Longitudinal Impact of the myPlan App on Health and Safety Among College Women Experiencing Partner Violence

Abstract

The objective of this study was to examine differences in change over time in health and safety outcomes among female college students randomized to myPlan, a tailored safety planning app, or usual web-based safety planning resources. Three hundred forty-six women (175 intervention, 171 control) from 41 colleges/universities in Oregon and Maryland completed surveys at baseline, 6- and 12-months from July 2015 to October 2017. Generalized estimating equations were used to test group differences across time. Both groups improved on four measure of intimate partner violence (IPV; Composite Abuse Scale [CAS], TBI-related IPV, digital abuse, reproductive coercion [RC]) and depression. Reduction in RC and improvement in suicide risk were significantly greater in the myPlan group relative to controls (p = .019 and p = .46, respectively). Increases in the percent of safety behaviors tried that were helpful significantly reduced CAS scores, indicating a reduction in IPV over time in the myPlan group compared to controls (p = .006). Findings support the feasibility and importance of technology-based IPV safety planning for college women. myPlan achieved a number of its objectives related to safety planning and decision-making, the use of helpful safety behaviors, mental health, and reductions in some forms of IPV.

Keywords

technology safety planning intimate partner violence college women reproductive health mental health

Introduction

Intimate partner violence (IPV) is a global public health issue and includes psychological, physical, and/or sexual violence (García-Moreno et al., 2013). While close to the same percentage of men (33.6%) and women (36.4%) will experience lifetime IPV, women’s prevalence of severe physical violence (21.4% vs 14.9%), injury (35.2% vs 11.5%), and rates of death are significantly greater (Smith et al., 2017; Smith et al., 2018; Velopulos et al., 2019). Physical health problems are significantly higher among women who have experienced IPV compared with those who have not (Amar & Gennaro, 2005; Fisher et al., 2002; Fisher et al., 2003). Research consistently demonstrates an association between IPV and increased rates of depression, substance misuse, and suicide (Beydoun et al., 2012; Devries et al., 2013; Rees et al., 2011).

Most women experience their first exposure to physical and/or sexual violence prior to age 25 in the context of an intimate relationship (Halpern-Meekin et al., 2013), with college-age women (ages 18–24 years) reporting the highest IPV rates (Desmarais et al., 2012; Smith et al., 2017). Yet, few colleges/universities offer evidence-based programs addressing IPV and the associated health and safety needs with students (Shorey et al., 2012). Young women weigh complex factors (e.g., health, academics, finances) in their safety decisions about abusive relationships. Rather than accessing formal support (e.g., campus security, counselors), most seek help from informal networks, turning to friends who often lack the knowledge or resources to effectively help (Amar & Alexy, 2005).

myPlan is an interactive decision aid and safety planning intervention that is free and accessible via a mobile app and website (myPlanApp.org), appropriate technology for college students (Glass et al., 2015; Glass et al., 2017). myPlan was created drawing upon foundational work in empowerment (Dutton, 1992), a previous trial of an internet safety decision aid for women of all ages (Glass et al., 2017), and the literature on safety planning (Campbell & Glass, 2009; Campbell et al., 2001; Davies & Lyon, 1998; Hardesty & Campbell, 2004; McFarlane et al., 2002, 2004). Evidence supports the use of decision aids to inform complex decision-making for people faced with difficult choices (e.g., end of life decisions, treatment course for chronic illness, safety in abusive relationships; Glass et al., 2012; Stacey et al., 2017). myPlan addresses three key factors: (1) protection, a focus on increasing safety; (2) enhancing decision-making around safety; and (3) reducing IPV to facilitate healing. myPlan was designed to assist diverse survivors to evaluate their relationship health and safety, consider their priorities, and design a safety plan tailored to their needs and priorities, with embedded links to connect directly to resources such as advocacy, mental and reproductive health services.

This article presents the findings from a two-group randomized controlled trial (RCT) to examine differences in change over time in decisional conflict, IPV, depression, substance misuse, suicide risk, and safety planning behaviors among college women randomized to using either the myPlan app or usual web-based safety planning. We hypothesized those in the intervention group would show greater improvement over time on each outcome. We also examined the association between IPV and number and usefulness of attempted safety behaviors, hypothesizing that the number and usefulness of safety behaviors will be associated with greater reduction in IPV among the intervention group.

Methods

Study Design and Participants

We collected data from intervention and control groups at baseline, 6- and 12-months from July 2015 to October 2017. Eligible participants were English-speaking women (including cisgender and transgender women), age 18–24, enrolled at least part-time in college/university in Oregon or Maryland, had safe access to email and a personal device with internet, and screened positive for IPV. IPV screening questions asked if a current or ex-partner had done any of the following (past 6 months): threatened to harm her physically; hit, kicked, punched, choked or otherwise physically hurt her; forced or coerced sexual activity; constantly called, texted, emailed or harassed or embarrassed her on social media; or made her feel unsafe in any way. Using the method safest for them, potential participants could provide instructions to be contacted (e.g., only call in the afternoon, do not leave a message) via email or phone by a research assistant (RA) or could contact the RA to complete eligibility, consent, and enrollment.

The study sample was drawn from diverse campus environments (three four-year public, one historically Black college, two four-year private, three community colleges) in Maryland and Oregon. Study recruitment was primarily online via advertisements across a range of platforms, including Facebook, campus bulletins, social media, and student listservs. Study advertisements were shared among students and staff at colleges/universities throughout the two states, resulting in participants from 41 campuses.

Ethics approval was obtained from the Johns Hopkins Medicine Institutional Review Board (IRB 00054334) and the Kaiser Permanente Center for Health Research IRB (Pro00004875) for the oversight of the research in Oregon. RAs were trained in human subjects’ research ethics, safe technology use, IPV-related referrals, and suicidality protocols. Detailed study and safety procedures are explained in the protocol publication (Glass et al., 2015).

Randomization

An automated algorithm randomly assigned enrolled participants to the intervention or control group. The randomization was stratified on state of residence, having children (child/no child in the home), and type of college/university. Random assignment was stratified by state to ensure the proportion of participants in each state assigned to one group or the other remained relatively constant.

