Domestic Violence Victimization Risk Assessment in Children and Adolescents: A Systematic Review

Abstract

Background: Risk assessment is the process of collecting information towards the goal of protecting the physical and psychological integrity of the victim, taking into account factors associated with violence to assess the severity of violence, protect victims, and prevent recidivism. This type of risk assessment is commonly used in situations of domestic violence and needs to be adjusted for the contexts of child and adolescent victimization. Objective: Resources and standardized criteria to guide a child-centered domestic violence victimization risk assessment are lacking. This systematic review aimed to evaluate the instruments, risk factors and outcomes identified in the literature for situations of domestic violence involving children. Methods: Following the PRISMA protocol, 313 articles from the EBSCO, Web of Science and PubMed databases were screened and 13 were identified for analysis. Results: An analysis of the characteristics of some instruments created to assess the impact of domestic violence involving children shows that caregivers’ risk factors are strong predictors of child abuse, highlighting the interrelationship with other factors, as well as warning about the cumulative risk, including child homicide. Conclusions: The literature confirms the importance of family system factors regarding the risk of the mistreatment of children in situations of domestic violence. Risk assessment must cater to the needs and specificities of individual children.

1. Introduction

While the assessment of the problems associated with experiencing domestic violence continues to evolve, there is still a lack of specific responses aimed at predicting and managing the risk for children and adolescents who are victims of this phenomenon (Fitz-Gibbon et al. 2019). This systematic review is the culmination of works that enumerate multiple factors in the risk assessment of victimization in children and adolescents who are victims of domestic violence, guided by defined criteria. Additionally, this systematic review aims to gather foundational information, identify factors and systems, and strengthen procedures in response to the need for a clear child-centered assessment scheme for children victimized by domestic violence.

Domestic violence is defined as any type of violence, attempt or physical threat that is perpetrated by an individual, whether female or male, against a person with whom they maintain or have maintained an intimate relationship (Baldry 2003; Sani and Benavente 2021). This definition includes all existing types of violence, such as physical, psychological, emotional, sexual, economic, and stalking. A close relationship between the aggressor and the victim is an increased risk factor for more frequent and severe violence, with a greater probability of occurrence when they cohabit or maintain frequent contact (Almeida and Soeiro 2010; Baldry 2003; Jongenelen et al. 2020).

Domestic violence against children occurs when a child, directly or indirectly, witnesses situations of violence between parents or parental figures (Sani 2018); however, this does not always translate into situations of child abuse (Shlonsky and Friend 2007). Domestic violence against children and child abuse can occur together or be a consequence of one another. For example, according to Ferreira et al. (2019), domestic violence is associated with patriarchal and adult-centered contexts, which emphasize the power of the man/adult over the child, leading to violent relationships between parents and children. On the other hand, child abuse can occur when severe forms of domestic violence trigger negative emotional activation and emotional dysregulation in parents, which, in turn, can contribute to the occurrence of disruptive parenting practices, increasing the likelihood of the child experiencing abuse (Ferreira et al. 2019).

The World Health Organization [WHO] (2022) defines child abuse as the “perpetration of physical, sexual and psychological/emotional violence and neglect of infants, children and adolescents aged 0–17 years by parents, caregivers and other authority figures, most often in the home but also in settings such as schools and orphanages” (p. 2). Research on children’s exposure to domestic violence confirms the negative impact of this experience on their overall adjustment (Sani 2020), with signs and symptoms manifesting through emotional, cognitive, physiological, behavioral, and social problems of different severity, in the short, medium or long term (Almeida et al. 2022; Artz et al. 2014). Over time, it has been proven that victimization due to exposure to domestic violence is a risk factor for the development of internalizing and/or externalizing problems (Malta et al. 2019; Sani 2006), such as depressive symptoms (Lv and Li 2023; Sá et al. 2009; Ximenes et al. 2009), a lack of self-control, impulsive behaviors (Barboza and Dominguez 2017; Pesce 2009) or substance use (Maia and Barreto 2012; Malta et al. 2019). For example, internalizing problems can develop up to 10 years after experiencing violence (Vu et al. 2016).

A child exposed to situations of domestic violence experiences various concurrent forms of victimization (Shlonsky and Friend 2007) and their impacts are identifiable markers of their exposure (Sani 2019). A child in the context of domestic violence is more likely to be exposed to, or be the direct target of, additional forms of victimization (Hanson et al. 2006; Malta et al. 2019; Sani et al. 2021). However, there is limited recognition of the impact of violence on children’s well-being, with an urgent need to consider the impact of violence on children (Sani 2019).

The literature emphasizes the importance of considering the variability in risks and each child’s individual response to violence in the family context (Fitz-Gibbon et al. 2019). The age at which violence is observed may have varying impacts on the likelihood of future perpetration or victimization (Bazargan-Hejazi et al. 2014). Furthermore, studies show that understanding the variability in children’s adjustment requires considering parental cognition, emotions and behaviors (Chiesa et al. 2018). For example, children’s emotional and behavioral dysregulation after exposure to violence may vary depending on the intensity of affection provided by the parents, as well as their strategies during marital conflict and its resolution (Bonache et al. 2019). Moreover, the type of marital conflict may also impact children’s adaptation to how they perceive violence (Grych and Fincham 1990).

Addressing the issue of child abuse extends beyond merely discussing safety and protection concerning the individuals who perpetrate such violence. For a comprehensive assessment and intervention to take place, it is imperative to consider the child’s developmental stage and the factors associated with domestic violence, while acknowledging the potential risk of long-term damage (Shlonsky and Friend 2007).

Furthermore, in the realm of risk assessment for victimization caused by domestic violence, a contentious issue arises when an adult victim and a child victim co-exist in a violent environment. The risk assessment cannot be guided by the same parameters in such cases. An adult victim and a child possess distinct needs and characteristics concerning risk. Although these situations may be connected, there are distinct factors to observe (Jenney et al. 2014; Stanley and Humphreys 2014), resulting in insurmountable failure when there is a clear disregard for the immediate need to conduct a risk assessment oriented towards the child.

With an increased awareness of the problem of domestic violence and the negative impact on victims, both in childhood and adulthood, there have been calls for decisive action. The recognition of the devastating impact that domestic violence has on children, directly or indirectly, has laid a strong foundation for heightening governmental awareness of the problem (Skivenes and Stenberg 2015).

