Longitudinal Changes in Self-Efficacy,Mental Health,Abuse,and Stages of Change,for Women Fearful of a Partner: Findings From a Primary Care Trial (WEAVE)

Abstract

Women seeking healthcare while experiencing intimate partner violence (IPV) often report a mismatch between healthcare received and desired. An increase in detection of women experiencing IPV through routine screening has not consistently shown a parallel increase in uptake of referrals or decreased abuse. This study investigates relationships between women’s stage of change (SOC), mental health, abuse, social support, and self-efficacy. This study used data from a randomized-controlled trial (RCT) of an intervention to improve outcomes for women afraid of their partners (n = 225; WEAVE). Women’s progress toward change was categorized into pre-contemplation/contemplation (pre-change SOC) or preparation/action/maintenance of change (change-related SOC). Characteristics of women ending the 2-year study in pre-change SOC were compared with those always in change-related and those ending in change-related SOC. Variables were analyzed using multinomial logistic regressions at baseline, 6, 12, and 24 months. Compared with women in pre-change SOC, women always in change-related SOC or ending in change-related SOC are significantly more likely to have higher levels of self-efficacy at 6 (AdjOR = 1.19, confidence interval [CI] = [1.08, 1.30]) and 24 months (AdjOR = 1.21, CI = [1.04, 1.40]). Women always in change-related SOC are always significantly less likely to live with an intimate partner. Women ending in change-related SOC are less likely to live with a partner at 12 (AdjOR = 0.30, CI = [0.12, 0.75]) and 24 (AdjOR = 0.22, CI = [0.06, 0.80]) months. Clinicians should focus on enhancing abused women’s self-efficacy, supporting them to create and maintain positive changes.

Keywords

domestic violence intimate partner violence self-efficacy depression mental health and Stages of Change

Introduction

Intimate partner violence (IPV) is defined as “behaviour by an intimate partner that causes physical, sexual or psychological harm, including acts of physical aggression, sexual coercion, psychological abuse and controlling behaviours. This definition covers violence by both current and former spouses and other intimate partners” (World Health Organization, 2013). IPV is a common form of violence against women, damaging the health and well-being of affected women, their families and communities (Garcia-Moreno et al., 2013; Garcia-Moreno & Watts, 2011). Women living with IPV seek health care more frequently than women in violence-free relationships but often report receiving unsatisfactory health care (Feder, Hutson, Ramsay, & Taket, 2006). A recent Cochrane review found that increased health care screening or detection of women experiencing IPV did not result in more women accepting referrals for specific IPV-related care (Taft et al., 2013).

The Transtheoretical Model and Stages of Change (SOCs)

The transtheoretical model, particularly the five SOCs (Prochaska, Redding, & Evers, 2008) component, provides a useful framework for understanding why some clinicians’ responses to disclosure may not be helpful for women experiencing IPV, and why women may not be ready to accept referrals when offered (Prochaska et al., 2008; Reisenhofer & Taft, 2013).

Within the SOCs, the first stage, pre-contemplation, is the point where an abused woman may not realize or acknowledge her relationship is abusive (Brown, 1997; Reisenhofer & Taft, 2013). Women in this stage may believe that what they are experiencing is normal; having grown up witnessing IPV or violence in relationships, a woman may not recognize abuse in her own intimate relationship. Alternatively, women may minimize the abuse or take responsibility for abuse, modifying their own behavior as the presumed cause, and thus not acknowledge IPV (Brown, 1997; Reisenhofer & Taft, 2013).

The second stage, contemplation, occurs when women identify and label their experiences as abuse and consider the pros and cons of change (Brown, 1997; Reisenhofer & Taft, 2013).

When women have made a decision toward change and commence preparations for change, they are considered to have moved into the third, preparation, SOC. In preparation, women undertake activities necessary to enact change. For example, women planning to end an abusive relationship may seek legal advice, police protection, or safe housing options, prepare documents, save money, or get a job/education. Women in preparation have a plan for change and are preparing themselves and/or their environment for change (Brown, 1997; Reisenhofer & Taft, 2013).

In the fourth stage, action, women enact their plan for change. Women may end an abusive relationship or seek assistance with their abuser to alter his or her behavior, for example, through the use of counseling or drug and alcohol services. Women may frequently move in and out of action (often back into contemplation or preparation) a number of times prior to successful transitioning to the final stage (Brown, 1997; Reisenhofer & Taft, 2013). In maintenance, the fifth and final SOC, women maintain change for a period of 6 months or more (Brown, 1997; Reisenhofer & Taft, 2013).

Within the five SOCs, there are 10 processes of change which describe thoughts/events often occurring at the different stages or enabling movement between stages. Table 1 provides an overview of the five SOCs and the 10 associated processes of change. This model may provide a framework for health care clinicians to consider when working with affected women. For example, a woman who is in early pre-contemplative SOC and does not recognize abuse may appreciate opportunities to explore positives and negatives in her relationship (consciousness raising), but be resistant to suggestions that she should report violence to the police. In contrast, a woman in a late contemplative stage may be ready to move into preparation/action and accept the offer of support to leave safely (helping relationships).

Table 1.

Stages and Associated Processes of Change.

Stage of Change	Description	Process of Change (Often Associated With Movement Between Stages; Burke, Denison, Gielen, McDonnell, & O’Campo, 2004; Prochaska & DiClemente, 1986)
Pre-contemplation	Unacknowledged abuse—Women may blame themselves for the abuse or not recognize/name abuse in their relationship	Consciousness raising (awareness of abuse and seeking information)
Contemplation	Abuse is named and women weigh up the pros and cons of change	Dramatic relief (speaking about the abuse)
		self-reevaluation and environment reevaluation (assessing the abuse and impact on self and environment—that is, on family)
Preparation	Women decide on and prepare for a course of action	Self-liberation (choosing and committing to change; increased self-efficacy for change)
		Helping relationships (trusting and accepting help from others)
Action	Planned action is implemented	Counter conditioning (establishing new behaviors)
		Reinforcement management (rewarding self for continuing new behaviors)
		Stimulus control (removing stimuli that might promote a relapse)
Maintenance	The planned action has been maintained for ≥6 months	Social liberation (new understanding of life away from abuse; often involves helping other abused women)

We propose that women in an abusive relationship may have three primary goals: (a) leaving an abusive relationship in an effort to escape the abuse; (b) remaining in the relationship where the abuser ceases to perpetrate IPV, and the woman and her family are safe; or (c) remaining within the relationship, knowing the abuse will continue but attempting to minimize harm and promote her personal/family well-being (Reisenhofer & Taft, 2013). Clinicians should identify which goal meets a woman’s personal needs when supporting her to achieve ongoing well-being and (ideally) safety.

The overarching goal is reducing women’s exposure to abuse, although achieving this is likely to be outside of her complete control. Thus, when considering outcomes of interventions, success may potentially be measured by women’s movement into, or remaining in, the later SOC, namely preparation/action and maintenance, rather than a reduction of ongoing abuse. Women’s movement toward later SOC is influenced by a variety of internal and external factors often weighed up by women as pros or cons of change and identified as their decisional balance (Bliss, Ogley-Oliver, Jackson, Harp, & Kaslow, 2008; Chang et al., 2006). Women may move through SOC in a non-linear manner, regressing or leap-frogging stages, using some or all of the 10 processes of change (see Table 1), as they take action to end abuse (Bliss et al., 2008; Chang et al., 2006; Reisenhofer & Taft, 2013).

Factors Influencing Women’s SOC

Many factors may influence women’s decisional balance, including her socioeconomic status, having children, perceived/actual social support, mental health status, level or type of abuse (Catallo, Jack, Ciliska, & MacMillan, 2012) and self-efficacy (Patzel, 2001). These factors may change over time, and there is often a turning point where the decisional balance shifts toward change (Chang et al., 2010).

