Gender-Based , Public Health Systems Approaches to Improving Women and Girls ’ Health : Results from the USA Office on Women ’ s Health , Coalition for a Healthier Community Initiative

Gender is an important social determinant of women and girls’ health; and can affect how women access and obtain quality health care. Yet, health issues specific to women persist around the globe. There has been limited research and programming to show how socio-cultural roles and expectations for men and women affect symptoms, access to care, and health outcomes. The USA Office on Women’s Health funded 10 communities around the United States through the Coalition for a Healthier Community initiative to improve public health systems through comprehensive, gender-based programming. A national evaluation of the initiative included interviews and surveys with program staff and public health systems partners in the 10 communities to gather information about policies implemented to reduce barriers to public health care and improve health outcomes for women and girls. Evaluation data show that these efforts were gender-sensitive, responsive, or transformative as self-rated by the coalitions. Coalitions generated 107 policies to improve health, including 51 gender-aware policies. This paper summarizes the highlights of the gender-based approaches to adapt and implement evidence-based programs, influence public health systems, and formulate policies that reduce gender inequities and improve the health of women and girls.


Introduction
For several decades, the World Health Organization (WHO) and Interagency Gender Working Group (IGWG) have urged researchers, policy makers, and program staff to consider gender (vs. "sex") as an important social determinant of women's and girls' health. Sex refers to the biologically-determined characteristics that define males and females (e. g., based on anatomy and physiology). Gender is the culturally and socially constructed differences in roles, responsibilities, attributes, activities and behaviors a given society assigns to males and females, and considered appropriate by that society for boys and girls, men and women. The social definitions of what it means to be a woman or a man vary among cultures and change over time. While biological differences between the sexes can contribute to the onset and progression of disease states, gender can affect how both women and men access and obtain quality health care. Although gender health disparities constitute important aspects of health around the globe, the United States has limited research and programming regarding gender and how socio-cultural roles affect symptoms, access to care, treatment, and health outcomes. Albeit the scientific community in the USA has made progress in elucidating the biological conditions that impact health among women and girls, the social conditions which contribute to disease are often overlooked. However, important advances have been made in the United States. The state of Minnesota's Department of Health acknowledged the World Health Organization's call to address gender disparities in power when implementing its health reform policies and programs. This led to a better understanding of how social constructs and disparities in power and opportunities between women and men influenced health risks, health-seeking behaviors, and health outcomes [1]. The National Kidney Foundation of Michigan used gender-based analysis and gender-responsive approaches to programming and policy development to address burdensome caregiving roles such as those mentioned earlier that interfered with women's self-management of chronic conditions such as diabetes [2].
However, more advancement is needed in the scientific research community examining health and well-being inequalities through a gender lens. According to the National Vital Statistics Report [3], in 2010, over one million women died in the USA and over half of those deaths were due to chronic health conditions. Unfortunately, many of the chronic health conditions which contribute to the morbidity and mortality of women are preventable [3]. There are both biological (sex-based) and non-biological (gender-based) factors that are root causes for these results. These root causes include individual-level factors such as lack of education on how to prevent the root cause of disease; and gender-based factors such as "gender-bias" and "gender inequities." 1 Gender bias and inequities can have major influences on leading health indicators related to the death and disability status of women. For example, women have higher health care costs than men [5], due in part to sex differences (reproductive health issues and middle life health care) and women's longer life expectancy [6], but also due to gender-informed factors such as inequity in salaries [7] that might limit women's access to insurance or affordable high-quality care. Also, the power dynamics between men and women due to socio-cultural norms which perpetuate men's dominance over women, put women at risk for dependence and harm from interpersonal violence. Research has shown that some intimate relationships suffer from more systematic male violence which is deeply rooted in the patriarchal traditions of men controlling "their" women [8] and that society imparts men with a sense of entitlement to control women within their intimate relationships [9]. Other research using a power inequality perspective indicates that an imbalance of economic resources within couples underlies the use of violence [10]. Better data are needed on gender-based disabilities resulting from factors such as reproductive health, depression, domestic violence, and violence against women. The Interagency Gender G\Working Group [11] has provided evidence that gender perspectives can improve reproductive health outcomes. The Institute of Medicine (IOM) [12] in the USA established a committee on understanding the biology of sex and gender differences to illuminate research and data gaps in this area. Although the committee's report concluded that sex does matter, it was not clear as the extent to which these differences were biologically-based or related to socially constructed factors [12]. Another IOM report, "Women's Health Research: Progress, Pitfalls, and Promise" [13], drew a similar conclusion. Both reports recommended research to increase the numbers of women in studies as well as single-sex research to better understand the sex (biologically-based) differences, and acknowledged the importance of understanding gender (socially-based) differences in health outcomes for women and men. For example, the IOM women's health report states that, "even when great progress has been made through scientific advances, other factors in or determinants of health can present barriers to improving women's health, such as the effects of societal beliefs or morals on the use of the vaccine for cervical cancer, social acceptance of and stigmas attached to depression, and use of and compliance with use of contraceptives to prevent unintended pregnancy" [13]. In addition, more information is needed about the integration of gender equity and gender-based approaches into programs and policies to effectively address social determinants that result in poor health outcomes for women and girls [13].