Several authors refer to child victimization by domestic violence as a pervasive issue that impacts modern society. It is reported to occur in 60% of households where such incidents are identified (Hamby et al. 2010; Morais-Gonçalves et al. 2018; Sani 2019). In terms of child abuse, Stoltenborgh et al. (2015) found that it is a global and prevalent problem, affecting millions of children. In 1993, in the United States of America (USA), child protection service agencies submitted approximately 2 million reports regarding alleged child abuse and neglect for evaluation (Camasso and Jagannathan 2000). According to Baird and Wagner (2000), the number of abuse and neglect allegations in the USA dramatically increased over the last two decades of the 20th century. Consequently, most child protection services were pressured to respond effectively to the issue, leading to legal action in most states, media exposure due to child fatalities, and growing concerns about the responsibility of professionals and services. This has spurred the desire to discover new strategies and resources to address the phenomenon. Damashek et al. (2014), in a study based on a documentary analysis of files from the Child Death Review Board in Oklahoma from 2000 to 2003, examined child fatalities due to inadequate caregiver supervision. They noted that approximately 30 to 50% of these victims or their families had previously been flagged to child protection services, whether through a report, investigation or otherwise.

Risk assessments and interventions have increasingly been used for the mitigation and eradication of domestic violence victimization (Baird and Wagner 2000; Sani and Lopes 2018). Risk assessment guides teams in making critical decisions within limited timeframes for child protection in healthcare services (Baird and Wagner 2000; Kearney et al. 2023), e.g., identifying, at an early stage, families at high risk of child maltreatment (Bisagno et al. 2023; van der Put et al. 2017). Predicting the onset of abuse can be better achieved through an actuarial assessment (using criteria and structured measures) than through clinical assessment (Baird and Wagner 2000; van der Put et al. 2017). The appropriate and prompt execution of interventions, considering risks, will be crucial to the prevention of recidivism (Morais-Gonçalves et al. 2018).

2. Risk Assessment: From Conceptualization to the Relevance of a Guided Practice

Risk assessment for domestic violence victimization is a process of collecting information towards the identification of signs and indications of violence, as well as protective factors. The aim is to prevent the recurrence or escalation of violence (Almeida and Soeiro 2010; Morais-Gonçalves et al. 2018). “Risk” can be defined as the likelihood of a violent act occurring, i.e., the prediction of a negative event. Violent acts vary, depending on the perpetrator’s motivation, their relationship with the victim or the nature of the violence enacted. Risk is multifaceted and, for this reason, the nature, severity, frequency, duration, and imminence of any form of future violence must be considered (Morais-Gonçalves et al. 2018). Risk factors can be static, when they cannot be modified through intervention, or dynamic, when they can be modified through intervention, resulting in changes in the level of risk (Antunes et al. 2020; Morais-Gonçalves et al. 2018). Protective factors such as the individual characteristics of the subjects, their interpersonal relationships, or the meanings constructed in relation to the experience must also be taken into account, as they can contribute to risk reduction (Sani and Pereira 2020).

When assessing risk, it is essential to first determine whether the child is in immediate danger, as this is the only way to prevent more serious harm from occurring. To address these immediate needs, various domestic violence victimization risk assessment instruments have been developed (Navarro-Pérez et al. 2023; van der Put et al. 2017; Vial et al. 2020b). Domestic violence victimization risk assessment can and should include risk management, i.e., in addition to evaluating the risk factors present, technicians must identify and determine the most appropriate risk management strategies, taking into account the needs and specificities of the victim and the type of violence experienced (Almeida and Soeiro 2010; van der Put et al. 2017). Most existing tools are generally designed to predict the risk of immediate harm or the risk of the recurrence of mistreatment (Shlonsky and Friend 2007).

Ideally, risk management should help minimize factors that contribute to undesirable outcomes (e.g., harm to children), whether they originate from parents or child welfare workers (Gambrill and Shlonsky 2001). The assessment should focus on aspects relating to the family context. For instance, maternal depression is a significant factor associated with child abuse. This mental health condition is prevalent in situations of domestic violence, and its evaluation can predict child vulnerability (Vial et al. 2020b). In addition to the context, the individual characteristics of the child, such as their gender, age, perceptions and interpretations of violence, play a moderating role in the relationship between exposure to domestic violence and the child’s adjustment (Soares and Sani 2016). It is also crucial to recognize that children may react differently to similar forms of violence (Shlonsky and Friend 2007).

Services targeting domestic violence have primarily focused on adult victims, aiming to ensure their protection and safety (Stanley and Humphreys 2014), often overlooking child victims. These procedures are generally justified by the notion that protecting the adult victim also ensures the safety of the child (Sani and Pereira 2020). However, it is important to recognize that children may have different needs to adult victims (Øverlien 2011; Stanley et al. 2012). Therefore, there is an urgent need for differentiated risk assessment, depending on whether the focus is on adult or child victims of domestic violence.

Vial et al. (2020b) argue that not all critical aspects are considered in the immediate domestic violence victimization risk assessment. These include emotional abuse, situations where caregivers cannot protect the child from harm inflicted by third parties, and instances where a caregiver’s psychiatric disorder poses a threat to the child (Sani and Lopes 2019). In Portugal, the risk arises merely from the child’s exposure to domestic violence, necessitating the assessment of the child as a legally recognized victim (Law no. 57/2021) (Portugal 2021). Child maltreatment is not assessed merely by inquiring whether the child was physically or emotionally abused during a domestic violence episode (Shlonsky and Friend 2007). Observable behavior in a child does not transparently reflect what the child is feeling, as internal symptoms such as anxiety cannot be detected solely through external behavior (Costa and Sani 2007). Context observations, such as those made during home visits, are important for assessing child abuse risks but are somewhat limited and need improvement (van der Stouwe et al. 2023). This improvement could include the addition of specific detection indicators, which could be incorporated into more structured observation grids.

Therefore, this systematic review aims to answer the following research questions: (i) What instruments for domestic violence victimization risk assessment have been considered in situations involving children? (ii) What risk factors associated with domestic violence victimization have been identified in situations involving children? and (iii) What consequences/impacts have been observed in child victims of domestic violence?

3. Method

3.1. Search Information

The systematic review presented (PROSPERO register CRD42024526694) includes studies assessing the risk and implications of victimization due to domestic violence in children and adolescents.

The search was expanded to studies published in all languages across three databases, as recommended (Donato and Donato 2019). Based on the theme and research area, we selected the following databases: EBSCO, Web of Science, and PubMed. EBSCO includes major reference citation indices such as Scopus and subject indices like APA PsycInfo to assess relevance and quality. Moreover, Web of Science is recognized as an excellent platform that offers bibliographic research databases to support scientific and academic research. Additionally, we included PubMed due to the presence of articles discussing the victimization of children and adolescents through domestic violence, which may have otherwise been overlooked in our research.