Self-efficacy is defined as a “core belief that one has the power to produce desired effects by one’s actions, [without which] . . . one has little incentive to act or to persevere in the face of difficulties” (Benight & Bandura, 2004, p. 1131). For women who have experienced IPV, self-efficacy may be seen as the “personal strength” which allows them to work toward change (Cluss et al., 2006; Patzel, 2001). “Perceived self-efficacy is concerned with judgments of how well one can execute courses of action required to deal with prospective situations” (Bandura, 1982, p. 122), and those with a stronger sense of self-efficacy may be more likely to attempt change and persevere in the face of obstruction (Bandura, 1982). Self-efficacy may be enhanced as women take successful steps toward change (Patzel, 2001). Conversely, erosion of women’s self-efficacy may be associated with feeling unable to undertake or maintain change (Reisenhofer & Seibold, 2012; Shurman & Rodriguez, 2006).

Women experiencing abuse report higher levels of depression and anxiety compared with non-abused women (Hegarty, O’Doherty, Astbury, & Gunn, 2012; Trevillion, Oram, Feder, & Howard, 2012). Depression may have a variable effect on women’s change behaviors (Edwards et al., 2006). For some, acknowledging depression is a stimulus for change (Shurman & Rodriguez, 2006); however, depression may also decrease women’s energy for change. A diagnosis of depression, by clinicians unaware of or without acknowledgment of abuse, may reinforce a woman’s sense of responsibility, resulting in her trying to “fix” herself rather than acknowledging abuse (Reisenhofer & Seibold, 2012). Similarly, anxiety may paralyze women or push them toward change (Shurman & Rodriguez, 2006).

Evidence considering impact of abuse type and severity on women’s SOC is currently emerging (Haggerty & Goodman, 2003; Reisenhofer & Taft, 2013). Shorey, Tirone, Nathanson, Handsel, and Rhatigan (2013) find no significant differences in abuse type for women across the five SOCs supporting the earlier suggestion that SOC may be used as a measure for outcomes for abused women separately to experience of abuse. Other researchers have found that women who acknowledged worsening abuse (type or frequency) are more likely to create change as a result of changes within their decisional-balance assessment and move toward preparation/action (Bliss et al., 2008; Chang et al., 2010; Khaw & Hardesty, 2007).

Although limited information has been found exploring the link between women’s social support networks and their SOC, real or perceived social support may influence women’s relationship satisfaction, particularly in earlier SOC where subjective norms reflect those from women’s networks (Shorey et al., 2013). Women with higher levels of social support may have higher quality of life, and social support may buffer women against adverse effects of IPV including depression (Beeble, Bybee, Sullivan, & Adams, 2009).

Having children living at home has a variable influence on women’s SOC reflective of their decisional balance. Women may stay in earlier SOC as they work to maintain the family unit (Burkitt & Larkin, 2008; Zink, Elder, Jacobson, & Klostermann, 2004). Alternatively, having children at home sometimes pushes women to leave because of (increasing) risk of harm to children (Khaw & Hardesty, 2007; Patzel, 2001).

Given this information, this article investigates the relationship between women’s SOC over time and their mental health, experience of abuse, social support, self-efficacy and cohort allocation (control or intervention) using data from a randomized-controlled trial (RCT) of a clinical IPV intervention in primary care over 2 years (WEAVE; Hegarty et al., 2010).

Method

This article reports a secondary data analysis undertaken on a subset of participants in the WEAVE cluster RCT; the protocol (Hegarty et al., 2010) and primary outcomes have been previously reported (Hegarty et al., 2013; Hegarty, O’Doherty, Taft, et al., 2013).

Participants

The WEAVE study enrolled 272 women who screened positive to being fearful of their intimate partners in the prior 12 months, attending one of 52 general practitioners (GPs) in Victoria, Australia. The 225 of 272 women who completed >1 survey (baseline + one follow-up survey) were included in this analysis.

Measures

Depression and anxiety were assessed using the Hospital Anxiety and Depression Scale (HADS) where a cutoff score of 8 equaled probable/actual anxiety or depression (Fosså, Dahl, & Loge, 2003; Zigmond & Snaith, 1983). The nature and frequency of abuse were measured using the Composite Abuse Scale (CAS; Hegarty, Bush, & Sheehan, 2005). In this scale, IPV is present with CAS ≥7 (Hegarty et al., 2005), and types of abuse are measured in three subgroups; severe combined abuse (SCA; sexual/physical abuse and emotional abuse and/or harassment), physical abuse only, or emotional abuse and/or harassment only (Hegarty et al., 2005).

Social support was measured through the Oslo Social Support Scale (Meltzer, 2003). Scores of between 3 and 8 were classified as poor support, and a score of 9 or above classified as moderate/strong support (Meltzer, 2003).

Self-efficacy was assessed using the Generalized Self-Efficacy (GSE) scale (Schwarzer & Jerusalem, 1995), and a score <30 was used to determine lower than normal self-efficacy (Schwarzer, 2005). Self-efficacy was only measured at the 6-, 12-, and 24-month time points. Self-esteem using the Rosenberg Self-Esteem Scale was measured at baseline; however, participants identified difficulty answering these questions (making these results unreliable), and thus the Self-Esteem scale was replaced with the GSE from the 6-month survey. Due to the unreliable Self-Esteem scale results and as self-efficacy and self-esteem are different constructs, only self-efficacy scores from the 6, 12, and 24 time points are included in this analysis.

Procedures

The 52 GPs in the study were randomly allocated to either invention or control cohort, and the women enrolled in the study were allocated to their GP’s group (cluster; Hegarty et al., 2010). Women in the control group (n = 135) received “usual” care as required, whereas women in the intervention group (n = 137) were offered six free counseling sessions with their GPs. Intervention GPs received training on IPV, focusing on woman-centered care and motivational interviewing techniques (Hegarty, O’Doherty, Taft, et al., 2013). Women were sent postal questionnaires upon commencement and at 6-, 12-, and 24-month time points with written and telephone follow-up if surveys were not returned (Hegarty et al., 2010).

Developing SOC categories

The baseline questionnaire used a mixture of open-ended questions and the Domestic Violence Survivor Assessment (DVSA) tool; a validated tool designed to assess SOC in women who have experienced IPV (Dienemann, Campbell, Landenburger, & Curry, 2002). For the 6-, 12-, and 24-month time points, the questionnaire was changed, removing the DVSA tool and adding further open-ended questions (Table 2) as participants recorded difficulty answering DVSA at baseline. All questionnaires asked open-ended questions about women’s experiences, plans for change, and types of activities undertaken to keep themselves safe (see Table 2). The new open-ended questions were created based on qualitative comments participants made in response to the DVSA questions in the baseline survey. These additional questions were specifically developed to match the constructs of the DVSA tool.

Table 2.

Opened-Ended Questions Used to Assess SOC.

Describe any changes you have experienced in your intimate relationships in the last 6 months (such as separating, taking a break, becoming closer to your partner, or starting a new relationship?

How has fear affected your past/current intimate relationships?^a

How attached are you to the partner(s) who made you feel afraid?^a

What sorts of things influence (or influenced) you to stay in an intimate relationship where you felt afraid?^a

What sorts of things do (or did) you do to avoid feeling afraid^a

What are your thoughts on being able to take care of yourself and live without the partner(s) who made you feel afraid?^a

Are there other things that you have done in the last 6 months to keep you safe from your partner or ex-partner?

What, if anything, could make you feel safe from your partner or ex-partner in the future?

Making Change Questions at all time points

What do you consider to be the most stressful aspect of your life right now?

Thinking about the issues with your partner or ex-partner, have you thought about making any changes within the next 6 months?

Thinking about the issues with your partner or ex-partner, have you thought about making any changes within the next 30 days?

I wish that my partner (or ex-partner) . . .

I wish that health providers including my GP . . .

I wish that I . . .

I wish that my family and friends . . .

Note. SOC = stage of change; GP = general practitioner; DVSA = Domestic Violence Survivor Assessment.

Open-ended questions inserted in the 6-, 12-, and 24-month surveys to replace DVSA used at baseline.

