Evaluation and Program Planning

Women with disabilities are more vulnerable to violence, including intimate partner violence (IPV), yet the majority of emerging IPV prevention programs fail to explicitly consider the needs of participants with disabilities. Women and men living with disabilities engaged with IPV prevention programs in four countries were interviewed to explore how disability shaped their experiences of gender, violence, IPV, and whether the programs met their disability related needs. In-depth interviews were conducted with 16 women and 15 men living with disabilities in Ghana, Rwanda, Tajikistan and South Africa. The data were analysed thematically and compared across the settings. Participants described experiencing disability-related stigma, discrimination, exclusion, and for women, increased vulnerability to IPV. Barriers to full participation in programs included limited accessibility, and lack of disability-speci ﬁ c materials, recruitment or outreach. Enablers of inclusion included recruitment and monitoring strategies aimed at people with disabilities, partnering with a local disabled people ’ s organization, training sta ﬀ in disability inclusion, and raising awareness of disability rights. The data encouragingly suggests that inclusion of women and men with disabilities in IPV prevention programs designed for the general population has bene ﬁ cial outcomes. Inclusion can prevent violence, promote their wellbeing, support economic empowerment, and challenge disability-related stigma and discrimination.


Introduction
Approximately 15 % of the world's population live with some form of disability, and 80 % of people with disabilities live in low and middle income countries (LMICs) (WHO/World Bank, 2011). In this paper, disability is understood according to the International Classification of Functioning, Disability and Health (ICF) whereby disability is defined by physical, mental and/or emotional impairments that interact with social prejudice and environmental barriers to restrict peoples' equal participation in society (WHO, 2001). People with disabilities are often marginalised from society by disability-related stigma for representing differences that are socially devalued (Campbell & Deacon, 2006). Gender, age, race, sexual orientation, cultural norms and socio-economic status intersect and interact with disability, sometimes varying by severity of impairment, and can further marginalise people with disabilities (Haegele & Hodge, 2016;Oliver, 2013). To support people with disabilities to meaningfully participate in and contribute to society, social stigma and other barriers need to be addressed, and suitable accommodations, equivalent rights, and accessible opportunities provided (WHO, 2001). One of the most extreme and severe manifestations of discrimination against people with disabilities is their increased risk of experiencing violence relative to people without disabilities, necessitating their purposeful inclusion in violence prevention programs.
Intimate partner violence (IPV) is one of the most common forms of violence worldwide, and affects at least 1 in 3 women during their lifetime (Devries et al., 2013). Considerable evidence suggests that women with disabilities are significantly more likely to experience all forms of violence than both women without disabilities and men with disabilities, with IPV being the most common form of violence they experience (Ballan et al., 2015;Krnjacki, Emerson, Llewellyn, & Kavanagh, 2015;Scherer, Snyder, & Fisher, 2016;Walter-Brice, Cox, Priest, & Thompson, 2012). Women with disabilities in LMICs are particularly susceptible to IPV. They are less likely to have secondary education, and more likely to be unemployed and live under conditions of extreme pover Fig. 1: Indashyikirwa Activism Posterty with limited resources, further increasing their risk of violence (Mitra, Posarac, & Vick, 2011). In studies across South and East Asia, women with disabilities were found to be twice as likely to experience violence by a partner or household member than women without disabilities ((Astbury, 2012) Astbury, 2012;Hasan, Muhaddes, Camellia, Selim, & Rashid, 2014). More recent research in LMICs suggests that women with disabilities are similarly twice as likely to experience IPV than women without disabilities (Dunkle, van der Heijden, Chirwa, & Stern, 2018). Despite this clear need for IPV prevention and response strategies that can serve the needs of women with disabilities in LMICs, little is known about how to deliver such programs effectively.
IPV prevention programs often include a focus on engaging men as well as women to shift hegemonic masculine norms that condone the use of violence against women. However, such programs may fail to meet the needs of men with disabilities, whose experiences of masculinity are often complex and contested (Shuttleworth, Wedgewood, & Wilson, 2012). For men with disabilities, performance of hegemonic masculine norms of strength, entitlement and independence may be complicated because of their impairments, and/or due to social reactions to their disability. It is critical to understand how men with disabilities experience mainstream IPV prevention programs that intervene with men primarily as potential perpetrators of violence against women and seek to engage them in gender-transformative processes. Understanding the extent to which such programs resonate with and meet the needs of men with disabilities, has the potential to better inform gender-transformative efforts to address harmful norms of masculinity. Literature from LMICs about men with disabilities' experiences of violence is limited, however, there is some evidence that men with disabilities face increased risk for violence compared to men without disabilities (Mitra & Mouradian, 2014;Saxton, 2017). The relative dearth of data on men with disabilities and their experiences of violence in LMICs means we have limited knowledge of how and why men with disabilities may perpetrate or experience violence in community, family, or relationship contexts, and how they can participate in and benefit from IPV prevention programs. This paper draws on data from a multi-country qualitative study conducted with women and men with disabilities in the context of IPV prevention programs in four LMICs (Ghana, Rwanda, Tajikistan and South Africa), all of which were evaluated as part of the What Works to Prevent Violence against Women and Girls Global Programme. A diversity of programs and contexts were purposefully selected for this comparative case study. What Works was established to assess best practices for prevention and response to IPV, including active engagement of men as both potential perpetrators of IPV and as agents of positive change. The present project was designed to supplement learnings from the Global Program through exploring perspectives of male and female participants living with disabilities who were engaged with or targeted by the programs being evaluated. This qualitative study explored how gender and disability intersected to shape experiences of violence, including IPV, and appraised whether and how the programs included individuals with disabilities and met their disabilityrelated needs. The aim was to learn from women and men with disabilities what is required to develop more disability-appropriate and accessible programs to prevent and respond to IPV. Although the programs, contexts and overarching evaluations were different, the actively collaborative environment provided by the What Works consortium encouraged teams to share programmatic and research tools, exchange perspectives on challenges and best practices, and create unique opportunities for cross-contextual analysis. The analysis points to common experiences for people living with disabilities across diverse settings, and reveals some of the barriers and enablers to their participation with different IPV programs, as well as how they may benefit from inclusion. Such insights can inform recommendations for more meaningful and effective disability-inclusive IPV prevention and response programs.

