Abstract
This paper describes the delivery of a therapeutic infant/mother group work intervention program called The Peek-a-Boo ClubTM, which ran from mid-2005 until early 2012. It examines the importance of intervening early with infants and mothers impacted by family violence. The intervention used an ‘infant led’ approach to facilitate the repair of relational ruptures in the infant/mother attachment as a consequence of experiencing family violence. It provides an overview of the intervention and work undertaken to enhance the quality of the attachment between mothers and their infants. In particular, it presents the demographic data of 128 mothers and their infants who participated over a specific period (2007–2011) and then the results of a small quantitative pre- versus post-pilot evaluation of 30 groups over this same time period. Further qualitative data is also included. The challenges and complexities involved in collecting data from this ‘hard to reach’, ambivalent and vulnerable client group are discussed. The results of the evaluation indicate some limitations in the methodology, however, overall The Peek-a-Boo ClubTM program was associated with improved scores on outcome measures assessing infant, mother and infant–mother functioning. Though only a small study, it supports intervening early to assist mothers and infants impacted by family violence in order to repair relational disruption, and encourage mother’s availability to respond sensitively to their infant’s efforts in managing affect regulation. A more comprehensive, tailored and systematic evaluation of such interventions is recommended.
Keywords
Introduction
The first two years of a child’s life is a time like no other. Human development is at its most rapid and most vulnerable. Notwithstanding the importance of adequate nutrition and shelter, the quality and experience of an infant’s relationship with their primary caregivers are the key determinants in how their future neurobiological, emotional and cognitive capacities will unfold (Schore, 2005; Siegel, 2012). Early and persistent traumatic relational experiences, when not ameliorated by that caregiving system, leave ‘scars that won’t heal’ (Teicher, 2002), negatively impacting on the growing child’s cognitive, language and learning capabilities and ability to form healthy attachments (Levendosky, Bogat, & Martinez-Torteya, 2013; Schechter & Willheim, 2009). One of the most significant protective factors shown to buffer infants against the detrimental impacts of having lived with family violence is the quality of their attachment with their non-offending parent (Hibel, Granger, Blair, & Cox, 2011; Martinez-Torteya, Anne Bogat, Von Eye, & Levendosky, 2009). Severe attachment difficulties can be a precursor to developing lifelong mental health, behavioural and interpersonal problems (Breidenstine, Bailey, Zeanah, & Larrieu, 2011; Lieberman, Chu, Van Horn, & Harris, 2011). Attachment patterns have been shown to transmit across generations, making attempts to intervene early and address mother infant attachment difficulties resulting from family violence urgent and important work (Johnson, 2013).
Early intervention can counter early childhood trauma (Carpenter & Stacks, 2009; Lieberman et al., 2011), yet there is a dearth of intervention programs specific to infants and their mothers who have been exposed to family violence. This paper describes a therapeutic infant/mother group work program, the Peek-a-Boo ClubTM (PABC) and the pre- versus post-pilot evaluation of 30 groups over a 5-year period (2007–2011). The complexities inherent in not only delivering, but evaluating such an intervention will also be explored.
The PABC was originally developed within the mental health program located within Melbourne’s Royal Children’s Hospital. It aimed to positively rework relational ruptures and attachment difficulties resulting from exposure to family violence. The program was primarily informed by Object Relations and Attachment Theory frameworks. The ‘object’ is typically understood to be the infant’s mother, as usually this relationship is the first intimate relationship the infant experiences and learns from. As a result of this experience, the infant forms an attachment over those first few years which is discriminating, has specific features and is subsequently enduring (Ainsworth & Bowlby, 1991). Equally, we were informed by ‘social justice’ practices that advocated for the right of the mother and the infant to ‘be safe’ and ‘kept safe’. Using a therapeutic play space, we encouraged positive relational and regulatory experiences as we visited past traumas.
Further to this was a rigorous embedding of an ‘infant led’ approach to this work (Morgan, 2007; Paul & Thomson-Salo, 1997). As facilitators we consistently modelled how we: honoured the subjectivity of the infants; followed their lead; and were overt in our curiosity about their meaning making. Being ‘infant led’ required “curiosity about just what the infant/child maybe thinking, imagining, expressing and feeling. Infants and children are not objects that we do things to, nor are they passive participants in the therapeutic process whom we work on. Rather, they are willing, able and available unique subjects who are communicating volumes to their external world about how their internal world is faring (Bunston, 2008, p. 335).