A qualitative review of all questionnaires was undertaken to establish SOC at each time. To assess women’s SOC, we categorized women’s goals as either (a) leaving an abusive relationship or (b) remaining in the relationship but staying safe (Reisenhofer & Taft, 2013). As most women in abusive relationships try to limit harm from abuse (Burke, Mahoney, Gielen, McDonnell, & O’Campo, 2009), this was not used as a goal for change within the study when women expected the relationship to remain abusive. We reviewed the literature to identify dominant behaviors aligned to each SOC for women experiencing IPV reflective of the two nominated goals above (Reisenhofer & Taft, 2013). Using these behaviors, coding was piloted on 10 questionnaires by four authors (A.T., K.H., J.V., and S.R.) blinded to the others’ results. After piloting and minor amendments (Table 3), SR coded all women’s questionnaires. To monitor interrater reliability, a second researcher blinded to the initial result (A.T., K.H., or J.V.) assessed 15% of surveys. In the 22/142 instances where a different category was nominated, the two researchers reviewed the questionnaire to reach consensus.

Table 3.

Women’s Beliefs/Actions in the Five Stages of Change.

Pre-contemplation

“The woman is not aware that she has a problem or holds a strong belief that it is her fault” (Hegarty, Taft, & Feder, 2008, p. 378)

Denying the problem, defensive, accepting the batter’s definition of the situation and trying to change/modify her behavior to prevent abuse (Brown, 1997)

Feel hopeless for any kind of change and accept the status quo (Frasier, Slatt, Kowlowitz, & Glowa, 2001)

Thinking the abuse is love, believing the abuser cares about her (Burke, Gielen, McDonnell, O’Campo, & Maman, 2001)

Denying severity or minimizing the abuse compared with what is experienced by others (Haggerty & Goodman, 2003)

Contemplation

“The woman has identified a problem but remains ambivalent about whether or not she wants to or is able to make any changes”(Hegarty et al., 2008, p. 378)

Aware of the problem, reevaluating relationship but not committed to action (Brown, 1997; Haggerty & Goodman, 2003)

Increasing awareness of pros and cons of change (Zink, Elder, Jacobson, & Klostermann, 2004)

Abuse may or may not be disclosed—if not disclosed hints may be provided (Zink et al., 2004)

Indefinite plans to take action (in the next 6 months; Frasier et al., 2001)

Wishful thinking (I wish he would change, stop drinking, etc. or I wish I could get him to stop . . .; Frasier et al., 2001)

Preparation

“Some catalyst for change has arisen” (Hegarty et al., 2008, p. 378)

Pros for change outweigh the cons in the “decisional balance” (Haggerty & Goodman, 2003; Prochaska et al., 1994)

“Ready for action and seriously planning to change in the next month” (Brown, 1997)

Seeking legal assistance (Frasier et al., 2001)

Making a plan to leave (Frasier et al., 2001)

Looking for a job or new house (Burke et al., 2001)

Action

“Plan devised in the previous stage is put into action” (Hegarty et al., 2008, p. 378)

Seeking protection or separation (Chang et al., 2006)

Negotiating powerfully with the abuser to stop the violence (Brown, 1997)

Maintenance

“Commitment to above actions is firm” (Hegarty et al., 2008)

Sees self as separate with own choices (Kramer, 2007)

Working to prevent a relapse (Brown, 1997)

If in a relationship: No abuse > 6 months

Separated from abuser > 6 months

SOC outcome categories

With Stata 11 (Stata, 2009) and using descriptive statistics, we summarized proportions of women in each SOC category at each time. Due to small numbers in some categories, these were collapsed into pre-change stages (pre-contemplation and contemplation) and change-related stages (preparation, action, and maintenance) to facilitate analysis of factors influencing change. Using a complex number system (Watson & King, 2007), women’s data for each time were categorized to create a numerical representation of their SOC position over the study. Using this number and with particular attention to their end SOC, women were categorized into the following multinomial outcome variables:

Always change related: being in change-related stages at each survey in which they took part,

Ending change related: ending in change-related stages; having been in pre-change stages at some point during the study, or

Ending pre-change: ending in pre-change stages; women who were always in pre-change stages or ended in pre-change stages.

Regression modeling

The outcome variable comprised of three categories. With women who ended pre-change as the comparison category, multinomial logistic regression was undertaken adjusting for GP cluster. We conducted univariate analyses of baseline characteristics including study cohort (Table 4). Initially, we incorporated only significant baseline variables (p ≤ .05) into a multivariate multinomial logistic regression model at each time point. However, as a sensitivity test, we added some theoretically important but not initially significant variables to our final models that led to the retention of group, income, and social support as they are theoretically linked to women’s ability to undertake change in abusive relationships.

Table 4.

Baseline Variables.

Categorical Variables	Total	End Pre-Change	Always Change Related				End Change Related
Categorical Variables	n (%)	n (%)	n (%)	Adjusted OR	95% CI	p Value	n (%)	Adjusted OR	95% CI	p Value
Age	225
≤25 (ref.)	21 (9)	3 (6)	9 (8)				9 (15)
26-45	155 (69)	37 (72)	76 (67)	0.68	[0.21, 2.19]	.5	42 (69)	0.38	[0.09, 1.64]	.2
≥46	49 (22)	11 (22)	28 (25)	0.85	[0.28, 2.59]	.8	10 (16)	0.30	[0.08, 1.15]	.08
RCT cohort (ref intervention)	225
Control group	109 (48)	26 (51)	54 (48)	0.88	[0.51, 1.52]	.6	29 (48)	0.87	[0.4, 1.89]	.7
Country of birth	225
Australian born (ref.)	189 (84)	43 (84)	95 (84)				51 (84)
OS born English speaking	25 (11)	4 (8)	14 (12)	1.58	[0.57, 4.42]	.4	7 (11)	1.48	[0.43, 5.05]	.5
OS born English additional language	11 (5)	4 (8)	4 (4)	0.45	[0.11, 1.93]	.3	3 (5)	0.63	[0.13, 3.09]	.6
Living status (ref no live-in IP)	225
With intimate partner	120 (53)	44 (86)	29 (26)	0.05	[0.02, 0.13]	<.001	47 (77)	0.53	[0.17, 1.65]	.3
No. of child at home	223
No children (ref)	84 (38)	18 (36)	42 (38)				24 (39)
1-2 children	111 (50)	27 (54)	55 (49)	0.87	[0.42, 1.81]	.7	29 (48)	0.81	[0.34, 1.92]	.6
3-4 children	28 (13)	5 (10)	15 (13)	1.29	[0.41, 4.03]	.7	8 (13)	1.20	[0.32, 4.48]	.8
Total income (ref pension)	211
Wages as main source of income	161 (75)	39 (80)	73 (66)	0.51	[0.21, 1.23]	.13	49 (86)	1.57	[0.57, 4.32]	.4
Child abuse (Ref no abuse)	225
Physical child abuse	115 (51)	28 (55)	51 (45)	0.68	[0.33, 1.36]	.3	36 (59)	1.18	[0.55, 2.56]	.7
Child sexual abuse	80 (36)	23 (45)	35 (31)	0.55	[0.26, 1.16]	.11	22 (36)	0.69	[0.34, 1.38]	.3
Perceived social support	222
Poor support (ref)	150 (68)	36 (24)	77 (51)				37 (25)
Moderate/strong support	72 (32)	15 (21)	43 (47)	1.06	[0.54, 2.07]	.9	23 (15)	1.49	[0.61, 3.63]	.4
Type of abuse	220
No abusive (ref)	28 (13)	2 (4)	21 (19)				5 (8)
SCA	72 (33)	21 (41)	37 (34)	0.17	[0.03, 0.95]	.04	14 (23)	0.27	[0.04, 1.6]	.15
Other abuse	120 (54)	28 (54)	51 (47)	0.17	[0.03, 0.88]	.04	41 (68)	0.59	[0.1, 3.47]	.5
Linear Variables	Median (IQR)	Median (IQR)	Median (IQR)	Adjusted OR	95% CI	p Value	Median (IQR)	Adjusted OR	95% CI	p Value
Depression	n = 223	n = 51	n = 113				n = 59
Depression	7 (4, 10)	9 (6, 12)	6 (3, 10)	0.36	[0.19, 0.67]	.002	7 (4, 10)	0.54	[0.3, 0.95]	.03
Anxiety	n = 223	n = 51	n = 113				n = 59
Anxiety	11 (7, 13)	12 (8, 14)	10 (7, 12)	0.36	[0.16, 0.79]	.01	12 (8, 15)	0.72	[0.27, 1.93]	.5

Note. OR = odds ratio; CI = confidence interval; RCT = randomized-controlled trial; IP = intimate partner; SCA = severe combined abuse; OS = overseas born; IQR = interquartile range.

