Health social research in a context of violence : a look from the ethics 1 La investigación social en salud en un contexto de violencia : una mirada desde la ética

During the 2006-2012 six-year presidential term, a battle against drug trafficking began, which has been continued by the current president and it still occupies one of the first positions on the political agenda of Mexico. This battle has mobilized the national security forces and its consequences have gone through the political sphere and reached society as a whole. This is just one of the scenarios faced by the country; historical community conflicts and an atmosphere of generalized violence fueled by kidnappings, homicides, and various kinds of crimes accompany it. This essay thinks through how social research, namely the fieldwork phase, has also been affected by this context. The paper shows some obstacles that people working in the field (researchers, students, interviewers, and pollsters) have to face. It examines how this context is affecting the social research carried out, seriously compromising the results, and thinks over the research ethics, pointing out the lack of protective measures for the personnel participating in this kind of studies.


Introduction
During the administration of President Felipe Calderón (2006-2012), one of the most important social crises in the Mexican post-revolutionary history began.This battle, with hints of war, which the Mexican government launched as a response to the increasingly apparent drug trafficking penetration into the social fabric, has resulted in alarming figures of violent deaths.The Office of the Mexican Attorney-General acknowledged that, during the first five years of the Calderón administration, the number of deaths associated with violence due to drug trafficking was 47,515 people (PROCESO, 2012), although some civil society organizations, scholars, and journalists point out underreporting in the official figures2 .
However, this was not the only problem society faced during the Calderón administration, other figures provide us with evidence of a scenario that is unstable and frightening, for instance, the index of the incidence of crime and violence3 ; according to the Research Center for Development, A.C. (CIDAC), in 2009, Mexico showed a range between 82.16 and 7.82, where Chihuahua was the state ranked first and Yucatán was the last one.Regarding intentional homicide, in the same year, Mexico was still among the countries with the highest violence level: it was ranked 16 th out of 115 (CIDAC, 2009).
In December 2012, Enrique Peña Nieto came to office, a president from another political party, and after more than one year into his administration, violence did not show a substantial decrease in the country.As in many Latin American countries, another major threat in the Mexican society is delinquency, regarded as one of the multiple forms of violence causing major damage to the social fabric (Jiménez, 2003).According to data for 2013, Acapulco was classified as the third most violent city in the world (CONSEJO CIUDADANO PARA LA SEGURIDAD PÚBLICA Y JUSTICIA PENAL A.C., 2014); the figures referred to above may explain the finding that 68% of the population aged 18 years and over living in urban areas reported a feeling of lack of safety in Mexico in December 2013, according to the Mexican National Survey of Urban Public Safety conducted by the Mexican National Institute of Statistics and Geography (México, 2014a).
Simultaneously, other forms of social conflict are observed in the country, disputes over land tenure (López, 2006), particularly between indigenous peoples, were not extinguished by the death of Emiliano Zapata 4 or the Agrarian Reform driven by the revolutionary presidential administrations, and they reflect, among other things, the inadequacy of the legislation that justifies the need for "informal" arrangements, i.e. justice imposed by the people facing a direct conflict.By way of one example, in December 2011, the Secretary of the Government of Oaxaca pointed out, during the state congress, there were 19 agrarian conflicts classified into the "high risk" group within the entity (Matías, 2011).
The purpose of resorting to the statistical data presented above is providing elements that allow us to set the scenario through which we will think over the social research work in the health field.

Health social research (fieldwork)
For the purposes of this paper, we mean by social research in the health field the studies whose design is aimed at exploring the cultural aspects that determine illness and disease processes and the health care provided as collective phenomena that, through various theoretical approaches from social sciences such as Anthropology, Sociology, or Economics, provide evidence of the close relationship between the populations' health conditions and their economic, cultural, and political characteristics (Fuentes;López, 2005).
Social research uses various methodological tools that enable it to get into such diverse social contexts as the various Mexican communities and regions.Even when it comes to primary research, one of the main characteristics of social research is, undoubtedly, direct contact to people, this is the main source of information in the task of retrieving their experience, viewpoint, and perceptions of the themes under study.