The search was limited to studies published between 2000 and 2022 in academic and peer-reviewed journals.

3.2. Search Terms

The search terms and expressions, adapted for each of the databases, were as follows:

EBSCO: TI (“risk factors” OR “risk assessment” OR “risk management” OR scales OR check-list OR instruments) AND AB (“child victims” OR “child protection” OR child OR “adolescent victims” OR adolescence OR “child protection” OR “child welfare”) AND AB (“family violence” OR “domestic violence” OR “intimate partner violence” OR “family abuse”);

Web of Science: ((TI = (“risk factors” OR “risk assessment” OR “risk management” OR scales OR check-list OR instruments)) AND AB = (“child victims” OR “child protection” OR child OR “adolescent victims” OR adolescence OR “child protection” OR “child welfare”)) AND AB = (“family violence” OR “domestic violence” OR “intimate partner violence” OR “family abuse”); and

PubMed: ((“risk factors”[Title] OR “risk assessment”[Title] OR “risk management”[Title] OR scales[Title] OR check-list[Title] OR instruments[Title]) AND (“child victims”[Title/Abstract] OR “child protection”[Title/Abstract] OR child[Title/Abstract] OR “adolescent victims”[Title/Abstract] OR adolescence[Title/Abstract] OR “child welfare”[Title/Abstract] OR)) AND (“family violence”[Title/Abstract] OR “domestic violence”[Title/Abstract] OR “intimate partner violence”[Title/Abstract] OR “family abuse”[Title/Abstract]).

3.3. Study Selection Criteria

The selection of studies was based on the following inclusion criteria: (a) samples involving children and adolescents under 18 years of age; (b) victims of domestic violence; (c) studies employing quantitative methods; and (d) scientific articles published in peer-reviewed journals. Correspondingly, the exclusion criteria were as follows: (a) adult samples; (b) victims of crimes other than domestic violence; (c) qualitative or theoretical studies; (d) systematic reviews and meta-analyses; and (e) books and/or book chapters.

Figure 1 describes the selection process of studies included in this systematic review. The search carried out in three databases yielded a total of 313 articles, published between 2000 and 2022. After excluding 96 duplicates, 217 articles remained for an analysis of their title and abstract. Of these, 185 articles were excluded based on the criteria. Finally, 32 articles were eligible for full-text reading, and of these, 18 were excluded. One additional study was not included due to inaccessibility. Ultimately, a total of 13 articles were included in the final study.

The quality of the articles (cf.

Table 1. Analysis of the quality of selected articles using the adapted Prisma Checklist 2020 (Page et al. 2021).

Articles	1	2	3	4	5	6	7	8	9	10	11	12	13	Total
Berens et al. (2019)	2	2	1	2	2	2	2	1	1	2	2	2	2	23 High
Bonfim et al. (2011)	2	2	2	2	2	2	2	1	2	2	2	2	2	25 High
Craig et al. (2021)	2	2	2	2	2	2	2	2	2	2	2	2	2	26 High
Diamond and Muller (2004)	1	2	2	2	2	2	2	2	2	2	2	2	2	25 High
Erolin et al. (2014)	1	2	2	2	2	2	2	1	2	2	2	2	2	24 High
Flander et al. (2009)	1	2	2	1	2	1	2	1	2	2	1	2	2	21 Average
Miyamoto et al. (2017)	2	2	1	1	2	2	2	2	2	2	2	2	2	24 High
Olszowy et al. (2013)	1	2	2	2	2	1	2	2	2	2	2	2	2	24 High
Reif and Jaffe (2021)	1	2	2	2	2	2	2	2	2	2	2	2	2	25 High
Sičić and Mužinić (2007)	0	2	2	1	1	1	2	2	2	1	2	2	1	19 Average
Sidebotham and Heron (2006)	2	2	2	2	2	2	2	2	2	2	2	2	2	26 High
Terelak et al. (2019)	1	2	2	2	2	2	2	2	2	2	2	2	1	24 High
Vial et al. (2020a)	1	2	2	2	1	1	2	2	2	1	2	2	2	22 High

Grades: 0—not reported/not specified; 1—unclear/reported to some extent; 2—adequately done.

) was analyzed using the Prisma Checklist 2020 (Page et al. 2021), adapted to the purposes of this study, and 13 analysis topics were considered [(1) title; (2) abstract; (3) rationale; (4) objectives; (5) protocol; (6) eligibility criteria; (7) information sources; (8) inclusion/exclusion criteria; (9) data collection process; (10) study design; (11) main measures; (12) summary of main results; and (13) conclusion and study limitations]. After applying this checklist, articles were evaluated on to a Likert-type scale in which 0 (not reported/not specified) indicated that the article lacked information on the topic, 1 (unclear/reported to some extent) suggested that the authors provided some information on the subject, and 2 (adequately done) showed that the article presented the topic exhaustively, with a complete methodological explanation.

The authors of this systematic review independently analyzed each article using the same checklist, and then combined their analyses to obtain the final values presented in Table 1. Subsequently, to verify inter-rater agreement, Cohen’s Kappa coefficient was calculated, resulting in a concordance value of 95%, indicating a high level of agreement (Landis and Koch 1977).

4. Results

In the following, studies with lower adapted Prisma scores should be viewed with more caution compared to studies with higher scores, which are methodologically robust. The majority (84.6%) of studies included in this review had a high score (Table 1).

4.1. Instruments for Assessing the Impact of Domestic Violence

Two studies included in this systematic review evaluated, respectively, the structure and validity of the Parent–Child Conflict Tactics Scale (CTSPC) (Bonfim et al. 2011) and the Childhood Psychosocial Adversity Scale (CPAS) (Berens et al. 2019).

The results of the exploratory and confirmatory factor analysis regarding the structure of the CTSPC revealed that the items may behave differently from that proposed in the original scale. This has been confirmed by international studies carried out with this same scale, e.g., in the exploratory factor analysis, factor 1 incorporated items linked to the degree of violence and related to physical abuse, whereas factor 2 included items related to a medium degree of violence, encompassing items of corporal punishment and violence considered disciplinary; factor 3 showed lower correlations and is more associated with non-violent conflict resolution tactics (Bonfim et al. 2011). In relation to the confirmatory factor analysis, factor 1, termed non-disciplinary abuse in the original scale, included items from the scales of physical abuse, corporal punishment and psychological aggression, whereas factor 2, originally termed disciplinary abuse, included corporal punishment and psychological aggression items, also incorporating an item referred to as physical mistreatment; factor 3 was defined as the use of non-violent disciplinary tactics and included only items from the non-violent discipline scale. No items that, if removed, could significantly improve the internal consistency of the scale were found (Bonfim et al. 2011). Considering the differences in how the factors were grouped, not only in this study, but in other international studies where the factors were organized differently from the original study, the authors state that domestic violence against children can be interpreted and evaluated in different ways (Bonfim et al. 2011).

