Introduction

More than a third of children involved with child protective services (CPS) live in homes with intimate partner violence (IPV) (Casanueva et al. 2014; Hamby et al. 2011). For these children, IPV in the home increases risk for experiencing direct victimization like physical abuse, other forms of adversity and trauma (community violence, poverty, and neglect), and repeated contact with CPS (Casaneuva et al. 2009; Colletti et al. 2008; Grasso et al. 2019; O’Dea et al. 2020; Stover et al. 2017b). Exposure to adversity and trauma can accumulate across childhood to impose detrimental effects on development and functioning that can persist into adolescence and adulthood (Grasso et al. 2016; Kitzmann et al. 2003).

There is substantial variability within and across CPS agencies regarding strategies to identify and address children’s exposure to IPV and a dearth of research into effective practices. Many existing CPS strategies tend to focus exclusively on maternal caregivers, who are most often the primary victim of IPV (Nixon et al. 2007). This limited focus fails to engage and hold accountable IPV offenders, most often the paternal caregiver (Douglas and Walsh 2010; Dunn and Powell-Williams 2007), and to acknowledge their important parenting role. Thus, mothers often must bear the sole burden of carrying out CPS requirements to protect children from ongoing exposure to IPV (Humphreys and Absler 2011; Strega et al. 2008). Perpetuating this disproportionate emphasis on mothers is the notion that men who use violence cannot successfully participate in or benefit from interventions. Indeed, several meta-analyses report near zero effect sizes of group batterer intervention programs (BIPs) available to fathers with a history of IPV in reducing violence (Arias et al. 2013; Babcock et al. 2004). Existing BIPs are typically delivered to all types of offenders in the same way. BIPs are psychoeducational in nature, can range from nine to 52 weeks and are delivered in group format by facilitators who may have received training in the specific intervention and in IPV but typically are not required to have a college education or clinical training (Stover and Lent 2014). BIPs focus on societal influences that condone violence against women, power and control and teach skills for anger management. There is limited ability to tailor the topics of the intervention to the specific needs of fathers or focus substantial time on parenting and fatherhood since fathers make up only a proportion of participants, which limits group cohesion around this topic. There is also evidence that antisocial participants have a contagion effect within the group, which reduces engagement and the potential benefit of other participants (Meis 2008; Murphy and Meis 2008).

Beyond negligible effects for BIPs, attrition rates, even for court-mandated programs, are high (30–60%), indicating a significant service need (Stover et al. 2009; Sturmey et al. 2017). Several risk factors for disengagement or drop-out include younger age of participants (Chang and Saunders 2002), un- or underemployment, less education, and greater substance misuse (Daly and Pelowski 2000; Stalans and Seng 2007). Younger age of fathers also is associated with younger age of children who are at increased risk for negative developmental trajectories due to living in homes with IPV (Stover et al. 2017b).

Even among intervention completers, recidivism rates for BIPs can be 20–35%, indicating high intervention failure (Cox and Rivolta 2014; Houston 2011). Given their limited impact, there have been calls for moving beyond the “one size fits all” approach to batterer intervention to include improved assessment of perpetrators to better identify and conceptualize co-occurring mental health and substance use disorders (mood disorders, anxiety disorders, psychosis), personality characteristics (borderline or narcissistic) and circumstances (unemployment, socioeconomic strain, stress, etc.) that may be impacting violence and family functioning. Comprehensive assessment allows for tailored interventions based on the unique needs of the individual (Aaron and Beaulaurier 2017; Arias et al. 2013; Stover et al. 2009). Consideration of their role as fathers and the impact the violence has on their children is an important area of assessment. There have been calls for increased engagement and development of appropriate interventions for fathers with a history of IPV by CPS (Gordon et al. 2012; Labarre et al. 2016).

Evidence suggests that tailoring IPV interventions specifically for men who are fathers may improve outcomes (Guille 2004; Litton Fox et al. 2001) because fatherhood has been identified as a motivator for change among men who perpetrate IPV (Perel and Peled 2008; Rothman et al. 2007). Fathers who have used violence have been shown to desire greater warmth, involvement, and connection with their children (Perel and Peled 2008) and fear of losing family relationships and worry about harm to their children can be influential in fathers’ decisions to change their behavior and engage in intervention (Silvergleid and Mankowski 2006).