Description of Programs and Context
All the programs included in this study aimed to prevent IPV among male-female couples, although the precise nature of the program varied by site. In Ghana, the Gender Studies and Human Rights Documentation Centre implemented the Community Based Action Teams (COMBAT) program over 18 months from July 2016 to Dec 2017, across 20 rural communities in two districts. Male and female participants were selected and trained on types, causes and impacts of violence against women, family laws, conflict resolution, advocacy and counselling. These teams were then supported to facilitate community sensitization strategies. In Rwanda, Indashyikirwa ("Agents for Change") was implemented by CARE International, Rwanda Women's Network (RWN) and Rwanda Men's Resource Centre (RWAMREC) from August 2014 through August 2018, in fourteen sectors in rural Rwanda. The program employed various strategies to reduce experiences and perpetration of IPV, including a couple's curriculum to support healthy and non-violent relationships, establishment of women's safe spaces for survivors of IPV, facilitated community activism with a sub-set of individuals who had completed the couples training, and training opinion leaders to support an enabling environment. In South Africa, Sonke Gender Justice implemented the CHANGE programme, a multi-level intervention including community mobilisation and advocacy, aimed at preventing men's use of violence against women and girls among men in urban South Africa. Sonke CHANGE was implemented in 18 peri-urban informal settlements in near Johannesburg from May 2016 to December 2017. In Tajikistan, International Alert worked with regional NGO partners across rural villages to build a multi-component project to address underlying factors that condone and contribute to IPV in E. Stern, et al. Evaluation and Program Planning 79 (2020) 101770 Tajikistan. The 'Living with Dignity' project was implemented from August 2016 to August 2017, and included a gender empowerment curriculum (adapted from Stepping Stones) with households and development of economic and business opportunities. Targeted households were supported through distribution of resources to start income generating activities, and provision of expertise in economic/business establishment, household management and income generating activities. Projects included were chosen from the larger What Works portfolio based on targeting adult participants in the general population, availability of staff for data collection, and interest from the program teams in participating in a supplemental enquiry into disability and IPV program participation. It is important to consider the context of disability rights and legislation across the four settings included for this study. All four countries are signatories of the United Nations Convention of the Rights of Persons with Disabilities (UN CRPD), which details the rights of people with disabilities and identifies areas where violations of these rights have occurred. All four countries also have disability laws and policies, intended to protect the rights of people living with disabilities and to eliminate social exclusion and discrimination. In Ghana, the Persons with Disability Act also spells out the rights of people living with disabilities including unrestricted access to public places, free health care, employment, education and transportation (Ocran, 2019). In Rwanda, the National Council of Persons with Disabilities (NCPD) is a forum for advocacy and social mobilization for issues affecting people with disabilities and strives to build the capacity of people with disabilities and ensure their participation in national development (Njelesani, Siegel, & Ullrich, 2018). In South Africa, the Department of Social Development's 2015 White Paper on the Rights of Persons with Disabilities is a call to action for government, civil society and the private sector to work together to ensure the socio-economic inclusion of persons with disabilities and an inclusive society where people with disabilities enjoy the same rights as their fellow citizens. In Tajikistan, Law 458 on Social Protection of the Disabled details state policy and outlines steps to implement the rights of persons with disabilities by striving to eliminate limitations and encourages them to participate in economic and social life (World Bank, 2002).

