violence against women : systematic review of the Brazilian scientific literature within the period from 2009 to 2013

Violence against women is a worldwide issue, and the number of publications addressing it tends to increase, due to its magnitude. This article analyzes the scientific literature on the issue, within the period from 2009 to 2013. This is a systematic review of articles indexed in the Virtual Health Library (VHL). It included studies freely available online and published in Brazil, in full text, in Portuguese. It excluded articles without a scientific standard, those published abroad, and government information. The studies were assessed regarding methodological aspects and those related to violence, and data underwent statistical analysis in the softwares SPSS and Excel. Most of the 148 texts were qualitative, published in 2011, by up to 3 authors, and the victims were individuals mainly assessed through Content Analysis. The violence types identified were physical, sexual, and psychological, with overlapping forms related to mental and physical consequences. The findings converge with the current trend to address violence in a qualitative way, in order to provide an in-depth analysis of the victims’ experiences. It is expected that this study contributes to raising awareness as for the need to approach the issue, with a view to promote women’s health.


introduction
Violence against women (VAW) is any act of violence based on gender, which results, or is likely to result, in physical, sexual, or psychological harm, as well as suffering, to women, even threat to commit such acts, coercion or arbitrary deprivation of liberty within a public or private environment 1 .Violence suffered by women may also be named domestic violence (DV) or gender-based violence (GV), it consists in an extremely complex phenomenon, which affects women in all parts of the world 2 and its roots stem from the interrelationship between biological, economic, cultural, political, and social factors 3,4 .In Brazil, this condition has a high prevalence and it makes VAW one of the top priority issues to be addressed by public health 5,6 and the human rights monitoring organizations 7 , as well as a challenge to the health industry 8,9 .
Despite being characterized as a major problem, VAW only gained greater notoriety in Brazil after the creation of Law 11,340/2006 -the socalled "Maria da Penha" law 10 .From then on, this violence type has been defined as a specific crime and potential changes in the way of punishing offenders emerged 11 .According to this law, VAW may be classified as physical, sexual, psychological, moral, or patrimonial 10 , in a unique or associated mode, and there are often 12 overlapping violence types 13 .
Currently, it has been observed that women are victims of harmful acts in a comprehensive way only because they belong to the female gender.This fact suggests that society still has a worldview associated with male superiority 14 .Also, GV has a different nature and various patterns of interpersonal violence regarding the offender because, although there are numerous perpetrators, usually they are within the family context, such as husband, father, stepfather, uncles, cousins, or others 14 .
However, violence by an intimate partner -within the home environment -is the most prevalent and endemic kind of violence against women.A man's right to 'discipline' his spouse is often culturally accept 15 .In this scenario, coping with violence implies the deconstruction of social standards and cultural patterns, both by men and women, which confirm, authorize, naturalize, and trivialize male domination over women 3 .
The literature describes several factors associated with domestic violence, which perpetuate this women's condition, such as family history of violence, use of alcohol by the spouse 16 , unemployment, poverty 17 , the victim's low socioeconomic status, poor social support provided to women 18 , and emotional dependence on the offender 19 .
By considering the victims' context, shame, fear, and ignorance of the legal framework which sets limits to violence is noticed.These factors prevent victims to seek health services 20 .Even when they are forced to seek these services, due to the presence of physical injuries, they tend to hide the problem and rarely spontaneous complain during appointments 8 .This provides GV with a stealth nature 5,8 , which is not something agreed, but given, because women do not fully take advantage of patriarchal power, such as men do 21 .In addition, the effects of violence, stemming from injuries and traumas triggered, are clearly perceived, either due to the economic costs of health care, or within the judicial and penal system, or because of the social costs derived from a decrease in productivity 22 .
Given the complexity related to the issue, addressing GV requires to take into account intersectoral and transdisciplinary actions.Such actions involve various industries, such as health, education, public safety, social work, the judiciary, as well as non-governmental organizations (NGOs).These services contribute to making decisions with a collective impact that create and strengthen health care networks in order to achieve greater problem-solving ability and provide victims with greater support 23,24 .From the same viewpoint, health services deserve to be highlighted, mainly those that work directly with victims 25 .However, both professionals and health services are not prepared and qualified to deal with the problems related to GV prevention, yet 26 .
The magnitude and impact of VAW in society have led to a growing number of scientific studies related to the theme.Nevertheless, due to the use of various methodological designs and instruments for measuring it, there still lacks a uniform view of its actual magnitude 5 .However, conveying the information highlighted by these studies may help us to grasp the circumstances under which the problem occurs and contribute to prevent and fight against VAW, besides provides means for public policies that address women's health in a comprehensive way.
Having this in mind, this study aims to analyze the Brazilian scientific literature on VAW within the period from 2009 to 2013, in order to characterize the trend of these studies.

