Rapid ethnographic assessment for potential anti-malarial mass drug administration in an outbreak area of Santo Domingo, Dominican Republic

Background In the Dominican Republic, a recent outbreak of malaria in the capital, Santo Domingo, threatens efforts to eliminate the disease. Mass drug administration (MDA) has been proposed as one strategy to reduce transmission. The success of MDA is contingent upon high levels of acceptance among the target population. To inform the design of future MDA campaigns, this rapid ethnographic assessment examined malaria-related knowledge and attitudes toward malaria MDA among residents of a transmission focus in Santo Domingo. Methods In October 2019, a rapid ethnographic assessment was conducted in the Los Tres Brazos transmission focus, which had not previously received MDA. National malaria programme staff conducted 61 structured interviews with key informants, recorded observations, and held 72 informal conversations. Using a grounded theory approach, data were analysed during three workshop sessions with research team members. Results Among those who had heard of malaria in the structured interviews (n = 39/61; 64%), understanding of the disease was largely based on personal experience from past outbreaks or through word-of-mouth. Community health workers (promotores) were trusted for health information and malaria diagnosis more so than professional clinicians. No participant (0%) was familiar with malaria MDA. After learning about MDA, almost all study participants (92%) said that they would participate, seeing it as a way to care for their community. Reasons for not participating in future MDA included not trusting drug administrators, feeling reluctant to take unprescribed medicine, and fear of missing work. Additional identified challenges to MDA included reaching specific demographic groups, disseminating effective MDA campaign messages, and managing misinformation and political influence. Conclusion Residents appear accepting of MDA despite a lack of prior familiarity. Successful MDA will depend on several factors: fostering relationships among community-based health workers, clinicians, community leaders, and others; developing clear health messages that use local terms and spreading them through a variety of media and social networks; and contextualizing MDA as part of a broader effort to promote community health.

actionable information about sociocultural dimensions of health or disease in order for health programs to improve delivery and care [12]. Here, we prefer the term rapid ethnographic assessment because it clearly references both the temporal and social science aspects of this methodology. In general, the approach relies on the usual suite of anthropological techniques, such as key informant interviewing, focus group discussions, and observations to characterize local beliefs, perceptions, or practices regarding certain diseases or health issues and then translate that knowledge into appropriate planning and delivery of health interventions [13]. These methods may also be combined with quantitative surveys [14]. Study participants may be sampled theoretically or purposively (in which recruitment is driven by a priori hypotheses). Unlike quantitative survey sampling, which seeks representative samples to make generalizable claims, sampling in qualitative studies seeks to ensure that the data are "saturated," or that all conceptual categories to explain the data have been identi ed and exhausted [15]. To do so, a breadth of key informants with personal knowledge, insights, and experiences about the phenomena of interest are recruited until no further theoretical insights emerge.

Study location
Fieldwork took place over seven days in October, 2019 in the Los Tres Brazos focus of Santo Domingo ( Figure 1). The area is comprised of 94 neighborhoods (barrios). The study was especially interested in community perceptions in areas of relatively high and low transmission, so neighborhoods were ranked according to malaria incidence in 2019 up to the time of the study. Six neighborhoods were selected based on transmission intensity (2 high transmission neighborhoods; 4 medium or low transmission neighborhoods) and ease of access to the neighborhoods. The six selected neighborhoods accounted for 60% of the total malaria cases in the Los Tres Brazos focus up to the week of the study. We intentionally over-sampled the neighborhood with the highest malaria incidence.

Recruitment
This study sought to obtain a cross-section of the general adult population within the Los Tres Brazos focus. Inclusion criteria were: age greater than or equal to18 years, currently live in the study area, have indepth knowledge about the general community and willingness to talk at length to study interviewers. Purposive sampling was undertaken as needed to obtain more responses from, for example, more men or younger people. Additionally, research assistants were instructed to pay particular attention to persons who said that they would decline participation in MDA and seek more detailed answers from them.
Contact with key informants in selected neighborhoods was facilitated by existing connections between CECOVEZ and district-level Health Area Directorates (Direcciones de Áreas de Salud; DAS) that serve Los Tres Brazos. Typically, a key gure at CECOVEZ would contact gatekeepers at the DAS o ce, or a community health worker (promotora) directly, in the barrio/sub-barrios where sampling was planned. Then, the research assistants would go to that o ce, meet the gatekeeper, and then be escorted into the community. Once contact with the rst few key informants was made, the individuals from the DAS o ce would then leave the assistants to conduct interviews. Afterwards, research assistants were free to use chain-referral "snowball" sampling after their rst interview, go on their own in the community, or ask the DAS o ce representative or promotora for additional assistance in locating another key informant.
The sampling strategy was modi ed on an ongoing basis based on collected data. At times, the team was directed to collect more interviews among under-represented groups or among those with insights about emerging themes following post-data collection de-brie ngs.