Data collection
In-depth qualitative interviews were conducted with 16 adult women and 15 adult men (31 individuals in total) with disabilities targeted by the What Works programs described above, as follows: Ghana (6 women, 5 men), Rwanda (4 women, 3 men), Tajikistan (8 women), and South Africa (7 men). Individuals with a disability who were at least 18 years of age were identified by program staff and invited to participate in the interviews. Program staff were asked to identify participants that were participating in the interventions and had reported mild to moderate to a lot of functioning difficulty on one or more items on the ICF measure of disability status in the baseline surveys conducted across the What Works evaluations, using the WHODAS 2.0 short item measurement of disability (WHO, 2011). Yet there were limitations to this recruitment procedure since not all participants had completed baseline surveys. Project teams could also identify a disability rights organization to help identify participants, and in some cases, participants living with disabilities known to program staff were also recruited to the study. Because most interviewees were already participating in program activities or otherwise known to program staff, it was assumed that their impairments would not hamper their cognitive or physical ability to participate in the research. Only women were interviewed in Tajikistan, as this program was tailored to the empowerment and inclusion of women. Only men were interviewed in South Africa, as the CHANGE program targeted harmful forms of masculinity and men's use of violence. In South Africa, there were challenges identifying and recruiting men who had actively participated in activities associated with the CHANGE program's model of community outreach. While all men interviewed resided in the targeted communities and were potentially exposed to program messaging, only one had actively participated in program activities. Participants were recruited and interviewed either once or twice until the second and coordinating author of the study was confident that data saturation had been reached, as similar themes were being elicited from the data in order to answer the exploratory areas of interest (Fusch & Ness, 2015).
Interviews used open-ended questions to explore participants' experiences of living with a disability, intimate partnerships, gender, disability related stigma and community safety. Open ended questions on lifetime experience and perpetration of different forms of violence, including IPV, incorporated adapted questions from the Adult Abuse Assessment tool, to probe for disability-specific violence (for example being denied care or help, interference with assistive devices, or control over money by caregivers) (McFarlane et al., 2001). Participants' perceptions of barriers and enablers to their participation in the programs, and the impact of their participation were also queried. Socio-demographic information was collected detailing age, type and nature of physical impairment, living arrangements, educational level, marital status, number of children and employment status.
Research coordinators from all countries attended a disability-sensitive research training funded by the What Works consortium, and experienced, local, gender-matched, qualitative interviewers external to the program were then employed to conduct the interviews, under the supervision of the trained research coordinators. The number of interviews conducted with each participant varied, with some participants having one interview, and others having two, in order to complete the scope of inquiry. Questions about violence were planned for discussion once rapport was established between interviewers and participants, either at the later part of the first interview or during follow up interviews. A total of 52 interviews were completed, whereby 21 individuals were interviewed twice, and 10 individuals were interviewed only once. Each interview lasted approximately 1.5 h. All interviews were conducted in participants' preferred language, audiorecorded with the consent of the participant, and translated and transcribed verbatim into English by in-country professional translators that had experience translating qualitative transcripts into English.
Initial analysis of the data from participants with disabilities suggested that the Indashyikirwa program in Rwanda was the most successful in actively engaging people with disabilities. To further understand this, we drew on additional qualitative data from the larger What Works qualitative process evaluation of Indashyikirwa, on which the first and third authors were principal investigators (anonymous citation). These data included in-depth interviews conducted at program endline with 12 community activists (individuals that completed the Indashyikirwa couples curriculum for 6 months and were subsequently supported and trained to facilitate activism in their communities) and 3 women's safe space facilitators (individuals that completed the Indashyikirwa safe space facilitator training for 2 weeks and were supported by program staff to run women's safe spaces), which asked them to reflect on their experiences with the Indashyikirwa programme (anonymous citation). Where these respondents describe program activities with people with disabilities, their interviews offer valuable perspectives on inclusion of people with disabilities from the standpoint of front-line practitioners.
We recognize that traditionally, qualitative research entails small samples, and is not multi-sited research. However, there is increasing recognition of the value of comparative insights generated through qualitative approaches across diverse contexts (Lambert & Wood, 2005). A comparative approach cautions against essentializing notions of context or place, and prioritizes examining processes of sensemaking or identity in distinct settings, with regards to a similar phenomenon (Bartlett & Vavrus, 2017). This approach is especially relevant to inform effective program planning and evaluation and has been applied to other data from the What Works portfolio (Ruark, Stern, Dlamini-Simelane, & Fidele Kakuze, 2017;Stern, McGhee, Ferguson, & Clark, 2018). To ensure rigour of the study, the COREQ qualitative research checklist guided the methodology and analysis (Tong, Sainsbury, & Craig, 2007).

Ethical considerations
This study received ethical approval from the South African Medical Research Council and from national or university institutional ethics committees in all affiliated study countries. Before each interview, information on the aims, risks, and benefits of the research were provided and accessible, informed oral or written consent was obtained from participants in adherence with the ethical approval guidelines given by the respective review boards. Interview settings were agreed where identified where disability needs of participants were accommodated, and where confidentiality and privacy could be ensured. All identifying information of participants have been removed for the presentation of the findings.