These gaps are not limited to the USA. The global strategy for women's and children's health [14] identified several areas for urgent action that were needed to confront these challenges internationally. 1. Gender bias refers to the unequal way that women are treated in relation to men (usually to women's detriment) based on customs, norms, or traditions. Gender equity is defined as equivalence in life outcomes for women and men, recognizing their different needs and interests, and requiring a redistribution of power and resources [4]. Therefore, in most cultures, gender inequity refers to the unequal treatment of men and women in favor of men that result in women's poorer status in the workplace, healthcare settings, their homes, and society at large.
These actions were aimed at increasing investments to ensure that women and children could access prevention, treatment and care, as well as improved public health systems for efficient and cost-effective health services.
Recommended actions included: "1) country-led health plans, supported by increased, predictable and sustainable investments; 2) integrated delivery of health services and life-saving interventions, so women and their children can access prevention, treatment and care when and where they need it; 3) stronger health systems, with sufficient skilled health workers at their core; 4) innovative approaches to financing, product development and the efficient delivery of health services; 5) and improved monitoring and evaluation to ensure the accountability of all actors for results" 2 [13]. In accordance, governmental and non-governmental agencies have pushed for structural changes (not just programs aimed at individual-level changes). These include gender transformative policies and practices "to enhance gender equality by changing the way communities view, value and assign roles to women and men" [15].
The USA Department of Health and Human Services Office on Women's Health (OWH) launched the Coalition for a Healthier Community (CHC) Initiative to assist in the development of a sustainable public health system that could provide comprehensive health prevention programs to women and girls in high-risk municipalities within the USA. The purpose of the CHC initiative was to incorporate gender into women's health programs that would enable communities to implement evidence-based or evidence-influenced prevention interventions, which address gender relations that affect the health of women and girls within their respective communities. The funding opportunity announcement sought applications from any public or private non-profit entity to assist in the identification of major gender-based health issues affecting women and girls in their communities. Thus, the CHC initiative was developed to identify gender-based solutions to address major health issues affecting women and girls in their local communities by: • Creating a public health system that provided gender-based, comprehensive and seamless health prevention programs • Improving access and determining cost-benefit of gender-appropriate programs for women and girls • Increasing consumer and provider awareness of the community's gender-based health issues and the coordination of resources • Improving overall community health policies and gender-based health care programs Preventive programs such as those supported by the CHC allow health care organizations to address the needs of their specific communities, design programs to decrease the prevalence of disease, and minimize the impact of adverse health effects that affect women and girls within the USA. Having access to preventive programs that focus specifically on women and girls has the potential to ameliorate the high cost of health care. Understanding the socio-cultural mechanisms surrounding gender bias is crucial in lessening the health disadvantages that girls and women face.
The CHC initiative was implemented in two phases. In Phase I, each coalition was required to conduct a needs assessment and gender analysis to identify gender-based health disparities having the most adverse effect on the health and well-being of women and girls within their respective communities. During this phase, each grantee also developed five-year strategic plans linked to the USA Healthy People 2020 objectives to address the particular health issue identified in their community. Most of the health issues identified by the coalitions related to mental health, domestic violence, physical activity, and obesity. In phase II, OWH funded 10 demonstration projects throughout the USA that would incorporate gender-aware, evidence-based strategies within their program framework. The CHC initiative required community public health systems to implement a strategic action plan to address gender-based health issues and improve performance on a minimum of four Healthy People 2020 objectives that target women and girls. Grantees were required to identify specific short-term outcomes that would support the achievement of Healthy People 2020 medical health and public health infrastructure objectives as well as to discuss how the outcomes would be achieved and measured. Each grantee was required to contract with at least two community-based organizations, with designated funds, to be a viable and contributing member to the public health system.
The CHC also funded local evaluations that examined differences between outcomes for participants in evidence-based interventions that were gender-based and comparison groups in non-gender-based programming. As part of the national evaluation, data were collected to examine the extent to which these gender-based public health systems approaches resulted in policies that were gender-aware and likely to reduce barriers to improved health and health care systems for women and girls. Grantees described their gender-based approaches and the nature of the policy changes that resulted. This paper is focused on answering the following study questions with respect to those grantee evaluation data: 1. What was the nature of gender-based approaches used by the CHC grantees?
2. Where did the CHC grantees fall along the gender equity continuum?
3. To what extent did coalitions implement policy changes to address barriers to public health care for women and girls identified in their gender-based analyses?