Procedures

Following enrollment and randomization, an automated email was sent to the participant’s safe email address containing instructions and links for safely accessing myPlan or usual safety planning. Those randomized to myPlan had two options: download the free, publicly available app from a major U.S. app store or access a web-based version via the study website. Users entered an assigned username and password the first time the app was accessed and created a private 4-digit security PIN code for future access, which could be reset by logging out and back in using the provided username and password. Participants randomized to the control group accessed safety planning via a secure study website. All personal identifiers (only collected at study enrollment) were saved in a physically separate database in a HIPAA compliant network environment. For the duration of the RCT, study data and personal identifiers were never combined.

Data collection was completed via the app/website, allowing participants to complete the instruments whenever convenient, eliminating data input error, and potentially reducing social desirability bias compared to face-to-face interviews (Johnson et al., 2001; Rhodes et al., 2003). Participants were compensated for their time (~3 hours) and expertise with $100 total (baseline $20, 6-month $30, 12-month $50), an amount consistent with other studies.

The myPlan intervention includes: (1) Awareness of myths (jealousy = love) and characteristics of a healthy relationship, (2) My Safety, a self-assessment with immediate feedback on red flags for an unhealthy relationships and the Danger Assessment (DA) or DA-Revised (for women abused by female partners), an assessment of risk for repeat severe IPV, (3) My Priorities, an interactive visual aid that allows users to make pairwise comparisons of the importance between priorities including privacy, feelings for partner, safety, well-being of children (as appropriate), academic success, and social support/status, and (4) My Plan, a personalized safety plan with information and referrals based on responses in the previous sections.

Usual safety planning received by the control group is guided by basic emergency safety planning information provided to students on-campus and found through IPV websites targeting young women in unsafe relationships. The usual safety planning site is different from the intervention in three important ways. First, usual safety planning does not include opportunities to assess the health and safety of a relationship, such as the validated DA or DA-R to assess risk for repeat and severe violence by the partner or ex-partner, the red flags assessment, and healthy relationships information provided in the intervention site. Second, the usual safety planning site does not include an interactive safety priority activity or visual aid to provide feedback on priorities. Third, the control site includes a static emergency safety plan only with two resources (a national IPV hotline and national LGBTQ hotline), rather than more comprehensive safety information and resources beyond just emergency planning, and the usual safety plan is not tailored based on user characteristics (e.g., wants to stay in the relationship vs leave, is LGBTQ) or to the user’s risk assessment and/or safety priorities as detailed above for the myPlan intervention. A detailed description of the intervention and control sites is provided in the published study protocol (Glass et al., 2015).

Primary and Secondary Outcomes

Study measures included participant demographics such as age, race, ethnicity, year in college/university, living situation, employment, demographics of the abusive partner, and characteristics of the relationship status (current or ex-partner) and length.

Primary outcomes included use and helpfulness of 34 safety strategies (Parker et al., 1999) at 12-months, with a response scale of have not tried, tried but not helpful, or tried and helpful. Safety behaviors are scored by taking the number of items found helpful, dividing by the number tried, and multiplying by 100 to get the percent of safety strategies tried that were helpful. Decisional conflict was measured immediately post-intervention, with questions adapted from validated subscales of the Decisional Conflict Scale (DCS; O’Connor, 1995; O’Connor, 1993/2010). The scale has 12-items on advantages and disadvantages of safety options and values related to them and uses a 5-point Likert response scale from “strongly agree” to “strongly disagree.” The DCS provides a total score (sample α = .91), with higher scores indicating a greater degree of decisional conflict (i.e., poorer decision process; Rhodes et al., 2003; Sullivan & Bybee, 1999), as well as four subscale scores (feeling informed, certainty about decision, values clarity, and support).

Secondary outcomes included past 6-months IPV and reproductive coercion (RC), past-week depression and suicide risk, substance misuse, and preparation for decision-making. IPV was measured using the Composite Abuse Scale (CAS; Beeble et al., 2008; Foshee et al., 1996), a 30-item validated measure of four independent types of abuse: (1) severe combined abuse (defined as one or more episodes of severe combined abuse, for example, use of a weapon, rape or attempted rape), (2) physical abuse with emotional abuse and/or harassment, (3) physical abuse only, and (4) emotional abuse and/or harassment only. To calculate a total score (sample α = .93), we used an adapted response scale: never = 0, only once = 1, several times = 2, or many times = 3. Additional IPV variables included seven “yes” or “no” items examining digital abuse, measured with items adapted from existing studies; “[Partner] shared (or threatened to share) content that was supposed to be private (e.g., a nude picture)” (Glauber et al., 2007; Muise et al., 2009; Tokunaga, 2011). This yielded a score of 0–7 with higher values indicating more forms of digital abuse (sample α = .72). Four items examined traumatic brain injury (TBI)-related IPV (beating to the head, choking/strangling), and a dichotomous score (yes/no) for any experience of TBI-related IPV was created.

RC was defined as a positive response to either or both items assessing if (1) partner tried to get her pregnant when she did not want to be and (2) participant prevented or ended a pregnancy and did not tell their partner out of fear. Depression was measured with the Center for Epidemiologic Studies Depression Scale, Revised (CESD-R; Eaton et al., 2004; Radloff, 1977), a 20-item self-report measure that assesses for depressive symptoms (sample α = .95). Items are rated based on frequency in the “past week or so” from 0 (Rarely or none of the time—less than one day) to 4 (Nearly every day for two weeks). Suicide risk was defined as a response of one or more days in the past week to the CESD questions “I wished I were dead” or “I wanted to hurt myself.”

The Monitoring the Future Drug and Alcohol Questionnaire (O’Malley et al., 1984) was used to measure alcohol and drug misuse. Participants reported how many occasions (0 = never to 6 = 40 or more times) they used alcohol in the last 30 days, got drunk in the last 6 months (1 = none to 5 = nearly all), and binge drank in the last 2 weeks (1 = none to 5 = nearly all), yielding three frequency scores. Participants were also asked how many occasions they used drugs (marijuana, club drugs/hallucinogens, stimulants/narcotics, prescription drugs, other) in the last 30 days (O’Connor, 1995; O’Connor, 1993/2010). Two variables were created: any use of marijuana and any drug use excluding marijuana in the last 30 days.