We also explored the mother’s past experiences of being parented, both negative and positive (Lieberman, 2007), and tied these together with their current experiences of being ‘the parent’. We used concepts such as ‘Watch, Wait & Wonder' (Cohen et al., 1999), to encourage mothers to be patient and available to their infant’s invitations to engage, and then reflect on and be curious about the meanings behind their behaviour. Similarly, we used the behaviours, interactions and dynamics that occurred within the room itself to identify and overt powerful opportunities for relational repair in the ‘here and now’ (Tronick & Beeghly, 2011). Although the groups consistently explored a number of themes, the intervention was not manualised, nor was it simply a set of principles; it was relational. Each group had a life of its own and it was the facilitation team’s job to enable a space for each group’s own unique ‘life’ to come alive and grow.
The model
The facilitation team consisted of up to four workers, two of whom were infant mental health trained clinicians. Facilitators who were not clinicians from the Mental Health Service were trained up in the PABC model either through ‘in-situ’ training or training workshops. The group participants were infants (0–4 years), and their mothers, who had been exposed to significant levels of family violence. On average, a PABC group consisted of four mothers and four infants, however, twins or a sibling under 4 were occasionally included. The PABC provided 11 sessions in all (8 weekly 2 hr groups, 1 reunion group and individual pre- and post-group sessions) and took a full day, consisting of: room preparation; group facilitation; writing process notes; follow up phone calls; attendance at clinical group supervision; and production of a weekly therapeutic newsletter.
The pre-group program of mother–infant assessment session (conducted in 2 hr with two facilitators) provided an opportunity to engage with both the infant and the mother, observe their interactions and collect information. The ‘Working Model of the Child' interview was adapted (Benoit, Zeanah, Parker, Nicholson, & Coolbear, 1997), with additional questions exploring the infant's and mother’s own experiences of violence, the mothers potential for violence and how both mother and infant may have felt about the father (often the perpetrator). Questions were also sensitively asked about the conception of the infant, and the mother’s feelings upon learning she was pregnant. Additionally, consent forms were completed, evaluation measures administered, demographic information collected, limited confidentiality discussed and our legal responsibility to report any child protection issues made clear.
The individual post-group session offered ‘reciprocal feedback’, further referral options and community support and the post-group reunion session was held 4–8 weeks after the final group. Group members were made aware that a reunion would be offered at the conclusion of the group. This, we felt, held the group psychologically whilst they prepared to say goodbye.
Video feedback techniques are often used in mother/infant treatments with high risk groups (Beebe, 2005; Puckering, 2004). Where this intervention differed is through relying on the immediacy of ‘real time’ interactions, and the subjectivity and agency of the infant themselves to bring about opportunities for therapeutic change. Infants are incredibly sociable, receptive and spontaneous within a safe therapeutic space. When seen and thought about as active participants in the group, and particularly in the mother/infant relationship, they are as powerful as any other participant in activating change.
The intervention
Whilst the model remained the same, each group itself was unique. Common themes and issues arose over the course of the intervention and moved through three distinct phases:
Beginning sessions: Weeks one to three – ‘Encouraging Engagement’
These three weeks focused on engagement and safety and we remained flexible about group members joining until week three. Along with personal introductions, we established group expectations and a shared understanding of the purpose of the group. The degree of structure was fluid depending on the facilitator’s levels of experience and comfort, however some aspects varied little. Rituals included a ‘hello’ and ‘goodbye’ song, and a mid-way shared morning tea. The room set-up remained the same, creating a clear, safe space (physically and emotionally) that was inviting and not overwhelming. The emphasis was on encouraging interaction, curiosity and reflection rather than overstimulation. Large floor cushions were arranged in a circle so that everyone was at eye level with the infants and a small number of simple toys were available.
Whilst the first session involved a sense of anticipation and anxiety for both facilitators and participants, the focus was on engagement and creating an emotionally safe space. Initially, a welcoming “hello” song was sung followed by a warm-up game inviting general introductions. This led onto discussions about the purpose of the group, following the lead of the infants, asking about member’s hopes and expectations and some volunteering of the dyad’s narrative of their experience of family violence. As part of setting up a culture of observation and reflection mothers were asked to reflect on what they thought their infants needed, to imagine what the infant’s might want to share with us, how their infant may be communicating with gesture, facial expression, vocalization or proximity and what both the mothers and infants would need from the group to be safe? Activities were often infant initiated and involved playing music, singing, movement, scarf play (such as peek-a-boo) and games with balls and/or dolls. For example, an infant playing with a Lion puppet led onto the group singing the song, “Leo the Lion”. Activities were also introduced by facilitators to continue to establish engagement, explore issues and encourage infant and mother attunement. For example, an activity used involved asking mothers to choose an animal that best represented themselves and another to represent their infants, then to imagine what animal the infant might choose for themselves and their mother and discussing the reason for their choices. This was extended to include: what the mother would prefer to be (and developed into a discussion around sense of self, perception of their identity as a mother and a reflection on hopes for the future); what animal they might prefer their infant to be; and choosing animals for other family members, including the infant’s father. This inclusion provided one of several entry points for reflecting therapeutically about the meaning of fathers in the lives of the infant and the mother (and how these potentially differed). Activities such as these allowed facilitators to make tentative reflections, interpretations or links as intra-psychic material emerged.