As for the validity of the CPAS scale, which measures cumulative risk and is designed to predict the levels of early childhood cognitive development (ECD), the results arrived at a final three-factor model with subscales relating to harsh discipline and abuse, neglect and caregivers’ emotional unavailability. Additionally, the final exploratory factor analysis obtained caregiver-focused items measuring depression, social isolation, intimate partner violence, physical violence, verbal abuse, family conflict, economic and community environment, family economic stress, and community adversity (Berens et al. 2019).

Berens et al. (2019) correlated the results obtained in the administration of this scale with the Edinburgh Postnatal Depression Scale (EPDS) and the Multidimensional Scale of Perceived Social Support (MSPSS). As expected, the results effectively demonstrated that physical violence perpetrated by an intimate partner, verbal abuse and family conflict were significantly correlated with maternal depression. The subscales related to prevention, punishment, restriction, economic and family stress were significantly correlated with the score obtained on the CPAS scale, revealing that severe parental discipline towards children has a negative impact on cognitive development in early childhood (Berens et al. 2019).

The greater the cumulative psychosocial risk in early (48–60 months) childhood (high CPAS scores), the greater the impact on child development (raw IQ scores); however, this was not confirmed for other age groups (Berens et al. 2019).

4.2. Parental Problems and/or Psychopathology as a Risk Factor for the Occurrence of Child Abuse

The majority of studies considered focus on the risk factors related to parental problems that reveal associations with the occurrence of child maltreatment and the emergence of problems in the short and long term (Diamond and Muller 2004; Erolin et al. 2014; Flander et al. 2009; Sičić and Mužinić 2007; Sidebotham and Heron 2006; Vial et al. 2020a). Parents with low academic performance and a history of childhood abuse and psychiatric illness were two to five times more likely to have their children reported to child protection services than the reference population (Sidebotham and Heron 2006). The impact of poverty on child abuse appears to be a mediated by the socioeconomic history of the parents. Maternal work appears to reduce risk, but not when other social and parental factors are considered. Sidebotham and Heron (2006) found that a negative social network increases the risk of registration with child protective services. Family characteristics, such as single parents, restructured families, parents who do not recognize positive attributes and describe their children in a negative way and/or who have premature children, also have a higher risk of being flagged in child protection services (Sidebotham and Heron 2006).

The incidence of child abuse in families with parental behavioral disorders showed a statistically significant difference between groups in terms of exposure to different types of emotional, physical, and sexual abuse or neglect. Participants who reported behavioral disorders in one of their parental figures were relatively more exposed to this abuse. However, no statistically significant differences were found between the types of abuse suffered when considering the presence of mental disorders in the parents (Flander et al. 2009). Furthermore, the results demonstrated that in relation to feelings of social and family loneliness, participants who reported the presence of disorders in at least one of their parents felt lonelier than participants who did not report mental disorders in their parents. This difference was not found in relation to feelings of loneliness with regard to love. These results align with previous studies showing that mental illness and parental alcoholism represent risk factors for different forms of child abuse (Flander et al. 2009).

Vial et al. (2020a), in a multidimensional network study, evaluated the interrelationship of risk factors associated with child abuse, both for children referred to child protection services (high risk) and those referred to community services (low risk). The factors were positively correlated in both networks, but were stronger in the high-risk sample than in the low-risk sample. Children referred to protection services were older, had a greater number of risk factors, had a higher risk of maltreatment in the future, and had immediate risk indicators, including factors associated with a history of domestic violence, a problematic relationship between caregivers, and experience of family conflict. The results of the study highlighted the strong interrelationships with other static risk factors, including those associated with a history of child abuse, domestic violence, and caregiver neglect. Confirming what is already known about transgenerational violence, these studies also showed the possibility of an exponential increase in the risk of child abuse in the presence of extra risk factors (Vial et al. 2020a).

Erolin et al. (2014) sought to understand the causes and effects of post-traumatic stress disorder (PTSD) in children and their families. Almost 60% of their sample reported symptoms of PTSD (DSM-V diagnostic criteria), of which around 90% had been exposed to situations of violence. Furthermore, of the children exposed to situations of violence, 65.5% reported continuous interparental violence, highlighting that the child’s exposure to family violence throughout their lives, along with exposure to violence in the community, increases the risk of developing PTSD or other disorders. These risks were higher still for females exposed to violence within the home (Erolin et al. 2014).

In a study of the long-term impact of witnessing domestic violence (Diamond and Muller 2004), the most common form of violence perpetrated by one spouse towards the other was psychological violence, encompassing behaviors such as speaking in a hostile manner, ridiculing in general, and ridiculing feelings. Regardless of whether the perpetrator was the mother or father, the most common forms of physical violence were slaps, kicks, pushing and throwing objects. The degree of association with children’s internalizing behaviors was relatively higher when their exposure to violence was characterized as ‘bigger’ (and vice versa). Furthermore, the development of psychopathology, including worsened symptoms of PTSD following exposure to physical and/or psychological violence, is significant, especially when perpetrated by the maternal figure (Diamond and Muller 2004). The impact on the development of PTSD symptoms and internalizing problems was greater when perpetrated by the father figure, but when the perpetrator was the mother figure, there tended to be greater externalizing symptoms. Finally, the results suggest that witnessing physical, verbal and psychological violence of a less violent nature has a long-term impact (Diamond and Muller 2004). However, victimization does not always lead to PTSD, for instance, in cases of sexual harassment, where adjustment disorder may occur (Marcos et al. 2023).

Externalizing behaviors as a result of exposure to domestic violence are common. Sičić and Mužinić (2007) studied the impact of risk factors on self-harm behaviors in adolescents. For adolescents who demonstrated self-mutilation behavior, approximately 20% were direct victims of family violence, and 31% had family members with some type of dependence, whether on alcohol or pills. Additionally, around 40% of the adolescents included in the study demonstrated intra and interpersonal aggressive behavior as a result of sexual, physical and psychological abuse, with physical and psychological abuse co-occurring twice, and with all sexual abuse occurring in an intra-family context. Furthermore, 54% of the adolescents in the study had already attempted suicide or had suicidal ideation behaviors. Finally, 80% of the adolescents lacked love, understanding, affection and support in the interpersonal relationships that were prevalent in the family, not having access to the positive characteristics that a child needs to develop in an adapted and healthy way (Sičić and Mužinić 2007).

