Reduction of Family Violence in Aboriginal Communities: A Systematic Review of Interventions and Approaches

Abstract

Many efforts to reduce family violence are documented in the published literature. We conducted a systematic review of interventions intended to prevent family violence in Aboriginal communities. We retrieved studies published up to October 2009; 506 papers included one systematic review, two randomized controlled trials, and fourteen nonrandomized studies or reviews. Two reviews discussed interventions relevant to primary prevention (reducing the risk factors for family violence), including parenting, role modelling, and active participation. More studies addressed secondary prevention (where risk factors exist, reducing outbreaks of violence) such as restriction on the trading hours for take away alcohol and home visiting programs for high risk families. Examples of tertiary prevention (preventing recurrence) include traditional healing circles and group counselling. Most studies contributed a low level of evidence.

Background

The United Nations Population Fund (UNPA) lists sixteen forms of violence directed against women. The most common are repeated assaults, physical and psychological, within families (UNPA, 2010). Family violence can include action that “endangers the survival, security or well-being of another person (College and Association of Registered Nurses of Alberta, the College of Licensed Practical Nurses of Alberta, and the College of Registered Psychiatric Nurses of Alberta, 2008). The Royal Commission on Aboriginal Peoples (RCAP) considered the infrastructure of family violence as “the serious abuse of power within family, trust or dependency relationship” (RCAP, 1996).

It is not only women who suffer. Infants in violent families experience adverse effects along with their mothers, and abuse of boys is common (Boy and Salihu, 2004). The term “family violence” reflects this suffering of all members of the family (including the perpetrators), and the children of these families are at increased risk of developing personality disorders, mental health problems, poor self esteem, and low educational achievement (Bair-Merritt et al., 2006).

Several interrelated factors contribute to high levels of family violence experienced in Aboriginal communities. These include poor socioeconomic conditions, high rates of alcohol and substance abuse, systemic discrimination and racism against Aboriginal Peoples, as well as the trauma and inter-generational cycle of violence resulting from the residential school legacy, and the impact of colonialism on traditional values and cultures (RCAP, 1996).

The Aboriginal-specific literature on family violence is sparse; a brief review summarizes what is known about the effectiveness of interventions in nonindigenous communities, and comment on the applicability of successful interventions in Aboriginal communities. The primary focus of this review is the relevance of initiatives to reduce family violence in Aboriginal communities in Canada.

Methods

The term “systematic review” refers to exhaustive searching, selecting, collating, appraising, interpreting, and summarizing data from original studies (Cook et al., 1997). These studies may be observational or randomized trials, qualitative or quantitative. Limited to published material, a well known difficulty of systematic reviews is that component studies are simply too different in the way they approach a problem for the information each offers to add up. When the quality of primary studies is poor, a systematic review can at best note the flaws; it cannot improve the data.

There is currently no agreed classification for interventions addressing family violence. Table 1 offers a standard prevention framework (primary, secondary, and tertiary prevention) including examples from the Human Rights and Equal Opportunity Commission in Australia (HREOC, 2006). Tertiary prevention of family violence receives most attention and research investment; it focuses on the protection and care of victims, and punishment and rehabilitation of perpetrators with the intention of reducing recurrence. Secondary prevention identifies those with risk factors and prevents their progress to overt violence through screening, counselling, or removal of the risk factor. Primary prevention receives less research interest and public funding; this seeks to avoid the genesis of risk factors for family violence.

Table 1.

Categories of Interventions to Reduce Family Violence

Primary Prevention: Reduce the Risk Factors for Family Violence Identity programs — designed to develop a secure sense of self-value or self-esteem; programs focus on culturally specific psychological or spiritual healing. Education and awareness raising — designed to raise awareness about family violence prevention; participatory self-education programs show promise. Structural intervention — designed to change the material position of women (for example, micro-credit).

Secondary Prevention: Reduce Breakout of Violence Where Risk Factors Exist Education and awareness raising — designed to develop skills within communities to resolve conflicts and identify the need for interventions with perpetrators. Behavioural change — men are frequently the subject of behaviour change strategies. Complementary preventive/intervention programs for youth. Reducing substance exposure — limit access to substances associated with violence and other antisocial behaviours. Examples include banning of alcohol (dry communities). Night patrols — night patrols, particularly in remote areas, can strengthen indigenous mechanisms for social control. They use traditional methods to control antisocial behaviour, minor infractions, and potentially serious incidents.

Tertiary Prevention: After Violence, Prevent Worst Consequences and Recurrence Screening Programs — these are usually implemented in health care settings that aim to improve the detection of domestic violence by primary care practitioners (doctors and nurses) dentists and faciomaxillary surgeons. Early childhood visiting — visits in the first two years of life enables detection of childhood injuries, and an opportunity to determine if the mother has been abused; it is an opportunity to provide counselling and advice. Support programs — these should be accessible and provide appropriate counselling and advocacy for family and community members who have been exposed to family violence. Refuges and Shelters — it is considered that refuges and shelters need to be coupled with proactive strategies targeted at the perpetrators of violence and other situational factors. Justice programs —typically, these are aimed at the perpetrators of violence, and the purpose is to mediate between people in conflict, designate culturally appropriate punishments, for example through “circle sentencing.” Dispute resolution — impartial members of the community act as facilitators and traditional dispute-resolution techniques are incorporated into mediation. Composite programs — these comprise elements of the above categories and target different forms of violence in the community, target different categories of offenders or victims, or employ different methods of combating or preventing violence.

In the characterization of interventions intended to reduce family violence, holistic refers to the fullness of interventions at any one of these levels, as they include spiritual, cultural, and other dimensions of indigenous ways of life.