Within the research process, a very important step, i.e. fieldwork, whose main objective is obtaining direct information, either in graphic, documentary, audio, and/or video format, at the actual study scenario.Fieldwork is the step within the research process where the social scientist and his team come in direct contact to the population (informants) in order to obtain evidence (testimonies) regarding the issue investigated.
Some obstacles that social researchers may face in field comply with requirements of the study design (e.g.survey application to scattered households; repeated sessions to take in-depth interviews); others are due to the cultural and structural characteristics inherent to the populations where the studies are conducted (e.g.limited access to media such as roads, telephones, internet; population's conceptions of the researcher and his task; community activities that do not correspond to the study schedule; language and worldview, in the case of indigenous communities); as well as other obstacles are the result of local conflicts, in addition to the context of violence and lack of safety experienced in the country as a whole, to which we have referred to above (e.g.territories occupied by drug traffickers, including 'ejidal' lands, communities, and groups faced with the support of political organizations).The first two types of obstacles are, to some extent, as predictable as inevitable and it is possible to prepare strategies to cope with them, ethnographic knowledge of the community under study and covering an extended period doing fieldwork, are two elements that can provide such strategies with a basis.
However, the obstacles linked to violence are more difficult to tackle and they are also increasingly usual in Mexico, although research institutions have not developed proper protocols to overcome them or trained fieldwork staff to pursue their work in this scenario of violence.This situation has significantly affected social research in general and the research team's safety in particular, something which makes us put into question the risk-benefit ratio of research, an aspect we think through below.

Research ethics
When referring to research ethics, we talk about complying with four primary principles to protect people who participate in a research: a) Autonomy (right to self-determination); b) No maleficence (do no harm); c) Beneficence (different from nonmaleficence, promoting good); and d) Justice (right to goods and services) (Luna, 2008).While these principles are recognized, the ethics applied to research is flexible and dynamic, since the contexts and social actors change, but an ethical minimum for a plural coexistence is needed, given the divergence (México, 2008), so that these principles have not a binding, but a guiding nature, as pluralistic reflections based on deontological ethical approaches could hardly be articulated.In other words, research ethics, understood as the critical application of moral arguments at the various times in the research process, is not a normative ethics, since it does not provide judgments or guidelines indicating moral obligations that apply universally; the principles on which it is based have a guiding nature5 and try to lead scientific work by monitoring compliance with the participants' rights, but non-compliance with it is not sanctioned.The Council for International Organizations of Medical Sciences (CIOMS) expressly states that its guidelines are intended "[...] to guide, especially poor countries, in defining national guidelines on the biomedical research ethics, applying ethical standards in local conditions, and establishing or redefining adequate mechanisms for ethical assessment of research involving human beings" (CIOMS, 2002, p. 14), thus, they can help establishing ethical guidelines and defining national policies and mechanisms for assessing the ethics of studies, recognizing an adaptation to the characteristics of various contexts.
Research ethics, then, is rather an applied ethics that goes beyond the traditional principles and it is set out, as expressed by Mondragon: "to identify, deliberate, and somehow mitigate dilemmas that arise in the knowledge generation process" (Mondragón, 2007, p. 27), considering the ethical conflicts that may occur are different due to the different research processes.In the Latin American context, it has been said there is a need to add notions such as respect for dignity, tolerance, inclusion, solidarity, and non-discrimination as guiding standards and practices (México, 2014b).The ethical judgment derived from this is aimed at the protection of people participating in research studies, but it comprises the possibility of various resources and/or pathways to get it.We share the idea of ethics as a dialogic and interdisciplinary exercise (Luna;Salles, 1996), applied to contexts and situations where science is built.
Understood this way, research ethics, let us say that the guarantee of these principles is closely linked to the scientific rigor of studies, i.e. to the extent that a research study is properly designed, it will minimize risks to participants and provide the best benefit possible, it will make apparent any conflict of interest, demonstrate a genuine commitment to the communities, always observe asking for informed consent, and ensure privacy and confidentiality of information.Ethics is a part of the methodological rigor of any scientific investigation.