The Preparation for Decision-Making Scale (Bennett et al., 2010; Graham & O’Connor, 1995) is a 10-item scale that assesses perception on usefulness of a decision support intervention in preparing to make a decision and has a response scale of 1 (not at all) to 5 (a great deal). Participants in both groups completed these items after first use of myPlan or usual safety planning and at 12-months follow-up. We examined individual scale items as we were interested in the specific ways the intervention informed decision-making.

Statistical Power Analysis

The power analysis was based on an effect size calculated from the results of an online IPV intervention (Eden et al., 2015) to enhance safety seeking, compared to usual safety planning. The effect size, f, was computed to be .58, a medium effect size equivalent to a mean change from baseline to follow-up of 14.4% for the intervention group compared to a change of 10.4% in the control group from baseline to 12-months. Assuming an autocorrelation of .3, 150 participants in the intervention and 150 participants in the control group, at attrition rate of 20%, the study has .99 power to detect a significant group-by-time interaction.

Statistical Analyses

Chi-square and t-tests examined if randomization achieved balance between groups at baseline. Mixed effect regressions were used to test group differences across time with Binomial, Gaussian, or Poisson distributions depending on the outcome. The main analyses included time, group, and time-by-group interaction accounting for the nesting of participants within colleges and time within participants. Mixed-effects regressions were used to examine differences between groups in decisional conflict immediately post-use of the intervention or control app and at 12-months. Mixed-effects regression with change from baseline to 12-months in IPV as the dependent variable and group, safety behaviors, and their interaction as independent variables tested the relationship in the number and usefulness of safety behaviors with decreased IPV. Analyses were conducted in STATA 16.0, using an intention-to-treat approach. All women were included in the main analyses as these models do not require complete data at each time point.

Results

A total of 353 women enrolled and completed the baseline survey (N = 177 intervention; N = 176 control) (Figure 1). Six-month retention was 98.3% for both the intervention (N = 174) and control (N = 173) groups. Twelve-month retention was 98.9% for the intervention group (N = 175) and 96.6% (N = 171) for the control group. Randomization achieved balance on all variables with the exception of age and RC (Table 1). The sample was racially diverse, primarily undergraduates equally divided across year in school, and mostly attending a state or private 4-year college/university and living off-campus. The majority reported their abusive partner was male, an ex-partner, and not living with them.

Table 1.

Demographic and Relationship Characteristics by Study Arm Measured at Baseline.

	Control (N=177)	Intervention (N=176)	P-value
Participant characteristics
Age, mean (SD)	20.61 (2.12)	21.23 (2.56)	0.029
Race
White	122 (68.16%)	117 (66.48%)	.736
African American	31 (17.32%)	38 (21.59%)	.309
Asian	25 (13.97%)	15 (8.52%)	.105
American Indian/Alaska Native/Native Hawaiian/Pacific Islander*	8 (4.67%)	8 (4.54%)	.991
Other	8 (4.47%)	11 (6.25%)	.456
Ethnicity
Latina	26 (14.53%)	22 (12.50%)	.577
Has children	14 (7.87%)	11 (6.25%)	.553
Year in school			.445
Freshman	43 (24.39%)	31 (17.61%)
Sophomore	41 (23.16%)	44 (25.00%)
Junior	37 (20.90%)	44 (25.00%)
Senior	44 (24.86%)	49 (27.84%)
Graduate student	12 (6.78%)	8 (4.55%)
Type of school			.783
Technical/community college	34 (18.99%)	28 (15.91%)
Four-year state college/university	79 (46.47%)	80 (48.78%)
Four-year private college/university	57 (33.53%)	56 (34.15%)
Currently live			.685
On-campus	60 (35.50%)	61 (37.65)
Off-campus	109 (64.50%)	101 (62.35%)
Relationship characteristics
Live with partner			.063
Yes	11 (6.32%)	21 (12.57%)
Not officially but spend most nights together	13 (7.47%)	18 (10.78%)
No	150 (86.21%)	128 (76.65%)
Partner gender is male	165 (92.70%)	160 (90.91%)	.540
Relationship status			.584
Boyfriend/girlfriend/married or married like*	29 (17.99%)	66 (24.58%)
Ex-boyfriend/ex-girlfriend/ex-married or ex-married like*	131 (71.91%)	118 (65.14%)
Casual/hook up/friends with benefits	10 (5.62%)	7 (4.00%)
Other/do not know	8 (4.49%)	11 (6.29%)
Alcohol and drug use
Frequency of alcohol use	1.69 (1.42)	1.82 (1.46)	.421
Frequency of getting drunk	2.60 (1.36)	2.43 (1.22)	.228
Frequency of binge drinking	2.02 (1.18)	2.01 (1.24)	.933
Any marijuana use	84 (47.46%)	90 (51.14%)	.489
Any drug use excluding marijuana	37 (21.51%)	43 (25.75%)	.358
Experiences of dating/partner and depression
Composite Abuse Scale (CAS), total score	28.62 (17.14)	28.23 (17.00)	.832
Traumatic brain injury (TBI)-related IPV	49 (27.68%)	50 (28.41%)	.879
Reproductive coercion (RC)	34 (18.99%)	52 (29.55%)	.020
Digital abuse	3.25 (2.14)	3.14 (1.99)	.489
Depression	31.41 (19.57)	30.98 (19.04)	.835
Risk for suicide	72 (41.4%)	79 (45.9%)	.393

Note. *Due to small cell counts, categories were combined for this table to ensure participants are not identifiable.

Over three quarters (75.9%) of women experienced severe combined IPV, 11.0% emotional abuse and/or harassment without other forms of IPV, 5.7% physical and emotional abuse/harassment, and .6% physical IPV alone; 6.8% screened eligible for the study but did not report IPV in the 6-months prior to the baseline assessment. A total of 28% reported TBI-related IPV and 24.2% reported RC. On average, women reported 3.19 (SD = 2.07) of 7 forms of digital abuse.

At baseline, the majority (77.9%) reported drinking alcohol in the last 30 days. The mean score for getting drunk over last 6 months was 2.52 (SD = 1.30), indicating participants were drunk between a few to half of the occasions they drank. The mean score of binge drinking was 2.02 (SD = 1.21), indicating binge drinking on a few occasions in the past two weeks. Nearly half (49.3%) reported marijuana use and 23.6% used other types of drugs in past 30 days (Table 1).