Data analysis
Thematic inductive analysis of the interview transcriptions was conducted using Atlas Ti (v10). Thematic analyses were conducted to uncover predominant themes to provide a rich, detailed, and holistic account of the data. The thematic coding framework included themes purposefully elicited from the interviews, including barriers and enablers to inclusion in IPV prevention programs, and also allowed for the identification of grounded issues from the data. The process of coding included reading the transcripts, identifying and labelling fragments of texts associated with the research objectives, searching for patterns and emerging themes among the codes, and defining and naming themes. Thematic codes with related information were then compiled into empirical findings, and connections, complexities, and contradictions within and between themes were integrated into the presentation of findings. The first and second authors conducted thematic coding separately using the same codebook, which enhanced triangulation of the analysis process. The data were split between the two authors, although a sub-set of the same data was coded by both authors to ensure strong consistency within the coding. The first author regularly workshopped the findings with the Indashyikirwa program staff to allow for their insights to the interpretation of the data and to validate programmatic insights. In Ghana, South Africa and Tajikistan, a similar process was done whereby principal investigators shared emerging findings from this study with the program staff.

Participants' characteristics
Participants ranged from 18 to 63 years old and the majority lived in contexts of extreme poverty and poor infrastructure. Participants reported low education levels, and typically informal or no employment. Most participants reported physical impairments that restricted individual mobility and dexterity and four participants reported visual impairments. Impairments were either congenital or acquired via illnesses such as polio, osteomyelitis, epilepsy, diabetes, aplasia, stroke, or traumatic brain injury, resulting in paralysis, amputation, scarring or sight impairment. Table 1 documents demographic details of participants.

Gendered disability stigma and discrimination
The majority of participants reported significant experiences of disability-related stigma and discrimination, including being blamed for or devalued because of their disability: Some insult me and talk to me as if I'm not a human being, I remember a guy telling me that my hand wouldn't have been amputated if I was a good person. (Ghana, male, physical impairment) This was extremely common across all countries, and included insults, mistreatment, and lack of support or acknowledgment in society: The way people generally treat people with disabilities, most of them don't give them any support, there are still some people who have the mentality that people with disabilities are of no importance. (Rwanda, male, physical impairment) Other forms of discrimination included not being paid for work or goods sold, which was attributed to one's disability: Because of my disability, people don't often pay me when they buy something from me because they know I can't take the money from them. (Ghana, male, physical impairment) Stigma around disabilities intersected with hegemonic gender expectations. Some men lamented how their disability affected their ability to obtain employment or to provide financially for their family, undermining their sense of manhood: Others are not disabled like I am. They have both hands, so they can work, and they won't also be discriminated against when they go searching for a job. (Ghana, male, physical impairment) In turn, limited employment opportunities in these settings was said to reinforce negative perceptions of men with disabilities: It is because we have no gainful employment to empower us like others or to make us recognized. We are marginalized in a way, hence it has paved the way for others to abuse us. We have no power to protect ourselves. (Ghana, male, physical impairment) Several participants noted how women with disabilities tend to experience more stigma and discrimination than men, especially regarding attracting a spouse: For men with disabilities, it is better compared to women with disabilities. You hear people saying that they can't marry a woman with disability. Sometimes people come and tell me: "we saw someone who is like you. For you, you were lucky to find a wife but for her, she won't get a husband!", so you can see that it is a problem… A girl with a disability suffers more. (Rwanda, male, physical impairment) This was said to be especially pronounced for women with physical disabilities, which was perceived to hinder their ability to perform domestic and care duties. One woman reported being married to a much older man because her family expected her to face difficulties finding a spouse: After graduating from school my dad thought no one would marry me because of my leg and he gave me to a man 20 years older than me. (Tajikistan, female, physical impairment) Stigma and discrimination were dependent on the type and severity of disability, and a few participants noted that those with cognitive or mental disabilities experience extreme marginalisation. Participants identified significant consequences of stigma and discrimination, including humiliation, anxiety, and sadness: What made me sad is the fact that he despised me because of my disability and then it made me very sad. That is when I realized that having disability is a very bad thing. (Rwanda, male, physical impairment)

Exclusion and isolation
Participants in all four settings reported being regularly isolated and excluded from social interactions and community affairs by intimate partners, family and community members: Some of the community members mock you… They often shout at us to keep quiet and sit down when we attempt to make a suggestion during community meetings. (Ghana, male, physical impairment) When going out, nobody notices you, responds to your question if you ask… Even when talking they behaved arrogantly, or turned their back to me. (Tajikistan, female, physical impairment and epilepsy) Exclusion was often linked to limited accessibility for people with disabilities: I suffer a lot when I go to the office, it looks like they didn't consider disabled people when putting up their structure. (Ghana, male, physical impairment) Several participants noted how they isolated themselves for fear of experiencing stigma or discrimination: I do not socialise often. I have a fear if I go to somewhere, they will not accept me, what would they say when I am there? They would say I am different and that I am disabled. (South Africa, male, physical impairment) My fear is that due to my impairment, my leg may twist and I may fall in the presence of visitors who do not know my situation, they might think I am drunk, that is why I feel ashamed to join a community gathering. Exclusion and isolation could have significant consequences, including anxiety, frustration, and lack of support. A few participants, predominantly men, expressed that exclusion and isolation from community meetings undermined their confidence and ability to advocate for their rights or benefits as people with disabilities: You may find that often people who have disabilities and who are very poor, are excluded because they were not able to attend meetings… they should attend so they can talk about their problems. (Rwanda, male, sight impairment) The participant is referring to the monthly community meetings held in Rwanda, where participants have a chance to express their concerns to local leaders and collectively identify solutions.