The HADS Depression and Anxiety Linear scales were log transformed prior to statistical modeling. We modeled anxiety, depression, and self-efficacy separately (univariate models) at each time due to potential collinearity prior to insertion together into the combined models. We also tested separate models using CAS as a linear scale or categorized variable of abuse type (experiencing no abuse, abuse [physical or psychological], or SCA), due to their strong correlation. Models using women’s abuse type were more reliable and included in final models.

Results

Baseline Characteristics

Of 272 female GP patients who commenced, 47 (17%) women completed only baseline questionnaires and were excluded. Their baseline characteristics (Table 4) and those who completed two or more surveys were similar, as were those between intervention and control groups. The majority were born in Australia (84%) and had children living at home (62%), half (53%) lived with an intimate partner and one quarter (25%) were on a pension as their main source of income. Approximately half (51%) had experienced physical child abuse, and one third (36%) experienced childhood sexual abuse. Nearly half (48%) experienced depression, and over two thirds (70%) experienced anxiety. At baseline, 28 (13%) women who reported fear of a partner within the last 12 months were not experiencing IPV as measured by CAS ≤7.

Baseline Univariate Analysis

Univariate regression at baseline only demonstrated a significant association between SOC outcome categories and women’s depression, anxiety, experiences of abuse, and whether they lived with their partners or not. Women’s age, country of birth, English as a primary language, study cohort (intervention or control), history of child physical/sexual abuse, and perception of social support were not significantly associated with SOC (see Table 4).

Upon commencement, 23 (10%) women were in pre-contemplation, 71 (32%) in contemplation, 10 (4%) in preparation, 50 (22%) in action, and 71 (32%) in the maintenance stage (Figure 1). Few women were in preparation, and while the proportion of women in maintenance nearly doubled (32%-62%), all other stages had lower percentages by study completion. Assessing the collapsed SOC variable (women’s movements from pre-change stages to change-related stages), the largest positive movement of women was between baseline and 6 months (Figure 2). There were 16% more women in change-related stages at 6 months than at baseline. This was followed by minimal change (−1%) from 6 to 12 months, and then a 6% increase of women in change-related stages between 12 and 24 months. Of women in change-related stages, only 12 to 14 women at any time point (9%-10%) were trying to remain within the relationship but stay safe, all other women ended their relationships.

Figure 1.

Women’s Stage of Change across the study.

Figure 2.

Women’s collapsed Stage of Change across the study.

Within the study, some women identified the consciousness-raising effects of participation, writing in their responses:

The survey has kept me reminded that his behaviour and therefore my responses are not normal and not OK.

These questionnaires make me think about how I could’ve and should’ve changed things.

This survey helped me realise the potential serious problems I could expect if my partner lost more control and became more aggressive.

Multivariate Analysis

Table 5 provides a summary of the median and interquartile range of participants in the variables of interest at the 6-, 12-, and 24-month time points. Table 6 details the multinomial logistic regression outcomes across the same time points. A summary of findings is discussed below.

Table 5.

Distribution of Associated Factors With Women in Differencing SOC at 6, 12, and 24 Months.

		6 Months				12 Months				24 Months
		Total	Ending Pre-Change	Always Change Related	End Change Related	Total	Ending Pre-Change	Always Change Related	End Change Related	Total	Ending Pre-Change	Always Change Related	End Change Related
Self-efficacy	n	185	41	92	52	194	43	94	57	163	33	79	51
Self-efficacy	Median (IQR)	29 (26, 33)	27 (22, 31)	30 (27, 34)	30 (28, 32)	30 (27, 33)	28 (22, 30)	30 (27, 36)	30 (28, 33)	30 (26, 33)	26 (23, 29)	31 (27, 35)	30 (27, 33)
Depression	n	191	41	95	55	195	42	96	57	162	35	77	50
Depression	Median (IQR)	6 (2, 10)	9 (5, 11)	5 (2, 9)	6 (3, 9)	7 (3, 10)	9.5 (6, 12)	6 (1, 9)	7 (4, 9)	7 (2, 10)	8 (6, 10)	6 (1, 9)	6 (2, 11)
Anxiety	n	191	41	95	55	195	42	96	57	163	35	78	50
Anxiety	Median (IQR)	10 (6, 13)	11 (9, 14)	9 (6, 12)	10 (6, 14)	10 (5, 13)	11 (8, 14)	9 (4, 13)	10 (5, 13)	9 (5, 13)	10 (8, 14)	7.5 (5, 12)	8 (6, 14)
Total abuse	n	186	41	92	53	189	43	90	56	161	35	76	50
No abuse	n (% in category)	63 (34)	8 (19)	41 (45)	14 (26)	70 (37)	6 (14)	46 (51)	18 (32)	72 (45)	3 (9)	46 (60)	23 (46)
SCA	n (% in category)	36 (19)	9 (22)	17 (18)	10 (19)	30 (16)	7 (16)	11 (12)	12 (21)	13 (8)	5 (14)	2 (3)	6 (12)
Other abuse	n (% in category)	87 (47)	24 (59)	34 (37)	29 (55)	89 (47)	30 (70)	33 (37)	26 (47)	76 (47)	27 (77)	28 (37)	21 (42)
Living status	n	191	41	95	5	194	42	96	56	166	35	80	51
Lives with IP	n (% in category)	95 (50)	34 (83)	27 (29)	34 (62)	94 (49)	34 (81)	26 (27)	34 (61)	76 (46)	28 (80)	21 (26)	27 (53)
Income status	n	216	49	110	57	216	49	110	56	216	49	110	57
Earn salary	n (% in category)	161 (75)	39 (80)	73 (66)	49 (86)	161 (75)	39 (80)	73 (66)	49 (86)	161 (75)	39 (80)	73 (66)	49 (86)
Social support	n	189	41	93	55	192	41	95	56	166	35	80	51
Poor support	n (% in category)	125 (66)	26 (63)	36 (65)	63 (68)	135 (70)	25 (61)	67 (71)	43 (77)	114 (69)	22 (63)	54 (68)	38 (75)

Note. SOC = stage of change; SCA = severe combined abuse; IP = intimate partner.

Table 6.

Multinomial Logistic Regression at 6, 12, and 24 Months.