4.3. Contextual Issues as Risk Factors for the Occurrence of Child Victimization

Terelak et al. (2019) studied the incidence of domestic violence in children and adolescents by considering the type of community in which they reside (rural, urban and urban–rural areas). Of the children or adolescents who suffered different forms of domestic violence, around 50% belonged to rural and urban–rural areas, and around 66% belonged to urban areas. The co-occurrence of three or more forms of violence in the family environment potentiated the development of the cumulative symptoms of pathological family life, which could put the health and even the lives of victims of domestic violence at risk. This climate of violence, concentrated in the family nucleus, affected around 14% of the rural, 12% of the urban–rural and 24% of the urban community. Psychological violence was predominant; physical violence was the second most common in the urban–rural community and the third most common in the urban community (Terelak et al. 2019).

Physical violence against children and adolescents was mainly in the form of beatings, using objects without leaving marks and/or beatings, causing bruises, and acts of pushing or stabbing. Although found in all three communities, negligence had a greater impact on the urban community due to the perpetration of behaviors associated with ignoring feelings and emotions, a lack of concern for the problems of children and adolescents, and the underestimation of children’s illnesses. Finally, economic violence and sexual assault were perpetrated in all three types of communities (Terelak et al. 2019).

4.4. Cumulative Familial Risk and Change in Genetic Pattern in Children

The main objective of the study by Craig et al. (2021) was to examine the extent to which the cumulative family risk and number of maltreatment experiences affect DNA patterns, specifically the SLC6A4 gene. In more than 90% of children who suffered more than one type of maltreatment in childhood, the index of cumulative exposure to family risk factors revealed a significant positive association with DNA mutation patterns in two different genes (SLC6A4 and CpGs). However, no significant correlation was found between the number of maltreatment experiences and DNA mutation levels. However, the DNA mutation was significantly explained by the number of experiences of maltreatment and the characteristics of the children, i.e., the score on the cumulative family risk index and the number of experiences of abuse were positively correlated with higher levels of DNA mutation. Witnessing interparental violence leaves epigenetic traces that remain even after children are removed from their caregivers and placed in foster care, suggesting that family risk factors play a critical role in the DNA mutations that occur in children (Craig et al. 2021).

4.5. Risk Assessment of Child Homicide

Different studies evaluated the occurrence of child homicide within violent interparental relationships by considering instances where individuals, family, and caregivers seriously abused children, resulting in hospitalization or death, among cases investigated by child protection services. Furthermore, the studies analyzed whether the presence of children within the relationships increased the likelihood of child homicide (Miyamoto et al. 2017; Olszowy et al. 2013; Reif and Jaffe 2021).

Miyamoto et al. (2017) found that children who were hospitalized or died were often male and had a significantly younger maternal caregiver. A structured risk assessment revealed that the families of children who had been hospitalized or who had died were more frequently assigned to moderate and high-risk categories. Correspondingly, the group of children who were neither hospitalized nor had died were more frequently marked as being at low or very low risk. The authors analyzed the possible differences between maternal and non-maternal caregivers. In relation to the former, the variables of the child being male, the family having a moderate or high risk score, and the family having more than three children under the age of five were significantly correlated with a higher risk. By contrast, variables relating to maternal involvement in intimate partner violence and the children’s enrollment in public health insurance were associated with a reduction in risk (Miyamoto et al. 2017). As with maternal caregivers, involvement in violence by an intimate partner and from a non-maternal caregiver is associated with a lower probability of serious maltreatment (Miyamoto et al. 2017). For non-maternal caregivers, variables relating to the younger age of the caregiver, the number of children under the age of five living in the house, and at least one biological child not living with either parent were significantly predictive of serious child abuse.

Another study (Reif and Jaffe 2021) found no significant differences between cases where domestic homicide occurred with or without children in the violent relationship, with both groups having the same number of risk factors. In cases with children, there was a higher percentage of significant risk factors associated with child custody, previous aggression against the victim during pregnancy, and a history of domestic violence in the current relationship. Risk assessments, risk management and safety planning revealed no significant differences between the groups. Within the group that included children, no statistically significant differences were found between cases where children were the targets of violence and were not the targets of violence (Reif and Jaffe 2021).

Olszowy et al. (2013) analyzed three instruments for assessing the risk of marital aggression to assess whether they were suitable for identifying cases with a high risk of child homicide. They found that the level of risk was similar and was identified by the instruments as high when children were victims of lethal violence or the target of an attempt on their lives, as well as in cases in which children were present in the home when their mother was the victim of homicide. Furthermore, the results revealed that if the maternal figure was identified as being at a high risk of homicide, there was a greater likelihood of child homicide also occurring (Olszowy et al. 2013). In general, the instruments did not reveal significant associations with the child homicide group or the group in which no child homicide occurred. However, both the Danger Assessment (DA) instrument’s assessment of previous threats to hurt the child and the B-SAFER instrument’s assessment of the aggressor’s interpersonal adjustment problems revealed significant associations with cases in which child homicide occurred or in which there was an attempted homicide (Olszowy et al. 2013).

Table 2. Summary of the characteristics of the studies included in the review.