In short, when we talk of research ethics, it is usually referred to the protection of people participating as informants in a study, be it a clinical trial, an epidemiological survey, and by extension, a social research; but nobody has referred to the risks to which research teams working in contexts of violence and insecurity are exposed, in order to contribute to science and public policies, aiming to improve population's health.This protection, we believe, should also be of interest to bioethics.

Risks and vulnerability
Some authors argue that the research team is not usually put at risk during fieldwork, unless in the case of a scientific study conducted in disaster situations (O´Mathúna, 2010), in order to be true, this statement should include the context of violence linked to drug trafficking as a kind of social disaster, otherwise it would not observe the risks posed by situations that go beyond nature.
If we analyze the conditions in which a researcher and her/his team go out to the field, we may identify, initially, a double vulnerability.The first of them has to do with their socio-personal characteristics, such as gender, socioeconomic status (there are also social classes among the researchers themselves and in relation to their team), or their labor affiliation (this has to do with the institution's prestige and its relationship, almost direct, with the researcher's revenue).This vulnerability, which we name "personal", is added to her/his vulnerability as "scientist", defined among other things by the funds available to conduct the research, which determine, inter alia, the resources available, human and material, to carry out a study.
Regarding this double sense, we claim that a social researcher is vulnerable, but international regulatory documents on research ethics, such as the Helsinki Declaration (AMM, 1964), the CIOMS standards (CIOMS, 2002) or the recommendations issued by the World Health Organization (WHO, 2001) refer only to participants' vulnerability, and this leads to recommendations of special protection for these populations (e.g.witnesses or translators, for obtaining informed consent), so that what is hitherto absent in these recommendations is, first, recognition of the vulnerability of those who investigate and, second, the identification of specific kinds of protection.
What we have referred to, herein, as research risks in the context of violence and lack of safety that prevail in the country are, actually, "dangerous obstacles" threatening the sustainability and validity of a study.These obstacles contribute, at the same time, to a person's vulnerability, every time the research team can be threatened or violated, and to scientific vulnerability, considering what affects the quality of information gathered.
The credibility (internal validity) and safety (reliability) of a research study, as scientific quality criteria, have been affected, because when informants are fearful and do not trust the research team, it is likely they do not agree to participate in the study or, if they do, they can provide shallow answers; under these circumstances, the team might decide to improvise techniques that do not guarantee safety regarding research standards, in any event, data lose density or theoretical wealth, and this makes possible conclusions weaker.
There are various gradients with regard to this vulnerability, as the risks faced when working with a captive population (hospitals, nursing homes, people benefiting from a government program) or middle and upper social classes are not the same faced when working with a free population under conditions of social marginalization.
The research team reaches the community as an unknown group of individuals, despite the formal introduction to local authorities or the ethnographic exercise that has been previously performed.Trust is a feeling that is built through actions, but this is difficult to gain it in a context of violence or crisis.From the look of a potential participant, this context takes away the guarantees of safety and protection with regard to the information she/he can provide and, enjoying her/his autonomy, she/ he chooses not to participate in a study.This mismatch of viewpoints has been usual in science, the informant sees a threat where the researcher sees knowledge opportunities, the difference is that in health social research taking place in a context of violence, both the researcher and the participant may feel threatened.
The informant asks about the researcher's origin (Who are you?Where did you come from?Is it the government?), and the researcher wants to know about people's perceptions and practices, their questions become a matter of doubt for the informant and this can be understood only in the context of uncertainty which we live in, where selling confidential data is possible, too.Obtaining information gets more complicated considering the participant's profile, lack of trust is linked to the socio-cultural history, migrants fear deportation, indigenous individuals are afraid of being treated unfairly, women are afraid of being harassed, the elderly fear abuse, people with a higher socioeconomic level fear kidnapping, so that the effect results in a significant decreased number of informants who agree to participate in a research, thus compromising the validity or reliability of the study.