Data collection
The study team was comprised of two study coordinators and three research assistants. Two of the assistants were malaria technicians (evaluadores) from CECOVEZ who each have more than twenty years of community-based malaria experience in Santo Domingo and had knowledge of the study site. The other assistant was not a liated with CECOVEZ but had extensive experience in previous research projects with one of the study coordinators (HK).
Research assistants were trained to collect structured interviews, record observations, and hold informal conversations. One of the research assistants already had extensive prior experience with qualitative methods. Each assistant was assigned a target of three structured interviews per day of eldwork. Interview guide questions (Supplementary material 1) captured general demographics; general features of daily life, such as work schedules, local organizations, and trusted leaders; knowledge about malaria, including transmission and prevention; sources of health information; trust in the health system and local authorities; and MDA, including existing familiarity with MDA and likely participation in a future MDA campaign. If the participant was unfamiliar with MDA, a brief, standardized de nition was provided by the assistant; this de nition was developed by CEOCVEZ leadership and through team discussions.
Interviews were typically held in participants' homes, at their place of work, or in public spaces like corner stores (colmados) or street corners. Data collection was conducted in the middle-late afternoons, when most people were assumed to be at home. Interviews were not audio-recorded; instead, assistants noted key words and phrases used by the participant in free text boxes following each question in the interview guide.
Second, assistants were trained to record observations. These could span general features of the surrounding area where the interview took place: presence of environmental risks such as mosquito habitats and trash; socioeconomic conditions and behavioral cues of the participant; and population density of the area surrounding the interview site.
Third, assistants recorded informal conversations during each day of eldwork. These conversations were guided by study objectives and sought to triangulate ndings from interviews and observations. For example, if an interview participant described distrust of the local neighborhood association (juntas de vecinos), then the assistant could, afterwards, make brief contact with a passer-by and quickly ask for his or her opinion of the juntas de vecinos. These conversations occurred through any informal contact with a general community member, such as on the street, at a bus or moto-taxi stop, or in a colmado or beauty salon. The substance of these brief interactions was also recorded in key words and phrases taken by the assistant.
Data quality control was undertaken by study coordinators through serial checks of completed interview forms, assessing completeness of recorded responses, and use of the participant's own phrasing and vocabulary in recorded responses. Feedback was provided to individual assistants as needed.

Analysis
Taking a grounded theory approach [16], data analysis took place during team analysis sessions, moderated by the study coordinators (HK, KU). The team was encouraged to think broadly about general themes related to malaria, health information, trust in local institutions, and MDA. In this way, assistants applied their own interpretive lens to the data.
Team analysis sessions were held after Days 3, 4, and 7 (the nal day) of data collection. Each analysis session began with one hour of team discussion moderated by the study coordinators (HK and KU). Then, the assistants wrote a one-page essay in response to several questions put forward by a study coordinator (Supplementary Material 2). These questions were formulated to capture core concepts, themes, and ideas emerging from the data and team discussions. For example, some questions were as follows: Why do people trust promotoras?; Why do people say that they would participate in MDA? New questions were developed over time as new insights arose from the data or group discussions.
To support their written answers, assistants were asked to elaborate using personal observations and quotations from participants. Afterwards, a study coordinator (HK) typed the written analysis summaries and raw eld notes from each assistant into Word documents for content analysis. The study coordinator also took daily notes to track the team's activities, challenges and areas for improvement, and general impressions and re ections to keep an audit trail.
Finally, participant demographic data and responses to key interview questions from the structured surveys were entered into an Excel spreadsheet for quantitative data analysis. No personally identifying information was recorded in any data collection activity.