Author(s) (Year)	Objective(s)	Sample	Method/Instruments	Results and Main Conclusions
Berens et al. (2019)	Describe the development and validity of the Childhood Psychosocial Adversity Scale (CPAS), which measures cumulative risk. The CAPS was designed to predict levels of early childhood cognitive development (ECD).	Recruited through two longitudinal birth cohorts of mother–child dyads: PROVIDE e Crypto. A subset was randomly selected. Recruitment for both cohorts was performed through visits to the homes of all pregnant women. Participants in the development of the instrument were recruited by inviting all individuals who showed up for scheduled appointments to complete the questionnaire (capturing children aged 18, 24, 48 and/or 60 months).	Participant Selection Questionnaire; Semi-structured interviews (n = 20, including 10 with mothers and 10 with field workers) and Focus Group (n = 8, each with 10 participants); Cognitive Interview for the Pre-Test; Depression Scale (EPDS) and Multidimensional Perceived Social Support Scale (MPSS).	The CPAS is a comprehensive measure of childhood psychosocial adversity designed for implementation and validation in global health settings. The data confirmed its usefulness as a research tool that assesses psychosocial risk factors in early childhood in relation to development among children.
Bonfim et al. (2011)	To study the factorial structure of the CTSPC scales in an urban population sample from the Brazilian Northeast.	1370 children from Salvador da Bahia. (53.4% boys; n = 731 and 46.6% girls; n = 639) Child’s age (4–7 years; M = 7.1; SD = 1.7) Age of mothers (18–56 years; M = 32.1; SD = 6.7).	Parent–Child Conflict Tactics Scale (CTSPC).	The factor analysis in this study revealed items that present high factor loadings in the respective scales and others that were different from the original. The results revealed that DV against children is part of a phenomenon that can be interpreted and evaluated in very different ways, with this result proved by the differences in how the factors were organized in this study in relation to the original or other studies.
Craig et al. (2021)	To examine the direct effects between cumulative familial risk and altered DNA patterns, specifically the SLC6A4 gene, and indirect effects between cumulative familial risk and the number of maltreatment experiences and altered DNA patterns and the SLC6A4 gene. To analyze the cumulative familial risk index in explaining DNA mutation patterns in abused children.	33 Children (20 female; 13 male aged between 8 months and 15 years: M = 8.26, SD = 3.63), with a history of abuse recruited from the Clinical Unit of Abuse of the IRCCS Scientific Institute “Eugenio Medea” from Brindisi (Italy). All participants were hosted in an Institution by the Juvenile Justice System.	Saliva sample for epigenetic analysis; Assessment of the Cumulative Family Risk Index (CFR); Maltreatment Classification System (MCS) to assess individual child maltreatment experiences and the type of maltreatment suffered.	In children with a history of maltreatment, the severity of family risk factors and multiple experiences of maltreatment contributed directly to, at least partially, explaining the level of DNA mutation, which may be affected by a combination of direct variables (maltreatment treatment in childhood) and proximal factors (close family risk factors). The strongest predictor regarding DNA mutation was the CFR index (cumulative familial risk index). Children’s age was also a significant predictor of DNA mutation. This study added new evidence of how the family environment can become biologically embedded through epigenetic changes, revealing that these changes remain even after children are removed from their caregivers and placed in foster care, suggesting that familial risk factors play a critical role in causing DNA mutation in children.
Diamond and Muller (2004)	To examine in detail the association between witnessing DV and long-term psychological adjustment.	351 young adults (287 women and 64 men) between 16 and 24 years old.	Sociodemographic questionnaire; Young Adult Self-Report (YARS), a self-report questionnaire with 110 items that measure emotional and behavioral problems and 14 items that measure socially desirable characteristics; Trauma Symptom Checklist-40 (TSC-40), a 40-item self-report questionnaire that assesses adult symptomatology resulting from traumatic experiences; Record of Maltreatment Experiences, self-report (ROME), retrospective measure of victimization occurring up to 17 years of age.	Children who witnessed ‘minor’ psychological DV demonstrated significantly less internalizing behaviors than those who witnessed physical DV, as well as those who witnessed ‘major’ psychological DV. Witnessing DV of a physical or psychological nature was correlated with higher levels of psychopathology, including PTSD symptomatology; this was more significant when the perpetrator of violence was the mother, being more associated with internalizing behaviors, both for male and female victims. When total PTSD symptoms were analyzed, results indicated that the father’s psychological DV towards the mother emerged as the only significant predictor of PTSD symptoms for male victims, while psychological abuse perpetrated by the mother emerged as a significant predictor of symptoms of PTSD for women.
Erolin et al. (2014)	Understanding causes and effects of post-traumatic stress disorder in children and families in Mexico: (a) examine individual and family characteristics of mother-child dyads in the presence/absence of PSTD and (b) potential ecological and familial risk factors for the presence of the disturbance.	87 mother-child dyads (50 boys and 37 girls). Mother’s Age (24–50 years; M = 35.1; SD = 6.6). Child’s age (7–14 years; M = 10.2; SD = 2.1).	UCLA PTSD, assessing the exposure of children aged 7–18 to trauma and PTSD symptoms; Post-traumatic Stress Diagnostic Scale (PDS), to assess exposure to trauma and PTSD symptoms in mothers; Composite Abuse Scale (CAS), to assess exposure to intimate partner violence.	58.8% of children reported at least 1 event that met DSM-V criteria for PTSD for different episodes of violence across their lifetime. In relation to DV, 87.4% of children were exposed to one or more forms of violence. The results indicate that the factor relating to the child’s exposure to family violence throughout their life increases the risk of developing childhood PTSD. Children who live in environments characterized by high levels of violence, both family and community, were at a high risk of PTSD and/or the exacerbation of ongoing disorders. Furthermore, cultural factors reveal a great influence on the expression and meaning of trauma symptoms in individuals.
Flander et al. (2009)	To analyze the relationship between the incidence of child abuse in families and parental behavioral disorders.	4191 high school graduates from Croatian schools (M = 18.13 years).	General questionnaire (sociodemographic data); Questionnaire on child abuse (Karloviÿ 2001) according to the Comprehensive Child Maltreatment Scale for Adults (Higgins and McCabe 2000); Emotional and Social Loneliness Scale (ÿubela Adoriÿ 2004).	Mental illness and parental alcoholism represent risk factors for the occurrence of different forms of child abuse. Participants reporting this type of disorder in at least one parent were more exposed to physical, emotional and sexual abuse and neglect. The study also revealed that participants who reported that at least one of their parents suffered from these disorders had much more significant feelings of family loneliness than those whose parents did not have disorders.
Miyamoto et al. (2017)	Identify individual, family and child caregiver risk factors for serious child abuse, resulting in hospitalization or death, among children and their families investigated by Child Protection Services.	234 cases of children who suffered fatal or serious non-fatal abuse (702 children and their caregivers).	Case-control study of Child Protective Services (CPS) Database analyzed using the Structured Decision-Making Risk Tool®.	Caregiver involvement in IPV, identified in research before the index maltreatment event, was associated with significantly lower odds of a subsequent serious child maltreatment event. This fact can be explained by the increased recognition and response to IPV by CPS and law enforcement authorities over the last two decades. The system’s response to IPV may effectively remove the perpetrator through arrest or restraining order, or it may mitigate the risk by requiring participation in diversion programs or placing children in protective custody. These responses can protect children from the serious outcomes of maltreatment.
Olszowy et al. (2013)	Examined the utility of applying commonly used risk assessment tools to child homicide cases to determine whether they would be a valuable resource in identifying high-risk cases.	40 cases of domestic homicide.	Danger Assessment (DA); Ontario Domestic Assault Risk Assessment (ODARA); Spousal Assault Risk Assessment (SARA).	The results of this study did not indicate significant differences between cases of child and non-child homicide (adult homicide) for the three risk assessment instruments. This result suggested that children living with a female intimate partner may also be considered at risk of lethality if a risk assessment indicates that the female intimate partner is at high risk.
Reif and Jaffe (2021)	Identify the risk of child homicide in the context of DV.	140 cases of domestic homicide.	Ontario Domestic Violence Death Review Committee (DVDRC) Database analyzed through DVDRC risk factor coding form.	In addition to recognizing the risk that children presented when mothers were at risk, professionals who work with families experiencing DV must also be aware of the increased risk posed to children when risk factors were identified, regardless of whether or not obvious forms of mistreatment. Identifying these child-specific risk factors would justify timely risk assessment, risk management, and safety planning practices with families experiencing DV, as these practices play a vital role in reducing their exposure to the associated risk. to DV and the prevention of domestic homicides.
Sičić and Mužinić (2007)	Determine how frequently the risk factors studied were present in adolescents who self-harm.	26 adolescents who were hospitalized or treated at the day inpatient service at the Children’s and Youth Psychiatric Hospital.	Specially designed questionnaire containing psychiatric, social, basic sociodemographic characteristics and other important characteristics of the interviewees.	The research results showed that family dynamics, interpersonal relationships, material circumstances and other risk factors influenced the normal and healthy development of an individual and that, in the absence of the above, several disorders can develop in a young person.
Sidebotham and Heron (2006)	Analyze the multiple factors that affect the risk of maltreatment in young children within a comprehensive theoretical framework.	Of the 14,893 mothers who enrolled, taking into account fetal or early childhood loss and attrition, a total of 14,256 children were followed beyond infancy.	A series of prenatal and postnatal questionnaires.	Reported DV showed a clear association with child maltreatment. However, the relationship did not persist after being introduced into the logistic regression model. This may suggest that both DV and child maltreatment have a similar relationship with family structures and other background factors, and that DV should be viewed as both an adverse outcome and a risk in itself. Although the importance of the association between DV and child maltreatment is not denied, this may be a reflection of underlying risk factors common to both, rather than a direct causal relationship.
Terelak et al. (2019)	Determine the extent of DV against school-age children and youth in different types of communities.	Last class of students from primary and secondary schools in three communities in the province of Western Pomerania, Poland.	The research concept adopted was carried out using the audit questionnaire interview technique, based on previous research, to measure the social scale of DV suffered by school-age children.	The co-occurrence of three or more forms of violence in the family environment seemed to reveal cumulative symptoms of a pathological family life, putting the health and even the lives of DV victims at risk.
Vial et al. (2020a)	This study explored the interrelationship of risk factors for child maltreatment.	Children who were referred to child protective services (high risk of maltreatment) and children who were referred to community services (lower risk of maltreatment).	Actuarial Risk Assessment Instrument Youth Protection (ARIJ).	A cross-sample network was constructed to examine the interrelationships between child maltreatment risk factors in both samples. This strongly interrelated network revealed positive relationships between all factors that were part of the network. Two distinct groups of factors could be identified in this network: (1) a group of particularly static risk factors and (2) a group of risk factors related to DV. The risk factors “History of DV”, “Emotionally absent caregiver” and “Caregiver abused as a child” were the most central risk factors.