Safety Behaviors, IPV, Depression, Risk for Suicide

Both groups improved over time in the percent of helpful safety behaviors used (Table 2), but the change over time did not differ between the two groups (p = .294). IPV decreased significantly over time for both groups as measured by CAS (b = −7.574; 95% CI − 8.56, −6.59; p < .001), TBI (b = −.966; 95% CI − 1.24, −.69; p < .001), digital abuse (b = −.457; 95% CI − .54, −.37; p < .001) and RC (b = −.296; 95% CI −.52, −.70; p = .010). The intervention group had a significantly greater reduction in RC than the control group (χ² = 8.16, p = .017). Depression improved significantly over time for both (b = −2.175; 95% CI −3.22, −1.23; p < .001), but the change over time did not differ between groups (p = .653). The intervention group had a significantly greater improvement over 12-months than the control group on suicide risk (b = −.799; 95% CI − 1.159, −.01; p = .048).

Table 2.

Means (SE) for the Outcomes Over Time by Study Arm.

	Control			Intervention			P-value (Time)P-value (Interaction)
	Baseline	6-Months	12-Months	Baseline	6-Months	12-Months
CESD	31.40 (1.48)	27.18 (1.47)	27.00 (1.58)	31.04 (1.44)	28.46 (1.41)	26.69 (1.39)	<.001 .648
At risk for suicide	.42 (.04)	.40 (.04)	.43 (.04)	.46 (.04)	.42 (.04)	.36 (.04)	.173 .046
% of Safety Behaviors Found Helpful	74.01 (1.59)	70.60 (2.34)	75.37 (2.26)	66.58 (1.94)	68.79 (2.46)	69.38 (2.53)	.380 .276
Composite Abuse Scale (CAS) ^p	28.77 (1.31)	16.24 (1.32)	12.53 (1.19)	28.37 (1.30)	19.07 (1.27)	14.03 (1.20)	<.001 .210
Traumatic brain injury (TBI) ^N	.37 (.05)	.25 (.05)	.09 (.02)	.43 (.06)	.22 (.04)	.15 (.04)	<.001 .229
Digital abusep	3.24 (.16)	1.94 (.14)	1.80 (.16)	3.14 (.15)	2.23 (.15)	1.84 (.16)	<.001 .200
Reproductive coercion (RC) ^L	.19 (.03)	.25 (.03)	.17 (.03)	.30 (.03)	.20 (.03)	.18 (.03)	.033 .019
Frequency of alcohol useP	1.69 (.11)	1.61 (.11)	1.58 (.11)	1.82 (.11)	1.76 (.12)	1.67 (.12)	.240 .933
Frequency of getting drunk	2.60 (.10)	2.30 (.10)	2.22 (.10)	2.43 (.09)	2.42 (.10)	2.29 (.09)	<.001 .079
Frequency of binge drinking	2.03 (.09)	1.99 (.10)	1.90 (.10)	2.01 (.09)	2.02 (.10)	1.96 (.10)	.428 .851
Any marijuana use^L	.47 (.04)	.43 (.04)	.48 (.04)	.51 (.04)	.50 (.04)	.52 (.04)	.233 .752
Any drug use excluding marijuana^L	.22 (.03)	.21 (.03)	.18 (.03)	.26 (.03)	.22 (.03)	.21 (.03)	.286 .876

Note. PPoisson model, ^LLogit model, ^NNegative binomial model.

Alcohol and Drug Use

Both groups significantly declined over time in the frequency of getting drunk (b = −.134; 95% CI −.23, −.04; p = .005), but did not differ on the rate of decline over time. The change over time in frequency of alcohol use and binge drinking, as well as marijuana or other drug use, was not significant.

Decisional Conflict and Decision Preparedness

Both groups had immediate improvement on all subscales of decisional conflict, change in total DCS for intervention group was greater than control group, and the intervention group women were significantly more positive on “helped you think about which risks and benefits of each safety option are most important to you” (3.82 vs 3.68, p = .038) than control group women. Difference by group was not significant by DCS subscales, informed (p = .513), clarity (p = .246), support (p = .897), and certainty (p = .116) after first use of the intervention. In examining decision preparedness at 12-months, intervention group women were significantly more positive on “prepared to make better safety decisions” (4.11 vs 3.89, p = .022) than control group women.

Relationship Between Safety Behaviors and Decrease in IPV

We hypothesized a stronger relationship between the number of safety behaviors tried and reduction in IPV for the intervention group as compared to the control group. The interaction of group by number of safety behaviors tried was significant (b = −.99; 95% CI − 1.63, −.35; p = .002), meaning the safety behaviors tried effected IPV differentially for the intervention and control groups. The number of behaviors tried was not related to reduction in IPV for the control group (b = .221; 95% CI −.26, .70; p = .364), but it was for the intervention group (b = −.764; 95% CI − 1.19, −.34; p < .001). As the number of safety behaviors tried increased, there was a greater decrease in IPV from baseline to 12-months for the intervention group.

Increases in the percent of safety behaviors found helpful (out of those tried) at 6-months significantly reduced IPV from baseline to 12-months in the intervention group compared to control (b = −.174; 95% CI −.30, −.05; p = .006). The percent of safety behaviors found helpful was not related to IPV reduction for the control group (b = .086; 95% CI −.01, .18; p = .065), but was for the intervention group (b = −.089; 95% CI −.17, .01; p = .039). As the percent of safety behaviors found helpful increases, there is a greater reduction in IPV from baseline to 12-months for the intervention group. The two groups were not different on the number of safety behaviors tried at 6-months (7.78 vs 7.56, p = .685) or the percent found helpful (70.50 vs 68.77, p = .580). Women perceive safety behaviors to be equally helpful in both groups, but only in the intervention group are safety behaviors associated with a reduction in IPV.

Discussion

Use of myPlan significantly reduced young women’s experience of RC, an often-undisclosed form of IPV with potentially life-changing consequences, and risk of suicide, a growing public health crisis for young women, compared to usual safety planning. Women in both groups reported increases in percent of safety behaviors used and helpful, as well as decreases in decisional conflict, misuse of alcohol, symptoms of depression, and IPV, however the changes over time were not significantly different by group. While these results do not support the effectiveness of myPlan over usual safety planning on all outcomes, they support the usefulness of technology-based safety planning for young women that have recently (past 6 months) experienced IPV by a partner or ex-partner.

