewees did not perceive mosquitos as very prevalent, especially during the dry season. Several women who identified as high C-class suggested they may not see as many mosquitos because they live on higher floors of buildings, have fumigation services in the community, or because they can close their windows due to access to fans and air conditioning: I always see mosquitos when I walk on the street . . . not after they spray [ fumigation] . . . but when I am in my apartment, we don’t have a lot [of mosquitos] in the house because . . . I don’t know because they don’t fly that high maybe? [apartment on 4th floor]. Or maybe because we close the windows to use our air conditioning so they can’t get in. (High C-class, 37 years old)
In contrast, low C-class respondents described their communities as full of mosquitos all year round, especially during the rainy season. Interviewees mentioned living with mosquitos without pause. Respondents who identified as low C-class additionally mentioned needing to utilize behaviors associated with facilitating mosquito breeding, such as storing water. Water storage was essential as piped water was inconsistent, purchased water was expensive, and rainwater could only be used for house cleaning, pets, and washing cars. Purchased water was for consumption only. Low C-class interviewees associated the high number of mosquitos to be as a result of the lack of cleanliness in the streets, piles of garbage in the neighborhood, or their neighbor’s unwillingness to tend to their home: I can clean my house all day and still mosquitos! I don’t think it is my house because I am cleaning cleaning cleaning . . . I think it is because of my neighbor or because there is so much trash in the streets that nobody picks up. Mosquitos like trash, I think. (Low C-class, 38 years old)
This quote shows the frustration commonly expressed from continuously coexisting with mosquitos despite frequent household cleaning. This frustration is intensified by external factors such as a neighbor’s unwillingness to engage in vector mitigation techniques or the cleanliness of the neighborhood.
Cleaning the house
The recommendations targeting mosquito breeding site mitigation were generic, with no description of actionable behaviors. This left people in the community to find ways to fulfill the recommendations while important gaps in knowledge existed, ultimately yielding ineffective results. An example of this is the communication focused on mosquito-based preventive measures (e.g., removing standing water, tending to the yard, eliminating trash) being promoted as “cleaning the home.” Recommendations instructed individuals to “clean your home,” focusing on hygiene as opposed to mosquito breeding site mitigation. This led women to assume general hygienic cleaning of the home (dusting, sweeping, removing garbage, washing floors) was the same as source reduction of breeding behaviors (use of larvicide, cleaning, or covering water containers). This suggests there exist gaps in knowledge of techniques needed to effectively reduce mosquito breeding sites. All interviewees could identify cleaning “bad” or “dirty” water as mosquito mitigation methods, yet could not explain how to clean to best prevent mosquitos. When asked where they learned this type of information, respondents credited their mothers and television. These hygienic behaviors were believed to be an easy, learned response to growing up with the vector: No, I don’t think removing dirty water is hard. You know you need to clean your house in all these ways . . . no dirty water, no bad plants, no garbage, nothing dirty . . . and you do this every day to have a good home . . . a clean home . . . so I think if you do these things you should do every day then maybe mosquitos won’t come. (High C-class, 26 years old)
Stigmatization associated with house cleanliness was an underlying tone throughout these interviews, especially in its relation to a woman’s ability to mitigate mosquitos through the cleaning of their house. It became evident there existed negative perceptions or biases toward certain groups of people described as “poor,” “uneducated,” and “dirty.” All women interviewed, regardless of class, expressed biased views toward at least one of these stigmatized populations: If a person is dirty or poor, I think they will be sick a lot because they are uneducated and lazy and don’t clean the house. (High C-class, 29 years old) If you see a dirty home you know the woman is uneducated and poor, so she will probably be sick and have sick children. (Low C-class, 23 years old)
Sickness, in reference to an adult or a child being infected with ZIKV, was commonly associated with negative characteristic such as poverty, a lack of education, or a lack of hygiene. This concept was found in interviews across socioeconomic class.