We developed a protocol to summarize and evaluate the intervention studies in indigenous communities. Initial literature scans identified published systematic reviews as a means of summarizing several hundred studies. We extracted relevant reviews of intervention studies conducted in Aboriginal and non-Aboriginal communities at this stage. Beverley Shea and Amy Nahwegahbow assessed all publications to identify relevant studies and included these in the review. We searched for primary studies published up to and including October Week 3 2009 using MEDLINE, PsycINFO, HealthSTAR, North American Indian Biographical Database, Violence and Abuse Abstracts 2001–2004, EMBASE, Global Health, and the Cochrane Library. Search terms varied by database.²

We did not limit the searches by study type. Beverley Shea and Amy Nahwegahbow read the full text of these studies. We confined our literature retrieval to studies that provided qualitative and quantitative estimates of prevention. Formal assessment of randomized trials used the SIGN 50 instrument (SIGN, 2010) and non-randomized studies relied on the Effective Practice and Organization of Care instrument (EPOC 2010). However, the broad range of the other study types required an informal approach to quality assessment and precluded any attempt at data pooling.

Results

The search yielded 506 citations up to October 2009. Combinations of search terms identified a smaller number of potentially relevant titles from each database: MEDLINE (93), EMBASE (26), and CENTRAL (28), psychINFO (96), Global Health (16), North American Indian Biographical Database (70) Violence and Abuse Abstracts (6). There was considerable overlap between the retrieval lists for these databases. Beverley Shea and Amy Nahwegahbow examined all titles and identified 27 articles describing school and community-based interventions aimed at decreasing family violence (see Figure 1). We excluded 12 studies that did not describe or test interventions to prevent family violence in Aboriginal communities (Appendix 1). A key inclusion criterion was that studies involve an indigenous population, whether the intervention was at individual or community level. One of eleven studies in the single systematic review (Whitaker et al., 2006) included Aboriginal participants (Table 2).

Figure 1.

Flow Chart for Included and Excluded Systematic Reviews and Primary Studies: PRISMA 2009 Flow Diagram¹

¹ From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(6): e1000097. doi:10.1371/journal.pmed1000097

Appendix 1.

Excluded Studies with Reasons for Exclusion

Study ID	Reason Excluded	Reference
Babcock, 2004	Not specific to Aboriginal communities.	Babcock, J., Green, C., and Robie, C. (2004). Does batterers’ treatment work? A meta-analytic review of domestic violence treatment. Clinical Psychology Review, 23(8), 1023–1053.
Beals, 2003	Not specific to domestic violence interventions.	Beals, J., Manson, S.M., Mitchell, C.M., Spicer, P., and the AI-SUPERPFP Team. (2003). Cultural specificity and comparison in psychiatric epidemiology: Walking the tightrope in American Indian research. Culture, Medicine and Psychiatry, 27, 259–289.
Bohn, 1998	Not a peer-reviewed journal but a chapter in a book.	Bohn, D.K. (1998). Clinical Interventions with Native American battered women. In: J.C. Campbell, ed., Empowering Survivors of Abuse: Health Care for Battered Women and their Children. Thousand Oaks, CA: Sage, pp. 241–258.
Duran, 2009	Not specific to domestic violence interventions but rather a study on prevalence and association with ADM disorders.	Duran, B., Oetzel, J., Parker, T., Malcoe, L.H., Lucero, J., and Jiang, Y. (2009). Intimate partner violence and alcohol, drug, and mental disorders among American Indian women in primary care. American Indian and Alaska Native Mental Health Research, 16(2), 11–27.
Chamberlain, 2008	This is a process evaluation.	Chamberlain, L. (2008). Ten lessons learned in Alaska: Home visitation and intimate partner violence. Journal of Emotional Abuse, 8(1), 205–216.
Edwards, 1984	Not specific to domestic violence interventions; integrating culture in social work/group therapy with American Indians.	Edwards, E.D. and Edwards, M.E. (1984). Group work practice with American Indians. Social Work With Groups, 7(3), 7–21.
Fawcett, 1999	Non-Aboriginal community.	Fawcett, G.M., Heise, L.L., Isita-Espejel, L., Pick, S. (1999). Changing community responses to wife abuse: A research and demonstration project in Iztacalco, Mexico. American Psychologist, 22(6), 714–719.
Gray, 2000	Not specific to domestic violence.	Gray, D., Saggers, S., Atkinson, D., Sputore, B., and Bourbon D. (2000). Beating the grog: An evaluation of the Tennant Creek liquor licensing restrictions. Australian and New Zealand Journal of Public Health, 24(1), 39.
Joseph, 2009	Not specific to Aboriginal communities. Looks at African American population.	Joseph, J.G., Ayman, A.E., Kiely, M., El-Khorazaty, M.N., Gantz, M.G., Johnson, A.A., Katz, K.S., Blake, S.M., Rossi, M.W. and Subramanian, S. (2009). Reducing psychological and behavioral pregnancy risk factors: Results of a randomized clinical trial among high risk pregnant African American Women. American Journal of Public Health, 99(6), 1053–1061.
Lafromboise, 1990	Not specific to domestic violence, but discusses psychological interventions and the importance of including culture.	Lafromboise, T.D., Trimble, J.E., and Mohatt, G.V. (1990). Counseling Intervention and American Indian tradition: An integrative approach. The Counselling Psychologist, 18, 628–654.
McDermott, 2005	Not specific to Aboriginal community.	McDermott, R. (2005). The great arch of unimagined bridges: Integrative play therapy with an abused child. Journal of Psychology, 15(2), 165–175.
Norton, 1995	Qualitative study of women experiencing domestic violence; characteristics of those who participated in the intervention.	Norton, I.M. and Manson, S.M. (1995). A silent minority: Battered American Indian women. Journal of Family Violence 10(3), 307–318.

Table 2.