Awareness of RC remains low and it may not be understood as an abuse/control tactic or one that negatively impacts health (Follingstad et al., 1999). When intervention group participants indicated RC, myPlan tailored their safety plan to include relevant information and safety behaviors; for example, accessing campus or community-based reproductive health services. The tailored plan may have underscored the importance of seeking health and other supportive services to reduce risk of repeat RC. This finding reinforces the role of college-based health providers to provide education about the association between IPV and health outcomes, allowing women a safe place to disclose RC and other forms of IPV and access trauma-informed care.

Engagement in safety behaviors at 6-months significantly reduced IPV from baseline to 12-months in the intervention group, but not in the control group. While IPV decreased overall in the control group, this reduction was not associated with trying more safety strategies or strategies tried to be helpful. Safety plans clearly have benefits, yet a tailored plan with strategies based on input may assist the woman to better understand her situation and take targeted safety actions (Messing et al., 2017).

The intervention group had a significantly greater reduction in suicide risk, but not in symptoms of depression compared to the control group. Participants scored substantively over the threshold of 15, indicating risk for depression with higher scores indicating depression or major depression (depending on how many days symptoms are reported). Depression and suicidality are strongly associated with IPV, and given the seriousness of the IPV in this sample and the evidence of increases in depression and suicidality among U.S. college students (American College Health Association, 2019), the level of depression and suicidality is not surprising, even if highly concerning. Women that used myPlan and who had been in the relationship for a longer period of time had a significantly greater reduction in depression symptoms and risk for suicide from baseline to 12-months compared to the control group. Without intervention, IPV is rarely a one-time event and generally increases in severity and frequency over time, and often leads to isolation from informal and formal support, likely impacting mental health (Johnson, 2006). Intervention group women had access to the DA, with immediate feedback that likely helped in understanding risk of repeat severe IPV (Campbell et al., 2003), identify priorities for safety, and access services that could have reduced exposure to severe IPV, isolation, depression, and suicidality.

Alcohol and drug use was common among women in this study and the results were consistent with 2018 national survey data of young adults (aged 18–25; Substance Abuse and Mental Health Services Administration, 2019). Both the intervention and control groups significantly declined over time in the frequency of getting drunk (p < .001), the groups did not differ on the rate of decline over time. Evidence indicates that alcohol and drug use are related to IPV for college men and women, with women reporting being under the influence of alcohol and drugs when they are psychologically, sexually, and physically victimized (Baker & Stith, 2008; Fossos et al., 2007; Giancola, 2002; Gidycz et al., 2007; Shorey et al., 2011). An adaptation of myPlan is planned to strengthen content on the harmful effects of substance misuse and association with IPV, with tailoring to campus and community-based substance abuse and counseling programs.

The majority of women (65% in intervention and 71% in control groups) reported that IPV in the past 6 months was perpetrated by an ex-partner. Safety planning is critical for women while in an abusive relationship and when they decide to end the relationship, as the use of violence does not necessarily stop when the relationship ends. In fact, ending a relationship can be very dangerous and could trigger the abuser to increase the severity of IPV, including threats and use of harassment, stalking, and physical/sexual violence, due to their feeling of loss of control of the relationship, and can result in woman returning to the relationship to try and reduce their exposure to IPV. As relationships are dynamic and change over time, myPlan provides women 24/7 access to reassess danger and safety priorities to tailor their plan to the situation.

Limitations

Limited geographic diversity (Maryland and Oregon colleges/universities) and stringent criteria for participation (IPV in past 6 months) limit generalizability. Although approximately 2/3 of the sample identified as white, there was a higher proportion of underrepresented minorities in the sample than in U.S. colleges/universities (African Americans, 19.5% vs 13% nationwide, Indigenous students 4.5% vs 1% nationwide). The proportion of Latina students (13.5%) in our study was lower than the 21% U.S. proportion (Postsecondary National Policy Institute (PNPI), 2018). The majority (92%) identified their abusive partner as male and an ex-partner. More effort is needed in collaboration with campus LGBTQ programs to increase participation and improve relevance of myPlan to sexual minority students. Baseline survey measures completed by both groups (e.g., CAS, safety strategies) may have biased findings, as the questions themselves may have functioned as an intervention increasing women’s self-awareness of IPV/RC, safety behaviors, and mental health. Despite randomization, there were differences between groups in participant age and RC experiences, which may have affected outcomes. There is no true control group, as it is unethical to provide “nothing” to women in abusive relationships. Reductions in IPV over time may have been the effect of providing safety planning to both groups, or regression to the mean.

Conclusion

myPlan is a free, secure, anonymous resource accessible 24/7 that facilitates women’s engagement with safety decisions and planning. myPlan is, to our knowledge, the first safety decision and planning intervention delivered via a mobile and web-based app developed specifically for college-aged women, including racial/ethnic and sexual minority young women. The study demonstrates the promise of technology-based interventions to reduce reproductive coercion and risk of suicide among women that had recently experienced IPV by a partner or ex-partner. Importantly, myPlan and other technology-based tools are part of the comprehensive campus and community response to IPV and should not be considered or used as a replacement for skilled advocacy and counseling provided survivor-centered services on campuses and surrounding communities.

Figure 1.

CONSORT flow diagram.

Footnotes

Acknowledgments

The study was only possible because young women provided their expertise and time, and college/university staff and student groups advocated for health and safety on campus.

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 study was funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) Grant Number 1R01HD076881-01A1.

ORCID iDs

Nancy E. Glass

Jill T. Messing

Karen Trister Grace

Carmen Alvarez

Trial Registration

This study is registered at www.clinicaltrials.gov NCT02236663.

Author Biographies

Nancy E. Glass, PhD, RN, is a professor at Johns Hopkins University. She has 23 years of experience in conducting clinical and community-based gender-based violence (GBV) prevention and response interventions and research with diverse populations across multiple settings globally.

Amber Clough, MSW, is a research program manager at the Johns Hopkins University School of Nursing. Ms. Clough is a Social Work Clinician with extensive experience working with vulnerable women.

Jill T. Messing, PhD, MSW, is an associate professor in the School of Social Work and the Director of the Office of Gender-Based Violence at Arizona State University. She is particularly interested in the use of risk assessment in evidence-based practice and with the development and testing of innovative interventions for survivors of intimate partner violence.

