• Open Access

Partner violence prevalence among women attending a Maori health provider clinic


Dr Jane Koziol-McLain, Auckland University of Technology, Private Bag 92006, Auckland, New Zealand. Fax: +64 9 921 9796; e-mail: jane.koziol-mclain@aut.ac.nz


Objective: To determine partner violence rates among women attending a general practice in Aotearoa, New Zealand.

Methods: This descriptive study was conducted in a hauora (Maori health provider general practice clinic) in one South Auckland community. Non-acute, English-speaking women who entered the huaora during 30 randomly selected clinic sessions in a five-week period in 2003 were eligible to participate. Research assistants (RAs) verbally administered a structured, brief questionnaire that included a partner violence screen (past 12 months), assessment of high danger risk, and lifetime prevalence. Of 148 women approached, 109 participated. Participants generally self-identified as Maori (74%) or New Zealand European (18%) and ranged in age from 17 to 82 years (mean 38.8).

Results: Twenty-three per cent (95% CI 15–31) of women screened positive for partner violence. Among the 25 women who screened positive, six (24%) had one or more high danger risk factors and 24 (96%) reported one or more children living in the household. Seventy-eight per cent (95% CI 70–86) of women reported a history of partner violence.

Conclusions: In this sample of mostly Maori women, direct partner violence questioning in a general practice setting yielded a high disclosure rate. Three out of four women disclosed violence by a partner; nearly one out of four disclosed violence by a partner in the past year.

Implications: Healthcare providers have the opportunity to identify and provide services to women and their children experiencing partner violence. Health care providers and the health care system also have a responsibility to join with the community in calling for non-tolerance of family violence.

Family violence is not a new phenomenon, however reluctant people have been to talk about it in the past. Nor is it a problem confined to a few “disturbed” families on the fringes of society. It is a problem that affects the family life of many people, causing distress for individuals and far-reaching consequences for our society.

—J. L. Robson, Social Development Council, 19801

Partner violence against women is recognised to have a significant social, economic and health toll internationally as well as in Aotearoa, New Zealand.2–7 The identification of family violence as a public health problem – one that can be prevented – has prompted numerous health policy documents in New Zealand over the past decade.8–12 Reducing interpersonal violence is one of 13 Ministry of Health public health priorities.13 The Ministry participated in the 2002 government family violence prevention strategy Te Rito14 and the subsequent Taskforce for Action on Violence within Families.15 Maori policy documents that have addressed family violence include Maori Family Violence in Aotearoa16 and the 2002 He Korowai Oranga: Maori Health Strategy.17

Surveillance and risk factor identification have been the focus of large population-based studies in New Zealand.18–20 Within the 2001 national crime survey, based on computer-assisted self-interview by more than 2,500 ever-partnered women, 21% reported they had been deliberately physically assaulted (see footnote 1) by a heterosexual partner at some time in their life; a small proportion (2%) reported a physical assault in the most recent year.19 Lifetime prevalence of physical assault varied significantly by ethnicity: 17% for Pacific women, 20% for NZ European women, and 42% for Maori women. In a more recent population-based study of more than 2,500 ever-partnered women between 18–64 years of age from two regions, 33% (Auckland) and 39% (Waikato) reported physical and/or sexual violence by a partner in their lifetime; 5% reported physical and/or sexual violence in the past year (see footnote 2).18

In addition to population-based figures, rates of partner violence among women in the healthcare setting are similarly important to inform health strategies. International research indicates that rates of partner violence among women in healthcare settings are higher than population-based figures.21 In the Aotearoa context, the over-representation of Maori women in partner violence statistics may translate to the healthcare setting. The purpose of this study was to document the prevalence of partner violence among women seeking health care in a Maori primary care setting, providing screen positive (past year), high risk and prevalence (lifetime) rate estimates. This information is necessary to inform the healthcare system response to family violence. The study also provided information about implementing a routine screening and brief intervention process for Maori women.