Intersections of disability, gender and violence
Several participants reported threats and experiences of community violence, related to their disability.
The other day I heard someone saying, 'sometimes I wish I broke that arm of yours! You show those ugly thin arms; I will break them one day!' I told myself, if someone dares to say this openly to me in addition to saying it in an indirect way, this means that I am a burden for them. (Rwanda, female, physical impairment) Several participants reported that experiences of violence intersected with community perceptions of people with disabilities as weak or unable to care or defend for themselves. A few men in South Africa emphasized how their disability compounded their already unsafe environments and perceptions of insecurity: Because they see you have a disability, that you cannot run away, they can mug you, especially these gangsters, and the murderers will come to you first. (South Africa, male, physical impairment) Several participants identified limited community awareness of disability-related stigma and discrimination as a form of violence: The community does not educate people on why you don't have to violate disabled people. Most people say things that they don't even know is violence or assault. (Ghana, male, physical impairment) Women living with disabilities were said to be particularly vulnerable to sexual or physical violence by a range of perpetrators, including partners, non-partners, and carers: Women with disabilities are mostly exposed to violence because people usually take advantage of their disability and have their way with them, and generally women are mostly the victims of violence. (Ghana, male, physical impairment) None of the men interviewed reported experiences of IPV (either victimization or perpetration), compared to the majority of women who reported experiencing varying types of IPV, including emotional IPV related to their disability. This often involved verbal insults and humiliation: He always blamed me that I was lame. He repeats it 10-15 times a day. His insults me which makes me sad. I asked him repeatedly not to insult me. He used to insult me even if I do nothing wrong. (Tajikistan, female, physical impairment) In Tajikistan and Ghana, a few female participants reported physical abuse from their partners, including being hit, slapped, and beaten. Emotional and physical forms of IPV generated shame, depression, and anger amongst female participants: He tells me he doesn't want to live with a sick person. He insults and tells me I am a dirty woman. That my disabled fingers look like an animal and that he is ashamed to walk with me. When he says that I also get angry. (Ghana, female, physical impairment) A female participant in Tajikistan noted that she would rather have her husband use physical violence rather than insult her about her disability, indicative of the negative consequences: I told my father…it would be better to marry me to an alcoholic or drug addict and let him beat me, but not remind me of my ailment all the time.
(Tajikistan, female, physical impairment) The experiences of emotional IPV from partners often led to despair and even suicidality for some female participants. However, women's isolation and dependency on partners for income and care needs, including care for their children, limited their options to end the abuse: E. Stern, et al. Evaluation and Program Planning 79 (2020) 101770 When hearing those bad words again, I was going to kill myself. But when thinking about my children, I promised myself never to do so. I endured for the sake of my children. (Tajikistan, female, physical impairment) In addition to experiencing different types of violence, participants with disabilities encountered numerous barriers to active engagement in the various IPV prevention programmes.

Barriers to inclusion in IPV programming
Barriers to participation in IPV prevention programs mirrored the barriers people with disabilities experienced in daily life. Overt stigma and discrimination, inaccessibility, and internalised stigma were said to limit their active involvement, and undermine the potential benefits of participating in the various programs: I participated at all sessions, but I did not participate in the role-plays because my lip stretched and I was ashamed...Iwould have participated if not ashamed. (Tajikistan, female, physical and sight impairment) We can't mingle with those who are not disabled in certain activities. We only sit back and watch. (Ghana, female, physical impairment) In Ghana, one man reported that discriminatory behaviors by community members towards people with disabilities dissuaded him from actively participating in community meetings and the COMBAT program: I got the understanding that if I dare to contribute to the discussion, I will also be shut down and forced to sit so I should just keep quiet, sit, observe, and listen to their considerations and go home peacefully when the meeting is closed. (Ghana, male, physical impairment) Accessibility barriers were also identified to hinder their participation in the programs: I have not been able to attend any of their meetings because of my wheelchair. (Ghana, male, physical impairment) One man in Rwanda emphasized that the program would have benefitted from training more service providers around disability to ensure accessibility and adequate care: Many people should have training about the rights of people with disabilities. Teachers, healthcare givers and staff from the sector offices should also receive training. (Rwanda, male, sight impairment) Another man in Rwanda emphasized that people with disabilities should be purposefully recruited to participate in IPV programs given the barriers to their inclusion: I was the only one who was lucky to win the random selection, though there are many who are vulnerable. I think it would have been better if there were other people who have a disability to be recruited to participate in the Indashyikirwa program, maybe 10 people. (Rwanda, male, physical impairment)

