A systematic review of parenting interventions used by social workers to support vulnerable children

This paper reports on the findings from a systematic review of parenting interventions used by social workers to support vulnerable children in the United Kingdom. The study focused on children from birth to 11 years and 11 months based on Munro's rationale for early intervention. From the 423 papers initially identified, twelve met the inclusion criteria for this review. Four common themes were identified: developing relationships, the effectiveness of parenting interventions, societal impact on families and health and psychological concerns. The importance of effective relationships between parents and social workers was identified as key to effective parental interventions but there was limited evidence of improved outcomes for children despite this. A common factor in the studies was the level of parental deprivation which in many cases was associated with a range of mental health issues frequently seen in association with drug and alcohol abuse and domestic violence. The review identified a number of successful outcomes across a range of parenting interventions. However, what was surprising was the limited input from the children themselves within this review. Applying our findings to practice, the authors recommend a number of ways to contribute to the development of parenting interventions.


Background
This article focuses on parenting interventions used in the United Kingdom by social workers and other agencies to support vulnerable children. Although the policies and practices described in this article are based on the United Kingdom literature the principles of caring for vulnerable children are universal. The fundamental principle of the Children Act 1989 is that the welfare of the child is paramount. Section 47 of the Children Act requires local authorities to make enquiries, or cause enquiries to be made when the authority has reasonable cause to suspect that a child is suffering, or is likely to suffer, significant harm). Keeping children safe includes protecting children from maltreatment; preventing impairment of children's health or development; ensuring that children grow up in circumstances consistent with the provision of safe and effective care and taking action to enable all children to have the best outcomes (HM Government, 2018, pp. 6-7).
The Children Act 1989 (The Stationary Office, 1989), emphasizes that children are best brought up within their own families and state intervention should be minimal. However, there are times when government agencies may have to intervene to protect vulnerable children. There is no legal definition of the term "vulnerable," but it is commonly used in the literature and government documents. We define "vulnerable" as a child in need or/and a child suffering, or likely to suffer, significant harm (The Stationary Office, 1989). We have used the term vulnerable to include both children in need and children at risk because as Munro (2011) rightly points out interventions into the lives of young children are not only about protecting children from harm but also improving their opportunities to flourish. In her review of child protection, Munro (2011) observed there was a developing body of evidence to support the effectiveness of early interventions with children and families. It is preferable that interventions have a preventative scope, as they are more likely to reduce overall harm (Munro, 2011). Early help is not simply about preventing or stopping, but also about improving life opportunities for children and young people (Munro, 2011).

Contemporary evidence on parenting interventions
The extensive adverse childhood experience (ACE) literature provides a wide range of evidence to support how early adverse experiences in childhood impact negatively in adulthood. The ACE questionnaire has been adapted to increase understanding of the needs of diverse populations whose outcomes are of particular concern for social work professionals, including children who experience domestic violence (McGavock & Spratt, 2014) and young people who commit suicide (Devaney et al., 2014) (for a comprehensive account of ACE research and implications for social work practices, see Davidson et al., 2010;Spratt et al., 2018).
The reasons identified for children being vulnerable included lack of parenting skills, parental drug and alcohol misuse, domestic violence, mental health concerns, children exhibiting behavioral disorders, and poverty. The now or never argument is particularly relevant to the risk of harm to young babies. Parent-infant interaction was reported to impact the infant's stress regulatory system. Without responsive parenting, exposure to stress impacted the infant's neurodevelopment and hormonal secretion. This influenced response to stress in later developmental stages (Gunnar, 1994 in Barlow et al., 2014b).
We can see there are multiple needs in vulnerable children and their families. Growing awareness of the frequency with which domestic violence, substance use and mental health problems co-exist, particularly in the context of child protection is an example of the interlinked structural complexity of working with vulnerable families. The use of the term "toxic trio" is used to highlight this complexity (Skinner et al., 2021).
Parenting programs have been shown to improve the psychological well-being of parents, parental mental health and parenting skills'. Engaging in a parenting program has resulted in a reduction in harsh parenting practices and child conduct problems in the short term, which impacts the quality of the parent-child relationship and therefore enhances the long-term psychological well-being of the child (Barlow et al., 2014a;Furlong et al., 2012). It is recognized that future research needs to address the reasons for this short-term change. Social workers are key professionals who provide long-term support to vulnerable families and are therefore well placed to provide such parental interventions. In addition, social workers have the professional knowledge and skills to identify the common aspects of parenting programs and interventions that enhance success; elements of which they can use in their practice with families at risk (Barth and Liggett-Creel, 2014;Stevens, 2014). Contemporary social work in the United Kingdom has sought to identify and develop more evidence-based approaches to social work interventions. Despite the focus on evidence-based practice the development of evidence in social work practice remains limited (Barlow et al., 2014b;Forrester, 2020).