Published Systematic Reviews of Interventions to Reduce Family Violence in Nonindigenous Communities

Authors	Topic/Setting	Interventions	Results/authors’ conclusions
Ramsay et al., 2002	Primary Care; women who have been abused.	Screening for domestic violence by health professionals.	In nine studies of screening compared with no screening, most reported a greater proportion of abused women identified by healthcare professionals. Six interventions used weak study designs and gave inconsistent results. Other than increased referral to outside agencies, little evidence exists for changes in important outcomes such as decreased exposure to violence.
Wathan and MacMillan, 2003	Primary Care; women who have been abused.	Women’s shelters; post shelter advocacy and counselling services.	Among women spending one night in a shelter, there is evidence that those who received a specific program of advocacy and counselling services reported a decreased rate of reabuse and an improved quality of life.
Nelson et al., 2004	Health care settings; women who have been assaulted.	Screening for intimate partner violence by health professionals.	Some screening instruments demonstrate internal consistency and some have been validated with longer instruments, but none have been evaluated against measurable violence or health outcomes. Existing intervention studies focused on pregnant women, and study limitations restrict their interpretation.
Coulthard et al., 2004	Dental and oral and maxillofacial practice; women who have been assaulted.	Screening for domestic violence.	Identified no eligible RCTs. There is no evidence to support or refute that screening for domestic violence in adults with dental or facial injury is beneficial nor that it causes harm.
Bilukha et al., 2005	Early childhood; children of families with domestic violence.	Home visits in the first two years by trained personnel who convey information about child health, development, and care; offer support; provide training; or deliver any combination of these services.	The Task Force found insufficient evidence to determine the effectiveness of early childhood home visitation in preventing violence to children, violence by parents (other than child abuse and neglect), or intimate partner violence.
Whitaker et al., 2006	Community — Primary Prevention in middle- or high-school aged students.	Brief educational interventions directed at all students, usually in a school setting.	Although a majority of studies were RCTs, study quality was generally poor with relatively short follow-up periods, high attrition rates, and poor measurement. Of the four studies that measured behaviour, two found a positive intervention impact. Conclusions about the overall efficacy of dating violence interventions are premature, but such programs are promising.
Smedslund et al., 2007	Community — Men who have physically abused their partner.	Cognitive behavioural therapy (CBT) seeking to change behaviour, but also targeting the thinking patterns and beliefs that are thought to contribute to violence.	Six trials, all from the USA, involving 2343 people, were included. A meta-analysis of 4 trials comparing CBT with a no-intervention control with 1771 participants, reported that the relative risk of violence was 0.86 (favouring the intervention group) with a 95% confidence interval (95% CI) of 0.54 to 1.38. This is a small effect size, and the confidence interval is so wide that there is no clear evidence for an effect. There are still too few randomized controlled effect evaluations to conclude about the effects of CBT on domestic violence.
Andersson et al., 2008	HIV prevention in southern Africa.	Primary prevention interventions that reduce gender based violence, in the context of HIV prevention.	Two RCTs in South Africa tested participatory learning interventions to reduce intimate partner violence (IPV). One study combined this education/awareness with micro-finance; it reduced the IPV risk by 55%. Participants in the second trial reported less GBV and fewer new HIV infections with exposure to participatory learning than those in the control arm.
Ramsay et al., 2009	Community – women experiencing abuse, role of empowerment and advocacy.	All of the interventions sought to empower the women by helping them to achieve their goals. They differed in duration (30 minutes to 80 hours), outcomes reported, and duration of follow up.	Ten trials involved 1,527 women. The studies compared advocacy with “usual care” and were conducted in a variety of settings both within and outside of healthcare. Participants were recruited from diverse ethnic populations and across a wide age range (15–61 years). Most were experiencing current, often severe, abuse. The evidence is consistent with intensive advocacy decreasing physical abuse more than one to two years after the intervention for women already in refuges, but there is inconsistent evidence for a positive impact on emotional abuse. Similarly, there is equivocal evidence for the positive effects of intensive advocacy on depression, quality of life and psychological distress. There is evidence that brief advocacy increases the use of safety behaviours by abused women.

We identified two randomized controlled trials (RCT) of interventions in indigenous communities (Duggan et al., 2004; Duggan et al., 2007) and fourteen nonrandomized studies (Tables 3 and 4). These interventions include: community injury prevention (role modelling, life span focus, accessibility, acceptability and active participation), restriction on the trading hours for take away alcohol, home visiting, parenting, traditional healing circle and group counselling provided in an Aboriginal health setting. The methodological quality of the two RCTs was 60% (Duggan et al., 2004) and 66% (Duggan et al., 2007) while that of the other studies ranged from 29%–45%.

Table 3.

Characteristics and Results of Qualitative Studies and Reviews of Prevention in Aboriginal Communities