Materials and Methods
The evaluation study on which this paper is based was reviewed and approved by the USA Office of Management and Budget (OMB), control number 0990-0443; and the Institutional Review Board (IRB) of The MayaTech Corporation (Federal Wide Assurance No. FWA00012366). This national, cross-site evaluation of the CHC initiative used the USA Centers for Disease Control and Prevention's framework for evaluation of public health programs [16] to implement a one-group, longitudinal design. A participatory Community Systems Analysis (CSA) mixed-methods approach was used to assess effects. CSA is a longitudinal approach developed by the evaluator to assess integrated effects of multi-component and/or multi-site initiatives at the system level [17]. For example, system-level changes might include changes in the community's norms or epidemiological data, policies, environments or other structural changes. This approach was guided by the national evaluation logic model and allowed us to analyze multiple sources of data from communities. The CSA includes descriptions of the grantees' gender-based approaches as well as the contextual factors (facilitators and barriers) that influence success. In addition, we can capture the possible effects at both the individual and system levels that selected approaches were able to realize over time. CSA is essentially a culturally competent approach to conducting embedded case study research. Specifically, we can generate a case study for the initiative overall as well as individual case studies for each community. Each grantee's target community was used as its own "comparison" (i.e., each community was compared to itself in terms of where the community was with respect to the state of gender-based programming and gender-aware policies prior to the CHC initiative and currently/post-CHC grant). In a previous contract with the Centers for Disease Control and Prevention (CDC), the CSA developers conducted focus groups and interviews with over 300 stakeholders across multiple communities in the United States to provide qualitative data for evaluation of the CDC's Minority AIDS Initiative. CDC analyzed the data to provide insights into stakeholders' perspectives on barriers to and facilitators of systems-level changes to address health inequities and social determinants of health influencing the impact of HIV/AIDS in communities of color [18,19].
The CSA evaluation approach allowed for the examination of the context in which interventions were implemented through interviews with key stakeholders; review of existing documents and reports about the initiative; and secondary analyses of reports on local evaluations. Data were extracted from site visit reports and document reviews of grantees' Quarterly Progress Reports (QPRs) and End-of-Year (EOY) reports. From these sources, data and information were collected on the following: • The public health problems on which coalition efforts were focused • Descriptions of gender-based approaches that were implemented • The policies that were generated by the coalitions This paper is based on compiled reports from grantees of their activities and accomplishments. As such, the data for this paper were extracted from site visit reports, grantees' Quarterly Progress Reports (QPR), End-of-Year Reports (EOY) for the first four years of their five-year grants, and the final reports. The QPRs included places for grantees to report on progress on performance measures as well as on implementing the gender-based approaches, evidence-based interventions and activities related to the policy changes. EOYs were used to gather data annually on accomplishments in the past year and the final report was based on all five years of the grant. The EOYs and final reports also included a place for grantees to highlight their policy achievements, the level at which policies were aimed and the status of each policy. Three sets of data were extracted from these reports: gender-based approaches used by the coalitions; a rating along the "gender equity continuum"; and an inventory of policy changes that resulted from each coalition's efforts. The operational definitions for each of these units of analysis appear below.

Gender-based approaches
All grantees were instructed to use OWH's definition of a "gender-based approach" as the basis for their responses. A CHC "gender-based program" was defined as: "A program which is designed to address the roles, behaviors, activities, and biological/psychological attributes that society assigns as 'appropriate' for men and/or women which impact treatment, access, and overall health" [20].