Enablers of inclusion in IPV programming: case study of Indashyikirwa
Despite the various barriers participants encountered, important lessons regarding enablers to inclusion of people with disabilities in IPV programming were identified, particularly from the Indashyikirwa program in Rwanda. Partway through program implementation, the Indashyikirwa team partnered with the National Council of Persons with Disabilities to train all program staff on disability inclusion. Rwanda Women's Network staff then delivered a version of this training to all Indashyikirwa community activists and women's safe space facilitators, and to 280 community members living with disabilities. One female community activist reflected on the value of this training: There is a certain language we used to use that we cannot use anymore.
For example, someone who can't hear and speak, we used to call him 'a deaf', someone we used to call albinos, we have learnt that is someone with a skin disability. Us as change agents, we have understood that that language should not be used anymore. (Rwanda, female, community activist) The Rwandan team also developed inclusive communication materials illustrating the intersections of gender, disability, and violence, and solicited input on these from the National Council of Persons with Disabilities. Materials such as Fig. 1 were used to guide facilitated community activism, raise awareness of disability rights, and to challenge disability related stigma and discrimination.
Indashyikirwa's proactive strategies around inclusion of people with disabilities were strongly encouraged and financially supported by the donor, DFID Rwanda. These included encouraging women's safe space facilitators to conduct home visits to people with disabilities in their communities and encouraging people with disabilities to engage with Indashyikirwa activism activities. A women's safe space facilitator reflected on the value of the home visits: We visit people with disabilities…we can tell a woman about power or domestic violence because people with disabilities don't attend the public discussions. There is one person I visited and she was like: 'I also have power just like someone who doesn't have any physical disability? I used to feel so small given the way I walk, I didn't wish to go where others are gathering. Thanks so much, I won't feel shy anymore to go where others are.' (Rwanda, female, safe space facilitator) One male activist with a disability noted the importance of specifically encouraging people with disabilities to attend the activism sessions: Even when we are conducting discussions we try to tell our fellow people who have a disability that they should not be swallowed by their sorrow, we tell them that when they have problems they should not think that it is over that; they should instead show up. (Rwanda, male, physical impairment) A few activists in Rwanda discussed how they monitored the number of people with disabilities attending the women's safe spaces and community activism to assess inclusion strategies.

Benefits of participation with IPV programming
The data suggests a range of benefits of including people with disabilities across the IPV prevention programs. One man emphasized that the program in South Africa offered the only local support for people with disabilities: 'There is a lot of us who are disabled in this area… I have never seen any other efforts besides this support group.' (South Africa, male, physical impairment) Some participants in Ghana noted how active inclusion in the program helped raise their self-esteem, confidence, and access to social support: I now feel we are also part of the human society, this change of feeling is as a result of the COMBAT intervention. I used to feel that people living with disability are regarded as nothing. COMBAT has made known our essence in society. (Ghana, male, physical impairment) We were all included, we were there during their meeting… I spoke a lot during that meeting and no one dared to shut me up. They brought me the microphone themselves hence I made my suggestion. I felt proud; I felt I was also accepted to talk in a gathering. (Ghana, male, physical impairment).
A woman in Tajikistan similarly noted how participating in the training led to changes in how her community behaved towards her: People started treating us differently. They talked to us and made orders for tailoring outfits. Before, a woman gossiped a lot about me, and one day she came to order an outfit from me… When finishing her order, she came to collect it and she apologized, and appreciated my work, saying E. Stern, et al. Evaluation and Program Planning 79 (2020) 101770 that despite her gossips I tailored her outfit very well. (Tajikistan, female, physical impairment) Several women in Tajikistan highlighted the benefits of the economic empowerment elements of the program. One woman in Tajikistan discussed the value of participating in a livelihood activity of renting out dishes, which was appropriate for her given her mobility impairments. In Ghana and Rwanda, some program elements targeted stigmatizing attitudes about disabilities and raised awareness of the rights and laws protecting people with disabilities: When we conduct discussions…we show them that people with disabilities are people like others, we tell them that they should use the language 'a person with disability' rather than calling him "Ikimuga 1 " otherwise it is excluding them. (Rwanda, male, physical impairment) They taught us that though we are disabled, no one has the right to commit violence against us. (Ghana, female, physical impairment) A participant in Rwanda expressed his surprise and appreciation of being elected as a community activist, as he had assumed he would have been excluded from this role given his disability. Another male activist asserted that by taking on this role, he could challenge discriminatory social norms of people with disabilities: When I conduct a discussion, they even learn from me and say, although he has a disability, he now does the work he did not used to do. He tries.
(Rwanda, male, physical impairment) Additionally, a few participants reported feeling better equipped to respond to abuse and violence because of their engagement in the programmes: [The facilitator] has made us, the disabled, to understand that we have people who speak on our behalf or who seek our welfare. Moreover, for now, I know I can report cases of abuse against me to the COMBAT in situations where I am not able to report to the police. (Ghana, male, physical impairment)