Mosquito repellent
Using mosquito repellent was universally viewed as important, yet it was not utilized. Probing found that women perceived repellent as an abnormal substance, that its use was uncommon, its odor was off-putting, and it was too expensive to purchase frequently. Participants believed that the idea of continuously utilizing mosquito repellent was unreasonable as this population has lived with mosquitos for generations with mostly mild perceived consequences. Women also mentioned that the smell of repellent may create the illusion that the user was ill or dirty: Yes, everyone tells us to use repellent, but we are not accustomed to it . . . and if you use it then you are smelly and sticky and that is not ok. Here in Brazil we always need to smell good and look good—if we wear repellent people would think we don’t shower . . . or are sick or something. (Low C-class, 31 years old)
Illustrated in this response are several cultural barriers to repellent use that stem from social norms, local hygiene narratives, and stereotypes around illness.
Feasibility of ZIKV recommendations for reproductive counseling
None of the participants were aware of the recommendation to receive reproductive counseling if pregnant or contemplating pregnancy, regardless of socioeconomic group. Interviewees who were currently pregnant or had been pregnant during the outbreak reported not receiving counseling services. Health care staff at the PHUs were also unaware of this recommendation. Study participants were confused by the concept of reproductive counseling as it is often used for women who struggle to conceive, not women who are already pregnant. In addition, the idea of planning a pregnancy was not perceived as normal. Meeting with a health care provider to design or schedule a pregnancy seemed unrealistic as pregnancies were often unplanned. When asked if reproductive counseling was a desired service, all respondents said no: Is that normal? I have never heard of that. I don’t think women plan a pregnancy unless they have trouble becoming pregnant. I don’t know about that . . . I have never had that . . . and why go if there is no cure for microcephaly? (Low C-class, 19 years old)
A key finding of this study is that seeking reproductive counseling was often seen as having no value since ZIKV, and microcephaly, had no cure.
Feasibility of ZIKV recommendations for sexual transmission
Women were confused or dubious as to why the question about sexually transmitted ZIKV was asked in the interview. No participant, regardless of socioeconomic group, was aware of the sexual transmission of ZIKV. Once the mechanism of transmission was explained, women continued to be confused and many became angry that they had not been given this information earlier. Women explained feeling betrayed by health authorities because they had been given incomplete information that described ZIKV as “just like dengue.” An overwhelming majority of respondents suggested men be told this information immediately as they play a major role in this behavior: Is this true? (angrily) Why didn’t anyone tell us? How are we supposed to be safe if men can make us sick? Men are always outside and never use repellent and are always complaining about mosquito bites . . . so then they get bites and we get Zika? That is not fair, you go tell them! (High C-class, 24 years old)
Condom access
Women in both socioeconomic groups identified strict gender norms as barriers impeding their ability to access and use condoms. Women believed it was crucial to adhere to these gender norms to preserve one’s social status. Condom use was conceptualized as entailing two major behaviors: obtaining the condom and using the condom. Most interviewees felt it was a man’s job to obtain the condom from the pharmacy as this demonstration of sexuality was more socially acceptable, and because men were believed to have more financial means to purchase condoms. A woman purchasing condoms was suggestive of promiscuity and could open her up to potential rumors: Condoms are everywhere . . . pharmacies have a lot of them I think . . . so boys can buy them there. He is the one who always wants sex and he is the one who works so he can buy it. If I go then I am a slut or people will think “why does her boyfriend not like her enough to get condoms, or maybe he has no job, she must be a bad quality girl” . . . and what if the pharmacist tells someone who tells my family? I am shy and I don’t want people saying bad things. (Low C-class, 18 years old) Men decide to use condoms or not. If I have condoms waiting for him what will he think? He will think another man bought them! (High C-class, 28 years old)
Condom use
The ease, necessity, and negotiation of condoms, however, varied by socioeconomic class. High C-class women acknowledged men’s resistance to condom use but stated feeling confident, comfortable, and motivated to negotiate condom use. This was because condoms were thought of as a tool to postpone pregnancy and protect her ability to continue her education and career. It was also noted that women of high C-class felt that if their partner respected them, he would be willing to grant their request to use a condom: Yes, you have to say “use it or I will leave” because boyfriends don’t want to use it . . . never. But if he is a good man then he will use it. He will complain but he will use it. If he is a good boyfriend and respects you then he knows he needs to use it. (High C-class, 33 years old)
Most women in the low C-class group reported not attempting condom negotiation or use as they felt the situation was out of their control. They reported that condom use was the partner’s decision as he had purchased the condom and needed to be the one to use it. Participants explained that advocating for condom use may lead their partner to suspect a woman’s infidelity or a sexually transmitted disease (STD). Condom negotiation was also perceived as precarious for the implicit accusation of her and her partner’s faithfulness. Interviewees were anxious that advocating for condom use could lead their partner to become agitated, violent, withhold affection, remove financial support, or even to abandon them: Oh, I don’t know . . . I am scared to try . . . I think he might become very mad and think that I am cheating on him. Men here get mad very quickly and it is very serious. Men will think you don’t love him or think you are a bad girlfriend. Maybe a man would hit the girl? Maybe he will leave? I don’t know . . . that would be the worst thing . . . if he were to leave me with no love and no money for the house and children. (Low C-class, 32 years old)
Captured here is the fear, anxiety, and discomfort many women expressed when contemplating advocating for their reproductive autonomy such as the desire to use contraceptives.