Each group concluded with the ritual of singing “Twinkle, Twinkle, Little Star” as infants sat or lay with their mothers on large cushions whilst a large, starry transparent fabric held by group facilitators was waved over the group participants, creating an image of stars twinkling in the night sky above. During these first weeks, the infants would show through their play how much they remembered people and activities within the group, which allowed opportunities to explore the concept of memory and highlighted the children’s capacity to recall and re-enact what they may have seen. This was sometimes a confronting experience as some mother’s held the belief that their infant was ‘too young’ to remember.
Toddler sessions were often more flexible, bringing in new activities or toys that engaged their interest, i.e. drawing, books and more physical games.
Middle sessions: Weeks four to six – ‘Encouraging Reflection’
During these weeks, the groups began to settle into a rhythm, sharing experiences and exploring issues in greater depth. Mothers would often bring questions, issues and dilemmas to the group, and in doing so, facilitate discussion and sharing. Questions/themes which arose included: How might you talk to your infant/toddler about their father: when separated; when a baby has no access; or when they do have access etc.? How do you talk to the infant about their experience of what they have witnessed? Do you separate the violent acts from the person? Can a person be loving and terrifying? What is family?
Family of origin issues were also often explored through conversations initiated by the mothers and sometimes by facilitators, recollecting who played with them, and/or sang to them as children, or how they parent compared to how they were parented. Intergenerational, cultural and gender issues relating to family violence were also explored. The role of the facilitators was not to provide answers, but make overt the complexities of these issues and the implications for how the infant thought about self and about other. The facilitators worked to bind the individual participants’ narrative to the group collective so that meanings could be shared and explored.
Remaining ‘infant led’ during these discussions was important as was making space for mothers to observe and think about what their infants were doing. How the infant interacts invariably raises issues about how they needed assistance in managing strong feelings and how to express these safely, as well as the challenges inherent in limit setting. The ‘watch, wait and wonder’ approach outlined previously continued to facilitate reflection and was also often explained overtly to the group. Activities encouraging observation, i.e. sharing gaze via a mirror, often opened up discussion about how the infants watched their mothers and others and prompted questions such as: What does the infant see when they look at themselves and/or their mother’s face in the mirror?; How do mothers feel they were watched, or kept in mind by their parents?
Blowing bubbles, playing peekaboo or hide and seek was often introduced and provided opportunities for shared delight and reciprocal interaction. These playful activities allowed for observation of the dyads, how they managed separations and reunifications, their capacity for mutual delight and opened up dialogue about mis-attunements and ruptures (perceived or real) in the infant mother relationship. Toddlers ‘playing at hiding’ was particularly common and often allowed the group to discuss and explore times when the mother and/or infants had been forced to hide.
During these ‘middle’ weeks facilitators began to talk about the group’s end, which often overted feelings of ambivalence, grief, sadness and/or anger for some families and consideration of work that still needed attention. Questions often asked included: What do we want to talk about before the group finishes? Have we talked enough about the impact and meaning of violence – others and ours? How do I talk to my babies about the violence? How do we deal with and tolerate our babies’ anger or aggression? How so we think we parent and have been parented? Whilst these themes commonly arose, each group presented with its own distinct material for discussion.
Often groups had unique ‘stand-out’ event/s or topics that defined that particular group. Such examples include: 1. A toddler accidently hurt another – this was a defining moment and provided an opportunity to unpack the meanings all group members gave this incident (intentional, unintentional and how conflict is enacted, managed), which provided a powerful entry into examining undercurrents of conflict between group members and broader realisations about the violence all group members had experienced. 2. Another group grappled with the thorny issue (consciously and unconsciously enacted) of rejection as one mother disclosed her decision to relinquish her older child to a foster family. This mother was swiftly judged by some group members but led to reflections on feelings of abandonment they had felt from their mothers and on occasions, their own ambivalence towards their infants. As facilitators it was important to allow permission for painful material to be made conscious, acknowledge its complexity and use the group space to safely explore its meaning.
Ending sessions: Weeks seven to nine – ‘Encouraging Consolidation’
Sessions became more reflective as the group entered a rhythm of responding to what presented itself in the room at the instigation of the infants and became more self-directed. Past activities were revisited anew, as infants and toddlers crave repetition, allowing opportunities for further exploration of play and infant–mother engagement, and reflecting on the group ending. In planning the ending, many groups chose to bring in special, culturally significant food, honoring the different backgrounds of group members. In the final group session, mothers were asked to also complete post-evaluative measures and a written feedback form. A memento of the group was also given to the infants and mothers in the final session, i.e. photographs taken, songs sung, or their own small piece of starry fabric. These were transitional objects to hold onto from the inside world of the group to the world outside the group.