Method
This is a systematic review carried out through a quantitative, retrospective, and documentary study, which surveyed scientific papers published within the period from 2009 to 2013 and available in the Virtual Health Library (VHL), which contains numerous databases 27 .
A systematic review consists in a research type that uses as its source of data the literature on a certain theme and it provides a summary of evidence related to a specific intervention strategy, by applying explicit and systematized search methods, critical assessment, and a synthesis of selected information 28 .
The strategy used to obtain the published papers took as its guiding principle these Brazilian health sciences descriptors (DeCS) provided by the VHL: "violence against women," "domestic violence," and "gender-based violence." For collecting data, these inclusion criteria were adopted: studies available online, published in Portuguese and conducted in Brazil, freely available in full text, published within the period from 2009 to 2013, indexed in the VHL, which encompasses the Latin American and Caribbean Center on Health Sciences Information (Bireme), Pan American Health Organization (Paho), and the World Health Organization (WHO) 27 .Thus, the following databases were used in this research: Latin American and Caribbean Health Sciences Literature (Lilacs), Scientific Electronic Library Online (SciELO), Coleciona SUS, Medical Literature Analysis and Retrieval System Online (Med-Line), Nursing Database (BDEnf), Brazilian Dentistry Library (BBO), CidSaúde (Healthy Cities), and the São Paulo State Health Secretariat (SES-SP/scientific-technical publications and Reference and Training Center on STD/AIDS).
The exclusion criteria were: medical guides, reviews, comments, technical and scientific reports, dissertations, articles published abroad in Portuguese, as well as ministerial and government information and other documents that, although addressing the theme, did not comply with the standard IMRDC (introduction, method, result, discussion, and conclusion) applied to scientific papers.At a second moment, articles related to violence against the elderly, children, and adolescents and obstetric violence at the institutional level, as well as repeated articles or those where the core theme was not VAW, were excluded.Articles associated with offenders were also excluded, due to the fact this study is primarily focused on victims of violence.
Thus, the process of searching for manuscripts in the VHL obtained 1,058 texts for the descriptor "violence against women," 8,283 texts related to "domestic violence," and 3,614 texts related to "gender-based violence."After establishing the inclusion criteria and doing a preliminary reading of titles and abstracts, 949 manuscripts were selected, i.e. 259, 355, and 335, texts related to the three descriptors mentioned above, respectively.
Then, the exclusion criteria listed above were applied.There was also a careful analysis and full-text reading of the articles.Therefore, 148 articles were chosen to provide this study with its analytical basis, according to the schematic flowchart in Figure 1.
Aiming to have access to the full text of each article, the link directly available in the VHL database was selected.However, not always the contents informed in the abstract matched the description contained in the article.Hence, for the preparation of this study, we chose to read all publications under analysis in full text.
For mapping the set of scientific papers, these variables were identified: • Journal's subject field or professional sector: according to the information provided in the beginning of the article, referring to authors or journal title; • Methodological approach or study approach: the studies were regarded as quantitative when they involved statistical inferences with mathematical description; qualitative, when addressing human relationships and activities represented by collective or individual reports; and quali-quantitative, when both approaches were used in a complementary way 29 ; • Violence typology: classified as physical, sexual, psychological, moral, and patrimonialaccording to the classification adopted by Law 11,340/2006 10 -and there may be overlapping of types, according to the description of articles under analysis; • Research subjects: defined as a study object, namely the victims of violence or professionals who provide women with care, both in the health and law and insurance industries; • Publication year of the manuscript: set within the period from 2009 to 2013; • Type of the study conducted: classified as cross-sectional, ethnographic, ecological, cohort, among others, according to the methodological description of articles; • Number of authors: described according to the number of authors indicated in the manuscript; • Consequences of violence: this variable was subdivided into harm to physical health, mental health, sexual/reproductive health, or the health of other individuals, either isolated or in combination.
Data were organized into a documentary form, based on the variables described above, and they were tabulated and submitted to descriptive statistical analysis using the statistical package employed in social sciences 30 , the so-called SPSS version 20.0 for Windows.
The search for articles, as well as the analysis of results and considerations, took place in April and May 2014.