Study	Participants	Study interventions	Evaluation methods	Results
Qualitative accounts of primary prevention
Kiyoshk, 2003		The Change of Seasons is a 28 session psycho-educational group counselling model combined with Aboriginal healing methods. Spiritual practices are integral.	Narrative description.	Some interventions have integrated practical rituals and ceremonies into psycho- educational group counselling models, as seen in the Changes of Seasons model.
Oetzel, 2004	American Indian and Alaska Native communities.	Review of interventions that address gender, age, socioeconomic status, alcohol, European colonization, and infrastructure. It describes multilevel interventions that are culturally appropriate for (AI/AN) communities.	Review uses a social ecological framework to organize the literature. The framework identifies proximal and distal factors related to IPV at 5 levels (individual, interpersonal, institutional or organizational, community, and policy levels).	The impact of mainstream interventions at any given level is small to moderate. The likely reason for this limited impact of any given intervention is that it does not address determinants at multiple levels.
Qualitative accounts of secondary prevention
Holkup, 2007	Native American Elders 26 families referred, 10 have participated.	Family Care Conference (FCC) FCC involves inviting family members, family-nominated supportive community members, a spiritual leader (if desired), and relevant health and social service providers to attend a meeting in which individuals bring to the forum concerns about the welfare of the elder.	FCC provides the opportunity for family members to come together to discuss and develop a plan for the well-being of their Elders. FCC has six stages: referral, screening, engaging the family, logistical preparation, family-meeting, follow-up.	Drawing on the values of interdependence and reciprocity among Native American kin, the FCC provides a culturally anchored and individualized way to identify a frail elder’s care needs and find solutions to meet those needs from among family members and available community resources.
Norton and Manson, 1997	Urban American Indian Over 25 women included in a peer- to peer counselling battered women program.	located in an urban Indian health centre that provides social service interventions such as housing, emergency clothing, and transportation to appointments. individual counselling and crisis housing. alternatives to traditional office-based interventions include home visits and a weekly domestic violence group incorporating traditions and values. staffed by 2 American Indian women.	Women’s sense of security increased with the group, they began to respond to each other with mutual advice and support. Establishing an alliance with the women through home visits, and informal atmosphere for discussion, including traditions such as the sharing of meals.	Interventions relevant to urban American Indian communities. The women were able to build a relationship with the counsellor, and benefit from active participation in a domestic violence program. Home visits significantly enhance care and the effectiveness of counselling.
Willmon-Haque, 2008	The recent published literature on poverty and historical trauma provides a contextual framework for understanding issues of violence and the resulting trauma, suicide, domestic violence, and PTSD shared by many AI/AN communities. Cultural adaptations of evidence-based treatments (see Chamberlain, 2008, this issue) are one attempt to integrate Western psychology with indigenous ways of the knowing. Indigenous people are seeking to regain their healing ways by looking around their circle and drawing toward them ways that work and ways to address violence.
Qualitative accounts of tertiary prevention
Heilbron and Guttman, 2000	First Nations and non-Aboriginal women survivors of child sexual abuse. Three Ojibway women and two non-Aboriginal women.	A traditional ceremony “healing circle” was included in a counselling group, combined with cognitive therapy. Group meetings weekly over a ten-week period for approximately two hours each session. Establish a spiritual component, traditional ceremony and Aboriginal beliefs, to the counselling process.	Group meetings were audio-taped group and transcribed. Open-ended evaluation forms were filled out at the completion of the ten-week group. Examine the healing circle ceremony and the sharing of Aboriginal beliefs or values for its influence on the group.	The healing circle influence provided a spiritual framework for addressing problems. First Nations women felt the teachings provided personal meaning for them and allowed them to develop a stronger devotion to Aboriginal healing methods. The teachings also promoted disclosure as they discussed how the stories were affecting them personally.
Mokuau, 2002	Indigenous Hawaiians	Two culturally based interventions incorporate cultural values and practices, and involve Native Hawaiian in the design and delivery of the interventions.	Narrative description.	Human services should continue to emphasize interventions that integrate “mainstream” and cultural-specific approaches.
Luna-Firebaugh, 2006	123 tribal programs evaluated in the USA, including Alaska.	The program entails partnerships between the tribal justice system and the nonprofit, non-governmental services. The evaluation assessed the impact of tribal programs aimed at reducing violence against women.	Used surveys; extensive, open-ended interviewing; and case studies. Data sets included material obtained from a mailed survey, telephone surveys, site visit interviews, and materials promulgated by the programs.	The STOP VAIW programs empowered tribes and tribal officials, program personnel, and women. The tribal programs enhanced the safety of Indian women in tribal communities.

Table 4.

Characteristics and Results of Quantitative Studies of Violence Prevention in Aboriginal Communities