Feasibility of ZIKV recommendations for avoiding pregnancy
Participants of both socioeconomic groups easily recalled the recommendation to avoid pregnancy during the ZIKV crisis as it had been something frequently advertised on the news and social media during the crisis. Although well-known, none of the women interviewed thought it was a serious recommendation. One woman stated, . . . it was serious? The government wanted all women to not get pregnant until you don’t know when . . . seriously? That is horrible! How can a whole country not have babies? Brazilians love babies! (High C-class, 24 years old)
All women interviewed stated that pregnancy avoidance was unrealistic. Participants felt it was impossible to control the timing of a pregnancy, that pregnancy was God’s decision, and that it was unfair for authorities to direct family life. Respondents emphasized that for women around 30 years old, it was important to not wait to have a child as their reproductive timeline was approaching its end. High C-class women felt more able to prevent pregnancy as they had more family planning options and motivation to postpone pregnancy at the prospect of continuing education or furthering a career. Both groups however, especially the low C-class women, felt pregnancy was an important aspect of life that was expected of all women. All women interviewed in this study felt social pressure to have children by family, friends, and peers. 2 This recommendation was deemed unreasonable as it contrasted with what was believed to be within the control and desires of women interviewed.
Discussion
Our findings echo results from other studies that suggest the ZIKV prevention recommendations were not formulated as health education or health promotion, but simply rephrased medical and epidemiological findings. The ZIKV health promotion put forth by Brazilian authorities did not satisfy the components of best practices in health communication as it appeared not to be informed by behavior change theories, was not tailored to the audience, was not actionable, was not instructive, and did not acknowledge the multilevel contextual factors that influence the barriers and benefits to ZIKV prevention.
While a key finding of our study is the inadequacy of health promotion for ZIKV and the failure to follow best practices in health communication, the study also highlights the consequence of that failure: the ignorance of the central role of socioeconomic status in exposure, knowledge, perception of risk, utility, and adherence to ZIKV prevention recommendations. Social class membership—C1 or C2—influenced how women perceived the presence of ZIKV in their communities, the risk of ZIKV, their ability to access prevention materials, and their sense of self-efficacy to adhere to the ZIKV prevention recommendations. High C-class women felt that their affluence removed them somewhat from the risk of contracting ZIKV, leading these women to have little motivation to attempt ZIKV prevention behaviors. Low C-class respondents felt there was a high likelihood of contracting ZIKV, yet generally felt unable to prevent ZIKV due to a lack of ability to access and use resources as well as a feeling of inevitability about mosquito bites and contracting the disease. These factors dissuade these participants from attempting to prevent ZIKV. Both groups had a low willingness to engage in ZIKV prevention, but for different reasons.
Women who self-identified as high C-class felt that they could avoid ZIKV if required. Although there was not a strong feeling of ZIKV risk, there was a strong feeling of being able to control whatever risk there was. Members of the high C-class felt that in future ZIKV outbreaks they would be able to adhere to mosquito mitigation recommendations. Women in low C-class had a diminished sense of self-efficacy to prevent mosquito bites due to a lack of access to mosquito avoidance products, the need to store water, condition of their house, unhygienic conditions in the community, and the high prevalence of similar endemic diseases such as dengue and chikungunya in their social networks.