Tina Bloom, PhD, MPH, RN, is an associate professor of nursing in the Sinclair School of Nursing at the University of Missouri. Her current research focuses on intimate partner violence and violence prevention interventions using collaborate community-based approaches.

Megan Lindsay Brown, PhD, is an assistant research professor in the School of Social Work at Arizona State University. Her research focuses on the use of information communication technologies (ICT) and the impact of mediated spaces on human development, specifically young adult women. Her work has examined young adult dating relationships, online harassment, interventions using ICT, and the ways ICT interaction impacts individuals’ perceptions during development. She is currently researching the impacts of technology-based abuse within intimate relationships.

Karen B. Eden, PhD, is a professor in medical informatics and clinical epidemiology in the Oregon Health and Sciences University (OHSU) School of Medicine and is a scientist focused on patient decision-making. The focus of her research career has been in translating evidence to help patients make informed decisions about their own health. She has extensive experience in translating comparative effectiveness information into patient-friendly interactive decision aids that inform patients and prepare them for an informed discussion with a provider. She is currently creating a mobile app to help OHSU students, faculty, and employees access resources to address discrimination, sexual harassment, and misconduct.

Jacquelyn C. Campbell, PhD, RN, FAAN, is professor and Anna D. Wolf Chair at Johns Hopkins University School of Nursing. Her major area of research is violence against women and associated physical and mental health outcomes, including intimate partner homicide. She has also developed and tested interventions for women exposed to violence in the health care, advocacy, and judicial systems in the United States and globally, including the danger assessment risk of domestic violence lethality instrument and interventions based on it.

Andrea Gielen, ScD, is a professor at the Johns Hopkins University School of Public Health. Her current research focuses on using mobile apps to communicate effective injury prevention messages as part of a comprehensive program; creating new pill dispensers to prevent diversion of prescription medications; identifying effective sexual assault prevention programming on college campuses; and conducting intervention trials that promote infant safe sleep practices, informed decision-making about opioid prescribing, and home and motor vehicle safety for children.

Kathryn Laughon, PhD, RN, FAAN, is an associate professor and director of the PhD program at the University of Virginia School of Nursing. She has focused her practice and research on issues of violence against women, in particular the intersection of gender, race, and trauma. Her work is particularly focused on creating structural change to improve health and safety for all women.

Karen Trister Grace is a postdoctoral fellow at the Johns Hopkins University School of Public Health, Adjunct Instructor at the Georgetown University School of Nursing and Health Studies, and practicing midwife. Her research interests include reproductive coercion and autonomy, unintended pregnancy, gender-based violence, and health disparities.

Rachael M. Turner, MSPH, is a senior research assistant at the Johns Hopkins University School of Nursing. She supports multiple intimate partner violence research projects.

Carmen Alvarez, PhD, CRNP, CNM is an assistant professor in the Johns Hopkins University School of Nursing. Her work focuses on supporting the unique needs of trauma survivors, particularly among underserved and ethnic minority women.

James Case, MS, is a senior systems engineer at the Johns Hopkins University School of Nursing. He has extensive experience in developing and implementing trials using technology and securing human subject data.

Jamie Barnes-Hoyt, MPH, is a senior research assistant at the Johns Hopkins University School of Nursing. She supports multiple intimate partner violence research projects.

Jeanne Alhusen, PhD, is an associate professor of nursing and assistant dean for research at University of Virginia School of Nursing. Her research is focused on improving maternal mental health and, consequently, improving early childhood outcomes, particularly for families living in poverty. Her research is funded by National Institute of Health (NIH), Health Resources and Services Administration (HRSA), and several foundation grants and seeks to elucidate the role of psychosocial risk factors on maternal, neonatal, and early childhood outcomes.

Ginger C. Hanson, PhD, is an assistant professor at Johns Hopkins School of Nursing. She is an industrial-organizational psychologist and statistician. She has more than 15 years of experience collaborating on and conducting statistical analyses for studies about interpersonal violence and workplace violence.

Nancy A. Perrin, PhD, is the director of the Biostatistics and methods core at Johns Hopkins School of Nursing. She is a Biostatistician with more than 20 years of experience designing and analyzing intervention and observational studies addressing gender-based violence, internationally.

References

Amar

A. F.

Alexy

E. M.

(2005, October–December). “Dissed” by dating violence. Perspectives in Psychiatric Care , 41(4), 162–171. https://doi.org/10.1111/j.1744-6163.2005.00032.x

Amar

A. F.

Gennaro

(2005, July–August). Dating violence in college women: Associated physical injury, healthcare usage, and mental health symptoms. Nursing Research , 54(4), 235–242.

American College Health Association. (2019). American College HealthAssociation-National College Health Assessment II: Reference Group Executive Summary Spring 2019 . https://www.acha.org/documents/ncha/NCHA-II_SPRING_2019_US_REFERENCE_GROUP_DATA_REPORT.pdf

Baker

C. R.

Stith

S. M.

(2008). Factors predicting dating violence perpetration among male and female college students. Journal of Aggression, Maltreatment & Trauma , 17(2), 227–244.

Beeble

M. L.

Post

L. A.

Bybee

Sullivan

C. M.

(2008, December). Factors related to willingness to help survivors of intimate partner violence. Journal of Interpersonal Violence , 23(12), 1713–1729. https://doi.org/10.1177/0886260508314333

Bennett

Graham

I. D.

Kristjansson

Kearing

S. A.

Clay

K. F.

O’Connor

A. M.

(2010, January). Validation of a preparation for Decision Making Scale. Patient Education and Counseling , 78(1), 130–133. https://doi.org/10.1016/j.pec.2009.05.012

Beydoun

H. A.

Beydoun

M. A.

Kaufman

J. S.

Zonderman

A. B.

(2012, September). Intimate partner violence against adult women and its association with major depressive disorder, depressive symptoms and postpartum depression: A systematic review and meta-analysis. Social Science & Medicine , 75(6), 959–975. https://doi.org/10.1016/j.socscimed.2012.04.025

Campbell

Glass

(2009). Safety planning, danger, and lethality assessment. In C. Mitchell (Ed.), Intimate partner violence: A health-based perspective . Oxford University Press.