	6 Monthsn = 170				12 Monthsn = 176				24 Monthsn = 147
	Always Change Related		End Change Related		Always Change Related		End Change Related		Always Change Related		End Change-Related
	Adjusted OR (95% CI)	p Value	Adjusted OR (95% CI)	p Value	Adjusted OR (95% CI)	p Value	Adjusted OR (95% CI)	p Value	Adjusted OR (95% CI)	p Value	Adjusted OR (95% CI)	p Value
Study cohort	0.49 [0.19, 1.26]	.1	0.53 [0.22, 1.27]	.2	0.55 [0.23, 1.32]	.2	0.62 [0.23, 1.67]	.3	0.73 [0.23, 2.31]	.6	0.73 [0.23, 2.35]	.6
Self-efficacy	1.19 [1.08, 1.30]	<.001	1.14 [1.02, 1.27]	.02	1.05 [0.94, 1.17]	.4	1.07 [0.97, 1.20]	.2	1.21 [1.04, 1.40]	.02	1.13 [1.00, 1.28]	.05
Depression	0.46 [0.18, 1.19]	.1	0.47 [0.18, 1.28]	.1	0.51 [0.19, 1.35]	.2	0.67 [0.22, 2.01]	.5	0.62 [0.23, 1.71]	.4	0.58 [0.20, 1.73]	.3
Anxiety	3.49 [0.43, 28.4]	.2	3.07 [0.56, 16.7]	.2	1.33 [0.43, 4.10]	.6	1.22 [0.33, 4.52]	.8	1.56 [0.27, 9.16]	.6	1.81 [0.28, 11.6]	.5
SCA^a	0.16 [0.03, 0.79]	.03	0.49 [0.09, 2.80]	.4	0.16 [0.03, 0.72]	.02	0.93 [0.19, 4.60]	.9	0.03 [0.00, 0.29]	<.001	0.31 [0.04, 2.65]	.3
Abuse^a	0.18 [0.05, 0.60]	.01	0.70 [0.24, 2.02]	.5	0.09 [0.03, 0.26]	<.001	0.31 [0.08, 1.20]	.1	0.05 [0.01, 0.37]	.01	0.13 [0.02, 0.69]	.02
Lives with IP^b	0.08 [0.02, 0.25]	<.001	0.38 [0.11, 1.30]	.1	0.06 [0.02, 0.17]	<.001	0.30 [0.12, 0.75]	.01	0.07 [0.02, 0.27]	<.001	0.22 [0.06, 0.80]	.02
Earn salary^c	0.17 [0.05, 0.66]	.01	0.49 [0.11, 2.25]	.4	0.53 [0.14, 2.03]	.4	1.26 [0.27, 5.86]	.8	0.40 [0.10, 1.68]	.2	0.87 [0.25, 2.98]	.8
Social support^d	1.13 [0.41, 3.12]	.8	0.87 [0.30, 2.51]	.8	0.61 [0.19, 1.95]	.4	0.41 [0.14, 1.20]	.1	2.45 [0.77, 7.78]	.1	1.21 [0.40, 3.66]	.7

Note. OR = odds ratio; CI = confidence interval; SCA = severe combined abuse; IP = intimate partner.

Compared with women not experiencing abuse.

IP—compared with women living alone.

Compared with women on a pension as their main source of income.

Perceived moderate/high social support compared with women with low social support.

Depression

At baseline, women ending in change-related stages were significantly less likely to have depression than women who were always in or ended in pre-change SOC (AdjOR = 0.54). Depression was significant in univariate modeling at 6 months but not in multivariate modeling (Table 6). Women who ended in pre-change stages maintained higher median depression score at baseline and 6 months, whereas women who ended change related or always change related had normal median scores. At the 12- and 24-month time points, depression was not significant in either model, and only pre-change SOC women had high median depression scores.

Anxiety

At baseline, all groups had median HADS anxiety scores >8, indicating that the majority of women were anxious, although women always change related were significantly less likely to be anxious than women in pre-change SOC (AdjOR = 0.36). At 6- and 12-month time points, women in all groups continued to have high median anxiety scores (>8); however, there were no significant differences between the groups. There were still no significant differences in anxiety between groups at 24 months, although women always change related now had lower median anxiety scores (7.5) and women ending change related had a borderline score of 8. Only women ending pre-change continued to have a high median (10) anxiety at study conclusion.

Self-Efficacy

As previously noted, self-efficacy was not measured at baseline. In univariate models at 6/12, always change-related women (AdjOR = 1.17) and women ending change related (AdjOR = 1.13) were significantly more likely to have higher self-efficacy scores than women who ended pre-change. At this time, self-efficacy was significant in the combined model with always change-related women (AdjOR = 1.19) and ending change-related women (AdjOR = 1.14) having higher self-efficacy scores. Women who ended pre-change had a median of 27 for self-efficacy (<30 = low self-efficacy), whereas the other two groups had medians of 30. At 12 months, always change-related women showed no significant difference for self-efficacy, while women who end change-related had significantly higher self-efficacy scores (AdjOR = 1.10) in the univariate model only. Women who end pre-change remained the only group with a low median self-efficacy score (28). Self-efficacy was significantly associated with SOC again at 24 months. Women always change related had higher self-efficacy in the univariate (AdjOR = 1.22) and combined (AdjOR = 1.21) models compared with women who ended pre-change. Similarly, women who ended change related had higher self-efficacy scores in univariate (AdjOR = 1.14) and combined (AdjOR = 1.13) models. Women ending pre-change were still the only group with a low median self-efficacy score (26).

Experience of Abuse

At baseline, there was no significant difference in abuse between women ending pre-change or change related, although women always change related were significantly less likely to be experiencing abuse (AdjOR = 0.17) or SCA (AdjOR = 0.17). This pattern continued at the 6- and 12-month time points. At 24 months, women always change related were still significantly less likely to experience any type of abuse than pre-change women (SCA AdjOR = 0.03 and abuse AdjOR = 0.05). Furthermore at 24 months, women who ended change related were now less likely than pre-change women to experience other forms of abuse (AdjOR = 0.13).

Living With an Intimate Partner

At baseline, 86% of women who ended pre-change, 77% of women who ended change related, and 26% of women who were always change related were living with an intimate partner. Women always change related were less likely to live with an intimate partner (AdjOR = 0.05) compared with women ending pre-change. These women remained significantly less likely to live with an intimate partner for the duration of the study (24-month AdjOR = 0.07).

There was no significant difference in living status between women ending change related or pre-change at the beginning of the study. While there was still no significant difference between women ending change related or pre-change at 6 months, approximately 15% of women ending change related had left their intimate partner at this time since the study’s commencement.

By 12 months, women who ended change related (AdjOR = 0.30) were significantly less likely to live with their intimate partner, and at 24 months these women remained significantly less likely to live with their intimate partner (AdjOR = 0.22). On study completion, 80% of women who ended pre-change, 53% of women who ended change related, and 26% of women who are always change related were living with their intimate partner. For women who end change related, there was a notable downward shift in living with an intimate partner; 77% were living with their partner at the start decreasing to 53% at study conclusion.

Other Variables

Women always change related were significantly more likely to earn a salary at 6 months compared with pre-change women (AdjOR = 0.17), although this was not significant in any other way during the study. Study cohort and perceived social support were not significant for any group at any time during the study.

Discussion

Of self-efficacy, depression, and anxiety, only self-efficacy remained significantly associated with abused women creating or maintaining positive change. Women actively making or maintaining change had higher levels of self-efficacy. Self-efficacy has previously been noted to regulate distress, stress, and anxiety, and Benight and Bandura (2004), and Lerner and Kennedy (2000) found a significant relationship between self-efficacy and depression. Baly (2010) found women who had identified abuse, prioritized their needs, and saw themselves as able to create change (self-efficacious) were more likely to end abusive relationships. Women in our study who had lower self-efficacy at 6 and 24 months ended in pre-change states. As self-efficacy was not measured at baseline in our study, this may also have had a significant influence in predicting outcomes for women, given the significance of diminished self-efficacy on women’s SOC later in the study.

Consistent with Zink et al. (2004) and Spangaro, Zwi, and Poulos (2011) who identified women’s enhanced awareness of IPV through being screened, the process of engaging in the questionnaire was consciousness raising for some women, which may have promoted change. Hegarty, O’Doherty, Taft, et al. (2013) suggested that answering the questionnaires may have had a Hawthorne effect, and this may account for the lack of significant difference between intervention/control groups. For women in both study arms, their shift to change making at 6 months may have occurred because of engagement with the study through consciousness raising, dramatic relief, and self-reevaluation; three of the early processes of change associated with the shift from pre-contemplation to contemplation or action.

There was a correlation between women moving into change-related stages and ending an abusive relationship. The majority of women who made changes did so in the first 6 months, usually by leaving their abuser. Women’s SOC positions plateaued at 12 months, and this was followed by a smaller wave of women making changes by 24 months. Women always change related, having made changes prior to entering the study were significantly less likely to live with an intimate partner during the study. This may be expected as women ending in, or maintaining change across the study, predominantly ended the abuse by leaving the relationship. Bybee and Sullivan (2005) also found that 81% of women (n = 124) in their 3-year longitudinal study were no longer with their abusive partner at the study’s conclusion.