For condom use, the high C-class group felt a slightly higher sense of self-efficacy, but neither group felt completely able to obtain, negotiate, and use condoms. This is consistent with findings from focus groups of women during the ZIKV outbreak in Brazil (Marteleto et al., 2017). Women additionally struggled with self-efficacy over pregnancy avoidance due to social pressure to start a family and perceptions of difficulty around planning a pregnancy. It is important to understand enabling and deterring influences on self-efficacy as it can heavily influence women’s willingness to attempt ZIKV prevention.
Barriers to ZIKV Prevention
Barriers to ZIKV prevention identified in this study include financial constraints, fear of partner’s reactions, stigma, judgment from community members, complacency, social norms, feeling prevention was out of their control, and the repeated theme of the inevitability of infection. The sense of inevitability of contracting a vector-borne disease dissuaded women from adopting ZIKV prevention behaviors. Participants felt certain of losing the battle against mosquitos due to the normalcy of cohabitating with the vector, the seasonal surge in exposure, neighbors not engaging in control measures, and the city’s failure to deliver services such as garbage disposal or fumigation. Participants’ perception of a lack of control was additionally a barrier in attempting to avoid sexual transmission of ZIKV and pregnancy. The barriers found in this study are consistent with other qualitative findings from Brazil during the epidemic (Center for Reproductive Rights, 2018; Elsinga et al., 2017; Linde & Siqueira, 2018).
This study asked women to retrospectively report their perceived risk of ZIKV at the time of the 2015–2016 crisis. It is important to understand this difference as it reflects how quickly individuals may forget the severity and susceptibility they felt when at highest risk. At the time of the crisis, both groups felt at risk of ZIKV, yet aside from pregnant women, there was very low perceived risk of ZIKV among women. The similarity of ZIKV to other arboviruses and the belief that symptoms were mild created a sense of low severity among participants. All participants felt their susceptibility had dropped since the end of the outbreak due to the decreased amount of ZIKV-specific media. These findings match results from qualitative work with ZIKV patients and nurses in Brazil during the epidemic (Tillman & Kristoffersson, 2017). Although ZIKV incidence has diminished, it is imperative to bolster ZIKV messaging to raise awareness as to the continued risk of the range of ZIKV-associated outcomes from Guillain-Barré Syndrome to CZS.
Developing a ZIKV Prevention Campaign
Health promotion during the epidemic depicted the main benefit of ZIKV prevention as the ability to lower the likelihood of vertical transmission of ZIKV during pregnancy, preventing CZS in newborns. If women were not pregnant, ZIKV prevention behaviors were not seen as beneficial. Engaging in mosquito control was perceived as beneficial as it mitigated the potential of contracting dengue and chikungunya, perceived as more serious. Cleaning the house was beneficial for aesthetic, hygienic reasons, and day to day functionality. Women interviewed, as well as health care staff at the PHUs, were unaware of the recommendation for women to seek reproductive counseling. This finding is consistent with other studies that have documented reproductive counseling as a not widely known component of the ZIKV response in Brazil (Wurth, 2017). Condom use was categorized as beneficial for pregnancy postponement, but not for ZIKV prevention. The many benefits of ZIKV prevention behaviors need to be better communicated so that prevention is seen as beneficial for a variety of reasons—not solely for pregnant women.
The decrease in ZIKV messaging was believed to be a signal to stop ZIKV prevention behaviors. ZIKV has not disappeared and there is still a need for health education, media coverage, and ZIKV awareness efforts. Innovative, actionable, and targeted vector control recommendations should be disseminated to communities to aid them in their battle with Aedes aegypti. Our findings echo results from other studies that suggest the repetition of familiar mosquito messaging recycled from previous dengue campaigns reduced engagement in health education during the epidemic (Ribeiro et al., 2018). The lack of messaging about the sexual transmission of ZIKV created gaps in knowledge, placing women at a higher risk of acquiring the disease. This study finds that messaging for ZIKV prevention should not assume that all women in Brazil, even just those in class C, are in the same position to adhere to health prevention behaviors. ZIKV messages also need to be targeted toward men and male partners as they too play a role in the transmission and mitigation of the disease. Messages need to be tailored to account for the many different socioeconomic, cultural, geographic, and educational contexts across a continent-size country such as Brazil.
