A/HRC/20/16 with a disability, or when the aggressor has abused any authoritarian power held in the family, work or educational environment. V. Conclusions and recommendations 103. While States have initiated various preventive programmes, there are numerous gaps in their efforts. Challenges include: a lack of overall societal transformation; inadequate provision of access to justice; the absence and/or insufficiency of the rights-based discourse when addressing the killings of women; and the blindness to structural inequalities and the complex intersecting relations of power in the public and private spheres, which remain the root causes of sex and gender discrimination. A holistic approach in preventing gender-related killings must be emphasized in all the measures taken by States to investigate and sanction violence, especially in crafting, implementing and evaluating legislation, policies and national plans of action. 104. Feminists have also identified what they believe are other challenges: the difficulty of translating social realities into claims based on rights; the narrow interpretation of rights within an international legal order; and the prevalence of discriminatory cultural stereotypes in the administration of justice. The formulation of rights-based claims by women remains an important strategic and political tool for women’s empowerment and for addressing human rights violations. 105. The weaknesses in information systems and the poor quality of data are major barriers in investigating femicides, developing meaningful prevention strategies and advocating for improved policies. The different frameworks, definitions and classifications used in the conceptualization of femicide often complicate the collection of data from different sources and could lead to documentation that may not be comparable across communities or regions.181 Studies on the subject have been conducted primarily using data from homicide databases. Information collected by official or State sources is frequently not harmonized or coordinated. Often there are incongruities between the data collated by the different facilities, including inconsistencies in the categories used to document the circumstances surrounding the crime, the victim-perpetrator relationship and the fact of pre-existing violence. The development of an effective evidence base requires improving the quality and comparability of data.182 106. The use of inexact categories for the classification of murders, such as the category “others”, results in misidentification, concealment and underreporting of femicides—in particular those that do not occur in a family situation. Another common practice is the use of stereotypical and potentially prejudicial categories, including “crime of passion” or “mistress”.183 107. In February 2009 the Statistical Commission, in response to a request from the General Assembly,184 adopted a set of indicators that can document the prevalence of violence against women. The indicators proposed include: degree, frequency, relationship with the perpetrator, victim’s age, and records of murder cases. 181 182 183 184 26 Supra note 19. Ibid., p. 4. Carcedo, A. ―Femicide in Central America 2000–2006‖. Presentation of a study co-authored by Lemus, G., Kennedy, M., Herrera, M., D‘Angelo, A., Hidalgo, Ana., Ungo, U., and S. Pola – in Conference Report supra note 20. See General Assembly resolution 61/143.