IPV is a risk factor for depression (Garcia-Moreno et al., 2013), and the WEAVE primary outcomes for 6- and 12-month time points found that women in the intervention group were less likely to be depressed than those in the control cohort (17% between-group difference; Hegarty, O’Doherty, Taft, et al., 2013). Our analysis suggests that women experiencing IPV with concurrent depression are less likely to make changes than women not depressed. The only significant baseline difference was lower depression scores for women who moved to change-related stages compared with those who ended pre-change. However, this finding was over ridden by the association with self-efficacy in the combined models at later points. There is a previously established link between depression and IPV, although a causal link is still being debated (Devries et al., 2013; Shurman & Rodriguez, 2006; Trevillion et al., 2012; Tsai, 2013). Our research supports studies that showed depression could inhibit women from making change (Edwards et al., 2006) and highlights the need for clinicians to assess women’s mental health for depression and anxiety (Tolman & Rosen, 2001; Trevillion et al., 2012) when providing care for abused women.

In our study, all women, except those who had taken action prior to study commencement (those always change related), had elevated median anxiety scores at all times. Always change-related women had significantly lower anxiety scores at baseline compared with pre-change women and were no longer anxious at 24 months. Previous studies have shown that women’s risk of violence (Campbell et al., 2003) and subsequently anxiety (possibly also related to loss of status and income, support and family structure) may be elevated for women with a history of IPV (Shurman & Rodriguez, 2006). SCA is more likely than other forms of abuse to be associated with higher anxiety (Hegarty et al., 2012). Women who had made change before study commencement were also less severely abused during the study, which may help explain their lower anxiety levels.

Bybee and Sullivan (2005) found women with social support networks less likely to be abused 3 years post intervention, this study found no association between women’s SOC and their perceived level of social support.

We note that there was no association between lower self-efficacy and ending pre-change at 12 months. We contend that this may reflect a woman’s changing experience. For example, it is possible that women who had left abusive relationships by 6 months had higher self-efficacy for change at that time. After a further 6 months, they may have been managing alone, potentially experiencing loneliness, financial hardship, and grief for the relationship or retaliative abuse; all of which may erode positive feelings of self-efficacy and increase levels of anxiety and depression (Ford-Gilboe et al., 2009). At 24 months, having maintained changes, women may feel more empowered and confident, resulting in higher self-efficacy levels. Similarly, Ford-Gilboe et al. (2009) found adverse health effects up to 20 months after an abusive relationship. The absence of significant findings at the 12-month time point is not new; Sullivan and Bybee (1999) only demonstrated significant results at 2 years. Women working to create change in abusive relationships may need support for a much longer period of time than clinicians may expect.

There are some overall limitations for the WEAVE study (Hegarty et al., 2013), while the lead author of this article was not involved in the original research we acknowledge that two of the authors (K.H., A.T.) were investigators in the original WEAVE trial. For this secondary analysis, only women who responded to two or more surveys were included in the analysis and due to small numbers of women in some SOC categories, collapsing the five SOCs into pre- and change-related categories may have hidden some women’s shifts across the study. The absence of GSE assessment at baseline is also a limitation.

Finally, the removal of the DVSA at baseline and insertion of qualitative questions eliciting the constructs present in the DVSA tool to assess women’s SOC for the 6-, 12-, and 24-month survey points are limitations, as this method has not yet been validated. However, the strength of this study lies in identification of women’s SOC position at 4 time points over a 2-year period and provision of further evidence of the important role of self-efficacy. This adds to the body of IPV knowledge highlighting factors associated with women’s ability to undertake change when living with abusive relationships.

Conclusion and Future Recommendations

The SOCs have been the focus of study within IPV research; a new lens for clinicians engaging with women living with abuse. In our study, depression is particularly associated with women in pre-change stages, whereas anxiety persists for women across all stages only lessening after women have been in maintenance for 1 to 2 years. We demonstrated that women with depression may be less likely to (successfully) make changes when living with IPV, and women usually leave their abusive partners in an effort to stop the abuse.

Our finding of the importance of self-efficacy may provide clinicians with a new emphasis in treatment and care when they support women experiencing abuse. There is an urgent need for research to validate the best methods for clinicians to enhance self-efficacy to support women experiencing IPV (Feder, Wathen, & MacMillan, 2013). This study demonstrates the desirability for clinicians to both identify depression and focus on enhancing women’s self-efficacy as a means of supporting them to make and maintain change with the ultimate goal of becoming free from abuse and finding safety for themselves and their children.

Footnotes

Authors’ Note

The funding body had no role in study design; the collection, analysis, and interpretation of data; the writing of the manuscript; or the decision to submit this manuscript for publication.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The WEAVE study is funded by the National Health and Medical Research Council (ID: 454532).

Author Biographies

Sonia A. Reisenhofer is an academic in the School of Nursing and Midwifery at La Trobe University and heads the International Development activities within the School. She is the course coordinator of the Schools Transnational Education Programs, including the bachelor of nursing conversion degree conducted in Singapore. She has a strong background in acute and emergency nursing care, and her current research interests include the areas of intimate partner violence, family violence, acute care, and international/transnational education. Her current studies focus on improving health care for women experiencing intimate partner violence with a particular focus on the use of the transtheoretical model for change and the benefits of improving women’s self-efficacy.

Kelsey Hegarty is an academic general practitioner who currently works as a professor and leads an Abuse and Violence in primary care research program. Her current research includes the evidence base for interventions to prevent violence against women; educational and complex interventions around identification of family violence in primary care settings, and responding to women and children exposed to abuse through primary care and through the use of new technologies. During the last decade, Kelsey has contributed at both national and international levels to the intimate partner violence field. She has developed a program of research in family violence, which commenced with her thesis. For this, she developed a new measure of domestic violence the Composite Abuse Scale, which is the first validated multidimensional measure of partner abuse. It has been used extensively globally and is available in 10 languages. She has coedited a book on “Intimate partner abuse for health professionals” and is on two Cochrane systematic reviews of screening and advocacy interventions for domestic violence. She played a significant role in the development of Royal Australian College of General Practice White Book on Abuse and Violence and a GP learning module. She has developed an innovative domestic violence curriculum for health practitioners, and she regularly teaches domestic violence and mental health issues to undergraduates and postgraduate medical and nursing practitioners. She is currently chair of the governance group of the Domestic Violence Resource Centre Victoria.

Jodie Valpied is a researcher within the Researching Abuse and Violence program, General Practice & Primary Health Care Academic Centre, The University of Melbourne. She is completing her PhD in psychology through the School of Psychological Sciences, The University of Melbourne, and specializes in sensitive issues, sensory-processing sensitivity, and introversion.

Lyndsey F. Watson is a biostatistician and has worked at the Judith Lumley Centre for over 20 years. She has been involved in analysing the intimate partner violence data collected as part of the Australian Longitudinal Study of Women’s Health, and also collected in an Australian Department of Health study conducted in Timor Leste. Other studies include depression after childbirth, IVF and cancer. Her PhD was a case control study of very preterm birth.

Mary-Ann Davey is a perinatal epidemiologist with a clinical background as a nurse, midwife, and maternal and child health nurse. In addition to her clinical qualifications, she holds a bachelor education, a postgraduate diploma in sociology, a master in epidemiology (major in biostatistics), and a doctorate in public health (perinatal epidemiology). She came to health research from senior clinical roles in nursing, midwifery, and maternal and child health. She worked as a senior research officer at the Royal Children’s Hospital Research Institute (now Murdoch) from 1993 to 1999, primarily as an investigator on community-based infant vaccine trials. She has worked at The Judith Lumley Research Centre since 1999 on a number of projects, including an RCT comparing caesarean section in caseload midwifery with standard care. She was awarded Australian Postgraduate Award to undertake a Doctor of Public Health program in 2002, investigating the events that follow induction of labor in otherwise uncomplicated first births. She currently fulfills an epidemiology/biostatistics support role at the center in addition to pursuing her own research interests. She also works part-time as an epidemiologist at the Victorian Consultative Council on Obstetric and Paediatric Mortality and Morbidity.