Rationale for focus on parenting interventions and choosing a systematic review
As a contribution to the development of the parenting interventions evidence base, we wanted to focus on vulnerable children as a contribution to existing knowledge such as in Scotland, Independent care review (2020) and to engage with the newly announced independent review on social care in England (HM Government, 2021). In particular, we see our contribution will link to both reviews' themes of safety and strengthening families (England) and family and scaffolding (Scotland).
We chose a systematic review as it was an appropriate tool to identify the evidence base in current social work practice. This systematic review, therefore, explores the impact of parenting interventions, undertaken by social workers, working with children and families in the United Kingdom. It is situated within the wider context of the independent review of children's social care that the authors offer the current review as a contribution to the review. It is helpful for international colleagues to mention also that the United Kingdom is decentralized across the four nations of England, Wales, Scotland, and Northern Ireland. Although each nation has a unique regulatory body that identified the requirements for social work education and practice, the four nations generally have a shared overall approach to social work.
This study is the third in a series of systematic reviews looking at the nature and quality of knowledge and interventions in social work (Boyle et al., 2019;Woodman-Worrell & Higgins, 2018).
Objectives. A preliminary search for existing systematic reviews and/or scoping reviews on the topic was conducted. A number of systematic reviews have been undertaken which explore the effectiveness and cost-effectiveness of both home-based and group parenting programs (Barlow et al., 2014a;Coren et al., 2018;Miller et al., 2011;Stevens, 2014). However, none of these reviews have focused specifically on the role of social workers in supporting parents to develop skills and strategies to enhance their parenting abilities and improve the relationships that they have with their child. The aim of the current systematic review is to provide a greater understanding of the parenting models, assessments and interventions being currently used by social workers in the United Kingdom. Following the population, intervention, control and outcome (PICO) framework, this systematic review aims to synthesize existing knowledge, identify gaps in the literature and provide recommendations for future research.

Methods
A systematic review is deemed the most suitable method due to the research question proposed and the literature it intends to include. The review followed the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) checklist (Moher et al., 2009) and the PRISMA for Protocols guidelines (Shamseer et al., 2015). For this review, we chose the age range of children from 0 to 11 years, 11 months. We chose this age range based on Munro's rationale for early intervention. Research suggests that even babies can experience ACEs (Barlow & Calam, 2011;Glaser, 2000), which highlights the importance of early intervention. The age range also links to our interest in parenting interventions with vulnerable children. The younger the child the more likely they are to be potentially more vulnerable "Families can be dangerous places for children and in particular for babies and young children" (Munro, 2011;United Nations, 2006: 35).
Eligibility criteria. Articles and reports related to the topic of parenting assessments, interventions and models currently used in social work were identified through searches using electronic databases. The purpose of the current systematic review was to synthesize all relevant available knowledge and provide a comprehensive overview of this research topic. All existing literature was included in the searches, for example, primary research studies, systematic reviews, meta-analyses, guidelines and websites. The search was limited to literature written in English and based on research undertaken in the area of social work in the United Kingdom. Searches started from January 2009. The rationale for starting in 2009 is that our principal research focal point is on the contemporary research evidence available on social work and parenting interventions. Adopting a relatively short and recent date provides the opportunity to identify research that is current for present-day social work practice.
Population, intervention, control, and outcomes. The PICO model was used to help frame the research question and define the inclusion and exclusion criteria.
Population: Parents of children aged from 0 to 11 years 11 months in receipt of statutory children's social care interventions, based in the United Kingdom.
Interventions: Parenting models, assessments and interventions used by social workers Control: none Outcomes: Impact of interventions used by social workers for families, with children identified as vulnerable aged up to 11 years 11 months; confidence in parents' skills, parent to child and child to parent communication skills, quality of the parent-child interaction, emotional and behavioral adjustments and emotional behavioral changes in children, behavioral adjustments and behavioral changes in parents.