In the individual post-group feedback session with each dyad and the concluding reunion group, the same rituals, space and toys were used. This gave a chance to reconnect and process what had happened since the group concluded. It was also an opportunity for facilitators to feedback what they had seen and what had changed, and for mothers to also provide feedback. As infancy is a time of such rapid development, there was much to capture and reflect on in these sessions.
As highlighted earlier, writing post session process notes and attending supervision were critical to this intervention, as both, in different ways unpacked and allowed facilitators to reflect on, process and digest what was always rich and intense intra-psychic material. This expatiated facilitator’s capacities to digest the material of the group, the dynamics unconsciously paralleled in the facilitation team itself and contain the often ‘crisis driven’ nature of work in order to enable ‘thinking’ to remain (Bunston, 2013a). Additionally, the therapeutic newsletter provided a powerful container and therapeutic thread for all participants, keeping their material in our mind, and the group in their mind (Bunston, 2013b). Complex material was able to be digested by facilitators and given back to the mothers in a palatable form, further refining an important tenet of the intervention, offering an “opportunity to take something enormous and terrifying from the outside world and make it into something smaller and perhaps a little less frightening within their internal world” (Jones & Bunston, 2012, p. 223). An example of such material is taken directly from a newsletter written in 2009;
Gretel (24 months – name changed) very quickly settled into the new group and joined Chloe (30 months – name changed) in her travels around the room. At first, she seemed to find the game Chloe was playing a little frightening. This game was a continuation of a game Gretel has played in earlier groups where she used a snake or the whale to pretend to scare each person in the room, moving towards them with the toy and going – ‘grrrrffff’. This was perhaps giving her a sense of control and playing with creatures that can be scary and turning them instead into something that can be enjoyed rather than feared. 
This is an interesting idea when thinking about the complex nature of people and how people who can both frighten us and give us pleasure. What was very curious to observe was that Gretel moved from finding these creatures, and the snakes in particular, frightening, to moving to playing with them and at one point even giving one of the rubber snakes a good whack with the music stick.
Methods
Program participants
Although the first PABC commenced in 2005, this paper specifically examines an evaluation undertaken between mid-2007 until late 2011 involving 30 groups, with 133 infants and 105 mothers. Mothers identified 65.5% of the infants as ‘Australian’. Whilst in 2010, two culturally specific groups were run, one indigenous and the other Sudanese, the PABC groups, overall, reflected the diverse range of cultures that make up Metropolitan Melbourne and in particular the outer Western Suburbs (Health, 2013, p. 5).
Participants were referred from across Melbourne with over half (17 of the 30 groups) facilitated in collaboration and ‘on-site’ with health professionals from outside of the RCH Mental Health Program. Referrals were largely from Maternal Child Health, Child Protection, Women’s Support Services, or when co-facilitating with an external service from their existing client base. Despite being faced with issues to do with homelessness, lack of resources and their infant’s health, only a small proportion (17%) of participants failed to attend the majority of the program
Demographic information
Descriptive characteristics of PABC infant–mother participants.
Mothers’ reports of mental health status of infant–mother participant and biological fathers.
Frequency of external agency involvement and family violence experienced by infant–mother participants.
Table 1 lists the demographic details of participants. This involved infant males (53.9%) and females (46.1%) with a mean average age of 20.4 months (SD = 11.31). The mothers ranged from 18 to 53 years (average age 30.24) (SD = 5.95) and reported the father’s average age as 34.51 years (SD = 7.15). Families were typically ‘stay at home’ single mothers and only 13.3% of families consisted of both parents. Most infants were naturally conceived with the majority born at term by normal vaginal delivery. Four mothers (3.1%) reported that the conception was through rape whilst 35 (27.3%) mothers did not disclose.
Table 2 lists mental health issues reported by mothers. The most common mental health issues identified were depression (35.2%), posttraumatic stress disorder (PTSD) (7.8%) or a personality disorder (6.3%), however, 48.4% of mothers did not disclose mental health issues. Paternal mental health issues, as reported by mothers, were depression (6.3%) and PTSD (3.2%), with 86.7% not disclosing. Substance use was self-reported by 12.5% of the mothers and, according to the mothers, 33.6% of fathers used substances. Few mothers indicated that their children had mental health issues (86.4%) but did identify issues with behaviour (10.2%), anxiety (2.5%) and their attachment (0.8%).