Angela Taft is a professor/director at the Judith Lumley Centre (JLC), La Trobe University, Australia and an Honorary Senior Fellow in the Department of General Practice, University of Melbourne. She is a social scientist using rigorous combinations of qualitative and epidemiological methods to answer urgent and complex questions about women’s health. Over the last 15 years, she has led a major competitively funded program of research at JLC on intimate partner/gender-based violence. This has included analyzing the health impact of partner violence over time in the Australian Longitudinal Study of Women’s Health, Cochrane systematic reviews and randomized trials of IPV interventions in general practice, and maternal and child health nursing. She conducts studies to prevent unwanted pregnancies and prevent abortion, including those evaluating emergency contraception, long-acting reversible contraception, and medical abortion options for women whose contraception fails. Her interests include studies to improve women’s health and reduction of violence in migrant and refugee communities and in the Asia Pacific, especially in Timor-Leste.

References

Baly

A. R.

(2010). Leaving abusive relationships: Constructions of self and situation by abused women. Journal of Interpersonal Violence, 25, 2297-2315. doi:10.1177/0886260509354885

Bandura

(1982). Self-efficacy mechanism in human agency. American Psychologist, 37, 122-147. doi:10.1037/0003-066X.37.2.122

Beeble

M. L.

Bybee

Sullivan

C. M.

Adams

A. E.

(2009). Main, mediating, and moderating effects of social support on the well-being of survivors of intimate partner violence across 2 years. Journal of Consulting and Clinical Psychology, 77, 718-729. doi:10.1037/a0016140

Benight

C. C.

Bandura

(2004). Social cognitive theory of posttraumatic recovery: The role of perceived self-efficacy. Behaviour Research and Therapy, 42, 1129-1148. doi:10.1016/j.brat.2003.08.008

Bliss

Ogley-Oliver

Jackson

Harp

Kaslow

(2008). African American women’s readiness to change abusive relationships. Journal of Family Violence, 23, 161-171. doi:10.1007/s10896-007-9138-3

Brown

(1997). Working toward freedom from violence: The process of change in battered women. Violence Against Women, 3, 5-26. doi:10.1177/1077801297003001002

Burke

J. G.

Denison

J. A.

Gielen

A. C.

McDonnell

K. A.

O’Campo

(2004). Ending intimate partner violence: An application of the transtheoretical model [Research Support, U.S. Gov’t, P.H.S.]. American Journal of Health Behavior, 28, 122-133. doi:10.5993/AJHB.28.2.3

Burke

J. G.

Gielen

A. C.

McDonnell

K. A.

O’Campo

Maman

(2001). The process of ending abuse in intimate relationships: A qualitative exploration of the transtheoretical model. Violence Against Women, 7, 1144-1163. doi:10.1177/10778010122183784

Burke

J. G.

Mahoney

Gielen

McDonnell

K. A.

O’Campo

(2009). Defining appropriate stages of change for intimate partner violence survivors. In Motivational interviewing and stages of change in intimate partner violence (pp. 251-272). New York, NY: Springer. doi:10.1891/0886-6708.24.1.36

10.

Burkitt

K. H.

Larkin

G. L.

(2008). The transtheoretical model in intimate partner violence victimization: Stage changes over time. Violence and Victims, 23, 411-431. doi:10.1891/0886-6708.23.4.411

11.

Bybee

Sullivan

C. M.

(2005). Predicting re-victimization of battered women 3 years after exiting a shelter program. American Journal of Community Psychology, 36, 85-96. doi:10.1007/s10464-005-6234-5

12.

Campbell

J. C.

Webster

Koziol-McLain

Block

Campbell

Curry

M. A.

. . . Sachs

(2003). Risk factors for femicide in abusive relationships: Results from a multisite case control study. Journal Information, 93, 1089-1097. doi:10.2105/AJPH.93.7.1089

13.

Catallo

Jack

S. M.

Ciliska

MacMillan

H. L.

(2012). Identifying the turning point: Using the transtheoretical model of change to map intimate partner violence disclosure in emergency department settings. ISRN Nursing, 2012, Article 239468. doi:10.5402/2012/239468

14.

Chang

J. C.

Dado

Ashton

Hawker

Cluss

P. A.

Buranosky

Scholle

S. H.

(2006). Understanding behavior change for women experiencing intimate partner violence: Mapping the ups and downs using the stages of change. Patient Education & Counseling, 62, 330-339. doi:10.1016/j.pec.2006.06.009

15.

Chang

J. C.

Dado

Hawker

Cluss

P. A.

Buranosky

Slagel

. . . Scholle

S. H.

(2010). Understanding turning points in intimate partner violence: Factors and circumstances leading women victims toward change. Journal of Women’s Health, 19, 251-259. doi:10.1089/jwh.2009.1568

16.

Cluss

P. A.

Chang

J. C.

Hawker

Scholle

S. H.

Dado

Buranosky

Goldstrohm

(2006). The process of change for victims of intimate partner violence: Support for a psychosocial readiness model [Research Support, Non-U.S. Gov’t]. Women’s Health Issues, 16, 262-274. doi:10.1016/j.whi.2006.06.006

17.

Devries

K. M.

Mak

J. Y.

Bacchus

L. J.

Child

J. C.

Falder

Petzold

. . . Watts

C. H.

(2013). Intimate partner violence and incident depressive symptoms and suicide attempts: A systematic review of longitudinal studies. PLoS Medicine, 10(5), e1001439. doi:10.1371/journal.pmed.1001439

18.

Dienemann

Campbell

Landenburger

Curry

M. A.

(2002). The domestic violence survivor assessment: A tool for counseling women in intimate partner violence relationships. Patient Education & Counseling, 46, 221-228. doi:10.1016/S0738-3991(01)00216-6

19.

Edwards

T. A.

Houry

Kemball

R. S.

Harp

S. E.

McNutt

L.-A.

Straus

. . . Kaslow

N. J.

(2006). Stages of change as a correlate of mental health symptoms in abused, low-income African American women. Journal of Clinical Psychology, 62, 1531-1543. doi:10.1002/jclp.20310

20.

Feder

G. S.

Hutson

Ramsay

Taket

A. R.

(2006). Women exposed to intimate partner violence: Expectations and experiences when they encounter health care professionals: A meta-analysis of qualitative studies. Archives of Internal Medicine, 166, 22-37. doi:10.1001/archinte.166.1.22

21.

Feder

G. S.

Wathen

C. N.

MacMillan

H. L.

(2013). An evidence-based response to intimate partner violence: WHO guidelines. Journal of the American Medical Association, 310, 479-480. doi:10.1001/jama.2013.167453

22.

Ford-Gilboe

Wuest

Varcoe

Davies

Merritt-Gray

Campbell

Wilk

(2009). Modelling the effects of intimate partner violence and access to resources on women’s health in the early years after leaving an abusive partner. Social Science & Medicine, 68, 1021-1029. doi:10.1016/j.socscimed.2009.01.003

23.

Fosså

S. D.

Dahl

A. A.

Loge

J. H.

(2003). Fatigue, anxiety, and depression in long-term survivors of testicular cancer. Journal of Clinical Oncology, 21, 1249-1254. doi:10.1200/JCO.2003.08.163

24.

Frasier

P. Y.

Slatt

Kowlowitz

Glowa

P. T.

(2001). Using the stages of change model to counsel victims of intimate partner violence. Patient Education & Counseling, 43, 211-217. doi:10.1016/S0738-3991(00)00152-X

25.

Garcia-Moreno

Pallitto

Devries

Stockl

Watts

Abrahams

. (2013). Global and regional estimates of violence against women: Prevalence and health effects of intimate partner violence and non-partner sexual violence. World Health Organization. Retrieved from http://www.popline.org/node/572931

26.

Garcia-Moreno

Watts

(2011). Violence against women: An urgent public health priority. Bulletin of the World Health Organization, 89, 2. doi:10.2471/BLT.10.085217

27.