Discussion
This study drew on the perspectives of people living with disabilities, which can critically inform violence prevention programs, including those that use gender transformative approaches. Interviewing men and women living with disabilities shed light on how gender intersects with disability related stigma, discrimination and experiences of violence. The findings indicate the importance of applying an intersectional analysis to appreciate how individuals experience stigma and discrimination on the basis of multiple identities (Crenshaw, 1989). The data indicates the systemic risk factors whereby women with disabilities experience overlapping discrimination based on their gender and ability status (among other identities), including heightened risk of experiencing IPV. The data also revealed the extent of stigma and discrimination men with disabilities experience. Indeed, men and women reported similar experiences of disability stigma and discrimination across the four contexts, including being blamed and devalued for their disability, denied access to employment opportunities, and excluded from community interactions. People with visible disabilities frequently experienced stigma during social interactions because of the prominent differences that distinguish them from others, as has been demonstrated elsewhere (Dunn & Burcaw, 2013;Phemister & Crewe, 2004). The visibility of impairments could precipitate verbal abuse and/or threats of violence that participants experienced. Physically inaccessible spaces also contributed to exclusion and isolation. Both men and women described a sense of failure to live up to ableist gender norms and social obligations in their communities and relationships due to the isolation, discrimination and stigma they experienced. For men, this related to challenges to providing economically for families, and lacking the strength or ability to protect themselves from violence or abuse. This confirms other literature suggesting that hegemonic masculinitiesand disabilityconflict with each other; disabilityis typically associated with being dependent and helpless, whereas masculinity is associated with being powerful and autonomous (Shuttleworth, Wedgwood, & Wilson, 2012). Women with disabilities were considered less attractive marital partners, which is congruent with literature suggesting that women with disabilities are less likely than men with disabilities to be married or have an intimate partner (Mitra et al., 2011). This could be a significant source of internalized and community stigma and discrimination for women. Both male and female participants shared experiences of community violence, however, women reported more severe forms of isolation and self-stigma, confirming existing literature that women with disabilities are often more socially excluded and isolated than men with disabilities (Brownridge, 2006). Many women reported experiences of physical and emotional IPV related to their disability; emotional IPV has been found to have significant health impacts, independently of physical and sexual IPV, including poor mental health, depression and suicidal ideation (Gibbs, Dunkle, & Jewkes, 2018). The findings confirm the literature establishing that women with disabilities are more likely to experience both severe and more subtle forms of abuse than women without disabilities and men with disabilities, and for longer periods of time (Ballan et al., 2015). None of the men reported victimization or perpetration of IPV. Yet, evidence suggests that abuse against men by female partners is often hidden and under-reported due to normative gender restrictions against men admitting to violence, especially IPV, lack of services for male survivors of abuse, and poor awareness of disability-related abuse among men (Ballan et al., 2015;Kolivas & Gross, 2007).
Despite participants' experiences of discrimination, stigma and violence related to their disabilities, participation in the What Works programs appeared to have significant and positive impacts. Inclusion in programs helped challenge disability-related social exclusion. Participants in the Rwanda and Ghana programs appreciated their integration of disability rights awareness via the platform of community activism. Such efforts raised awareness of rights and laws protecting people with disabilities, potentially helping people with disabilities to better respond to violence and abuse. Active engagement in programs and involving people with disabilities in visible programmatic roles, such as the community activists in Rwanda, helped challenge stigmatizing perceptions around disabilities. Barriers to accessing the What Works programs were identified across all countries, including limited accessibility, not having disability inclusive mandates, materials, spaces or outreach. The fact that participants were not for the most part engaged with the program in South Africa demonstrates the difficulty of reaching people with disabilities through general awareness campaigns and the need for specific outreach to reach people with disabilities. These insights lend to recommendations for supporting more disability inclusive IPV programming, which we turn to now.