This descriptive study measured screen positive (past year) and prevalence (lifetime) partner violence rates among women in a Maori health provider general practice. It was conducted in a South Auckland Raukura Hauora O Tainui hauora (health clinic) that served approximately 4,000 enrollees. During the data collection period, hours of operation were 9am to 4.30pm Monday through Friday; the hauora was staffed by two doctors, two nurses and one receptionist.

Research Team Kaupapa (philosophy)

This study was conducted by a bi-cultural research team. While it does not represent kaupapa Maori research (research by Maori for Maori),22 we attempted to educate ourselves about things Maori23 and to use processes aimed at reducing the institutional hegemony evident in our institutions and research methodologies.22 The principles of the Treaty of Waitangi – partnership, protection and participation – were considered throughout the project. Our partnerships with Kawa Whakaruruhau Komiti (University committee to monitor things Maori) and clinical (Raukura Hauora O Tainui) and community (South Auckland Family Violence Prevention Network) agencies began early in the process. The safety and well-being of women and children were paramount in our planning and conduct of the study. Our team kaupapa called for respecting all women, whakawhanaungatanga (connectedness) and forging a path for when the research was over. We also had an appreciation for time (being willing to take the time to listen to what women have to tell), an orientation affecting the quality of the interaction more so than the actual time taken.

Our processes called for bringing kai (food) whether we were meeting with our partners or collecting data with women in order to reciprocate their good faith and gift of knowledge. We sought to have discussions with our partners kanohi ki te kanohi (face-to-face) whenever possible. We were challenged during a pilot run in the hauora when a daughter of a client requested she be included in the study. Our team acknowledged that the hauora was there to serve the community, regardless of whether someone was presenting to the hauora on that occasion as a client or not. We therefore adapted our target population to all women who crossed the threshold of the hauora.

Sampling plan

This was an ‘all comers’ study that mimicked a partner violence routine inquiry policy. All women, both clients and support persons, who entered the hauora during 30 randomly selected morning or afternoon hauora sessions during a five-week period beginning in October 2003 were eligible to participate. We included women regardless of sexual identity and partner history, excluding only women who were acutely ill, non-English-speaking, or those who had already been entered in the study. The number of shifts selected was based on the hauora's usual clinic volume with the desire to enter 162 women (providing 95% CI ± 5% on the screen positive point estimate based on expected rate of 13%). Clinic staff and clients did not have foreknowledge of the selected shift schedule. The study was approved by the Auckland Ethics Committee and the Raukura Hauora O Tainui Clinical Governance Board.


Study procedures were similar to those used in a previous study measuring partner violence prevalence in an emergency department (ED) setting24 and consistent with Ministry of Health guidelines and resource documents.9–11 Research assistants (RAs) were nurses who had attended a six-hour training session; most had also collected data in the ED prevalence study and none were employed at the study site. Training, devised by our bi-cultural team and informed by the literature and practice experience, included a discussion of the socio-cultural context of family violence for Maori women and the role of colonisation.16,25,26 Two of six research assistants were Maori. Interactions with women were based on an empowerment framework27 focusing on positive coping strategies and women's strengths; acknowledging the goal as living violence-free, not necessarily leaving; and that achieving the goal is a process rather than an event.

A log was kept of all women who entered the hauora during selected shifts to allow calculation of response rates. RAs described the study with women privately, then verbally administered the study questionnaire to those who consented. Women could choose to have a support person present during questionnaire administration. The information sheet included a commitment that the RA would involve the hauora staff if she identified a risk of imminent serious harm. All women were offered a partner violence brochure and local family violence specialist referral number whether they chose to participate in the study or not, or screened positive or negative. Referrals included Maori provider services. The screening and intervention procedure is shown in Figure 1. Women who screened positive were given supportive messages followed by an assessment for high danger risk. Women who responded affirmatively to any of the high danger risk questions were then involved, along with hauora staff, in developing a safety and referral plan that may have included a referral to the community specialist family violence agency; the police; or Child, Youth and Family (New Zealand Government child protection agency). The principal investigator made routine visits to the hauora during which she monitored the data collection process (speaking with hauora staff and RAs regularly) and research documentation (shift log sheets, consent and data collection forms).