Data Summary
Over seven days of site visits, 61 structured interviews and 72 informal conversations were conducted with community members across six sampled neighborhoods of Los Tres Brazos (Table 1). As shown in Table 2, the structured interview participants (n = 91) were balanced between men (51%) and women (49%). Average age of participants was 44.6 years (range, 18-82). Nearly half were married or in a civil union (49%). The average residency time in Los Tres Brazos was 23.6 years (range, 3 months to 60 years), indicating a well-established population. Almost half the sample (49%) had no more than a primary education. Common job functions included working as a store owner, being a homemaker, construction, informal work, students, lottery ticket booth sales, carpenters, and salon beauty shop workers.

Results
Based on the analysis sessions, the ndings are organized into the following core domains: 1) malaria knowledge, perceptions of healthcare, and other health concerns; and 2) perceptions of anti-malarial MDA, institutional trust, and health messaging.  (7) 4 (7) 4 (7) 2 (3) 2 (3) 19 (31) 1. Malaria knowledge, perceptions of healthcare, and other health concerns Overall, most interview participants (64%) a rmed knowledge of malaria (answered yes to the question, "Do you know what malaria is?"; Supplementary Material 1). In follow-up questions, participants gave a generally accurate description of the disease as transmitted by mosquitos; causing fever, headache, or body aches; and preventable through the use of mosquitos nets, repellant, fumigation, and/or the elimination of breeding sites (criaderos). Research team members remarked that although participants connected malaria with mosquitos, some confused malaria with dengue, another mosquito-transmitted febrile illness common in the Dominican Republic. For example, when asked how malaria is contracted, a 60 year-old woman in Guaricano said: "From mosquitos that reproduce in clean water," likely recalling health messages about dengue (transmitted by Aedes mosquitoes that favors small containers, tires, or natural holes that ll with rain water for breeding).
In fact, participants appeared more familiar with dengue than they did malaria; in group discussions, this was said to result from greater exposure to health messaging about dengue than malaria and because of heightened fear of dengue. One assistant said that there are likely political reasons underlying the lack of awareness about malaria. The Dominican Republic is a popular tourist destination, and the country's political leadership and tourism industry have a vested interest in preventing news about malaria from disseminating beyond transmission areas. Simply put, malaria frightens tourists. There was urgent need to control outbreaks, "before people [in the community] start talking to the press," one assistant halfjoked. All of this was said to contribute to poor dissemination of information about malaria.
Knowledge about malaria was generally shaped by 1) personal experience or word-of-mouth in past outbreaks, in which a participant and/or his or her family member or neighbor had contracted the disease and dealt with symptoms, care-seeking, and diagnosis rst-hand; and 2) health activities led by community health workers (promotores). In general, there was a positive perception of promotores sponsored by CECOVEZ and the Ministry of Health. Promotores are individuals who live within the community and were sometimes said to be the only source of health-related information or initial pointof-care.
Doctors, too, were trusted, although negative experiences in past malaria outbreaks had sown mistrust. For example, numerous stories recounted how a sick family member or neighbor who had fallen ill was misdiagnosed by a doctor, sent home, and diagnosed later by a health promotora after the patient's condition had worsened. When malaria was suspected (as opposed to other diseases), people often trusted the CECOVEZ eld team and community health workers more than doctors, usually based on past experience or overheard anecdotes in which patients were nally diagnosed by a promotora. One 38-yearold man in Los Tres Brazos said that, "hospitals don't cure malaria." Still, many people advised that in general, they would still seek care at hospitals or clinics, likely because they had few options anyway and because medical doctors were still respected for their knowledge: El que sabe es el medico-"he who knows is the doctor," said an older man in the Simon Bolivar neighborhood. Decisions for seeking healthcare tended to rest with female heads-of-households (amas de casa), such as a mother, wife, or grandmother.
Malaria was linked to trash, contamination, and inadequate sanitation. People cited the nearby canals (cañadas) that were clogged with foul water, emitting a bad odor also said to cause disease (akin to miasma). A few people thought that malaria could be transmitted through water itself. Community residents valued cleanliness, mentioning that prevention of not just malaria but disease in general could be brought about by attention to hygiene (limpieza) of the home and surrounding area. This was in turn linked to environmental contamination in their communities, which surfaced in explanations of disease.
"What worries us most is what they will do about this trash here," said a resident of Las Lila, a particularly marginalized sub-barrio within Los Tres Brazos (Fig. 2). Indeed, upon returning from Las Lila, the research assistants were visibly taken aback by the extent of contamination and abject poverty found there: "there were kids playing in the trash, and people preparing meat to cook [while] standing in trash," said one team member. "Sometimes, the trash truck won't come for [up to] two weeks," a young man in Sabana Perdida was noted to say.
Two-fths (39%) of interview participants said that asymptomatic malaria infection was possible, usually understood from personal experience or word-of-mouth. For example, one woman in Los Tres Brazos said that, "I heard it said that one neighbor here [had] malaria and [he] said that he didn't feel anything." Another woman in Capotillo said that "There are many people who are sick and still go about their day as if nothing is wrong." In an evocative remark by a young man in Los Guandules, asymptomatic malaria was described as something hidden: Si esta escondido, es algo que esta todavía tapado -"If it is hidden [not causing symptoms], then it must be something that is covered up," or undetectable.
The majority (58%) said that malaria could not present asymptomatically. "I think that it isn't possible to not have symptoms, I think [you] must feel something," said an older man in Los Guandules.
Finally, common health problems cited by community members included gastrointestinal illnesses (vomiting, diarrhea) as well as infections, such as HIV/AIDS, dengue, the common cold or other viral syndromes known broadly as gripe, and papilloma. Chronic diseases mentioned by participants included diabetes, high blood pressure, and cancer. The overwhelming support for MDA appears attributable to two key factors: concern for one's personal health and the desire to help the broader community. "We do not know if tomorrow we [will] fall ill," said a 55-year-old woman in Los Tres Brazos, expressing a need for prevention. In a way, participation in MDA was one way to circumvent malaria's unpredictability: "I understand that prevention is better than [trying to] cure the disease," said a 24-year-old woman in Los Tres Brazos.
Participating in potential MDA was also seen as an act of service. "I would take the medicine to prevent malaria in the neighborhood," said a 28-year-old woman who worked in a banca selling lottery tickets.
Connecting individual participation to community-level bene ts was widespread across all sampled communities. In essence, participation in MDA communicated a sense of solidarity and care for the community. To illustrate the vocabulary and phrasing that participants used to communicate this sentiment, exemplary quotes (accompanied by original Spanish) are compiled in Table 3. It is a social act that deserves support.