results
The main findings related to the studies under analysis are displayed in Tables 1 to 3 and in Graph 1.
The results show that most of the 148 articles analyzed were published in 2011, totaling 36 articles (24.3%).However, the number of publications for each year remained constant, with an average of 30 manuscripts.As for the number of authors, it was noticed that most studies, 81 (54.7%), had up to 3 authors; only 1 article (0.7%) was published by more than 8 authors.
As for the study approach, 81 articles (54.7%) resorted to a qualitative approach, 63 (42.6%) employed a quantitative approach, and 4 (2.7%) used both approaches.This data is displayed in Table 1.
To address the issue, the authors used different research or data analysis methods, which are provided in Table 2 and described in the very texts.So, 25 manuscripts (16.9%) used the Content Analysis method, 18 (12.2%)were cross-sectional studies, and 14 (9.5%) were descriptive and exploratory studies, among other types.In 11 manuscripts it was not possible to determine what method was used, because the authors failed to emphasize this information in their study.
The knowledge fields more frequently related to the number of publications were: Nursing, Medicine, and Psychology, with 62 (41.9%), 22 (14.9%), and 19 (12.8%) articles, respectively.However, many publications were written by authors from 2 or even 3 areas together, totaling 17 papers (11.4%).Other areas, such as Social Work, Occupational Therapy, Statistics, and Social Sciences -whose authors published alone or in a joint effort with Nursing and Medicine practitioners -were grouped into the item 'others,' in order to make understanding easier, because each of them had only one publication.This is shown in Graph 1.In only 19 studies (12.8%) it was not possible to to identify the authors' areas of expertise.
The main forms of violence identified and shown in Table 3 were: physical, sexual, and psychological, in 39 articles (26.3%), followed by physical and psychological, in 22 papers (14.9%), and only physical in 19 manuscripts (12.8%).
In most articles, physical violence was observed as the main expression of GV, either isolated or associated with other forms.However, in 12 studies, the authors addressed VAW without defining its typology.
As a result of the harms, physical and mental health assault was the most frequently referred to, 29 papers (19.6%).Many articles, i.e.50 (33.8%),did not identify any consequence suffered by the victims exposed to violence.

Discussion
The issue of VAW is seen as a social problem whose relevance has gained visibility in the society as a whole 3 .There is a growing trend towards recognizing its importance as a public health problem that requires interventions 31 , although there still lack sufficient elements to deal with the entire complexity of this phenomenon 3 .This theme is closely related to feminist movements and it has been a study object since its inception, in the 1970s 32 .
In this study, most articles were published by more than 3 authors, a finding different from that obtained by another paper 33 , where there was a prevalence of up to 3 authors.The most notable knowledge fields regarding the number of publications were Nursing, Medicine, and Psychology.This is corroborated by another study 32 that, by analyzing the journals' knowledge areas that publish articles related to GV, found that the medical sciences stand out, because the health team works since the very first moment to provide the victim with care.In this context, Nursing is highlighted, which plays a significant role both by dealing with physical injuries and reporting cases of violence observed in the clinical practice.Due to its broad range of services provided to society and the need for it in health care facilities, the investments in research and publication in this field have increased, with a view to discussing and proposing measures to cope with VAW 33 .
In this study, an article was regarded as multiprofessional when written by authors from more than 2 different industries, whether within the health field or not.Such information is noteworthy, because GV is analyzed and addressed by more than one knowledge field, it is key the various sectors that participate in the coping networks communicate in order to contribute to an increased problem-solving ability.
The accounts of the victims identified in the articles under analysis were obtained directly, through interviews, or indirectly, through compulsory notification forms, medical records data, expert reports, and case bulletins.Despite scientific studies carry out research both with the victims of violence and the professionals who deal with this clientele, there is a tendency that scientific papers better assess the victims' views.This trend may be justified by the fact that subjects who experience this phenomenon can explain a reality of their only more clearly -from a perspective that enables discoverying and grasping life experiences 32 .Thus, it is expected that more key findings are brought to reflection by those who devise policies and services.
Despite the study approaches are quantitative, qualitative, or a combination of them, the papers analyzed in this study show a prevalence of the qualitative design.This was also observed in a similar study, which found that about 49% of papers adopted this approach 33 .A recent trend to address VAW through qualitative methodologies and a focus on victims' experiences has been observed, as mentioned above.Nevertheless, there are many manuscripts that use a quantitative approach because it is related to the need for information on the prevalence of GV and its risk factors, a finding still not obtained in various regions of the world to support public policies 32 .
Besides, the adoption of various research methods reflects the multifaceted and complex nature of violence as a phenomenon, by bringing along with it the need for different approaches to achieve some understanding, which range from reflections through essentially theoretical contents to a dialogue concerning this reference, having empirical data collected from the reality of social subjects 3 .Taking this into account, the use of both qualitative and quantitative approaches may enable a better and expanded view of the complexity involved in a reality observed, because it brings advantages by evaluating the theme proposed and sheds some light on key aspects aimed at changes in the victims' context 32 .
The results also show a small portion of authors who have not disclosed the kind of study involved in their articles.This data corroborates that found by another study, where most of the texts under analysis did not inform its typology 33 .Failure to classify the study type makes it more difficult to identify the actual trend of studies addressing women's health and violence.
Regarding the typology of violence, the association between physical, sexual, and psychological forms as the most prevalent mode is pointed out, unlike other surveys, where sexual violence had a higher representativeness 33 , physical violence was more significant 34 , psychological violence was more expressive 35 , or the physical form associated with the psychological form was the most frequent 31 .
Not every physical assault leads to an injury that requires health care.This was shown by a study conducted with a sample of 117 victims, where 24% of the cases needed medical proce-dures, but only 7% led to hospitalization 36 .However, even when the violent act does not result in a physical injury, it causes psychological damage 37 hard to identify, because the parameters for definition are the limits and the rules of coexistence, i.e. very subjective 38 .
Episodes of violence may be severe, recurrent, and overlapping.Furthermore, different types of abuse can coexist in the same relationship 39 .And there is also the fact that violent relationships tend to occur again and again, following a progressively more severe scale.These relationships can also emerge in a transgenerational way 40 , over the years of a violent relationship, whose onset involves verbal abuse, then it becomes physical and/or sexual until reaching the cases of death threats and murder 41 .
Regarding the consequences of violence, which range from a slight injury to death, the most common injuries affected the victims' physical and mental health.These results are confirmed by a study 40 , but they differ from the findings of another 33 , where there was a greater impact on mental health.These consequences are not limited to immediate physical damage, but also to long-term effects, such as depression, suicide attempts, unintended pregnancy, etc. 42 .These effects may evolve to a chronic process and require adequate treatment and support, not only by health professionals, but by the family, and friends, i.e. a support network 33 .Harm to the health of other individuals, identified in the results, refer to abortion situations or low birth weight, reported in articles referring to women who have suffered violence within the prenatal period.These studies reveal that the problem has affected not only the woman, but also the fetus in her womb.