Study ID	Participants	Study interventions	Evaluation methods	Study Outcomes	Results
Brewin 2004	North Island of New Zealand. Participants were nonrandom selected from local tribe groups. 254 completed questionnaires in the first survey and 222 in the second survey.	Intervention incorporated Maori situations and concepts. Road safety, alcohol and drug programs, family violence, and playground safety audit. To address family violence, two focus areas: 1) identifying and affirming traditional care and protection practices, 22 education sessions, information packs and ongoing training of resource people. 2) regional promotional programs to address domestic violence and child abuse to tie in with national campaigns.	The evaluation used Maori cultural frameworks. Survey and injury surveillance data for the period 1996–99. 12 key informant interviews every six months. Utilization-focused evaluation used to ensure that the information would be useful to community. Impact evaluation data obtained through pre-/post-intervention surveys and participant observation.	Awareness of injury prevention, injury rates, and behaviours.	Proportion who encountered threatening situations outside the home decreased (pre 73%, post 66%) and an increase in % who always/mostly walk away from a threatening situation at home (pre 27%, post 31%). Key informant interviews supportive of the emphasis placed on affirming traditional care and protection. Increased awareness of injury prevention (pre 17% and post 25%). Injury morbidity statistics for the period 1996–99 showed a significant decrease in injury rates for all age groups, compared with the comparison community.
Douglas 1998	Halls Creek, small town in Western Australia. Population 1,200, 63% are Aboriginal.	Decreased availability of alcohol through restricted trading hours: no packed liquor sold in the town before midday, cask wine only sold between 4 pm and 6 pm; one case to any one person on any day. Other programs along with the intervention included school education program.	Patterns of alcohol use, incidence of crime, and outpatient data at the local hospital.	Alcohol consumption, incidence of crime and cases of domestic violence at the local hospital.	Decrease in alcohol consumption, crime, and alcohol-related presentations to the hospital. Fluctuations in the domestic violence data, without any evidence for trend. Overall, the intervention had a positive impact on the community.
Duggan 2004 RCT	643 families were enrolled and randomly assigned to intervention and control groups. 34% of those in the HSP group (N=373) were Native Hawaiianor Pacific Islander. 33% of those in the control group (N=270) were Native Hawaiian or Pacific Islander.	This RCT focused on Hawaii Healthy Start Program (HSP) sites operated by three community-based agencies from 11/94 to 12/95. The HSP program involves home visiting to prevent child abuse in families targeted as at-risk for abuse of their newborns. The HSP model has two parts: (1) population-based screening and assessment to identify families at-risk of child abuse and neglect; and (2) home visiting of identified at-risk families.	RCT with outcomes (child abuse and neglect) measured by observed and self-reported parenting behaviours, hospitalizations for trauma and for conditions where hospitalization might have been avoided. Annual interviews (88% follow-up of families); observation of the home environment; and review of CPS, HSP, and paediatric records.	Preventing child abuse and neglect over 3 years using a broad range of indicators.	The HSP and control groups were similar on most measures of maltreatment. HSP group mothers were less likely to use common corporal/verbal punishment but this was attributable to one agency’s reduction in threatening to spank the child. Abusive parenting behaviours –AOR (95% CI); Psychological aggression .76 (.54, 1.07); Severe physical abuse 1.30 (.89, 1.88); Corporal/verbal punishment .59 (.39, .88); Assault on child’s self-esteem .90 (.67, 1.20); Hitting with an object 1.22 (.88, 1.68); Extreme physical abuse 1.26 (.66, 2.41); Shook child .94 (.57, 1.54); Revised neglect category .72 (.54, .96). The program did not prevent child abuse or promote use of non-violent discipline.
Duggan 2007 RCT	325 families randomized to intervention and control groups. 23% (n=162) in intervention were Alaska Natives. 20% (n = 163) in the control group were Alaska Natives.	This RCT focused on 6 Healthy Families Alaska (HFAK) programs aimed at preventing child maltreatment by promoting positive parenting and child health and development.	RCT. Follow-up data were collected when children were 2 years old (85% follow-up rate). HFAK records to measure home visiting services. Analysis limited to families with a baseline interview. Student’s t test and chi-square were used to assess sample representativeness and baseline comparability between the treatment and control groups.	Outcomes included maltreatment reports, measures of potential maltreatment and parental risks, for example, poor mental health, substance use, and partner violence.	There was no overall program effect on parental risks for child maltreatment. There was no impact on outcomes for families with a ‘high dose’ of home visiting. No significant program impact on malleable parent risks for child maltreatment. Association of parental risks and behaviours - AOR (95% CI); Severe physical assault 2.3 (0.9, 6.1); Assault on esteem 2.9 (1.5, 5.7); Neglect 2.9 (1.3, 6.2); Poor quality home environment 3.0 (1.4, 6.6). Process measures (All sites, Range); Discussion of risks in active, risk-positive families - Domestic violence at any time (36%, 6–60%); Referral for risks in active, risk-positive families - Domestic violence (24%, 14–60%).
Wolfe 2003	Sample size N=191 (92 boys and 99 girls) 14 to 16 year old at risk of abusive relationships based on their history of maltreatment 85% white, 8% First Nations, 3% Asian, and 4% African Canadian.	An 18-session program used a health-promotion approach to preventing violence in dating relationships by focusing on positive alternatives to aggression based interpersonal problem-solving and gender-based role expectations. It includes education about healthy and abusive relationships, conflict resolution and communication skills, and social action activities. Experimental (intervention vs. control). Comparison group activities: Standard services offered by CPS Study quality.	Experimental study Recruited via past experiences of maltreatment were assessed with the Childhood Trauma Questionnaire (CTQ). Measures of abuse and victimization with dating partners, emotional distress, and healthy relationship skills were completed at bi-monthly intervals when dating someone.	Abuse perpetration and victimization, emotional distress, healthy relationships skills, trauma symptoms, and hostility. Trauma Symptom Checklist-40. Adolescent Interpersonal Competence Questionnaire was used. Hostility sub-scale of the Symptom Checklist-90	Growth curve analysis showed intervention effective in reducing incidents of physical and emotional abuse and symptoms of emotional distress over time. Findings support involvement of youth in reducing the cycle of violence as they initiate dating in mid-adolescence. Abuse perpetration and victimization: INC at 16 month follow up (effect stronger for girls than for boys). Healthy relationship: All five sub scales: null Trauma symptoms: INC Hostility: null Unconditional growth models showed significant reduction in all forms of victimization over time. TIME −.007; −.006; −.006, p.01, for physical abuse, emotional abuse, and threatening behaviour, respectively.
Becker 2008 Explanatory trial	106 children between the ages of 3 and 17 and their non-offending parent (guardian). Participants from diverse ethnic back- grounds, Asian or Pacific Island heritage.	12-week, community-based, culturally influenced group program for children and adults exposed to family violence. Weekly child intervention through 90 min support group. Weekly intervention through a simultaneous parenting support group.	56/106 parents completed a baseline questionnaire. The intervention involved group sessions. Parents completed the Child Behaviour Checklist post-intervention.	Child functioning: domestic violence-related skills and behaviour. Parent functioning: violence-related skills and parenting – 28-item conflict questionnaire from Conflict Tactics Scale – 11 item counsellor rating check-list of violence-related skills	Statistically and clinically significant improvement for children and parents over the course of the treatment. Children and parents’ domestic violence-related skills, the primary target, improved. Following the intervention, children showed improvements in internalizing and externalizing difficulties, despite the lack of direct symptom-focused intervention in these domains.

Studies in Nonindigenous Communities

Table 2 lists nine systematic reviews that together summarize much of the large literature on the prevention and reduction of family violence in non-indigenous communities. Despite the international importance of the subject, few high quality intervention studies appear in the literature. Every systematic review and task force report concludes with a call for more and better studies on family violence.

Two RCTs in the review by Andersson and colleagues (2008) point to promising ways to reduce interpartner violence in Africa: a randomized controlled trial of income enhancement and gender training resulted in reduced gender violence and HIV risk behaviours (primary prevention), and a trial of a participatory learning program reported a reduction of male risk behaviours.