Los Guandules
For the people of the neighborhood, because they need it.

Los Tres Brazos
To help my neighborhood with this evil disease.
Para ayudar a mi pueblo con esa enfermedad mala.

Los Tres Brazos
To cooperate with others.

Los Tres Brazos
That way, other people in the community won't get sick.

Male, 21 years old, Sabana Perdida
Because what I want for myself, I want for others.
Por que lo que quiero para mi, yo quiero para los demás.

Female, 38 years old, Sabana Perdida
If it bene ts the community, I see it as very important.
Si es bene cial a la comunidad, lo veo muy importante.

Female, 27 years old, Guaricano
However, several key reasons emerged for the small proportion of those who would refuse participation in MDA. One reason was feeling unsure if the drug administrator was trained competently or doubting whether the medicine was prescribed by a doctor. In one informal conversation, a resident of Los Tres Brazos said that he or she would take the medicine only, "if the public health [team] is trustworthy." Another in Capotillo said that, "There are people who won't take [the medicine], if it isn't prescribed by a doctor." An older man in Simon Bolivar said that he would participate, but emphasized that, "I don't take medicine if [it] is not prescribed by a doctor." Another potential reason for declining MDA is that some people would rather decline the MDA in order to observe if others in the community succumb to side-effects. This re ects caution and skepticism towards MDA and underscores the importance people place on their own observations, rather than reassurance by health workers.
Perhaps one of the most obvious hurdles to overcome in a successful MDA is coordinating the MDA with the work schedules of community residents. Work was clearly valued by residents because of precarious economic circumstances. Thus, although they may see participation as a bene t to themselves and their community, their need to earn income may be prioritized above the perceived risk of contracting an otherwise rare disease (  Older people, too, expressed unique reasons for hesitating or refusing to participate, such as concern about drug-drug interactions and possible side-effects, because they may already be taking prescribed medicines and conceptualize their health as too fragile for MDA. Another interesting point was that those with upper-and middle-class socioeconomic status may be less inclined to participate in MDA because they have generally had little prior exposure to malaria outbreaks and may feel as though malaria is not a problem for them. To note: this point surfaced in a team analysis session after visiting more well-to-do neighborhoods. Neighborhood associations (juntas de vecinos) are widespread throughout the Dominican Republic and are common entry-points for public health programs to begin community outreach [4,17]. Presumably, community engagement for a future MDA in Los Tres Brazos would also involve working closely with juntas de vecinos in the transmission focus. However, most people in this study had a negative opinion of local neighborhood associations, which were seen as too political and self-serving. "An evil rat," said one resident in reference to his junta de vecinos in Los Guandules; ladrones (crooks), said another in the Sabana Perdida neighborhood; "they convene us only when they need something," said another in Simon Bolivar. Juntas were also said to be overly political; "the junta here resolves problems only for their political party," said a resident of Simon Bolivar. Finally, the assistants described that some residents seemed to be unsure if their community even had a junta de vecinos, re ecting a lack of community-level organization.
Some hypothetical scenarios in which local politics could adversely affect MDA were raised in team analysis sessions. Some politically aligned neighborhood associations, for example, may seek to capitalize on any large-scale provision of free services in order to gain support among the voting populace and thereby distort MDA campaign messages, such that MDA and ancillary activities appear to be a liated with a given political party. In contrast, other political parties may seek to undermine MDA activities by spreading misinformation such as, for example, "Political party XYZ is funding the MDA in order to poison this community, because the people here belong to the opposition party." However, although trust in juntas de vecinos may have been low, some people mentioned certain key gures in their communities who do help to resolve issues. These individuals were often referred to by their nicknames and were said to be available (dispuesto) for anything (cualquier cosa). Participants also referenced churches as trustworthy sources for information about health and as sources of support in the community. However, sometimes people said that there was simply nobody, or no formal group or organization, in whom they could trust or turn to resolve problems.
Health messaging and health information were closely related to issues of trust as well, carrying additional implications for MDA. Common sources of health information included the promotores, nearby hospitals or clinics, and even 9-1-1 ambulance staff. Some cited mass communication, such as television and radio, although one research assistant observed that few people actually had televisions (due to their economic circumstances); even if they did, electricity in the community was unreliable. In one analysis session, one potential idea for spreading messages was identi ed: the small gua-guitas (little trucks) that blast announcements (perifoneos) about upcoming events, such as happy hour drink specials at a nearby discotheque or discounted prices for fruits and vegetables.
In addition to spreading information about malaria, word-of-mouth among community members and educative talks (charlas) led by promotores were seen as important ways to learn about health issues in general. An older man in Simon Bolivar neighborhood said: Well, here, whenever there's a disease, we know about it from others in the neighborhood talking about it, that's how it was with dengue.
One woman in Capotillo opined that the best way to get health information was, "from the health promotores, because sometimes the news says fake things." Social media, especially among young participants, were also mentioned as sources of health information.
However, assistants detected a general disconnect between the content of health information that penetrated these communities and the ways in which community members described malaria or disease in general. This tended to be framed in light of the precarious socioeconomic conditions of the community. For example, although many interview participants cited contaminated water, including the nearby Ozama and Isabela rivers, as responsible for disease, they seemed unsure as to exactly how: "There is a disconnect between information channels [how people obtain information] and their social and economic reality," wrote one assistant during an analysis session.