interfaces identified to design further research
When determining an overview of the texts analyzed by this study, some gaps to be bridged are identified, due to poor approaches provided so far, namely: GV within a context of homosexual relationships; the typology of VAW conveyed by the media; violence suffered by handicapped and Indian women; the participation of other offenders -besides husband/companion; and GV from the offenders' perspective.
Moreover, despite GV is less frequent in other environments along with the victim's home, there is a need for further studies that report GV in other spaces, such as workplaces, schools, churches, etc., associating it with other variables to be determined.
Perhaps, most Brazilian studies on violence and health choose to address the clientele of health services due to methodological ease.This may lead to an overestimation of the incidence of violence over the population because, although many victims do not seek health services -because of fear or ignorance, as already reported -there is a significant number of women who seek health care facilities due to their lesions 32 .This finding corroborates the results of this research, since most of the studies under analysis were conducted in public hospitals, where the clientele provided with care usually consists of people with a low socioeconomic status, generating some bias related to low educational level and high rates of violence and poverty.We suggest further research addressing other populations, in order to establish comparisons between various socioeconomic classes, so that it is possible to increase understanding on the phenomenon of violence.

Conclusion
Considering the large number of articles evaluated, it was possible to provide a mapping of the research addressing VAW as a theme.This mapping may facilitate the access of professionals to the knowledge produced within the study period.However, this systematic review did not cover the whole Brazilian literature within the health field, because many Brazilian researchers publish in international journals and many journals are not registered in the VHL, yet.Nevertheless, it is suggested to carry out studies using other variables and different periods of analysis, because the study profile may differ depending on its time frame.
Finally, it is expected that the findings of this study contribute to raise awareness among managers, professors, practitioners, and Higher Education students in the health field -as well as other knowledge areas concerned with the theme -about the need to address the issue with political commitment, in order to promote women's health, provide it with greater visibility, and devise public policies to change the reality highlighted herein.

Collaborations
LEL Silva contributed to the project conception and design, manuscript writing, data analysis and interpretation, as well as critical review of the relevant intellectual content.MLC Oliveira contributed to critical review of the relevant intellectual content and approval of the final version to be published.

Table 1 .
Variables related to the year of publication, number of authors, and methodological aspects of papers published within the period from2009 to  2013.Brasília, 2014.variable

Table 2 .
Variables related to the study type, method, or data analysis identified in papers published within the period from 2009 to 2013.Brasília, 2014.

Table 3 .
Variables related to the violence type and its consequences to women's health.Brasília, 2014.Journal's knowledge field or authors' area of expertise of papers published within the period from 2009 to 2013.Brasília, 2014.