To successfully meet the individual needs of families, the Connecticut Department of Children and Families (CT DCF) implemented a family-focused program to address IPV that involves a comprehensive family assessment and treatment options including an empirically-supported psychosocial intervention (i.e., Fathers for Change). The current study reports on the initial implementation evaluation following 4 years of statewide implementation.

Fathers for Change (F4C)

F4C was developed to fill a gap in evidence-based interventions for men who use family violence. It has a dual focus on IPV and child maltreatment and is an intervention that can be offered by CPS to assess individual family needs and provide services to fathers, an area of significant need in the field (Gordon et al. 2012; Labarre et al. 2016; Maxwell et al. 2012). It is also in keeping with calls for interventions that will work with families impacted by IPV in different ways to address the needs and wishes of survivors (Arroyo et al. 2017). F4C’s central premise is that a focus on men’s roles as fathers provides motivation to change maladaptive patterns of communication and aggressive or unhealthy interactions in relationships. In F4C, the clinician employs motivational strategies early by discussing the father’s conceptualization of fatherhood and how he hopes his relationships with his children and coparent will change. It utilizes a family systems frame to examine multigenerational patterns and experiences. These initial motivational sessions serve to increase engagement in the subsequent reflective, emotion regulation and communication-focused sessions of F4C.

F4C was designed to be offered to fathers with a history of IPV who have children under the age of 12. Combining family systems, attachment and cognitive behavioral approaches, F4C addresses nine individually-focused core topics, four coparent focused topics, and 5 father-child focused topics in 60-min weekly individual therapy sessions with master’s level clinicians over 18–24 weeks to achieve cessation of family violence through improved emotion regulation and reflective functioning (see Fig. 1).

Fig. 1
figure 1

Theoretical Model of F4C

Full size image

Figure 2 provides an overview of the flow of F4C treatment and sample topics. F4C focuses on identifying, understanding and managing emotions to reduce emotion dysregulation related to partner and father-child interactions. It is hypothesized that improvement in emotion dysregulation and reflective functioning (RF) in turn lead to reduced family violence (Stover et al. 2017a; Stover et al. 2018). RF is a person’s capacity to understand their own as well as their child’s mental states (Slade 2005). The ability to understand the mental state of oneself or others underlies overt behavior. RF is impacted by strong emotions and is significantly associated with emotion dysregulation. Both poor RF and emotion dysregulation have been associated with violence including IPV (Oliveros and Coleman 2019; Stover and Spink 2012; Stover and Coates 2016). RF focused treatments have been successful with populations of individuals with emotion dysregulation and violent behaviors (Asen and Fonagy 2017; Bateman and Fonagy 2004; Fonagy 2003).

Fig. 2
figure 2

Fathers for Change Intervention Components

Full size image

Following the individually focused topics, F4C allows for optional joint coparent and child participation in sessions. Conjoint sessions with the mother target communication skills pertinent to coparenting, but only if the clinician determines through ongoing assessment with the mother that doing so would be safe and appropriate. If so, conjoint sessions must follow a minimum of two assessment sessions and nine individual treatment sessions with the father and two assessment sessions and a minimum of three individual sessions with the mother. Sessions with mothers are parallel and enable her to engage in the program if she chooses. Sessions with mothers focus on safety planning, inform her on the progress of her partner’s treatment progression, and allow mothers the opportunity to understand the skills the father is learning and how they apply to their coparenting relationship. If it is unsafe or not possible to hold conjoint coparent sessions, fathers can work individually with their clinician to improve their co-parenting skills; e.g., practice how to give compliments, use active listening, make I statements, and problem solve.

Father-child sessions focus on reparations that benefit children including: the father taking responsibility for his violence, making an apology and sharing what he is learning in treatment to change his behaviors (Lamb et al. 2018). Clinicians work closely with the father to help him craft specific developmentally appropriate language of what he will say to his child about his responsibility for past behaviors. Clinicians only go forward with father-child sessions if they feel they will be beneficial to the child and the father is fully prepared to take responsibility for his past behavior, not seek forgiveness during the session and understands this is a first step in facilitating a strong relationship with his child that may take time to build.