Table 3 lists forensic and violence demographic data. Of the families involved in the program: 52.3% had police involvement due to intimate partner violence (IPV) (33.6% reported no police involvement and 14.1% did not respond); 53.9% had Family Violence Orders (FVO) in place (32.0% reported no FVO and 14.1% did not respond); and 19.5% had Family Law Orders (FLO) in place (57.8% reported no FLO and 22.7% did not respond). The types of violence reported were physical (89.1%), verbal (87.5%), emotional (67.2%), financial (36.7%) and sexual (21.9%). Mothers typically reported having experienced two or more types of violence (91.4%) and approximately 19.0% of mothers reported experiencing all of the above forms of violence. The identified perpetrator of violence was largely reported to be the father of the infant (74.2%), whilst a smaller portion of mothers identified that the violence was perpetrated by both the mother and the father of the infant (13.3%). Nearly half of the families (43.0%) required Child Protection involvement (48.4% reported no Child Protection involvement and 7% did not respond), and only 6.3% of infants were currently or had previously been placed in foster care (7.0% did not respond).
Program evaluation
The capacity to create a wait-list control group was problematic. Small referral numbers, the ambivalence of mothers in seeking help and the urgency of intervening took precedence over evaluative demands. It is acknowledged from the outset that the inclusion of a control group would have enhanced our evaluation, however, this proved untenable to achieve in the real-life delivery and demands of this unique program. Poor literacy, anxiety about information being collected and the critical priority of engagement meant that not all pre-measures were successfully completed, and as such the effectiveness of the program may be underestimated, or skewed. Notwithstanding the very real difficulties inherent in obtaining this information, it remains important to present what data we were able to collect about a cohort that is severely under-represented in the literature, largely because they are so difficult to reach.
Data collected
Standardised measures
Three measures were used to assess the functioning of mother/infant attachment and selected because of reliability, affordability and ease of use.
Infant functioning
The Brief Infant Toddler Social Emotional Assessment (BITSEA) (Briggs-Gowan & Carter, 2002) is parent-completed tool that screens emotional/behavioural problems and socio-emotional competence in one- to three-year-old children (Briggs-Gowan & Carter, 2002). The BITSEA consists of 42 items that are summed to produce a Problem subscale (31 items) and a Competence subscale (11 items), with high scores indicating higher levels of domain. The Problem subscale assesses emotional/behavioural problems, such as aggression, withdrawal, negative emotionality and anxiety. The Competence subscale assesses areas of socio-emotional competence, such as prosocial behaviours and compliance. The BITSEA has known test retest reliability (r = .79–.92) and internal consistency (Cronbach’s α = .79 for problems scale and α = .65) (Briggs-Gowan, Carter, Irwin, Wachtel, & Cicchetti, 2004).
Mother–infant attachment
The Maternal Postnatal Attachment Scale (MPAS) (Condon, Corkindale, & Boyce, 2008) is a self-report scale that quantitatively measures the quality of parent infant attachment. The scale consists of 19 items on a 5 point scale, with ‘1’ indicating low attachment and ‘5’ indicating high attachment. The items are summed to provide an overall global score of attachment and subscales for Quality of Attachment, Absence of Hostility and Pleasure in Interaction. Higher scores indicate higher levels of each measure. The MPAS has a high test retest reliability (r = .86) and an internal consistency (α = .78; Condon et al., 2008).
Clinician rating of carer–infant functioning
The Parent–Infant Relationship Global Assessment Scale (PIR-GAS; Zero-To-Three, 2005) is a clinician rating of the global adaptive status of the relationship between the primary caregiver and the infant. The PIR-GAS uses a continuous qualitative scale that contains nine anchor points, ranging from “severely disturbed” (10) to “well-adapted” (90). Little is known on the psychometrics of the PIR-GAS (Müller et al., 2013); emerging research has reported inter-rater reliability (r = .83; Aoki, Zeanah, Heller, & Bakshi, 2002) and intra class correlation (r = .86–.90; Salomonsson & Sandell, 2011). A pre-intervention rating for each dyad was determined by clinical observation of the interactions and relationship between the mother and infant dyad in their first week of the group and by interview and observational information obtained at initial assessment. A post-intervention rating was determined in the last session of the group program.
Evaluation of design and challenges with data collection
Whilst the demographic information pertains to the 128 program participants, the quantitative evaluative outcome data is based on only a third to half of this number (BITSEA (n = 38), MPAS (n = 62) and PIR-GAS (n = 50)). Both the MPAS pre-group self-report questionnaire and the BITSEA pre-group screening tool (for infants 12–36 months) were completed during the pre-group assessment with the respective post-group measures completed during the last session of the group program. When necessary, interpreters were used with families from culturally and linguistically diverse (CALD) backgrounds, assisting mothers to fill in the forms. During the final session of the PABC, mothers also completed a qualitative participant feedback form.