summarizes the results and main conclusions, as well as the objectives, type of samples, and instruments used for each study selected.

5. Discussion

In order to obtain a better understanding of the victimization risk assessment of children in situations of domestic violence, a systematic review was carried out on studies that could guide the identification of factors, consequences and instruments able to achieve this objective.

When assessing risk, it is essential to first assess whether the child is in immediate danger, thereby enabling the possibility of preventing more serious damage from occurring. Some risk assessment tools have emerged to respond to these immediate needs (Navarro-Pérez et al. 2023; van der Put et al. 2017; Vial et al. 2020b), but none of them are targeted at children. Olszowy et al. (2013) analyzed the effectiveness of existing instruments for assessing the risk of domestic violence and homicide in children exposed to this violence. They found that certain items could predict child homicide, including those listed in the study by Reif and Jaffe (2021), including factors associated with the parents’ inability to establish and maintain a relationship, explaining the occurrence of a separation based on conflict.

The existence of scales (e.g., CTSPC and CPAS) to assess the impact of domestic violence revealed inconsistencies in the way the items are described. These inconsistencies are explained by cultural differences in interpretation, grouping together taboo ideas associated with social desirability, and the avoidance of conflict for fear of reprisals (Bonfim et al. 2011). CPAS, while supporting the assessment of the prevalence of specific forms of psychosocial risk factors, may also contribute to the identification of early intervention priorities. However, it is necessary to consider that the social environment has a complex relationship with biology and therefore may not accurately predict the impact of domestic violence (Berens et al. 2019). As Fitz-Gibbon et al. (2019) point out, it is important to take into account the variability in risks and each individual child’s response to violence in the family context. For these reasons, and allowing for the selected convenience sample and the small number of participants, the difference in the items compared to the original scale is justified (Berens et al. 2019).

Other included studies focused on contextual, family and psychopathological issues to assess the risk and impact of domestic violence on children and adolescents (Craig et al. 2021; Diamond and Muller 2004; Erolin et al. 2014; Flander et al. 2009; Sičić and Mužinić 2007; Sidebotham and Heron 2006; Vial et al. 2020a). These studies found that the presence of psychiatric problems prior to pregnancy, a low parental age, low academic performance, the socioeconomic environment of families and the prematurity of children can all determine behaviors that influence parenting patterns, increasing the risk of mistreatment (Sidebotham and Heron 2006). This is because these factors are associated with high levels of stress, feelings of isolation and a lack of social support (Berens et al. 2019), and relate to the categories of dynamic and static risk factors for domestic violence. As Vial et al. (2020a) revealed, risk factors are interrelated and, therefore, the discrepancy between the contribution of each of them to the occurrence of domestic violence can be explained by how certain it is that these risk factors lead to an exponential increase in child abuse. This confirms the need to take into account the intergenerational factors associated with domestic violence, especially parental history (Chiesa et al. 2018; Fitz-Gibbon et al. 2019; Shlonsky and Friend 2007). Chiesa et al. (2018) showed that understanding the variability in children’s adjustment requires considering parents’ cognitions, emotions and behaviors. Finally, differences in the incidence of domestic violence depending on the area of residence (e.g., rural, urban) are explained by the different ways in which it is experienced and even recognized by third parties, but this requires further studies (Terelak et al. 2019).