F4C Evidence Base

F4C has an emerging evidence base with three studies showing promising results (Stover 2013, 2015; Stover et al. 2019). Completion rates in these studies ranged from 67 to 80%. In an outpatient sample of fathers with co-occurring IPV and substance misuse, those randomized to F4C were significantly more likely to complete treatment than those randomized to an evidence-based substance use treatment, Individual Drug Counseling (IDC; Mercer and Woody 1999). F4C fathers showed greater reductions in IPV and significantly greater gains in their video-recorded interactions with their children. Men who received F4C showed less intrusiveness during free-play interactions and more consistency in their interactions post intervention than men who received IDC (Stover 2015). Additionally, F4C participants had higher mean scores on a measure of satisfaction compared to IDC with significantly higher scores on the following: (a) “met my treatment needs” and (b) “helped me deal more effectively with my problems”. Men reported liking all three phases of the intervention (individual, coparent, and father-child) and were particularly positive about the focus on coparenting and their roles as fathers.

Implementation of F4C within Community Agencies

Beginning in July 2015, the Connecticut (CT) Department of Children and Families (DCF) contracted with 6 agencies, one in each of the six DCF regions, to provide F4C as part of a larger Intimate Partner Violence - Family Assessment Intervention Response (IPV-FAIR) project with state funding. IPV-FAIR allows for assessment, case management, and treatment for the entire family. F4C was the intervention selected to offer fathers who were referred due to perpetration of IPV. The 6 agencies were selected through a competitive bid process after submitting a written application detailing their experience providing services to families impacted by IPV, their cultural and linguistic competency, community connectedness, and plan for implementing F4C. The selected agencies were not-for profit, community mental health agencies with a track record of providing services to children and families, particularly victims of IPV. Agency clinicians (master’s level and above) providing F4C were eligible for CT clinical licensure. Clinicians at each agency (3 to 4 per agency) were trained in F4C by the model developer using a two-day didactic and experiential training curriculum (Stover 2015) followed by twice monthly consultation calls and yearly booster trainings to ensure fidelity to the model.

The Current Study

The current study is an evaluation of the implementation of F4C by six agencies funded by DCF to provide F4C. The goals of the study were to explore a) the feasibility of statewide implementation of F4C by community mental health agencies, b) barriers to retention of families in F4C, and c) the impact of F4C on psychosocial, IPV, and children’s exposure outcomes. The following specific hypotheses were tested: 1) agencies will successfully retain fathers in F4C; 2) risk factors for BIP dropout (i.e., greater severity of past IPV, more severe substance misuse problems, and younger father age) will not be associated with attrition in F4C; 3) there will be a significant reduction in IPV and children’s exposure to conflict from pre to post F4C based on mothers’ reports; and 4) fathers who complete F4C will show significant improvements in emotion dysregulation, anger/hostility, RF, and negative mood from pre to post-treatment.

Method

Participants

De-identified data from all families enrolled in the CT DCF F4C program between January 2016 and February 2020 were included in the current study. The analytic sample included 373 fathers and their female coparents (see Fig. 2). Although the program allows for same-sex couples, no male same-sex couples were referred during the study period. Fathers’ average age was 33.59 (SD = 8.2). Ninety-nine percent of the fathers were biological fathers. Most fathers identified as Non-Hispanic white (37%), followed by 30% Hispanic/Latinx, 22% Non-Hispanic African American/Black, and < 10% multiracial or ‘other’ race/ethnicity. Female coparents of participating fathers were on average 30.7 (SD = 6.9) years old with 42% identified as Non-Hispanic white, 38% Hispanic/Latinx, 15% Non-Hispanic African American/Black, and 5% multiracial or ‘other’ race/ethnicity. The mean number of children per referred family was 2.2 (range 1–8; SD = 1.24) and the average age of children was 8.6 years old (range 0–18 years-old; SD = 4.0). More than three quarters (77.4%) of families were referred with a child under the age of three. Fifty-five percent of the fathers were married or cohabitating, 36% were single and 9% were divorced or separated. Most fathers were employed (56%). Thirty percent did not complete high school, 37% completed high school, and 33% reported having vocational training or education beyond high school.