Gender equity continuum
Grantees were also asked to rate their coalition along the "gender equity continuum" as described by one of five scenarios. Each of the scenarios depicted one of the places along the continuum that corresponded, in principle, to harmful/exploitative, neutral/blind, sensitive, responsive, or transformative, with the latter being at the optimal end of the continuum and the first-listed (harmful) anchoring the suboptimal end of the continuum. The continuum was adapted from synthesizing the WHO's action framework [21,22] and the IGWG's gender equality continuum [23,24]. The actual scenarios were adapted from the CARE gender toolkit [25] which used five stages of the gender equity continuum: stage 1. Harmful, stage 2. Neutral, stage 3. Sensitive, stage 4. Responsive, and stage 5. Transformative. CARE describes each stage as follows: 1. "Harmful: Program approaches reinforce inequitable gender stereotypes or disempower certain people in the process of achieving program goals.

Neutral:
Program approaches or activities do not actively address gender stereotypes and discrimination. Gender-neutral programming is a step ahead of harmful on the continuum, because such approaches at least do no harm. However, they often are less than effective, because they fail to respond to gender-specific needs.

Sensitive:
Program approaches or activities recognize and respond to the different needs and constraints of individuals based on their gender and sexuality. These activities significantly improve women's or girls' (or men's or boys') access to protection, treatment or care. But by themselves they do little to change the larger contextual issues that lie at the root of gender inequities; they are not sufficient to fundamentally alter the balance of power in gender relations.

Responsive:
Program approaches or activities help men and women examine societal gender expectations, stereotypes and discrimination, and their impact on male and female sexual health and relationships.

Transformative:
Program approaches or activities actively seek to build equitable social norms and structures in addition to individual gender-equitable behavior" [25].

Policy changes
Grantees were provided training on the definition of "policy, system, and environmental changes" at annual grantee meetings. We used the CDC's CHANGE Action Guide [26] definitions of these concepts that are as follows. Policy changes are "laws, regulations, rules, protocols and procedures designed to guide or influence behavior (e.g., legislative or organizational)." Systems changes refer to "changes that impact all elements (including social norms) of an organization, institution, or system" (and may include policy or environmental change). Environmental change refers to "physical, social or economic factors designed to influence people's practices and behaviors." 3 For the purposes of this paper, all of these changes were referred to as policy changes, and grantees were further trained on additional issues that CDC [26] has indicated refer to policy change. For example, 1) "policies generally operate at the systems-level and can influence complex systems in ways that can improve the health and safety of a population; 2) a policy approach can be a cost-effective way to create positive changes in the health of large portions of the population; and 3) there are several types of policy, each of which can operate at different levels (national, state, local, or organizational)" [26]. We only report on the number of overall policies and the policies that are gender-aware (informed by gender disparities/inequities) or gender-based (related to programming). Policies were self-reported by grantees as gender-aware, but subsequently re-rated by two independent raters (procedure described below) using the definition of gender aware policies as those that "take into account the different social roles of men and women that lead to women and men having different needs" [27,28]. Policies that merely took into account universal approaches (e.g., everyone will have access to healthy food choices in school cafeterias) were not rated as gender aware.

Procedures
Two members of the evaluation team reviewed the QPRs, EOYs, and final reports. In addition, two gender experts conducted content analyses of QPRs and provided guidance to assess grantees' integration of their gender analyses into their reporting. Final reports of the grantees were also each reviewed by two members to rate the policy changes as gender-aware. Two senior researchers culled the information on gender-based approaches from the grantees' EOYs and used this information to develop a short description of each project. The self-ratings were pulled from the final reports and frequency counts for each stage of the continuum were reported. OWH staff compiled the policy inventory from the grantees' QPRs and the national evaluation team augmented this inventory with additional policy changes reported in the final reports. Two members of the evaluation staff then reviewed each policy listed in the inventory, and confirmed the policy listings by comparing the inventory to the policies listed in the QPRs. These two coders also coded the levels at which the policies were aimed and whether the policies were gender-aware and/or sensitive. Gender-aware/sensitive policies "take into account the different social roles of men and women that lead to women and men having different needs" 4 [27,28]. Inter-rater reliability was established for the coding by comparing the ratings of the two coders for 64 of the policies (set received by the end of year 4). For 63 of the 64 policies, the coders agreed on the rating (.99 reliability rating). The other policy rating was achieved by consensus after discussion between the two coders.