Figure 1.

Study procedure algorithm.


A standardised data collection form was used to collect screen and lifetime partner violence information, high-risk indicators (for women who screened positive) and selected demographic variables (such as age, self-identified ethnicity and children living in the household). The questionnaire was the same used in the ED prevalence study.24 Brief screening questions included: “Within the last year, have you been hit, slapped, kicked or otherwise physically hurt by someone” (if yes, “by whom”), “Within the last year, has anyone forced you to have sexual activities you didn’ t want” (if yes, “by whom”) and “Is there a current or past partner that is making you feel unsafe?” If a woman responded “yes” to one or more of the three questions, referenced to a current or past intimate partner, she was considered to have a “positive screen”. The same questions were asked replacing “within the last year” with “in your adulthood, since you were 16”. Lifetime prevalence was a “yes” response to one or more of the three questions. High danger risk questions included: “In the past three months has your partner (or ex-partner) threatened to kill you or someone in your family such as the children?”, “Are there children or elderly in your household who are in danger of being harmed?” and “Is it safe for you to go home?” and “Are you thinking of harming yourself?” Data were entered in SPSS and following accuracy checks, partner violence rates (crude) were determined with 95% confidence intervals, then compared by ethnic identity and age.


One hundred ninety-seven women entered the hauora during selected shifts. Some women were ineligible (n=28); a research assistant (RA) missed contacting some women due to periods of high clinic volume (n=28; see Figure 2). Among 148 women approached by a RA, 109 consented to participate. The participation rate (for women who were approached) was similar for hauora clients (72.5%) and support persons (73.6%). Some women gave a reason for not participating: “not enough time” was the most common, several women said, “I want to leave it [partner abuse] in the past”.

Figure 2.

Sample recruitment.

Three-quarters (74%) of participants self-identified as Maori; 18% as New Zealand European; and 8% Samoan, Cook Island Maori, Nuiean or Indian. Participants ranged in age from 17 to 82 years (mean 38.8 years); most (62%) were less than 40 years of age. Seventy-two per cent of women reported that there were one or more children or young persons (less than 17 years) living in their household.

Twenty-three per cent (95% CI 15–31) of women screened positive for partner violence, with physical violence and feeling unsafe most commonly reported (see Table 1). Thirty-two per cent of women who screened positive responded affirmatively to two or more of the questions (physical, sexual, unsafe). Partners were most often identified as ‘boyfriend’; no one reported an abusive lesbian relationship. Among the 25 women who screened positive, six (24%) reported one or more high danger risk factors: five reported that there were children or elderly living in the household who were in danger of being harmed, four reported their partner had threatened to kill her or someone in the family in the past three months, two stated they were thinking of harming themselves and one felt it would not be safe to go home. Only one screen positive woman associated her hauora visit to her experience of partner violence. Nine (36%) of the screen positive women were separated or in the process of separating from their partner and almost all (n=24, 96%) reported that there were one or more children living in the household.

Table 1.  Partner violence among women in a Maori health provider clinic (n=109).
 n%(95% CI)
Screen (past year)
 Physical violence1513.8 
 Sexual violence21.8 
 Feeling unsafe1614.7 
Lifetime (adulthood)
 Physical violence7568.8 
 Sexual violence3330.3 
 Feeling unsafe6156.0 

Seventy-eight per cent (95% CI 70–86) of women reported a history of partner violence (see Table 1). Physical violence and feeling unsafe were most common (69% and 56% respectively), but a significant proportion of women (30%) reported forced sex by a partner. A majority (68%) of women who reported a history of partner violence responded affirmatively to two or more of the three questions (physical, sexual, unsafe).

We examined prevalence by hauora status (client or support person), age and ethnic identity (see Table 2). The sole statistically significant association was that younger women were more likely to screen positive for partner violence. While there was a trend for Maori women to be more likely to screen positive, the relationship was not statistically significant and there was no association between partner violence prevalence and ethnicity.