As in many other areas, usually in health social research, the study groups are often cataloged as vulnerable (elderly, migrants, indigenous, sexual minorities, etc.), which already face disadvantages due to the mere fact of belonging to these groups, but this may become worse in extreme situations, such as economic crises or increased violence.In our field experience, we have observed how, in these extreme situations, mistrust and refusal to participate in research studies also increase, in part because they do not want to be exposed to any risk that could lead them to a worse situation.
Generally, these vulnerable groups live in contexts where violence has turned the locations into marginalized zones with poor infrastructure and public services, there are not any kind of guarantees to ensure protection for the research team involved in the field survey.In this scenario, researchers try to meet all objectives of their studies and they are committed to conclude them with the highest scientific rigor possible, but the context of lack of safety can restrain them to make decisions that could jeopardize the validity of the study, as well as the safety of potential participants; they can make hasty decisions with no methodological foundation, such as shorten observation time, wandering, ethnographic record, and remaining within the field; unintentionally, they may expose the participants who accept to help to risks, as the latter may be asked by other members of the community; they may even push the collection of information when it does not work properly, something which can lead to insufficiently grounded conclusions and recommendations.It could be argued that, when designing a study, the research team has the freedom to choose the context where it will take place, thus researchers would be able to avoid exposure to situations such as those described above, however, in a country where violence has spread throughout its territory, the alternatives are significantly decreased and a peaceful locality or region selected for the study could become violent between project approval and completion of fieldwork, besides the very interest of a social research, to be able to document various health situations, precisely where violence becomes a determining factor, so it is not always possible to change the study scenarios.
That is why we identified a significant link between the vulnerability of those who investigate and research ethics, whose interests so far has fo-cused on the protection of participants.We propose that international organizations, at the same level of those involved in drafting documents such as the Helsinki Declaration, provide some space for thinking through the protection of research teams, considering that a study cannot be ethical if it has to endanger the life or integrity of those conducting it in order to generate results.

A case example
The study described below is only one among many examples of research study in the health field that are carried out by facing situations of violence and lack of safety.This is a study conducted in Mexico, in municipalities within the states of Jalisco, Guanajuato, Puebla, y Oaxaca, whose objective was identifying the influence of transnational migration between Mexico and the U.S. in order to solve health problems of migrants and their families in the communities of origin; the study was approved by the Ethics, Research, and Biosafety committees of the Mexican National Institute of Public Health.
The original criteria to choose locations to carry out a study referred to theoretical and empirical information related to the theme: intense migratory flow, belonging to traditionally migratory regions, in the central-south area, rural communities, and those having a marked ethnic diversity were considered.The outcome of this election has led to states and municipalities where there was also violence associated with drug trafficking and community conflicts, and this had increased within the period between project approval and fieldwork, thus, for safety purposes, three out of four municipalities had to be replaced, choosing others whose characteristics were as close as possible to the original ones, but without known conflicts, and to elect the new municipalities an internet search for recent news related to violence and drug trafficking were performed, a physical visit was made to the communities, and the researchers spoke to the local political and health authorities.So, the first point to emphasize here is the fact that the context of lack of safety and violence is an element that is involved in decision making, going beyond the methodological criterion.
For fieldwork organization, three major issues affecting logistics were faced: a) the need for providing evidence on expenditure in rural communities, where it is very difficult to find facilities that issue invoices for lodging and/or shopping; b) difficulties in obtaining a suitable vehicle, given the terrain characteristics where we had to access (landslides, dirt roads, holes); and c) delay or inability to obtain support material for proper institutional identification, needed for the context where the study would take place, such as shirts and vehicle with distinctive logos, radiolocation systems, since they are dispersed communities.The issues mentioned are, to some extent, usual in the research study, but in contexts of high uncertainty, these problems are magnified, because evidence on expenditure forced researchers to return to the municipalities where the provision of shops and services is greater, and they have move again along the paths that are recognized as hazardous, thus increasing exposure to risk.The lack of a suitable vehicle not only affects access to communities, but also ensures leaving them, something which may be unexpected.And the credentials and badges can contribute to increase population's trust.
At the time of field survey, both in Jalisco and Puebla, increased violence had not been observed at the community level, therefore it was possible to conduct the survey in a timely and accurate manner, without major complications.Although one year after study completion, in the municipality selected in Jalisco, clandestine graves with several corpses were found.