Analysis plans
Descriptive statistics and content analyse were used to describe the gender-based approaches, tabulate the self-ratings along the gender equity continuum, and compile the policy inventory.

Gender-based approaches
All grantees implemented evidence-based approaches that were culturally adapted to their locations based on the target population and health conditions targeted. Gender-based approaches varied widely and ranged from simple consideration of gender to more specific inclusion of gender-based policies, interventions, and outcomes assessments. Some examples of the gender-informed approaches included (some grantees utilized more than one approach): 1. Trauma-informed care for African American women and Latinas in the Women, Infants, and Children (WIC) program for pregnant and postpartum women with children up to age two.
2. HIV prevention and substance abuse treatment coupled with advocacy training to navigate the judicial system for commercial sex workers.
3. Healthy eating and active living for the community at-large with a focus on adult African Americans and Latinos (adult men and women), coupled with health education classes with discussions about sex, sexual health and healthy relationships for high school youth/adolescents.

Girls
Circle: Small-group discussions and support for elementary and middle school girls at risk for, or survivors of, child sexual trauma.
5. Talk Story: Small-group discussions with women telling their stories of intimate partner or interpersonal violence and men working through issues of perpetration.
6. Social media and curricula to promote 5-2-1-0-delivering messages and curriculum to promote 5 fruits and vegetables per day, 2 hours of screen time or less, 1 hour of physical activity and 0 (zero) sweetened beverages, with emphasis on teaching women alternate ways to use leisure time, and providing safe spaces to walk and do other exercise (e.g., in workplaces).
7. Cognitive behavior therapy to reduce stress, promote mental health, improve healthy eating and active living, and eliminate or prevent smoking for women in public housing and 8. An expecting fathers program to provide skills, education, and support for men in preparation for fatherhood. This program is one of many in which men are increasingly involved and is related to the anti-violence movement.
The evaluation team's review of grantee documents supported grantees' claims that the approaches were gender-based [20], in all but one case. In one instance, the grantee's approach appeared more sex-based, in that the approach was to simply point out biological differences in men and women and not expound on differences due to socially constructed roles for men and women or what constituted male and female norms.

Grantees' self-assessment on gender equity continuum
Grantees rated their own coalitions on the gender equity continuum, and the evaluation team reviewed grantee materials to further document the ratings. Grantees rated their coalitions as responsive for the most part (six grantees), with two grantees each at sensitive (an earlier stage of progression) and transformative (a more developed stage of progression). None of the grantees perceived that their coalition is at either stage 1 (harmful/exploitative-which reinforces inequitable gender stereotypes or disempowers people in the process of achieving program goals) or stage 2 (neutral/blind-which does not actively address gender issues).
Based on subsequent document reviews by the evaluation team, the two grantees that perceived their coalition as at the sensitive stage were either serving commercial sex workers referred through the judicial system for substance abuse and HIV prevention, treatment or care; or serving the general public in rural counties in the south of a mid-western state to promote healthy hearts and lifestyles. These coalitions implemented approaches or activities that recognized and responded to different individuals' needs and reduced barriers to access to care based on gender and sexuality. However, they were not further along the continuum, because they did not yet address community-level norms (e.g., stigma) and other issues (e.g., health care provider awareness of gender-based issues for these women) that are the root causes of gender inequity.
In general, the six responsive coalitions were characterized as implementing approaches and policies that helped both women and men to examine societal gender expectations, stereotypes, and discrimination, and the impact of gender on women/girls' and men/boys' health. These grantees were focused on diverse issues and populations such as: 1) adolescents' healthy relationships and the public's healthy eating; 2) women in the USA's Women, Infants and Children's (WIC) healthy food subsidy program to reduce obesity; 3) healthy lifestyles for adults at risk for diabetes and kidney disease; 4) USA-born and immigrant women at risk for obesity; 5) adult overweight/obesity prevention and healthy lifestyles; and 6) girls at risk for childhood sexual trauma. These grantees implemented evidence-based approaches to address individual needs as well as introduced policies within organizations to reduce gender inequities. However, their efforts were not yet directed at community-level systems changes (e.g., transforming gender-based norms that perpetuated negative stereotypes).
The two grantees that perceived their coalitions at the transformative stage had programs that actively worked on policies to build equitable social norms and structures at the community level. These included one community in the northeast USA that used cognitive behavior therapy for women in an urban public housing development; and a Pacific Islander community off the mainland USA that used an intervention known as Talk Story for small group discussions about the issues related to and possible solutions to intimate partner violence. In the northeastern community, efforts included educational hubs in community supermarkets as well as a formal data sharing agreement between the public schools, public housing, the coalition, social service agencies, and the community-based participatory researchers/local evaluators. This agreement was the first of its kind in the community; and allowed for better tracking of progress on sex-disaggregated health outcomes for women in the coalition's program and the health and education of their children. The Pacific Islander coalition's efforts resulted in legislation from the Governor's office to allow emergency contraception after Intimate Partner Violence (IPV) involving sexual assault.