Lessons learned
Valuable lessons for disability inclusive IPV programming were learned through this comparative qualitative approach. Firstly, programs should be both gender and disability transformative in terms of challenging stigma and discrimination related to both identities. Prevention of violence against women and men with disabilities should not focus on functional impairments, but rather challenge ableist gender norms and explicitly address local, culturally specific manifestations of disability-related stigma. Program content should draw upon and incorporate the local knowledge and experiences of women and men living with disabilities. This is particularly important with respect 1 This Kinyarwanda term translates into English to mean 'one who is disabled' but is considered to be a derogatory term in Rwanda. Ikimuga also refers to cracked pots designed for storing traditional beer. E. Stern, et al. Evaluation and Program Planning 79 (2020) 101770 to how disabilities interact with dominant gender norms, and how these intersectional identities can worsen isolation and vulnerability to violence, discrimination and other human rights violations. Program materials and approaches should be designed in collaboration with people with disabilities to complement existing guidelines, protocols and tools for IPV prevention and response. For example, it was highly valuable that the Indashyikirwa program in Rwanda partnered with the National Council of Persons with Disabilities, as they have knowledge and expertise on local forms of disability related stigma, discrimination, violence, and rights. Programs should also incorporate disability awareness training among people with disabilities, program staff and service providers. Targeted training for people with disabilities and program staff, as was the case with Indashyikirwa, is critical to raise awareness of the intersection between disability and the various types of violence and discrimination they are susceptible to, and inform people with disabilities of their rights to live free from violence. Furthermore, programs would have benefitted from training formal service providers in their referral networks in disability awareness and disability-specific responses to IPV. Programs should also foster and highlight the resilience and capabilities of women and men with disabilities and ensure people with disabilities are recruited into roles with status and high visibility, such as community activists. It would be helpful for programs to draw on existing international and national legislation and rights offering protection for people living with disabilities.
Given the extent of exclusion, internalized stigma, and accessibility barriers experienced by people with disabilities, programs need to explicitly and proactively target people with disabilities for inclusion. Programs should further explore ways to ensure that people with disabilities know about, can physically access, and will feel welcome at all program activities. Outreach activities should strive to overcome isolation through strategies such as home visits and partnering with disability organizations to aide recruitment. The data suggests the need to better understand how IPV prevention programs can include and meet the needs of different types of disabilities and address different barriers to inclusion dependant on the type of disability. IPV prevention programs should monitor inclusion of and accessibility to people with disabilities. Guidelines and trainings to ensure more comprehensive and sensitive monitoring are warranted to ensure that not only people with visible disabilities are included and properly accommodated. Finally, donors should prioritize disability inclusive IPV prevention programs and provide funding opportunities for disability inclusive program development, costing, provider training, implementation, monitoring and evaluation. For Indashyikirwa, it was especially critical that the funder prioritized inclusion of people with disabilities.

Strengths and limitations
Participants with severe and/or invisible disabilities, including psychiatric and cognitive disabilities, were not included in the study. This is especially limited as some participants noted how more invisible forms of disabilities, including mental illness, are more discriminated against and stigmatized. This speaks to the need to properly train field staff to safely identify people living with a diversity of disabilities for participation in research, including those with less visible forms of disabilities, or multiple, co-existing disabilities. Due to limited time and budget, full double checking of the accuracy of the translations of the transcripts into English was not done however we did do spot checks and verify translation of key phrases. We were also unable to validate the analysis with participants themselves, which we recognize as further limitations. We recognize the limitation of not assessing programs that had specific disability-related programming (apart from Indashyikirwa), or among participants who had not been exposed to a particular program (in the case of South Africa). However, the aim of this study was to assess the experiences for men and women living with disabilities in mainstream IPV programmes. The findings are not intended to be generalizable given the small samples across diverse contexts, yet common insights regarding the stigma and discrimination for people living with disabilities, and how this related to their exclusion and inclusion in IPV programs is a strength of this paper. Indeed, this study has significant value for exploring and identifying similar themes across four diverse evaluations of ongoing IPV prevention programs.

Conclusion
This study makes an important contribution to the limited evidence on delivering IPV programming for people with disabilities in developing countries. Participants across diverse countries reported similar experiences of violence, disability stigma and discrimination, isolation, and physical barriers to participation in the violence prevention programs, affirming the need for such programming to be more sensitive to and inclusive of people with disabilities. Our data encouragingly suggests that including women and men with disabilities in programs targeting the general population may mitigate their risks for violence and promote economic empowerment and well-being, while challenging disability-related stigma and discrimination. Valuable lessons were identified to improve disability-inclusive strategies for IPV prevention and response. More attention to this area is warranted to further develop guidelines for disability-inclusive development, implementation, and evaluation of IPV prevention and response efforts, and to develop strategies to ensure that people with the most severe disabilities can be effectively reached by IPV prevention programs.

Funding
This piece of research was funded by DFID-UK as part of the Global What Works to Prevent Violence against Women and Girls Programme.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Dr. Erin Stern is an Honorary Assistant Professor with the London School of Hygiene and Tropical Medicine, and the study coordinator and lead qualitative researcher of the evaluation of Indashyikirwa in Rwanda. Her background is in qualitative research and uptake, with expertise in prevention of gender-based violence and gender transformative programming evaluation. She has a PhD in Public Health from the University of Cape Town and an MSc in Health, Community and Development from the London School of Economics & Political Science. Dr. Stern is a Senior Associate with the Prevention Collaborative.
Ingrid van der Heijden is a post-doctoral fellow and senior scientist at the South African Medical Research Council, with a strong interest in gender, disability and gender-based and disability-based violence in low-middle income countries. She was the principle investigator of the evaluative qualitative multi-country study on disability and violence for the What Works to Prevent Violence against Women and Girls Global Programme (2016-2019).
Dr. Kristin Dunkle is a Chief Specialist Scientist at the South African Medical Research Council. Dr. Dunkle is a social epidemiologist who has been working on gender-based violence and health for over 20 years, and has published widely on: connections between intimate partner violence and HIV among both women and men; links between gender-based violence, transactional sex, and poverty; intimate partner violence; sexual violence and health among men who have sex with men (MSM) and others in LGBTQIA communities. Dr. Dunkle oversees the What Works portfolio on understanding and preventing gender-based and sexual violence among people living with disabilities. E. Stern, et al. Evaluation and Program Planning 79 (2020)