Other interventions that one might expect to help in practice have not done so in well conducted studies. Screening female patients by health professionals, for instance, increased referral for counselling but did not reduce subsequent episodes of abuse. The creation of temporary sanctuaries for women does not have a lasting effect on occurrence of family violence unless combined with postshelter advocacy and counselling. There is no evidence that early childhood visits, though beneficial in other ways, reduce the rates of subsequent assaults on mothers and children. Abuse is well established by the time detection occurs and interventions at this stage are probably insufficiently powerful to change the behaviour of perpetrators (Table 2).

Studies in Indigenous Communities

Eight studies described interventions relevant to the prevention of family violence in Aboriginal and indigenous communities (Table 3). We found no published reports of these interventions tested in a controlled setting. We also identified six quantitative studies (Table 4) that examined prevention including role modelling, life span focus, accessibility, acceptability and active participation; secondary prevention such as restriction on the trading hours for take away alcohol, home visiting programs (healthy start program) and parenting; and tertiary prevention including traditional healing circles and group counselling.

Primary Prevention: Reduction of the Risk Factors for Family Violence

Qualitative publications on primary prevention

Relevant to primary prevention, Kiyoshk (2003) discussed interventions that integrated spiritual practices and ceremonies into psycho-educational group counselling models. Their “Change of Seasons model” included a 28 session psycho-educational group counselling approach that sought to integrate cultural healing methods for Aboriginal men. Spiritual practices and ceremonies included smudging, the talking circle, and the sweat lodge. The author claimed their model was accepted by Aboriginal clients “because it fits with their community’s spiritual beliefs and worldviews” and proposed that “Systems Thinking” could be applicable to Aboriginal communities. The description offers no evidence of impact on family violence.

Oetzel and Duran (2004) summarized several studies on the determinants of and interventions for intimate partner violence (IPV) in American Indian and/or Alaska Native (AI/AN) communities. Using a social ecological framework that identified proximal and distal risk factors, they described multilevel interventions that addressed a variety of determinants including age, gender, socioeconomic status, alcohol, European colonization, and infrastructure. “While the social ecological framework makes intuitive sense, there are few multilevel interventions to address IPV and none in AI/AN communities.” They offered no quantitative evidence of impact.

Secondary Prevention (Stopping the Risk Factors Becoming Violence)

Qualitative studies and reviews on secondary prevention

Holkup et al. (2007) described the Family Care Conference (FCC), an elder-focused, family-centred, community-based intervention for the prevention and reduction of elder abuse. The intervention involved inviting family members, family-nominated supportive community members, a spiritual leader, and relevant health and social service providers to attend a meeting in which individuals bring forward their concerns about the welfare of the elder. The authors claimed that families accepted and appreciated their intervention, and attributed its success to the community’s long history of respect for elders and preference for mediation over confrontation:

Drawing on the values of interdependence and reciprocity among Native American kin, the FCC provided a culturally anchored and individualized way to identify a frail elder’s care needs and to find solutions for meeting those needs from among family members and available community resources.

The report did not provide quantitative measures of impact in reducing elder violence.

Norton and Manson (1997) described a family violence program in an urban US Indian health centre that provided a range of services for American Indians including counselling, crisis housing, and transportation to appointments. To increase participation, the program used home visits and a weekly family violence group that incorporated traditions and values (talking circles, sharing of meals). The home visits tended to build an emotional connection with the counsellor. A weekly domestic group provided an informal atmosphere for sharing problems and concerns over a potluck dinner. Though not supported by numeric data, the authors reported that

Many of these women were able to build a relationship with the counsellor, and benefit from active participation in a family violence program. Home visits significantly enhance care and the effectiveness of counselling.

Willmon-Haque and Bigfoot (2008) reviewed the published literature on poverty and historical trauma, including a discussion on oppression and hegemony (authority and power). They described research on violence and the resulting trauma, suicide, family violence, and post-traumatic stress disorder. Since very young children can also be affected by traumatic events, the authors called for early intervention and awareness of family violence. Recent efforts to influence mental health care provided to AI/AN children and families included: training in advocacy and community mobilization, increasing community capacity, promoting culturally relevant services, enhancing knowledge and awareness on the issue, and involving communities in research.

Quantitative studies of secondary prevention

Brewin and Coggan (2004) evaluated the Ngati Porou Community Injury Prevention Project (CIPP) in a predominantly Maori rural community on the east coast of the North Island of New Zealand. Participants included Maori selected at a forum (tribal gatherings, language nests, and sports venues). The CIPP used the World Health Organization (WHO) Safe Community Model for injury prevention, which acknowledges that community injury problems are best addressed by those who live in that community. Two activities addressed family violence: 1) identify and affirm traditional Ngati Porou care and protection practices for whanau (families) through 22 education sessions, information packs, and ongoing training to resource people; 2) coordinate and implement regional programs to address family violence and child abuse to tie in with national campaigns. The initiative included a gathering on emotional and physical violence, and a family violence-free concert to raise awareness. Community surveys indicated that the proportion who encountered threatening situations outside the home decreased (pre 73%, post 66%) and there was a small increase in those who always/mostly walk away from a threatening situation at home (pre 27%, post 31%). There was a significant increase in awareness of injury prevention among Ngati Porou whanau (pre 17% and post 25%).

Douglas (1998) considered an intervention to restrict availability of alcohol in a small town in Western Australia (population 1,200); 63% were Aboriginal. Other programs implemented along with the intervention included a school education program. Overall, the incidence of crime declined. Alcohol-related presentations to the hospital and presentations resulting in family violence decreased relative to the period prior to the intervention (odds ratio 0.43, 95% CI 0.31–0.60). The authors reported no change in domestic violence.