Policy changes
A total of 107 policy changes were reported from September 2011 through September 2016, including 51 gender-aware policies. 5 Policies were rated as gender-aware by two independent evaluators who confirmed that the policies did "take into account the different social roles of men and women that lead to women and men having different needs" [27,28].
Policies were formulated to address community-level barriers to improving women and girls' health. Grantees focused most of their efforts on organizational-level policies. In one community, the university-based grantee and its coalition focused efforts on obesity prevention and reduction, implementing an academic policy to increase the number of curricular hours for women's health for university medical students. In the western region of the country, another coalition was involved in the enactment of a chamber of commerce-implemented lactation policy, which was part of a pilot worksite wellness project. The grantee in the Pacific Islander community, a domestic violence action center, passed an organizational policy to integrate partner violence prevention education into all community programming. The CHC partnership in the northeastern community enacted an organizational-level policy in which the state's largest children's mental health provider received a Medicaid license to provide mental health services for mothers.
Multiple local-level policies were also implemented by grantees. One example is a city government's workplace breastfeeding friendly policy. Grantees reported on 11 transformative policies focused on state-level changes to reduce gender disparities. These included introduction of a State Bill (SB265), which was passed by lawmakers to establish a victim-centered sex trafficking ban. Another example was the enactment of a local and state policy that provided housing vouchers to mothers with mental illness in danger of losing custody of their children.

Discussion
The CHC data contribute to the growing body of findings that shows that integrating gender components into interventions and advocacy efforts can result in gender equity in programming and policy changes [15,20,29]. Although, the gender-based approaches in this study vary with respect to culture, religion, target populations and targeted health issues, the CHC coalitions appeared to be implementing consistent gender-based approaches. As a group, the study highlights how communities can use gender-based approaches to adapt and implement evidence-based programs; and formulate and advocate for policies that reduce gender inequities and improve women and girls' health. Practitioners and policy makers can benefit from lessons learned to implement gender-based programs and system-level strategies. The lessons learned to date should also inform "health in all policies" approaches for women and girls' health. For example, policies do not necessarily have to be aimed at changing public health systems as a whole or health care systems per se, but can be aimed at reducing organizational barriers as well as community barriers (such as negative norms and stereotypes or lack of access to healthy choices). There is now increasing emphasis in other arenas on gender mainstreaming [24] that will involve formal written policies in organizations that seek to institutionalize the gender equity gains they are making. Future evaluation activities are needed to explore the extent to which policies are implemented, enacted, enforced, or mainstreamed at the organizational and community levels. Therefore, the list of gender-based approaches and policies generated by this sample serve as examples of what is possible with collaborative, gender-based, public health system approaches to eliminate gender disparities in women and girls' health. Additional research is needed to examine the extent to which gender equitable programs/practices and policies result in sustainable and cost-effective health outcomes for women and girls.

Limitations and Bias
The data in this report reflect secondary analyses of information that were based primarily on grantee project directors' and local evaluators' input and reports. Although sites were across the USA, grantee communities were located in the mid-/north-Atlantic (Delaware, Maryland, Pennsylvania, Connecticut), Western/Pacific Island (Montana, Utah, Washington, Hawaii), and Midwest (Illinois, Michigan) regions of the USA. Thus, several geographic regions of the USA were not represented nor did the funded sites represent the full programmatic or policy contexts in communities/states where their programs and policies were implemented. Although some data relied on input from other key persons' discussions and observations at site visits, these reports were not subjected to verification procedures, triangulation, or other approaches to determine their reliability and validity. No comparison communities were included, only comparison groups of individuals for the tested interventions. Also, there were no retrospective comparison data at the community or organizational level to determine what policies existed prior to CHC funding, even though there were needs assessment and gender analysis data collected in Phase I by grantees. Finally, sample sizes were not large enough to draw conclusions for specific subgroups of women across grantee sites.