Haggerty

L. A.

Goodman

L. A.

(2003). Stages of change-based nursing interventions for victims of interpersonal violence. Journal of Obstetric, Gynecologic & Neonatal Nursing, 32, 68-75. doi:10.1177/0884217502239802

28.

Hegarty

K. L.

Bush

Sheehan

(2005). The Composite Abuse Scale: Further development and assessment of reliability and validity of a multidimensional partner abuse measure in clinical settings. Violence and Victims, 20, 529-547. doi:10.1891/0886-6708.2005.20.5.529

29.

Hegarty

K. L.

Gunn

J. M.

O’Doherty

L. J.

Taft

Chondros

Feder

. . . Brown

(2010). Women’s evaluation of abuse and violence care in general practice: A cluster randomised controlled trial (WEAVE). BMC Public Health, 10, Article 2. doi:10.1186/1471-2458-10-2

30.

Hegarty

K. L.

O’Doherty

Astbury

Gunn

(2012). Identifying intimate partner violence when screening for health and lifestyle issues among women attending general practice. Australian Journal of Primary Health, 18, 327-331. doi:10.1071/PY11101

31.

Hegarty

K. L.

O’Doherty

L. J.

Chondros

Valpied

Taft

A. J.

Astbury

. . . Feder

G. S.

(2013). Effect of type and severity of intimate partner violence on women’s health and service use findings from a primary care trial of women afraid of their partners. Journal of Interpersonal Violence, 28, 273-294. doi:10.1177/0886260512454722

32.

Hegarty

K. L.

O’Doherty

Taft

Chondros

Brown

Valpied

. . . Gunn

(2013). Screening and counselling in the primary care setting for women who have experienced intimate partner violence (WEAVE): A cluster randomised controlled trial. The Lancet, 382, 249-258. doi:10.1016/S0140-6736(13)60052-5

33.

Hegarty

K. L.

Taft

Feder

(2008). Violence between intimate partners: Working with the whole family. British Medical Journal, 337, Article a839. doi:10.1136/bmj.a839

34.

Khaw

L. B. L.

Hardesty

J. L.

(2007). Theorizing the process of leaving: Turning points and trajectories in the stages of change. Family Relations: An Interdisciplinary Journal of Applied Family Studies, 56, 413-425. doi:10.1111/j.1741-3729.2007.00470.x

35.

Kramer

(2007). Stages of change: Surviving intimate partner violence during and after pregnancy. Journal of Perinatal & Neonatal Nursing, 21, 285-297. doi:10.1097/01.JPN.0000299785.22505.92

36.

Lerner

C. F.

Kennedy

L. T.

(2000). Stay–leave decision making in battered women: Trauma, coping and self-efficacy. Cognitive Therapy and Research, 24, 215-232. doi:10.1023/a:1005450226110

37.

Meltzer

. (2003). Development of a common instrument for mental health. In Nosikov

Gudex

(Eds.), Developing common instruments for health surveys (pp. 35-60). Amsterdam, The Netherlands: IOS Press. Retrieved from http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.452.1919&rep=rep1&type=pdf#page=42

38.

Patzel

(2001). Women’s use of resources in leaving abusive relationships: A naturalistic inquiry. Issues in Mental Health Nursing, 22, 729-747. doi:10.1080/01612840152712992

39.

Prochaska

J. O.

DiClemente

C. C.

(1986). Toward a comprehensive model of change. New York, NY: Springer.

40.

Prochaska

J. O.

Redding

C. A.

Evers

K. E.

(2008). The transtheoretical model and stages of change. In Glanz

Rimer

Viswanath

(Eds.), Health behaviour and health education theory research and practice (4th ed., pp. 97-118). San Francisco, CA: A Wiley Imprint.

41.

Prochaska

J. O.

Velicer

W. F.

Rossi

J. S.

Goldstein

M. G.

Marcus

B. H.

Rakowski

. . . Rosenbloom

(1994). Stages of change and decisional balance for 12 problem behaviors. Health Psychology, 13, 39-46. doi:10.1037/0278-6133.13.1.39

42.

Reisenhofer

Seibold

(2012). Emergency healthcare experiences of women living with intimate partner violence. Journal of Clinical Nursing, 22, 2253-2263. doi:10.1111/j.1365-2702.2012.04311.x

43.

Reisenhofer

Taft

(2013). Women’s journey to safety–the transtheoretical model in clinical practice when working with women experiencing intimate partner violence: A scientific review and clinical guidance. Patient Education & Counseling, 93, 536-548. doi:10.1016/j.pec.2013.08.004

44.

Schwarzer

(2005). Everything you always wanted to know about the Self-Efficacy Scale but were afraid to ask. Assessment, 18, 242-251. Retrieved from http://userpage.fu-berlin.de/~health/materials/faq_gse_220405.pdf

45.

Schwarzer

Jerusalem

(1995). Generalized Self-Efficacy Scale. In Weinman

Wright

Johnston

(Eds.), Measures in health psychology: A user’s portfolio. Causal and control beliefs (pp. 35-37). Windsor, UK: Nfer-Nelson.

46.

Shorey

R. C.

Tirone

Nathanson

A. M.

Handsel

V. A.

Rhatigan

D. L.

(2013). A preliminary investigation of the influence of subjective norms and relationship commitment on stages of change in female intimate partner violence victims. Journal of Interpersonal Violence, 28, 621-642. doi:10.1177/0886260512455520

47.

Shurman

L. A.

Rodriguez

C. M.

(2006). Cognitive-affective predictors of women’s readiness to end domestic violence relationships. Journal of Interpersonal Violence, 21, 1417-1439. doi:10.1177/0886260506292993

48.

Spangaro

J. M.

Zwi

A. B.

Poulos

R. G.

(2011). “Persist. persist.”: A qualitative study of women’s decisions to disclose and their perceptions of the impact of routine screening for intimate partner violence. Psychology of Violence, 1, 150-162. doi:10.1037/a0023136

49.

StataCorp. (2009). Stata statistical software: Release 11. College Station, TX: Author.

50.

Sullivan

C. M.

Bybee

D. I.

(1999). Reducing violence using community-based advocacy for women with abusive partners. Journal of Consulting and Clinical Psychology, 67, 43-53. doi:10.1037/0022-006X.67.1.43

51.

Taft

O’Doherty

Hegarty

Ramsay

Davidson

Feder

(2013). Screening women for intimate partner violence in healthcare settings. The Cochrane Database of Systematic Reviews, 4. Retrieved from http://www.thecochranelibrary.com

52.

Tolman

R. M.

Rosen

(2001). Domestic violence in the lives of women receiving welfare mental health, substance dependence, and economic well-being. Violence Against Women, 7, 141-158. doi:10.1177/1077801201007002003

53.

Trevillion

Oram

Feder

Howard

L. M.

(2012). Experiences of domestic violence and mental disorders: A systematic review and meta-analysis. PLoS ONE, 7(12). doi:10.1371/journal.pone.0051740

54.

Tsai

A. C.

(2013). Intimate partner violence and population mental health: Why poverty and gender inequities matter. PLoS Medicine, 10(5). doi:10.1371/journal.pmed.1001440

55.

Watson

L. F.

King

J. F.

(2007). A new formula for summarising the pregnancy history. Australian and New Zealand Journal of Obstetrics and Gynaecology, 47, 475-476. doi:10.1111/j.1479-828X.2007.00782.x

56.

World Health Organization. (2013). Responding to intimate partner violence and sexual violence against women: WHO clinical and policy guidelines (9241548592). Retrieved from http://www.who.int/iris/handle/10665/85240#sthash.f1Zzgmcx.dpuf

57.

Zigmond

A. S.

Snaith

R. P.

(1983). The Hospital Anxiety and Depression Scale. Acta Psychiatrica Scandinavica, 67, 361-370. Retrieved from http://www.health.fgov.be/internet2Prd/groups/public/@public/@dg1/@acutecare/documents/ie2divers/19073947.pdf

58.

Zink

Elder

Jacobson

Klostermann

(2004). Medical management of intimate partner violence considering the stages of change: Precontemplation and contemplation. Annals of Family Medicine, 2, 231-239. doi:10.1370/afm.74