Discussion
Conducted over seven days, this rapid ethnographic assessment generated contextual understanding of health and healthcare in the Los Tres Brazos focus of Santo Domingo, particularly community perceptions of malaria and MDA as a potential intervention to interrupt transmission and contain an outbreak currently affecting the area. Salient ndings and their relevance for MDA planning, implementation and community engagement are summarized in Fig. 3 and Box 1. This work further demonstrates the value of rapid ethnographic assessments to inform community engagement [12,18] and adds to previous research on motivations and barriers to MDA for malaria on Hispaniola [19] and elsewhere [20][21][22][23][24].
The population appears to have a basic understanding of malaria, trust community health workers Much of the knowledge about malaria is due to personal experience-of either one's own malaria infection or that of an immediate family member or acquaintance. Community promotores were the source of additional knowledge about malaria. Although many participants who reside in the transmission focus still trust medical professionals-advising that they would seek care at medical centers should they fall ill in general (perhaps because they have few options anyway)-they trust community-based health workers for accurate malaria diagnosis and information. However, there was an apparent contradiction: participants expressed, on the one hand, mistrust of doctors regarding diagnosis of malaria, but still felt hesitant or skeptical of MDA if they were not assured the medicine was prescribed by a doctor. This discrepancy will need to be carefully considered when conceptualizing the roles of medical professionals and community health workers for potential MDA implementation.
Greater familiarity with dengue may due to a few reasons. First, dengue is more widespread than malaria, affecting 29 of the country's 32 provinces. Second, the year 2019 witnessed a dramatic rise in cases nationwide compared to 2018: up to epidemiological 41 (the week of this study), there were 542 cumulative cases nationally, compared to 35 at the same time point in 2018 [25]. Furthermore, annual deaths from dengue have consistently outnumbered deaths from malaria in the Dominican Republic: in 2019, there were 53 deaths from dengue, while there have been fewer than 10 deaths per year from malaria since 2012 [26]. Taken together, these features of dengue, along with public health sensitization campaigns, may have contributed to more awareness and fear of dengue in the general population.
No participant in the study was familiar with the term "mass drug administration" (administración masiva de medicamentos). This result is not surprising because the only recent anti-malarial MDA campaign in the country was in 2005 in the El Seibo province, located several hours' drive from the capital. MDA campaigns against lymphatic lariasis have also taken place in several regions of the country, including annual treatment in the nearby Santo Domingo de Guzman neighborhood of La Ciénaga from 2004-2006 [27]. After providing a formal de nition of MDA, nearly all agreed that they would participate.
Although desirability bias cannot be excluded, inspection of eld notes indicated strong themes of solidarity. Given this nding, one way to ensure high participation in MDA would be to implement MDA as a part of a larger effort to improve community health [23,24,28,29]. Campaign sensitization could use salient idioms and phrases used by participants in this study to build on the idea of participation in MDA as a way to care for and take pride in one's community (see Table 3 above).
This study revealed low levels of institutional trust in current civic associations and the potential danger for political manipulation of health campaigns. In addition, the COVID-19 pandemic has highlighted the growing trend of medical misinformation. "Fake news" about the pandemic circulates on social media in the country driving demand for ineffective therapies such as hydroxychloroquine [30]. Mistrust of political authorities, counter-narratives to o cial explanations, and medical malpractice and poor oversight seem to be ourishing. These issues must be proactively addressed to ensure high community participation in MDA. One potential way forward to build trust in affected communities is to conduct MDA in conjunction with other services, such as free medical screening for hypertension and diabetes and donations of children's school materials, chlorine tablets for home water storage, and iron tablets and prenatal vitamins. Such ancillary activities have been shown to improve participation in MDA in other contexts [23,24,29] and would help