Table 2.  Partner violence prevalence by clinic status, age and ethnicity.
 nScreen positivepAdulthoodp
  1. Note:

  2. Number total less than 109 due to missing values (four for age and three for ethnicity).

Clinic status  0.57 0.47
 Client7921.5% 79.7% 
 Support person3026.7% 73.3% 
Age group (years)  0.02 0.17
 16–241435.7% 78.6% 
 25–395133.3% 88.2% 
 40–592910.3% 72.4% 
 60+110% 63.6% 
Ethnicity  0.39 0.99
 Maori7826.9% 79.5% 
 New Zealand European1915.8% 78.9% 
Pacific Island or Indian911.1% 77.8% 


In this all-comers study of women in a South Auckland Maori health provider general practice clinic we found a partner violence screen positive rate (23%) similar to the rate among women surveyed in an emergency department in the same region (territorial authority) using the same study procedures (21%).24 Three out of four (78%; 95% CI 70–86) women had experienced violence by a partner ever, higher than the proportion reported among women in the emergency department study (44%; 95% CI 37–52).24 That three out of four women reported partner violence confirms that for women attending the hauora, whether Maori or non-Maori, partner violence is a common experience.

The partner violence prevalence rate in this study was higher than that found in the emergency department24 as well as in other Australasian general practice studies. Partner violence rates, however, are influenced by both research design and sample characteristics. With regard to design, definitions (physical, sexual and psychological), selected screening items (and scales), referent partnerships (current or any) and time periods (12 months or ever), data collection methods (e.g. written, computer, face to face), interviewer (e.g. research staff or clinician) and the context in which the questions are asked (e.g. family violence specific or general health risks) are all likely to influence reported prevalence rates.28 For example, in a survey of 1,836 women attending 20 general practices in Brisbane in 1996, 37% of the women reported having “ever been abused in an adult intimate relationship”.29 On the other hand, among 2,543 New Zealand general practice patients (two-thirds of the sample were women) who completed a self-administered, written general health risk screen, only 5% answered affirmatively to the question “Is there anyone in your life whom you are afraid of, who hurts you in any way or prevents you doing what you want?”30

Prevalence rates higher than 50% among women in primary care are not unprecedented. In a South Carolina (United States) study, 1,401 insured women 18–65 years of age were interviewed in university-affiliated family practice clinics using an established screening tool and two standardised instruments;31 55% reported a lifetime history of partner abuse. In another study, 312 Native American women in a tribally operated Women Infant and Children's Nutritional Program (WIC) clinic in Oklahoma completed a self-administered survey that included a modified Conflict Tactics Scale;32 59% reported lifetime physical or sexual partner abuse.33

While our statistics are sobering, there is evidence that many women do successfully make the transition from living with violence to living violence free.31,34 Three-quarters of the women in this study had experienced violence, yet only one-quarter (23%) had experienced it in the past year. The high proportion of women who screened positive that were separated or in the process of separating from their partner (36%) would indicate that leaving the relationship is often necessary. This has important implications because separating from an abusive partner is an especially high-risk time.35,36 Whether women continue to be exposed to violence or not, evidence suggests that they will suffer long-term ill-health effects such as chronic pain and depression.5,6,37,38 Their children are also at risk for emotional and behavioural problems, as well as being at higher risk for being abused themselves.39–43 In the current study there were 52 children living in the households of the 25 women who screened positive for partner violence.


The findings of the study should be interpreted with caution. This study was conducted in a single hauora located in a community that is disadvantaged – based on the 2001 Census (http://www.stats.govt.nz) – in education (16% of people over 15 years had post-school qualification, compared with 27% in the region and 32% for New Zealand); income (8.5% of persons in the community had incomes above $40,000 compared with 18% for the region and 19% for New Zealand); and employment (80% employed, compared with 90% for the region and 93% for New Zealand). The deleterious effects of colonisation (including urbanisation and racism) for Maori,16,25,26,44,45 and immigration for Pacific peoples46,47 are also likely to contribute to higher partner violence rates in the study community. Because we sampled a single hauora and did not collect individual (such as socio-economic status, partner employment and substance misuse3,48,49) or macro-level correlates in order to keep our survey brief,50 we are unable to generalise our study findings. In addition, our sample was small (95% CI ± 8) and our participation rate (74% among contacted eligible women) leaves room for error. However, if women who refused to participate had all either experienced partner violence or not, the prevalence rate, 84% and 57% respectively, would still indicate that for women seeking care in the hauora, violence exposure is significant.