Statistical analysis
The demographic data was analysed using the SPSS (Statistical Package for the Social Sciences) version 19.0. Preliminary data screening was performed to ensure the suitability of the data for analysis. Incomplete data sets on five infant–mother participants were removed; analysis was performed on N = 128 infant–mother participants. All evaluative measure outcome data were screened prior to analysis to ensure that there were no statistical violations and that the data was normally distributed. Based on skewness and kurtosis ratios, only scores on the Quality of Attachment, Absence of Hostility, and Pleasure in Interaction subscales of the MPAS were significantly negatively skewed. As a result, these subscales were transformed using a reflect square root transformation. All transformed variables were reanalysed and found to be normally distributed, and unless otherwise indicated, the transformed values were used in the analyses.
To determine whether changes in the outcome measures used were clinically significant, a reliability change index (RCI) was calculated for each score and subscale of each participant. The RCI was calculated by dividing the magnitude of change from pre- to post-intervention by the standard error of the difference score (Jacobson & Truax, 1991). Internal reliability was calculated for each of the measures in the current study. Participants were categorised on their individual RCI score as having shown significant improvement (RCI > 1.96), no change (RCI ≥ 1.96 and ≤ 1.96) or significantly deteriorated (RCI < −1.96).
Results
Pre- and post-intervention outcome measures of infant functioning, infant maternal attachment and infant–mother functioning from mothers informants and clinician ratings.
Infant functioning (BITSEA)
A paired t test was conducted to compare pre- and post-program scores on mother’s reports of infant functioning. Results show that mothers reported that their infants were significantly more socially competent post-intervention (M = 17.42, SD = 3.49) than at pre-intervention (M = 16.47, SD = 3.65), t(37) = −2.05, p = .047. The results also show that mothers reported that their infants displayed significantly less problematic behaviours post-intervention (M = 15.55, SD = 6.59) than at pre-intervention (M = 20.68, SD = 8.97), t(37) = 4.18, p < .001.
Infant maternal attachment (MPAS)
Paired-sample t tests were conducted to compare pre- and post-program scores on the Quality of Attachment between the infant and mother. The results show that at post-intervention the score on overall global attachment (M = 76.72, SD = 9.68) was significantly higher than at pre-intervention (M = 73.55, SD = 13.13), t(61) = −2.30, p = .025. At a subscale level, results showed significant improvements post-intervention for Pleasure in Interaction (pre-intervention M = 17.98, SD = 5.21 versus post-intervention M = 21.42, SD = 3.37), t(61) = 4.71, p =< .001, Quality of Attachment (pre-intervention M = 36.37, SD = 5.68 versus post-intervention M = 37.03, SD = 5.56), t(61) = 7.65, p =< .001 and Absence of Hostility (pre-intervention M = 17.11, SD = 4.73, versus post-intervention M = 17.60, SD = 4.97), t(61) = 7.65, p =< .001.
Clinician evaluation of infant–mother relationship using PIR-GAS
A paired-sample t test was calculated to determine whether there were any differences in the clinicians rating of adaptive functioning between the mother and infant at pre- and post-intervention. The results show that clinicians reported better adaptive functioning post-intervention (M = 53.25, SD 13.88) than at pre-intervention (M = 49.62, SD 16.60), t(49) = −2.05, p = .046.
Reliable change index results
Analysis of the RCI showed that around 10% of participants had significant improvement post-intervention in their scores on MPAS Global functioning and MPAS Pleasure in Interaction, with 90% having no significant change. Only 3% of participants reported significant improvements post-intervention in their scores on MPAS Quality of Attachment (94% showed no change and 3% had significant deterioration). Participants demonstrated either no change (92%) or significant deterioration (7%) in Absence of Hostility based on their RCI score.
Based on their reliability change scores, 8% of participants demonstrated significant improvement post-intervention in their clinician ratings (86% showed no change and 6% had significant deterioration).Whilst no significant improvements were found for problem behaviours, 16% of infants showed significant improvement in the social competence scores (79% showed no change and 5% had significant deterioration).
Qualitative feedback form
A participant satisfaction survey was completed during the last group program session, which consisted of six questions: What was the best thing about coming to the PABC? What was the worst thing about coming to the PABC? In what ways do you think your relationship with your baby has improved? How has coming to the PABC helped your baby? Have your feelings/thinking about yourself as a mother changed since coming to the PABC? Have you noticed anything about yourself or your baby over the weeks of coming along to the PABC? This method of evaluation had over 80% compliance and consistently indicated a high level of satisfaction. The most common response to the question “what was the worst thing?” was that the group was not long enough. Additional specific feedback included:
What was the best thing about coming to PABC?
“Support, information, understanding, watching x interact with others” “Watching my child with others, facilitators really know how to help with the kids” “You treat me from the heart, you treat me like your sister, you helped me with the bond with my child” “Getting out of the house, watching x play and have fun with other kids”
In what ways do you think your relationship with your baby has improved?