to align MDA with community health concerns. "Beyond health concerns, [people] see solutions to basic problems, like electricity, potable water, and lack of work as important", wrote one assistant in an analysis essay; another assistant wrote simply that, "They seek to obtain a better way of life." Because people also seem attuned to issues of sanitation and pollution, MDA could occur alongside trash clean-ups, elimination of mosquito habitats, bed net distribution, and residual spraying. Medical screenings could be staffed, for example, by local medical and nursing students supervised by resident physicians. The presence of medical doctors on site could also reassure people who may be skeptical of taking medicine without a doctor's prescription. Ancillary activities like these would signal that MDA is part of a larger concern for community health and would address issues raised by participants in this study.
These ndings provide substance to enact key recommendations for community engagement strategies highlighted in other studies [8,24,31]. In general, these studies emphasize early formative research about the community; strengthening bonds between communities and MDA program staff; being responsive to community needs and concerns; and fostering shared responsibility. These strategies could become operationalized in this context through, for example, conducting more formative research into themes of community solidarity, social networks, and key community gures; holding community-based educative talks (charlas) by trusted promotores to raise awareness about malaria; contextualizing MDA as part of a broader effort to improve community health; and incorporating ancillary medical care alongside MDA activities. Even though 64.0% of participants appeared to know about malaria, it will still be crucial to launch educative campaigns that use clear, non-technical language through diverse media and forums.
For example, the vocabulary of malaria as being "hidden" (escondido) or "covered up" (tapado) may be more effective at communicating ideas about asymptomatic infection than biomedical language.
Appropriate venues for transmitting information about malaria and MDA could include in-person meetings at community gathering spaces such as churches, youth clubs, colmados, and clubs and even through networks of gang leaders or drug dealers in marginalized areas. Involving women's groups would also be particularly worthwhile because women had considerable decision-making power over health matters in the home. It is important to recall, too, that in this setting, forms of collective organization and mutual aid go back centuries [32]. For example, the convite is a form of cooperation among rural neighbors who join together to accomplish tasks [33], while lines of credit ( ao) are commonplace at corner store colmados for neighborhood residents in need [34]. To have greater local resonance and generate trust, MDA activities could be modeled on these longstanding cultural practices of mutual aid and reciprocity.
Other important media to consider for transmitting health messages include digital platforms such as YouTube, Facebook, Instagram, and WhatsApp. Word-of-mouth will be important in spreading health information and countering unsubstantiated rumors. Precarious economic circumstances are widespread throughout the transmission focus; a large number of people may lack radios, televisions, and smartphones (or electricity in general). In these cases, it will be important to develop tailored messaging strategies that circulate through reliable intermediaries and other trusted but less formal social networks. It will be necessary to identify, recruit, and supervise these intermediaries, who could spread messages about MDA and counter rumors. Another potentially promising way to spread information about MDA would be through clustered friendship connections [35], which deserve more research in this setting.
Health messages should be harmonized between community and clinical spheres. People can become confused (as they appear already) when they hear from promotores that malaria is important and deserves their attention, but are then told by doctors at local clinics that malaria is not a problem.
CECOVEZ, community health workers, and clinical staff should all collaborate together on developing a coherent, clear, standardized message in the months leading up to MDA.
Finally, working with local neighborhood associations (juntas de vecinos) may be problematic given the mistrust that community residents appear to have for these institutions. The malaria program will need to balance the need for collaboration and the risk of appearing too politically aligned or showing favoritism.