Campbell

Sharps

Glass

(2001). Risk assessment for intimate partner homicide. In C. U. Press (Ed.), Clinical assessment of dangerousness: Empirical contributions (pp. 136–157). Cambridge University Press.

10.

Campbell

J. C.

Webster

Koziol-McLain

Block

Campbell

Curry

M. A.

Gary

Glass

McFarlane

Sachs

Sharps

Ulrich

Wilt

S. A.

Manganello

Schollenberger

Frye

Laughon

(2003, July). Risk factors for femicide in abusive relationships: Results from a multisite case control study. American Journal of Public Health , 93(7), 1089–1097. https://doi.org/10.2105/ajph.93.7.1089

11.

Davies

Lyon

(1998). Safety planning with battered women: Complex lives/difficult choices . SAGE Publications.

12.

Desmarais

S. L.

Reeves

K. A.

Nicholls

T. L.

Telford

R. P.

Fiebert

M. S.

(2012). Prevalence of physical violence in intimate relationships, Part 1: Rates of male and female victimization. Partner Abuse , 3(2), 140–169.

13.

Devries

K. M.

Mak

J. Y.

Bacchus

L. J.

Child

J. C.

Falder

Petzold

Astbury

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. https://doi.org/10.1371/journal.pmed.1001439

14.

Dutton

M. A.

(1992). Empowering and healing the battered woman: A model for assessment and intervention . Springer Publishing Co.

15.

Eaton

W. W.

Smith

Ybarra

Muntaner

Tien

(2004). Center for Epidemiologic Studies Depression Scale: Review and revision (CESD and CESD-R). In M. E. Maruish (Ed.), The use of psychological testing for treatment planning and outcomes assessment: Instruments for adults (pp. 363–377). Lawrence Erlbaum Associates Publishers.

16.

Eden

K. B.

Perrin

N. A.

Hanson

G. C.

Messing

J. T.

Bloom

T. L.

Campbell

J. C.

Gielen

A. C.

Clough

A. S.

Barnes-Hoyt

J. S.

Glass

N. E.

(2015, April). Use of online safety decision aid by abused women: Effect on decisional conflict in a randomized controlled trial. American Journal of Preventive Medicine , 48(4), 372–383. https://doi.org/10.1016/j.amepre.2014.09.027

17.

Fisher

B. S.

Cullen

F. T.

Turner

M. G.

(2002). Being pursued: Stalking victimization in a national study of college women. Criminology & Public Policy , 1(2), 257–308. https://doi.org/10.1111/j.1745-9133.2002.tb00091.x

18.

Fisher

B. S.

Daigle

L. E.

Cullen

F. T.

Turner

M. G.

(2003). Reporting sexual victimization to the police and others: Results from a national-level study of college women. Criminal Justice and Behavior , 30(1), 6–38. https://doi.org/10.1177/0093854802239161

19.

Follingstad

D. R.

Bradley

R. G.

Laughlin

J. E.

Burke

(1999, Winter). Risk factors and correlates of dating violence: The relevance of examining frequency and severity levels in a college sample. Violence & Victims , 14(4), 365–380.

20.

Foshee

V. A.

Linder

G. F.

Bauman

K. E.

Langwick

S. A.

Arriaga

X. B.

Heath

J. L.

McMahon

P. M.

Bangdiwala

(1996, September–October). The Safe Dates Project: Theoretical basis, evaluation design, and selected baseline findings. American Journal of Preventive Medicine , 12(5 Suppl.), 39–47. https://www.ncbi.nlm.nih.gov/pubmed/8909623

21.

Fossos

Neighbors

Kaysen

Hove

M. C.

(2007, September). Intimate partner violence perpetration and problem drinking among college students: The roles of expectancies and subjective evaluations of alcohol aggression. Journal of Studies on Alcohol and Drugs , 68(5), 706–713. https://www.ncbi.nlm.nih.gov/pubmed/17690804

22.

García-Moreno

Pallitto

Devries

Stöckl

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.

23.

Giancola

P. R.

(2002, March). Alcohol-related aggression during the college years: Theories, risk factors and policy implications. Journal of Studies on Alcohol and Drugs, Supplement , (14), 129–139. https://doi.org/10.15288/jsas.2002.s14.129

24.

Gidycz

C. A.

Warkentin

J. B.

Orchowski

L. M.

(2007). Predictors of perpetration of verbal, physical, and sexual violence: A prospective analysis of college men. Psychology of Men & Masculinity , 8(2), 79.

25.

Glass

Clough

Case

Hanson

Barnes-Hoyt

Waterbury

Alhusen

Ehrensaft

Grace

K. T.

Perrin

(2015). A safety app to respond to dating violence for college women and their friends: The MyPlan study randomized controlled trial protocol. BMC Public Health , 15, 871. https://doi.org/10.1186/s12889-015-2191-6

26.

Glass

N. E.

Messing

J. T.

Koziol-McLain

Thaller

Bloom

T. L.

Eden

Campbell

Reinhart

(2012). IRIS Project: An internet based safety decision aid for survivors of intimate partner violence . Futures without Violence National Conference on Health and Domestic Violence, San Francisco, CA.

27.

Glass

N. E.

Perrin

N. A.

Hanson

G. C.

Bloom

T. L.

Messing

J. T.

Clough

A. S.

Campbell

J. C.

Gielen

A. C.

Case

Eden

K. B.

(2017, May). The longitudinal impact of an internet safety decision aid for abused women. American Journal of Preventive Medicine , 52(5), 606–615. https://doi.org/10.1016/j.amepre.2016.12.014

28.

Glauber

Randel

Picard

(2007). Tech abuse in teen relationships study . https://www.breakthecycle.org/sites/default/files/pdf/survey-lina-tech-2007.pdf

29.

Graham

O’Connor

(1995). User manual-preparation for Decision Making Scale . Ottawa Health Research Institute.

30.

Halpern-Meekin

Manning

W. D.

Giordano

P. C.

Longmore

M.A.

(2013, February). Relationship churning, physical violence, and verbal abuse in young adult relationships. Journal of Marriage and Family , 75(1), 2–12. https://doi.org/10.1111/j.1741-3737.2012.01029.x

31.

Hardesty

Campbell

(2004). Safety planning for abused women and their children. In Jaffe

Baker

Cunningham

A. J.

(Eds.), Protecting children from domestic violence: Strategies for community intervention (pp. 89–101). The Guilford Press.