Procedure

Families were referred to community providers by a DCF Intimate Partner Violence Specialist (IPVS) within DCF area offices. Child protection caseworkers met with the IPVS to determine case appropriateness based on the following criteria: 1) a caregiver in the home had experienced IPV (psychological, verbal, physical or sexual) in the last 6 months; 2) there was a child in the family aged 12 or younger; 3) the family had a case open with DCF due to the determination that a child maltreatment incident had occurred (which may have included exposure to IPV); 4) the DCF case plan was to either keep the children in the home or to reunify the children with the parents if they had been removed. Cases were considered inappropriate for referral if: 1) DCF planned to go forward with termination of parental rights; 2) the father had s a full no contact protective order pertaining to both mother and children; 3) severe IPV had resulted in significant physical injury requiring hospitalization or use of weapons during an incident; 4) untreated psychosis, mania or suicidality; or substance dependence that required detoxification, in which case the family could be referred following detox and re-evaluation.

IPVS specialists referred the family to the F4C contracted agency. Initial contact between the family and the F4C provider occurred with their DCF caseworker present. This provided an opportunity to differentiate the F4C treatment program from DCF, set the limits of confidentiality and explain the goals of the intervention. Following this initial meeting, the F4C clinician met with the father and mother separately for pre-assessments. If following these assessments, the father was deemed appropriate for F4C based on the criteria described above, treatment commenced. Following completion of treatment, clinicians administered post-treatment assessments and completed discharge summaries.

De-identified data collected by clinicians using Qualtrics, a HIPAA-compliant web-based survey program, at pre and post treatment assessments with fathers and their female coparents were extracted for data analysis. This study was reviewed and deemed exempt by the Connecticut Children's Medical Center Human Subjects Review Board. Treatment completion and dropout status were determined as follows: 1) fathers were considered to have completed the program if they finished a minimum of the 9 core topics, 2 co-parent topics and 2 father-child focused topics in sessions with their clinician; and 2) a father was considered to have dropped out of the program if he completed the assessment process and subsequently withdrew from the program or was administratively discharged for missed appointments.

Because data were collected by clinicians, data acquisition was influenced by clinical demands, which consequently resulted in missing data. Since computerized, tablet-based assessment prevented missed items, missing data reflected the omission of an entire measure. Among 272 treatment completers 46 (16.9%) to 82 (30.1%) participants were missing data on a measure, with fewer missing on key measures (see rates presented in Table 1).

Table 1 Pre- to Post-Treatment Changes in Abusive Behaviors and Symptoms
Full size table

Measures

The Abusive Behavior Inventory (ABI; Shepard and Campbell 1992) is a 30-item self-report tool designed to measure the frequency of abusive behaviors in the previous six-month period. It includes subscales for physical violence (e.g. My partner pushed, grabbed, or shoved me), psychological violence (e.g. My partner told me I was a bad parent), and sexual abuse (e.g. My partner physically forced me to have sex). Mothers responded based on a 5-point Likert scale (0 = not at all to 4 = very frequently). Items were summed to create a total use of all forms of violence score. The ABI has established good reliability and construct validity (Shepard and Campbell 1992). Cronbach’s alpha for mothers’ rating of the fathers’ behaviors were 0.95 at pre-treatment and 0.94 at post-treatment.

Mothers’ reports on the Children’s Exposure to Conflict subscale of the Co-parenting Relationship Scale (CPRS; Feinberg et al. 2012) is a 5-item measure of coparenting conflict in front of children that uses a 7-point Likert scale (0 = Never, 6 = Very often). The CPRS has demonstrated good reliability and strong stability (Feinberg et al. 2012). Mothers’ reports on the Children’s Exposure to Conflict had acceptable Cronbach’s alpha of 0.64 at pre-treatment and .62 post-treatment.