We found that a large proportion (74%) of women in an Auckland general practice hauora were willing to answer sensitive questions regarding partner violence. The rates of both screen positive (23%, past year) and lifetime (78%) exposure to partner violence were significantly higher than those reported in population-based studies for both Maori and non-Maori women. These findings from a hauora, along with our previously reported emergency department findings,24 lend support to healthcare settings being an important site for violence identification, intervention and prevention. It must be acknowledged that while the prevalence of partner abuse among women in healthcare settings is significant, there is insufficient evidence supporting intervention effectiveness.51–54 Sugg, however, cautions us against inactivity, saying at a Center for Disease Control sponsored workshop: “Are we really doing what we should do? That's my question, not should we do it all.”55


“Dispelling the illusion that whânau violence is normal and acceptable” is a fundamental task in the Transforming Whânau Violence26 framework. In a recent World Health Organization multi-country study, 86% of physically abused women surveyed in Samoa who did not seek formal help believed their abuse was “normal or not serious”.3 Healthcare providers and the healthcare system have the opportunity – and the responsibility – to join with the community in responding: “No, violence against women is not normal; and it is serious.”

The healthcare system can be a positive force in changing our society from one that colludes with violence against women to one that does not tolerate it. The status quo, reflected by health professionals’ hesitancy to screen56 and judgemental, victim-blaming attitudes when they do,57 is an action in itself. Although perhaps a subconscious response, it allows perpetrators to continue to wield power and control, violating women's human rights, all behind a wall of silence. As Klein states: “Whether there is room for neutrality is debatable. Doing nothing may often function as de facto support of the perpetrator.”58 As with other marginalised ills (social taboos) such as HIV, for violence against women, “silence = death”.59 This is certainly the case for far too many women in Aotearoa; in one recent six-week period alone, six women were murdered by their partners.60 It is time to move beyond a singular focus of identifying women who are or have experienced abuse to becoming part of a co-ordinated community response61,62 that challenges the attitudes and structures that sustain family violence. At risk is that another 25 years will pass and J. L. Robson's quote1 will continue to be resonant.

While the adulthood prevalence rates in our study were equally high among Maori and non-Maori, that does not mean the solutions should be the same. Westernised models seek to empower the victim and criminalise the batterer within an ethos of individualism, although it must be said with limited success.53,63 In contrast, Maori initiatives such as Project Mauri Ora and Amokura64 seek whanau, hapu and iwi well-being within an ethos of whakapapa (ancestral identity), more consistent with a community (versus individual) model of intervention.26,62 Aotearoa New Zealand family violence public health initiatives guided by an explicit kaupapa and acknowledging our bi-cultural context are likely to serve us well as we increase awareness of violence against women and its link to ill-health, improve healthcare services for those affected by violence, collaborate with community-based programs and together defend our human right to live violence-free.


We thank the women who exhibited generosity, grace and courage in sharing their experiences of violence in order to help others. We also thank our team members from Raukura Hauora O Tainui, Trish King, Tane Taylor, Dale Te Iwimate, Katrina Wahanui and all the hauora staff; from South Auckland Family Violence Prevention Network, Rosemary Nash and Chislane Theodore; and from Middlemore Hospital Emergency Care Department, Pam Batty. We also thank research assistants Puti Nichols and Carol Fryer. Funding for this study was provided by the Auckland University of Technology Contestable Research Fund and the Ministry of Health.


  • 1

    While a more inclusive ‘physical violence’ variable is provided in the report, the item limited to a physical assault is reported here to facilitate comparison across studies.

  • 2

    Rates by ethnicity have not yet been published.