32.

Johnson

M. P.

(2006, November). Conflict and control: Gender symmetry and asymmetry in domestic violence. Violence Against Women , 12(11), 1003–1018. https://doi.org/10.1177/1077801206293328

33.

Johnson

A. M.

Copas

A. J.

Erens

Mandalia

Fenton

Korovessis

Wellings

Field

(2001). Effect of computer-assisted self-interviews on reporting of sexual HIV risk behaviours in a general population sample: A methodological experiment. AIDS , 15(1), 111–115.

34.

McFarlane

Malecha

Gist

Watson

Batten

Hall

Smith

(2002, November–December). An intervention to increase safety behaviors of abused women: Results of a randomized clinical trial. Nursing Research , 51(6), 347–354.

35.

McFarlane

Malecha

Gist

Watson

Batten

Hall

Smith

(2004, March). Increasing the safety-promoting behaviors of abused women. American Journal of Nursing , 104(3), 40–50. https://doi.org/10.1097/00000446-200403000-00019

36.

Messing

J. T.

O’Sullivan

C. S.

Cavanaugh

C. E.

Webster

D. W.

Campbell

(2017, March). Are abused women’s protective actions associated with reduced threats, stalking, and violence perpetrated by their male intimate partners? Violence Against Women , 23(3), 263–286. https://doi.org/10.1177/1077801216640381

37.

Muise

Christofides

Desmarais

(2009, August). More information than you ever wanted: Does Facebook bring out the green-eyed monster of jealousy? CyberPsychology & Behavior , 12(4), 441–444. https://doi.org/10.1089/cpb.2008.0263

38.

O’Connor

(2010). User manual: Decisional Conflict Scale . Ottawa Hospital Research Institute. https://decisionaid.ohri.ca/docs/develop/User_Manuals/UM_Decisional_Conflict.pdf

39.

O’Connor

A. M.

(1995, January–March). Validation of a decisional conflict scale. Medical Decision Making , 15(1), 25–30. https://doi.org/10.1177/0272989x9501500105

40.

O’Malley

P. M.

Bachman

J. G.

Johnston

L. D.

(1984, July). Period, age, and cohort effects on substance use among American youth, 1976–1982. American Journal of Public Health , 74(7), 682–688. https://doi.org/10.2105/ajph.74.7.682

41.

Parker

McFarlane

Soeken

Silva

Reel

(1999, February). Testing an intervention to prevent further abuse to pregnant women. Research in Nursing & Health , 22(1), 59–66. https://doi.org/10.1002/(sici)1098-240x(199902)22:1<59::aid-nur7>3.0.co;2-b

42.

Postsecondary National Policy Institute (PNPI). (2018). Fact sheet: Latino students in higher education . https://pnpi.org/wp-content/uploads/2019/10/Latino-Students-Fact-Sheet-FINAL.pdf

43.

Radloff

L. S.

(1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement , 1(3), 385–401.

44.

Rees

Silove

Chey

Ivancic

Steel

Creamer

Teesson

Bryant

McFarlane

A. C.

Mills

K. L.

Slade

Carragher

O’Donnell

Forbes

(2011, August 3). Lifetime prevalence of gender-based violence in women and the relationship with mental disorders and psychosocial function. JAMA , 306(5), 513–521. https://doi.org/10.1001/jama.2011.1098

45.

Rhodes

S. D.

Bowie

D. A.

Hergenrather

K. C.

(2003). Collecting behavioural data using the world wide web: Considerations for researchers. Journal of Epidemiology and Community Health , 57, 68–73.

46.

Shorey

R. C.

Stuart

G. L.

Cornelius

T. L.

(2011, November). Dating violence and substance use in college students: A review of the literature. Aggression and Violent Behavior , 16(6), 541–550. https://doi.org/10.1016/j.avb.2011.08.003

47.

Shorey

R. C.

Zucosky

Brasfield

Febres

Cornelius

T. L.

Sage

Stuart

G.L.

(2012, July). Dating violence prevention programming: Directions for future interventions. Aggression and Violent Behavior , 17(4), 289–293. https://doi.org/10.1016/j.avb.2012.03.001

48.

Smith

S. G.

Basile

K. C.

Gilbert

L. K.

Merrick

M. T.

Patel

Walling

Jain

(2017). National intimate partner and sexual violence survey (NISVS): 2010–2012 state report . National Center for Injury Prevention and Control, Centers for Disease Control and Prevention. https://www.cdc.gov/violenceprevention/pdf/nisvs-statereportbook.pdf

49.

Smith

S. G.

Zhang

Basile

K. C.

Merrick

M. T.

Wang

Kresnow

Chen

(2018). The National Intimate Partner and Sexual Violence Survey (NISVS): 2015 Data Brief—Updated Release . https://www.cdc.gov/violenceprevention/pdf/2015data-brief508.pdf

50.

Stacey

Légaré

Lewis

Barry

M. J.

Bennett

C. L.

Eden

K. B.

Holmes-Rovner

Llewellyn-Thomas

Lyddiatt

Thomson

Trevena

(2017). Decision aids for people facing health treatment or screening decisions. Cochrane Database of Systematic Reviews . https://doi.org/10.1002/14651858.CD001431.pub5

51.

Substance Abuse and Mental Health Services Administration. (2019). Key substance use and mental health indicators in the United States: Results from the 2018 National Survey on Drug Use and Health (HHS Publication No. PEP19-5068, NSDUH Series H-54) . https://www.samhsa.gov/data/

52.

Sullivan

C. M.

Bybee

D.I.

(1999, February). Reducing violence using community-based advocacy for women with abusive partners. Journal of Consulting and Clinical Psychology , 67(1), 43–53.

53.

Tokunaga

R. S.

(2011). Social networking site or social surveillance site? Understanding the use of interpersonal electronic surveillance in romantic relationships. Computers in Human Behavior , 27(2), 705–713.

54.

Velopulos

C. G.

Carmichael

Zakrison

T. L.

Crandall

(2019, April 1). Comparison of male and female victims of intimate partner homicide and bidirectionality: An analysis of the national violent death reporting system. The Journal of Trauma and Acute Care Surgery . https://doi.org/10.1097/ta.0000000000002276