“I now understand why certain things are happening so I can find ways to deal with them. We are happier as things are calmer as issues have been worked through”. “Taking more time to sit down on the floor and play with x, much more than I did before Peek-a-Boo” “Showed me to talk to x, I now understand that it’s important to talk to him whether he understands or not” “We are closer”
How has coming to the PABC helped your baby?
“Due to recommended resources x now sleeps and is happier and calmer and we enjoy our time more” “I think she has really enjoyed it and has looked forward to coming every week”. “X really enjoyed being at the group, it was really great bonding time for x and myself”. “Developed relationship with others adults, the relationship with (male facilitator) will help him relate to men- he had never had that chance”
Discussion
Information collected about those participating in the PABC intervention provides some insight into the status of infants and mothers (predominantly from greater Western Metropolitan Melbourne) affected by family violence from 2007 to 2011. The large number of single mothers (78.9%) participating may be attributed to their being no partner to prevent them from seeking outside support. The significant number of fathers (albeit mothers report) ‘using substances’ (33.6%) does correlate with other findings (Easton, McMahon, & Moore, 2011). However, physical violence features at a significantly higher rate than in other studies measuring frequency and types of violence used (Coker, Smith, McKeown, & King, 2000). The data supports the extent with which violent acts are used within family violence to exert control, with 70% of the women reporting that they experienced three or more forms of violence (Kelly & Johnson, 2008).
Of all respondents, 13.3% of mothers acknowledged that they also used violence. Whilst this violence may be understood as reciprocal, we suspect this rate is much higher and not always reciprocal. Our belief is that the shame associated with women using violence and the social debate associated with the prevalence of men’s violence, silences this discussion. We concur with Cho and Wilke (2010) that “attempts at understanding the nature of female perpetrated IPV should not be influenced by fears of a backlash from a male dominant social structure. Instead, it should lead to better understanding of the dynamics of IPV that is critical to better serve victims” (Cho & Wilke, 2010, p. 399). In this instance, the ones most silenced and less served are the infants.
Whilst 70% of the women reported that conception was not forced, only 3% reported rape, with 27.3% not disclosing. Our experience is that these disclosures would be significantly higher if collected again post-program. There is great shame, ambivalence, and for some resentment, which surrounds the conception of many infants born into a relationship where there is IPV (Gee, Mitra, Wan, Chavkin, & Long, 2009). This is an important clinical area for exploration for those working with mothers and infants affected by family violence, exacerbating already painful and distressing maternal feelings towards the infant.
Research has found a strong correlation between IPV and high levels of maternal depression (Levendosky, Bogat, Huth-Bocks, Rosenblum, & von Eye, 2011), this correlation is supported by 35.2% of the mothers in this study reporting depression. Involvement with Child Protection was also reported as significant with 43% of mothers disclosing their contact over concerns about their children. Australian child protection figures show children under the age of one are most likely to have those concerns substantiated, followed by those aged between one and four years. Further still, over half the mothers reported contact with police. This speaks to the severity of the violence experienced, with research indicating that women are more likely to contact police when the violence is severe and life threatening (Lee, Park, & Lightfoot, 2010).
Overall, the PABC intervention was associated with improved scores on outcome measures assessing infant, mother and infant–mother functioning. The RCI analysis showed that only some of the improvements were clinically significant. Although the absence of a control group makes it difficult to draw definitive conclusions as the effectiveness of PABC, the outcomes combined with qualitative reports of the mother, suggests improvements. Post-intervention, analysis showed that infants had improved socio-emotional competence and had less challenging internalizing, externalizing and dysregulating behaviours. This combined with mother’s reports of improvements in the infant’s gaze, levels of affection, pleasure, pro-social interactions and compliance would suggest positive shifts were found in infant’s functioning. More importantly, it shows a mother’s capacity to reflect, notice and delight in their infant’s capacities to relate; factors that were not always evident at the outset of their involvement. The results suggest that infants appeared to have developed new ways of regulating and modulating their behaviours, whilst mothers were more attuned and available to assist the infants to contain strong emotional responses.
Maternal perceptions of overall attachment showed improvement in the MPAS scores. At a subscale level, the results showed an improvement in ‘Quality of Attachment’ and a ‘Pleasure in Interaction’ as well as a reduction in ‘Hostility’ between mothers and their infants. This indicates a triggering of strong protective factors in the mother–infant relationships, suggesting a possible reworking of maternal representations; an important step in interrupting the transmission of intergenerational violence.