Miyamoto et al. (2017) reported that domestic violence is associated with a lower probability of serious child abuse, but this result was explained by the recognition, increase in and mobilization of responses to domestic violence by child protection systems, facilitating the removal of the aggressor (Miyamoto et al. 2017). This does not mean that mistreatment does not occur, but rather that more serious mistreatment tends to be more easily avoided.

Since children exposed to domestic violence are victims of different forms of victimization that occur simultaneously (Shlonsky and Friend 2007), there is a negative on their overall development (Sani 2020). Consequently, for a detailed assessment of the development of long-term damage and related interventions, it is necessary to take into account the child’s level of development and the various factors associated with domestic violence. Therefore, working on and intervening in child abuse cases is more complex than just discussing issues of safety and protection concerning the perpetrators (Shlonsky and Friend 2007).

The intergenerational nature of violence confirms that domestic violence has an impact on the social and emotional behaviors of adolescents (Sičić and Mužinić 2007). The aggressive behaviors of adolescents can be explained by exposure to domestic violence, which is the only way they know how to resolve conflicts and externalize the anger they feel, explaining the development of internalizing and externalizing symptoms and behaviors (Sičić and Mužinić 2007). As Malta et al. (2019) and Sani (2006) have pointed out, victimization through exposure to domestic violence is a proven risk factor for the development of internalizing and/or externalizing behavioral problems. Furthermore, long-term symptoms in male victims exposed to psychological domestic violence are more evident when the perpetrator is the mother, and can be better explained by the deterioration that occurs in the mother–child relationship than by the occurrence of domestic violence per se (Diamond and Muller 2004). The probability, severity, expression, and development of post-traumatic stress disorder varies depending on the prevalence of ecological, family and contextual factors, explaining the increase in symptoms and the impairment of psychosocial functioning, especially in females (Erolin et al. 2014).

Flander et al. (2009) found that emotional and social issues occur in children who are victims of domestic violence associated with the existence of psychopathology and parental alcoholism. High levels of feelings of family and social loneliness, associated with high levels of stigmatization, result in children isolating themselves from their peers, having difficulty in establishing social relationships, and greater likelihood for suffering other forms of victimization (Hanson et al. 2006; Malta et al. 2019; Sani et al. 2021). However, children and adolescents exposed to different types of mistreatments can still develop strong emotional bonds, seeking emotional support in relationships (Flander et al. 2009).

Craig et al. (2021) showed that risk factors and impacts of exposure to domestic violence can explain the variance in DNA mutation levels. Proximal family risk factors are associated with high levels of parental stress and variations in early-stage care, resulting in altered neural, hormonal, cognitive and behavioral responses in children, with the degree of impact dependent on the age at which mistreatment occurs (Craig et al. 2021).

An increase in the frequency and severity of domestic violence is associated with an increase in the probability of child abuse, which may culminate in the death of the child. This can be explained by attacks that occur during pregnancy, custody disputes over the child, or jealousy, with homicide being the result of attempted or broken relationships (Reif and Jaffe 2021).

This systematic review targeted studies that list multiple factors in the assessment of risk in children and adolescents who are victims of domestic violence, subject to our inclusion/exclusion criteria. Although we used three recognized databases that cover current and high-quality articles, allowing for the broad coverage of relevant scientific literature, we could have also included other databases to increase the number of studies for analysis in order to obtain a broader and more robust response to the research questions. We believe that one of the limitations of our study is the conceptual definitions surrounding violence against children. The chosen concept of domestic violence encompassed, in several studies, notions that could be expressed in different ways (e.g., intimate partner violence, interparental violence, child abuse), so we opted for a comprehensive concept that involved the victimization, whether direct or not, of the child, and which would lead to a potential negative experience and consequences for their development. Therefore, studies that did not consider children or that addressed child abuse but not in the family context of their caregivers were not included in this review. This decision was the result of many difficult discussions between researchers, and the viewpoint of other colleagues might differ.

The choice of a more quantitative study design was, to a large extent, linked to the search for objective measures that would validate a more structured (actuarial) assessment, which has been identified as allowing for a better prediction of mistreatment. However, studies with other designs can make an important contribution to the assessment of risk and impact. The literature revealed the interrelationship of risk factors, the possibility of them being mediated by others, and the possibility of a significant exponential increase in this risk, thereby underscoring the importance of considering the simultaneity of these indicators and the need for other studies that deepen these interrelations, with other methodologies. Reviews focusing only on existing instruments that assess the risk of child abuse can sometimes appear to be generic, but there remains a scarcity of measures focused only on the problem of domestic violence between caregivers and involving the presence of children. Acknowledging attempts to control for possible geographic and cultural differences, there is still a need for transnational studies involving two or more countries; this would allow for larger samples and comparisons that would lead to the development of more universal instruments. It is important that such studies are carried out in the future. In any case, this review contributes, in an integrated way, to the identification of the main criteria that could be the basis of a child-centered risk assessment.

6. Conclusions

The study of child abuse requires the consideration of various contextual and individual risk factors and the multifactorial impact of victimization, making it more complex than just taking into account safety and protection issues in relation to perpetrators. Domestic violence requires a simultaneously integrated and differentiated reading. On the one hand, multiple contextual–situational or individual factors (parents and/or child) must be integrated into the assessment, whether static (caregivers’ history of victimization) or dynamic (families’ socioeconomic environment); these interact and impact on the consideration of the risk. On the other hand, given the multifactorial impact of victimization by domestic violence in the short, medium or long term, it is important to differentiate the experience and (estimation of) risk of this victimization in children and adults.

It is imperative to break down the adult-centered approach to addressing domestic violence to highlight the importance of conducting a child-centered assessment of the risk of domestic violence victimization. This literature review gathers sufficient empirical evidence such that—in situations of domestic violence among adult caregivers—the victimization of the child is undeniable. The empirical literature supports the importance of various factors in the life history of caregivers. It also provides a framework for instruments that identify and measure the key elements to consider when assessing the risk posed to children in cases of domestic violence. This underscores the need for a comprehensive assessment that goes beyond solely evaluating the adults involved.

This systematic review confirms the intergenerational transmission of violence and child abuse, and alerts us to the cumulative and exponential risk that can arise from a partial assessment that is not centered on the child and which, for various reasons, disregards their protection and their rights. It also allowed us to uncover, in addition to the material screened and analyzed, the body of literature that has emerged on the victimization risk assessment of children in situations of domestic violence. We consider it important to make efforts to improve existing instruments based on the knowledge that has been gained in recent decades and to develop more structured resources for the earlier prediction of risk, thereby supporting child protection services in minimizing or even preventing child abuse.