Strengths And Limitations
This study was limited by its primary focus on general health and anti-malarial MDA, whereas a traditional malaria knowledge, attitudes, and practices (KAP) survey would shed more light on the spectrum of malaria knowledge, prevention, and treatment in the focus. Therefore, the next step for the research team is to conduct a KAP survey based on a representative sample across a wider segment of the focus. Additional limitations to this study include response bias, in particular to the question of MDA participation because the de nition of MDA did not mention the possibility of side-effects and was given by CECOVEZ-a liated interviewers. The question "Do you know what malaria is?" is also fraught with issues of desirability bias (saying "yes" to a question in biomedical language posed by public health personnel). Furthermore, all three-eld assistants were men, which may have in uenced rapport with female participants. The shortened timeframe prevented more in-depth training of team members, sampling in other neighborhoods, obtaining a larger sample, and more time for analysis and re exivity.
Nonetheless, certain strengths helped address these limitations. First, the team was composed of locally hired assistants, two of which had years of experience in the malaria program, were familiar with the study site, and had already established rapport with community members. The other assistant was also Dominican and had extensive prior qualitative research experience. Second, the questionnaire was developed at length in team discussions; the team decided that the question "Do you know what malaria is?" would be the most comprehensible to people living in the focus and could be veri ed by the follow-up questions regarding symptoms and transmission. Third, the built-in analysis sessions allowed the team members to re ect upon their own experiences and observations and interpret emerging themes through shared discussions, which helped identify bias. Fourth, the study included multiple methods to help triangulate ndings (structured interviews, informal conversations, personal observations, and team discussions) to con rm or offer alternative explanations. Lastly, one study coordinator (HK) took daily notes as re exive practice to track sources for bias and how key ndings arose in team discussions.
Longstanding relationships among all team members and with community members further facilitated team cohesion, lively analysis sessions, and rapport with study participants.

Conclusion
This rapid ethnographic assessment yielded rich contextual information crucial to evaluating the feasibility of potential anti-malarial MDA. The ndings reveal pathways to ensure that MDA be conducted in a socially responsible way. The larger project in which this study was based was reviewed and approved by the Dominican Consejo Nacional de Bioetica en Salud (CONABIOS) and the Ethical Review Board of the University of Amsterdam.
The study was reviewed by Emory Institutional Review Board (IRB00099347) and given a non-research determination. All participants provided verbal informed consent.

Consent for publication
Individuals appearing in photographs provided consent for publication.

Availability of data and materials
The dataset used and analyzed in this study is available from the corresponding author on reasonable request.  Scene within Las Lila with metro system cable cars overhead and Rosario Sánchez bridge in distance.

Figure 3
Major factors likely to in uence the acceptability of MDA in Santa Domingo.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. SupplementaryMaterial12.docx