An RCT by Duggan and colleagues (2004) assessed the impact of home visits in the prevention of child abuse and neglect in the first three years of life in families identified as at-risk of child abuse through population-based screening at the child’s birth. The study focused on Hawaii Healthy Start Program (HSP) sites operated by three community-based agencies. The intervention had two parts: (1) screening of medical records by paraprofessionals to identify indicators of families at risk of child abuse and neglect; and (2) home visiting of these identified at-risk families. The home visiting aimed to prevent child abuse and neglect by improving family functioning in general and parenting in particular. Home visitors were paraprofessionals working under professional supervision. Home visits included both direct services and linkage with community resources. Direct service included providing emotional support to parents, encouraging them to seek needed professional help, teaching about child development and role-modelling parenting skills and problem-solving techniques. Intervention group mothers were less likely to use common corporal/verbal punishment but this was attributable to one agency’s reduction in threatening to spank the child. Intervention group mothers also reported less neglectful behaviour, related to a trend toward decreased maternal preoccupation with problems and to improved access to medical care for intervention families at one agency. The intervention and control groups were similar on most measures of maltreatment. The program did not prevent child abuse or promote nonviolent discipline. From the reader’s viewpoint, it seems the intervention could be tightened up considerably; HSP records rarely noted home visitor concern about possible abuse.

A second RCT by Duggan et al. (2007) assessed the impact of a para-professional home visiting program in preventing child maltreatment and reducing parental risks for maltreatment. This focused on six Healthy Families Alaska (HFAK) programs aimed at preventing child maltreatment by promoting positive parenting and child health and development. There was no overall program effect on parental risks for child maltreatment or on outcomes for families with high dose home visiting. Home visited mothers did report using mild forms of physical discipline less often than did control mothers, although the groups were similar in their use of more severe forms of physical discipline. Mild physical assault of the child was less common in the HFAK group only in the subgroups of multiparous mothers and for those not in a violent relationship at baseline. These positive findings must be balanced against deficiencies of the study: home visitors discussed parental risks for maltreatment in less than one half of active, risk-positive families; most risk-positive families were not referred to community services raising an issue of viability of the intervention.

The systematic review by Whitaker et al. (2006) included one study involving First Nations youth (Wolfe et al. 2003). This dating violence prevention program included middle or high school students with behavioural outcomes including conflict in adolescent dating relationship inventory (perpetration and victimization), healthy relationships skills, adolescent interpersonal competence questionnaire (emotional support, negative assertion, self-disclosure, conflict management, conflict resolution, trauma symptoms (trauma symptom checklist) and hostility. Interventions consisted of education about healthy and abusive relationships, conflict resolution and communication skills, and social action activities. Only 8% of participants were First Nations youth; the positive results declared by the study may also apply to them, but this is not presented in subgroup analysis.

Tertiary Prevention (Reduction of the Worst Effects or Recovery from Family Violence)

Qualitative studies on tertiary prevention

Heilbron and Guttman (2000) report interviews with five women exposed to traditional Aboriginal healing practices and the sharing of beliefs in the therapy process for First Nations and non-Aboriginal women survivors of child sexual abuse. A counselling group combined a traditional “healing circle” with conventional cognitive therapy. The intervention involved weekly group meetings over a ten-week period for approximately two hours per session. The healing circle provided a spiritual framework for addressing problems in the group, and may have influenced participants in several ways: spiritual focus heightened participant motivation for confronting issues; participants felt empowered to speak; they felt comfortable and supported; the ceremony was conducive to encouraging a warm, caring, and relaxed environment; and a safe, supportive and spiritually nurturing environment was encouraged. The sharing of Aboriginal teachings and stories in the group may have provided personal meaning for the First Nations women and enabled them to develop a stronger devotion to Aboriginal healing methods:

At first she blamed herself for the abuse that occurred. It was through aboriginal teachings that this woman learned she was not to blame.

Mokuau (2002) examined culturally based interventions for Native Hawaiians that incorporate cultural values and practices, and involve Native Hawaiian participants in the design and delivery of the interventions. This study described two culturally based interventions: The first, ho’oponopono (“to set right”), was a family-focused holistic approach for maintaining and restoring relationships with family members and others in a spiritual realm. It was facilitated by a respected elder, family member, or community leaders, and included a spiritual component. The second intervention, aloha ’aina (“caring for the land”) involved activities that reflect the Native Hawaiian connection to the land. For example, working in the taro field entailed physical discipline, cognitive attention, emotional reflection, and spiritual openness. While working, participants learned about the cultural past through storytelling by elders and other caregivers. Participants learned values such as cooperation and reciprocity, engaged in self-reflection on cultural identity and cultural pride, and explored their spirituality:

Native ancestors have left the United States a legacy of cultural values and traditions that can have a powerful impact on resolving some of the issues of child abuse.

Luna-Firebaugh (2006) reported the evaluation of the Services-Training-Officers-Prosecutors Violence Against Indian Women (STOP VAIW) Program, a partnership between the tribal justice system and the nonprofit, nongovernmental service programs. Indian tribal governments received financial assistance to implement strategies to address violence against Indian women, and to develop and enhance services provided to Indian women who are victims of violent crimes. Tribal approaches include coordination of services to ensure better response to violence against women; enhancement of law enforcement response; innovative approaches to tribal prosecution; new approaches to shelter; and implementation of the full faith and credit provision of the Violence Against Women Act (with reference to protection and restraining orders). The authors claimed that STOP VAIW programs had a significant impact in Aboriginal communities. The program helped raise awareness among tribal leadership and the community, as well as promote various approaches to confronting the problem of violence against women. Most tribal communities felt that it was essential to combine “culturally compatible law enforcement and prosecution components with victim’s assistance and outreach.” Tribes and tribal officials, program personnel, and women were reportedly empowered by involvement in the development and enhancement of tribal programs that address the safety of Indian women. This supported tribal sovereignty. Across all programs, the authors claimed that the development of coordinated community-wide responses to the problem of family violence was successful in addressing family violence, sexual assault, and stalking. Other results they claimed included:

• Tribal police officers and their departments were more efficient in their police work and more thorough in their investigations.