The results from the PIR-GAS found clinician assessed improvements in the adaptive status and dynamic of the mother/infant relationship. In particular, the infant and mothers’ overall functional levels improved; the levels of distress and conflict in the relationship reduced; the levels of resolution in the relationship improved; more adaptive flexibility in the relationship was observed; and the infant’s developmental progress was more positively influenced by the improved quality of the infant–maternal relationship.
The RCI was used to evaluate the effectiveness of the PABC. Results indicate that only a small percentage of participants (8–15%) reported statistically reliable improvements on some of the outcomes measures used. Interpretation of the change found in participants of the current study is difficult because of a lack of studies in infant mental health, and in family violence in particular, that report the RCI. The small change could be attributed to the use of measures inappropriate for this population and/or not sensitive enough to detect change. Additional research is needed to investigate, for example, what is the optimal number of sessions for interventions that targets infants and mothers affected by family violence and can improvements be sustained?
These outcomes suggest that dyadic group interventions can assist mothers and infants who have experienced significant relational disruption – creating opportunities in the ‘here and now’ for the infant/mother relationship to rework inhibitive maternal representations and encourage availability to respond sensitively to their infant’s efforts in managing affect regulation. The MPAS scores generally reflected that the mothers presented with either overly negative or idealised representations of their infant(s) pre-group, echoing concerns expressed by others that such rigidity in maternal thinking “constitutes a risk to future parenthood among high-risk mothers” (Flykt et al., 2012, p. 135). Significantly, post-program these scores reflected mothers holding a more realistic picture of their infant and their relationship, and an improvement in their attachment experience overall. Further, adopting an ‘infant-led’ approach can be reparative in attending to disruptions in attachment. A large proportion of the mothers attending the program had left violent relationships suggesting a desire to actively protect their infant from harm. The earlier a mother leaves a violent relationship the greater the opportunity there is to develop a secure attachment with their child (Levendosky et al., 2011).
This evaluation certainly highlights the difficulty in obtaining data and sourcing appropriate measures. Re-collecting demographic details of participants at the conclusion of the program would have been illuminating, and in hindsight, also a valuable indicator of the impact this intervention. Engagement and service delivery took precedence over collecting data. Had we more time and resources, further measures could have been included, however, the challenge remains as to how to collect data without compromising engagement. Additionally, the age range of some assessment tools proved problematic. Ideally, using measures targeting infants and toddlers 0–5 would allow for maximum data collection. The BITSEA can only be administered on infants aged 1–3 years, so information for infants under 12 months (not reported here) was collected using the Crying Patterns Questionnaire (James-Roberts & Halil, 1991).
Notwithstanding the difficulties presented in undertaking this small evaluation, there is sufficient evidence to indicate that the program was beneficial and justifies a more methodologically sound study. The short term nature of the intervention precludes any ability to lay claim to interrupting the transmission of intergenerational violence. However, neither does this eliminate the possibility that such early work, undertaken at such a pivotal time in the life of the infant/mother relationship has as much potency for repair as does family violence for relational rupture.
Recognition of the high costs that family violence wrecks on the physical, social and economic fabric of our society is not unique to Australia (WHO, 2013). What these costs are, have been easier to identify than how to address them. Further still, how this then translates into supporting interventions which endeavour to reach this sometimes hidden, often suspicious and particularly vulnerable client group is complex and not necessarily assisted by economically driven agendas to roll out, across all sectors, blanket responses that risk missing the target group altogether. The complexity of how to both reach and then research this client group needs urgent attention, thoughtfulness and time. Smaller, community driven projects being undertaken by service providers are worthy of both financial and research support. As (Breckenridge & Hamer, 2014) notes, “Traditionally, quantitative research methodologies grounded in the natural sciences (with the randomised control trial as the ideal model) have tended to dominate understandings of what is accepted as the ‘best’ or ‘gold standard’ evidence. However, criteria for gold standard evidence are not easily implemented in the complex arena of DFV practice and do not fully encompass the importance of the worker–client relationship” (p. 1).
The results of this evaluation are encouraging and indicate that not only is it possible to engage this vulnerable client group in treatment but also to impact positively on their functioning and attachment. Given the overwhelming evidence that exposure to family violence does impair the neurobiological, psychological and social functioning of infants, the present findings add support to the urgency of intervening early. Focusing on improving the functioning of mother–infant relationships may break the cycle of poor attachments as well as mitigate future mental health issues.
Footnotes
Acknowledgements
We would like to recognise the contributions and talents of other members of the PABC program including Robyn Sketchley, Meredith Banks, Katie Waters and Tim O’Leary as well as the amazing support provided by Susannah Duncan. Finally, we are indebted to the Sidney Myer Fund and The Grosvenor Foundation, through the Victorian Women’s Trust, for funding this program. For any further information about any of the data, samples or models used in this research please contact Kathy Eyre: kathy.eyre@rch.org.au, or download material about the AFVP on 
.