The Alcohol Use Disorders Identification Test (AUDIT) is a 10-item self-report scale designed by the World Health Organization to measure alcohol intake, dependence, and adverse consequences (e.g. How often during the last year have you found that you were not able to stop drinking once you had started?). AUDIT uses a 5-point Likert scale (0 = Never, 4 = Daily or almost daily) with scores ranging from 0 to 40, with a generally accepted cut-off score of 8 for identifying a potential alcohol problem (Saunders et al. 1993). The Cronbach’s alpha for fathers’ reports in the current study was 0.86.

The Drug Abuse Screening Test (DAST-10; Skinner 1982) is a brief, 10-item self-report measure of problematic substance use consisting of a series of 10 yes/no questions (e.g. Have you had blackouts or flashbacks as a result of drug use?). A score of 3–5 indicates moderate problems, 6–8 is substantial, and a score of 9 or 10 indicates severe problems related to drug use. Cronbach’s alpha for father reports in the current study was 0.95.

The Depression Anxiety and Stress Scales (DASS-21; Lovibond and Lovibond 1995) is a self-report measure designed to assess depression, anxiety, and stress reactivity. For this study, fathers reported on the shortened, 21-item version. Each construct is assessed by 7 questions that measure the frequency and severity of negative emotions over the previous week on a 4-point scale (0 = did not apply to me at all, 3 = applied to me very much, or most of the time). Scores are summed to determine severity for each of the subscales. Items included questions such as “I couldn’t seem to experience any positive feeling at all” and “I found it difficult to work up the initiative to do things.” Because the depression, anxiety, and stress subscales were highly correlated (rs > 0.8), a composite score was created by calculating the z-score for each scale and then calculating the average, to create a measure of “negative emotionality” (Lovibond 2018). Cronbach’s alpha for fathers’ negative emotionality was 0.93 at both timepoints.

The Difficulties in Emotion Regulation Scale (DERS; Gratz and Roemer 2004) is a 36-item self-report scale used to measure fathers’ emotion regulation. The DERS uses a 5-point Likert scale (1 = Almost never, 5 = Almost always) with items such as “When I’m upset, I lose control over my behaviors” and “I have difficulty making sense out of my feelings.” The DERS has demonstrated high internal consistency, good test-retest reliability and sufficient construct and predictive validity (Gratz and Roemer 2004). Cronbach’s alpha for fathers’ DERS total score in the current study was 0.84 at pre-treatment and .76 post-treatment.

The Multidimensional Anger Inventory (MAI; Siegel 1986) consists of 38 items that measure the following dimensions of anger: anger arousal (which reflects frequency, duration and magnitude of anger responses), mode of anger expression (including the dimensions of anger-in and anger-out), anger-eliciting situations, and hostile outlook. Anger arousal was measured by 11 items (e.g. It is easy to make me angry), mode of anger by 6 anger-in items (e.g. I harbor grudges that I don’t tell anyone about) and 6 anger-out items (e.g., When I am angry with someone, I let that person know), hostile outlook is 6 items (e.g., People bother me just by being around), and anger-eliciting situations is 9 items (e.g. I get angry when people are unfair). Participants choose the extent to which each statement describes them using a 5-point scale (1 = completely undescriptive; 5 = completely descriptive). Cronbach’s alpha for fathers’ Anger score at pre- and post-treatment was 0.90 and 0.84, respectively.

The Prementalizing subscale of the Parental Reflective Functioning Questionnaire (PRFQ; Luyten et al. 2017 was utilized as the measure of RF because high scores on this scale indicate unhealthy parental reflective functioning (e.g. When my child is fussy he or she does that just to annoy me). It is a 6-item scale using a 7-point Likert scale (1 = Strongly disagree, 7 = Strongly Agree). Cronbach’s alpha at pre- and post-treatment were .65 and .58, respectively.

Analytic Strategy

Analyses were conducted using IBM SPSS software (Version 25). Distributional properties of dependent variables were examined for non-normality and non-parametric tests were used for those dependent variables with excessive skew or kurtosis (> ± 2). First, we conducted chi-square and bivariate associations to identify correlates of drop-out, followed by logistic regression to identify driving predictors of drop-out. Next, paired sample t-tests were used to examine pre-post differences in IPV (ABI), emotion dysregulation (DERS), parental RF (PRFQ) Anger (MAI), and depression/anxiety (DASS) scores. Wilcoxon signed rank tests were used for outcomes that were not normally distributed. Participants with missing data on a measure were excluded from analyses that included that measure (see Table 1).