• Fifty-seven tribes developed mandatory arrest policies that included a holding period in their family violence codes and protocols.

• With respect to protection orders, 30% of tribes implemented a no-drop policy for the prosecutor, and 50% allowing for victimless prosecution to occur.

• Calls for service and the number of arrests increased by more than 10 times during the years 1995–1998. Family violence complaints increased more than 400% during that time period.

• A number of tribal programs entered into cross-jurisdictional agreements with surrounding law enforcement agencies, while others developed protocols and ordinances that eased jurisdictional issues.

Quantitative studies of tertiary prevention

Becker and colleagues (2008) reported a methodologically constrained trial; only one half of the participants completed the baseline questionnaire and the authors only analyzed those who completed the treatment. They examined outcomes of a 12-week “culturally influenced” group intervention for children and adults exposed to family violence. The intervention involved 90 minutes of support through a psycho-education children’s group provided — Haupoa meaning “make the ground soft for planting.” A simultaneous weekly support and education group exposed parents to what their children were learning and addressed parenting in the aftermath of family violence. Notwithstanding the methodological flaws, the authors claimed statistically and clinically significant improvements in children and parent violence-related skills. Children receiving the intervention showed improvements in internalizing and externalizing difficulties despite the lack of direct symptom-focused intervention in these domains. More than one-half (53.2%) of children exhibited clinically significant internalizing difficulties at pre-intervention, compared to 21.3% at postintervention. Participants were of diverse ethnic backgrounds, most reporting Asian or Pacific Island heritage; this cultural diversity raises questions about how culturally appropriate the intervention could have been.

Discussion

Few studies relate to Aboriginal and indigenous communities and those that do exist contribute only weak empirical evidence. We could not identify quantitative evidence of primary prevention (reduction of risk factors for family violence). Most authors equate tertiary prevention (recovery and reduction of the worst consequences) with prevention of family violence. Our extensive search retrieved a very small number of published reports of controlled studies.

In considering protective factors against suicide of Aboriginal youth, Chandler and Lalonde (1998) considered that restoring and rebuilding community culture could be protective. The concern for “cultural continuity” resonates with almost all the interventions described in this review, which took into account Aboriginal beliefs and values, holistic concepts of health, spirituality, traditional ceremonies, and healing practices. Many authors on interventions to reduce family violence recognize that strong social support networks should underwrite prevention efforts and change community perception about the issue of family violence. Authors also recognized that community involvement in the development, design, and implementation of family violence interventions should empower the community, make the interventions better received by its members, and contribute to the overall success of the intervention.

Despite this almost uniform recognition of the importance of cultural relevance and participation, there is not much hard evidence of impact in reducing family violence in Aboriginal communities. We see at least two reasons for this. First, the current literature rests on qualitative approaches that generate insights into cultural dynamics and participant perspectives. These studies can be pivotal to understanding what might work and how it might work (Carriere, 2007; Olesen, 1994). With this now in hand, we look forward to a next generation of research that start from these qualitative studies, going on to quantify the impact of interventions that reduce family violence.

Second, mirroring the investment in nonindigenous communities in Canada and internationally, most family violence research focuses on tertiary prevention (prevention of recurrence). There is a growing body of work on secondary prevention (reduction of violence outbreaks where the risk factors exist). We found only two narrative reviews relevant to if not actually directed at primary prevention, the prevention of emergence of risk factors that could lead to family violence. Yet this is exactly the upstream prevention where indigenous strengths are mostly likely to play out. Two programs in Africa emphasized the importance of participatory learning and structural interventions that change the position of women — and therefore, their risks to family violence.

Conclusion

The causes of family violence are complex and deeply rooted. Once established, the cycle is difficult to break. Interventions most likely to be effective are those designed to prevent family violence rather than, once established, to reduce its frequency and severity.

This review highlights the need for high quality research to inform interventions that reduce family violence in Aboriginal communities. Barriers to prevention in this context include the instability of programs, funding, and lack of capacity. Even with a potentially effective intervention, these implementation factors will need consideration. Future research priorities include the development and implementation of evidence-based interventions tested in pragmatic randomized controlled trials.

Biographies

Bev Shea has a PhD in epidemiology and is a researcher at Community Information Epidemiological Technologies (CIET) and co-director of the Anisnabe Kekendazone (AK) Network Environments for Aboriginal Health Research (NEAHR), administered by CIETcanada. Over the past few years she has been working in Aboriginal health and international research leading to the production of systematic reviews and in the areas of HIV/AIDS and childhood immunization. She has contributed to the creation of new knowledge in the field of quality assessment and qualitative data analysis. During the past ten years she has worked in improving the field of systematic review methodology. She also mentors in systematic review courses through CIET and the University of Ottawa.

Amy J. Nahwegahbow is a research coordinator with Community Information and Epidemiological Technologies (CIET). Amy is a member of the Whitefish River First Nation and a graduate of Trent University in the Honours Bachelor of Arts Program with a major in Native Studies. Since 1999, she has worked for National Aboriginal Organizations such as the National Association of Friendship Centres, the National Aboriginal Health Organization and the Assembly of First Nations on a variety of First Nations health issues such as youth at risk, suicide prevention, injury prevention, research ethics, environmental health and public health issues of concern to First Nations. She has developed several toolkits for First Nations communities on ethical research practices, the principles of OCAP, and Aboriginal health indicators. Amy brings a broad knowledge of First Nations health and wellness issues, as well as experience in conducting on the ground research with communities. She is pursuing her education and a life long career in epidemiology to help improve the health, social, and personal prospects of indigenous communities.

Dr Neil Andersson is the executive director of CIET and adjunct professor in the Faculty of Medicine at the University of Ottawa. He has three decades of experience designing, implementing and managing evidence-based health planning initiatives. A medical epidemiologist, for the last 15 years he has supported training of researchers in more than 200 Canadian First Nations, Métis and Inuit communities.

www.ciet.org