In Guanajuato there were obstacles for conducting fieldwork, due to increased distrust deriving from the wave of violence.The study communities are very close to the state of Michoacán, one of the entities where drug trafficking violence has increased most in recent years, also during stay in the field, two events that could affect population's participation took place: 1) a group of migrants had left the U.S. in cars towards the community and they had not arrived or there was no information about them; 2) a criminal gang consisting of inhabitants from the municipality had been arrested by the police authorities.
In the Oaxacan Mixteca, the study area has a strong indigenous presence and it was surrounded by two regions with violent conflicts between communities and political groups ('Mixtecos' against ('Mixtecos' and 'Triquis' against 'Triquis') linked to land tenure.We had communication problems because they are communities that speak indigenous languages and we found little presence of the federal government both in order to provide police services that could prevent violent acts and to assist in the peaceful resolution of conflicts through negotiation tables.In addition to an unfavorable environment permeated by the cacique's political power, cooptation of civil society organizations (CSOs) and dishonesty and betrayal of political and social leaders, abuse of migrant savings and managers, alcoholism and violence against people, limited access to media (roads, telephones, deficient internet), presence of variants in the 'Mixteco' language, and scattered households.Given this reality, the research team met very suspicious, fearful, and apathetic people, drunken and aggressive men on the streets, minimal support from local authorities, and there were problems when communicating to inhabitants, as well as inadequate knowledge of community activities.
Faced with these obstacles, the results could be expected: many of potential respondents (returning migrants) did not meet the schedules and commitments agreed, and there was no help by men, especially young ones, regarding the focus groups organized for gathering information.
The working team decided to change the strategy originally proposed in order to overcome obstacles, e.g. for recruiting people, the information gathering technique was replaced by interviews, support was asked to the staff from the official primary health programs, as well as CSOs, native translators were hired 6 , the observation time in the communities was increased by interviewers, including weekends, and there was a field monitoring by people from the community.
It is true that some of the obstacles mentioned in this example are linked to the very methodological strategy and they may be solved through sound 6 The original protocol relied on the participation of translators who were not members of the community, but they knew the location.
decisions having the same nature, however, there are others that, by virtue of their nature, go beyond the researcher's decisions, and they put at risk both team's safety and data gathering and protection.

Final remarks
In recent years, journalists and human rights advocates have been among the most visible victims of the context of lack of safety in Mexico, observed in various scenarios.To inform the population, they have even paid with their lives by pursuing their profession.It is hard to particularly blame anyone in such a situation, but we cannot allow social scientists and their teams to become the next victims.
The institutions where research studies are carried out and government administrations have made a major effort in recent years to observe compliance during investigation with the minimal ethical standards regarding respect and protection to study participants.The research ethics committees, responsible for this task, are the instances that often have the first contact to the projects or studies that are about to be done, and they may be the starting point so that researchers and their institutions become aware of the risks involved in carrying out health social research in a context of violence.By way of one example, we may mention that, faced with the apparent lack of safety in the country, the Ethics Committee of the Mexican National Institute of Public Health has produced and distributed a document to all researchers from the institution, warning them about the risks posed to the fieldwork staff, and it urged them to comply with some preventive measures.Then, some researchers developed and implemented various training strategies for their fieldwork teams, in order to help them fulfill the tasks assigned and simultaneously identify and protect themselves against risks related to the current violence in the country.Hence, ethics committees may contribute to recognize such risks, not only by asking that an investigation study is modified, if the researchers responsible for it do not provide the personnel conducting the study with all guarantees, but also urging them to seek strategies to protect the fieldwork staff.Such risks should be part of the risk-benefit ratio that every research protocol must observe and every committee should include them as part of its recommendations.Every ethics committee that assesses such research studies should ensure the protection of fieldwork equipment, recognizing these risks -which sometimes even the researchers themselves are not able to identify, because it is often difficult to take away from one's own work and commit to signaling them.This effort must be undertaken in situations where risks are observed and violence has become another actor that plays an undeniable leading role.