Results

Means, standard deviations, and minimum and maximum scores for all study measures at pre- and post-treatment are presented in Table 1.

Treatment Retention

As shown in Fig. 3, 13% of fathers who came in for at least one assessment appointment did not complete their assessments and were ineligible to continue F4C. Of the 373 fathers who began intervention sessions, 27% dropped out statewide. The dropout rate for the six providers ranged from 22 to 32%. There were no significant differences across sites.

Fig. 3
figure 3

Study Case Flow

Full size image

Hypothesized predictors of drop-out were examined first at the bivariate level using correlations and chi-square analysis. Race and ethnicity were examined (Non-Hispanic Black, Hispanic, White Non-Hispanic) along with drug misuse (DAST score < 6, DAST score ≥ 6), alcohol misuse (AUDIT score < 8, AUDIT score ≥ 8), paternal age, and maternal report of father’s IPV severity on the ABI at pre-assessment, employment and education level. There were no significant associations between these variables and drop-out (all ps > .05). Next, race, drug misuse, alcohol misuse, paternal age and maternal report of IPV severity were entered together in a logistic regression model predicting drop-out. Only drug misuse significantly predicted drop-out (OR = 2.27) controlling for all other correlates (see Table 2).

Table 2 Predictors of Treatment Drop-out
Full size table

Intimate Partner Violence and Children’s Exposure to Conflict

At pre-treatment, most families (78.1%) reported having the police called to their home in response to an episode of IPV, with 74% involving the presence of children. Over three quarters of fathers (76.5%) reported being the subject of a court order of protection (limited or residential stay away order) and 65.5% reported previously attending a batterer intervention or anger management program. The average mother-reported Danger Assessment (Campbell 2004) reflected a high level of risk, with almost half of mothers (57.6%) having scores in the moderate or high range, and 11% having scores reflecting severe danger and extreme risk. Maternal report of fathers’ abusive behaviors on the ABI indicated high levels of violence, with 59% exceeding the cutoff score of 9, which indicates an abusive relationship.

As outlined in Table 1, Wilcoxon signed rank tests indicated statistically significant decreases in maternal report of fathers’ IPV on the ABI and mother’s reports of children’s exposure to conflict on the CRS from pre to post-treatment for F4C completers and non-completers with small to medium effect sizes. Reductions for non-completers were statistically significant, however mean ABI scores post-treatment were still in the abusive range, while scores for completers were well below the abusive score of 9.

Paternal Symptoms and Functioning

Fathers presented with high levels of reported difficulties in emotion regulation on the DERS, anger arousal, responses to anger eliciting situations, hostile outlook on the MAI and maladaptive parental RF at baseline. Significant pre- to post-treatment change was observed for emotion dysregulation on the DERS, anger arousal, responses to anger eliciting situations, hostile outlook on the MAI, negative emotionality on the DASS, and pre-mentalizing on the PRFQ (see Table 1). Effect sizes ranged from 0.10–0.45 indicating small to moderate effects.

Discussion

Results of the current study provide support for F4C as a community-based intervention for fathers with a history of IPV, with low overall drop-out rates (27%) following initial engagement. The program demonstrated medium effect sizes for reductions in IPV behaviors and children’s exposure to coparental conflict, and a medium effect for reductions in difficulties in emotional regulation, anger arousal, and RF. Although the current study was not designed as a randomized controlled trial, it yields effects that reflect a ‘real-world’ implementation of a novel intervention across an entire state welfare system using six major community mental health providers with state contracts.