Limitations
While the larger study was funded by Brazilian government grants, this qualitative study was funded by the John Snow Inc. Fellowship for Doctoral Research. There were no constraints put on the research questions, research team, analysis, nor dissemination of findings by the funder. This study, although critical of the national response, is meant to assist authorities in preparing for a future outbreak. Qualitative methods as conducted are vulnerable to interviewee bias (i.e., social desirability bias) as well as interviewer bias. We attempted to address these potential biases through careful piloting and feedback on the research instruments and discussions with the clinics, other ZIF project workers, and women attending the clinic. These biases are also addressed through the research team’s long presence in the clinic, familiarity with the research context, and training in qualitative methods. While sampling was not designed for generalizability, but rather to select women who ZIF staff thought would be open to additional interviews and forthcoming, we are comforted by the fact that the findings of this study echo results from other studies in the region (Center for Reproductive Rights, 2018; Elsinga et al., 2017; Linde & Siqueira, 2018; Marteleto et al., 2017; Ribeiro et al., 2018; Tillman & Kristoffersson, 2017). There are limitations associated with qualitative methods, yet the benefit of qualitative research is the essential, rich data it yields, which can give a deeper insight into how to create, tailor, and disseminate interventions, risk communication, and health education.
Conclusion
Recommendations made by health authorities during the ZIKV epidemic perhaps never intended to be a health communication program, yet assumed that they could influence behavior change. Not using an evidence-based approach to health promotion creates generic, incomplete messaging for communities with little targeting and specific actionable health prevention behaviors. Ineffective messaging can lead to complacency about the campaign and recommendations, lack of adherence, and the generation of alternative knowledge of transmission and prevention of diseases as has been documented in ethnographic research from other outbreaks such as Ebola (Hewlett & Hewlett, 2008; Richards, 2016). In Fortaleza, Brazil, best practices in health education and promotion were overlooked, creating ineffective ZIKV prevention campaigns, and subsequent gaps in knowledge and behaviors. Outbreak and epidemic responses, due to their need to be developed under emergency conditions, rarely utilize best practices in health communication in their design, creating missed opportunities for holistic approaches that consider enabling, deterring, and contextual factors, which may impact adoption and continued use of prevention behaviors.
There now exists a new cohort of individuals exposed to ZIKV, arboviruses, and other viruses waiting to take center stage, such as the COVID-19 pandemic. Developing effective strategies to capture and use community knowledge to design effective health promotion remains critical. Similar to ZIKV, currently health authorities around the globe are struggling with COVID-19 health promotion for mask use, physical distancing, and handwashing as they strive to incorporate rapid scientific discoveries, everchanging recommendations, and population level outbreak fatigue. From outbreaks to pandemics, we see parallel failures when health authorities do not utilize best practices in health communication and promotion. It is recommended that public health authorities adopt behavior change that is theory informed, evidence-based, context-specific health communication and promotion approaches to better serve the diverse populations affected by continuing and future public health threats.
Footnotes
Acknowledgements
The authors thank the women, health care staff, families, and communities whose patience and input greatly aided this research. Although we are critical of the public health education strategy that addressed women with a simple and direct admonition that they should not get pregnant and avoid mosquito bites, we also acknowledge the many strengths in Brazil’s approach to ZIKV control and prevention. The surveillance and information infrastructure already in place through Sistema Único de Saúde (SUS), the transparency in publicizing health information, and the strong connection between public health information and research in the country permitted Brazil’s rapid discovery of the link between microcephaly and ZIKV and the ongoing real-time monitoring of the epidemic. The knowledge and skills embedded in Brazil’s scientific base and public health infrastructure rapidly led to a coordinated epidemiological, scientific, clinical, and laboratory response to ZIKV.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by CNPq (440778/2016-6), CAPES (88881.130806/2016-01; 88887.130795/2016-00), FUNCAP (3898920/2017), and the John Snow Inc. (JSI) Fellowship for Doctoral Research.