Characteristics like father age, severity of IPV, severity of alcohol misuse, race/ethnicity, employment and education were not associated with F4C attrition. Significant drug misuse was the only factor associated with greater likelihood of program attrition. This suggests broad acceptability of the intervention across demographic characteristics, which is important given high attrition rates are considered a main reason BIPs are ineffective (Eckhardt et al. 2013) and the same factors that contribute to attrition are also associated with recidivism (Jewell and Wormith 2010). Dropout rates were lower than other programs for IPV (Stover et al. 2009; Jewell and Wormith 2010) and CPS recommended programs nationally which can have drop-out rates over 65% (DePanfilis and Dubowitz 2005; Gomby et al. 1999). F4C was recommended by DCF and fathers may have felt pressure to attend, but it was not court-ordered, unlike most BIPs, where non-attendance can result in criminal penalties.

Unlike BIPs, father age, education and severity of IPV were not associated with attrition. This suggests F4C may be engaging fathers differently than standard BIPs, which focus primarily on gendered use of violence and anger management in a group format. Consistent with other BIP studies, fathers with higher levels of drug misuse were less likely to complete the program (Daly and Pelowski 2000). Future studies should examine ways to further engage fathers with drug misuse. Other programs have used motivational interviewing strategies for substance misuse that could be incorporated into initial F4C sessions (Schumacher et al. 2011). Greater focus on how drug use impacts the family may prove helpful as coordinated interventions for IPV and substance misuse have been found to improve outcomes (Easton et al. 2018) and when paired with substance misuse treatment, F4C has been shown to reduce relapse to drug use (Stover et al. 2019).

Fathers who completed the program demonstrated significantly reduced abusive behaviors and co-occurring symptoms and impairment. Mothers reported a significant reduction in fathers’ abusive behavior from pre to post intervention with mean scores going from an abusive level to a non-abusive level. Children were exposed to significantly less violence and conflict between parents over the 6 months of the intervention. Limiting IPV and children’s exposure to violence and conflict is important to children’s recovery especially given that the severity of IPV and children’s exposure was moderate to high in the sample indicating high risk for symptoms and difficulties.

Fathers reported significant reductions in their emotion dysregulation, RF difficulties, and hostile outlook, which have all been found to be associated with IPV and child maltreatment (Bateman et al. 2016; Denson et al. 2011; Finkel et al. 2009; Stover and Coates 2016; Stover and Kiselica 2014). Emotion dysregulation, in particular, has been consistently associated with abusive behaviors and IPV (Denson et al. 2011; Finkel et al. 2009), making it an important target for successful intervention. Future randomized trials could further examine these constructs using non-self-report measures and a broader measure of RF that goes beyond child focused RF to include self-RF. It is also possible that other components of the intervention are equally important (e.g., motivation to change, communication skills, improved father-child relationship) to fathers’ reductions in violence. Further, there is a subset of violent individuals who present as highly controlled and not emotionally dysregulated (Holtzworth-Munroe and Stuart 1994; Rossi et al. 2020). Future studies of F4C could include an assessment of these traits to better understand for which kinds of offenders the intervention is working best, and which components of the intervention are most helpful in reducing IPV.

Limitations

The current findings must be considered in the context of the limitations of a single group program evaluation. F4C was provided by six different agencies across the state. Given sample size limitations, site was not included in modeling. Future studies with multilevel modeling will allow for more careful control for site level differences. Although all clinicians received the same two-day training and ongoing consultation from the developer of F4C, formal fidelity review using video recording or blind coders was not feasible. Collection of data from clinicians as part of treatment resulted in some incomplete assessments and missing data. While dropout rates were low, post assessments for clients who dropped out were rare, making posttest comparisons between clients who completed and those who did not difficult. Since this was not a randomized controlled trial and lacked a comparison group, it is not possible to determine whether F4C is more or less beneficial than other available treatments. A randomized controlled study comparing F4C to standard BIPs would be an important step in the evidence base for F4C.

Conclusion

Findings of this initial program evaluation of the implementation of Fathers for Change within 6 DCF funded agencies across CT indicate the intervention was feasible to implement with lower dropout rates than some standard BIPs and DCF funded programs. Mothers reported significantly reduced IPV and children’s exposure to parental conflict over the 6 months of treatment. Fathers reported significant reductions in emotion dysregulation, hostile outlook and RF from pre to post treatment. F4C may be an effective intervention for fathers with IPV that warrants further study.