Inclusion criteria
• Parenting models used by social workers in the United Kingdom are referenced in the literature and validated by the current literature, parenting assessment and interventions, therapeutic relationships between the practitioner and client. • Families with child/ren 0-11 years 11 months of age.
• Social workers involved with families with child/ren identified as vulnerable.
• Carers who are not caring for the children as part of their professional vocation, that is, they must be unpaid, parents, grandparents or other family members. • Models and interventions improving emotional and behavioral adjustments. The age range limiter or 0-11 years 11 months was identified because early development ends at the point when children complete their primary education (junior school) and move into secondary education.
Families identified as vulnerable (see "Introduction" section for definition adopted for the term vulnerable).

Exclusion criteria
• Parenting programs, assessment, and interventions used in areas other than social work. • Families with child/ren aged over 11 years 11 months.
• Research/models used with families outside the United Kingdom.
• Paid vocational carers employed to care for child/ren, for example, foster carers or residential workers.
Databases searched. Search strategy. Parenting AND model * AND Assessment * AND "therapeutic relationship * " AND "social work * " AND UK OR United Kingdom Limiters: January 2009-April 2020; English Language Data management. Data for analysis were extracted from the included studies and managed in an Excel spreadsheet. A data extraction sheet was developed and tailored to the requirements of the review. The data extraction sheet was tested independently by all three authors on two included papers and, where necessary, it was revised to ensure it could be reliably interpreted and capture all relevant data from different study designs.
Selection process. Study selection (both at title/abstract screening and full-text screening) was performed by three reviewers, independently. Any disagreements were resolved by consensus or by the decision of all three reviewers. After eliminating the duplicates, an initial screening of titles, abstracts, and summaries was undertaken to exclude records that clearly did not meet the inclusion criteria. Each record was classified as "include" or "exclude" or "maybe" with comments to identify relevant and exclude irrelevant literature. The researchers were inclusive at this stage and, if uncertain about the relevance of a publication or report, it was left in. The full text was obtained for all the records that met the inclusion criteria (based on the title and abstract/summary only). In a second step, all the full-text papers were screened against the inclusion/exclusion criteria. Studies that did not meet the inclusion criteria at this stage are listed with the reasons for exclusion (see Table 2 characteristics of excluded studies). Studies that did meet the inclusion/ exclusion criteria are listed with a brief description in (Table 1: Characteristics of included studies). An adapted PRISMA (with meta-analyses) flow-chart of study (1) Initial time for parents and children to interact.
(2) Parent educational program 1 hour; children engaged in different activities.
(3) Remainder of day spent together, e.g. working on routines, play activities.
(4) Parents required to manage mealtimesoften a period of conflict (1) Family and case characteristics.
(4) Both parents and SW were asked their views regarding parental involvement. 18 were interviewed (3 declined and 9 unavailable from the 30).
Of the 18, 6 interviewed in a group and 12 individually.
A convenience sample of 4 family center staff participated in a discussion 57 parents in the study population: 93% Female, mean age 30, only 5% in paid employment and 45% lived in a lone adult household. Howe's (1996) two-part concept of surface and depth Intervention: (1) Group or individual interview, notes were taken by the researcher during this.
(2) Questionnaire which collected demographic data and the forms of help parents had accessed Descriptive statistics Thematic analysis from interview notes Three main themes: (1) Resources available: Practical help, e.g., key skills and social support. The how to approach was seen as surface.
(2) Staff approach: Parents felt monitored, perceived as unhelpful. Seen as surface. Parents who were praised trusted workers.
(6) SAPAS.   This paper explored the role of fathers in decision making within the family by social workers in four European countries; it is not a parenting model.

Douglas and Johnson, 2019
The Solihull approach 10-week program: a randomized controlled trial This is a parenting program, but it was not facilitated by SW. Macmillan et al., 2009 Interventions to prevent child maltreatment and associated impairment.
Review of current programs, difficult to discern any in date range, by social workers in the UK. Also, it is a Canadian paper.
Lucas, 2011 Some reflections on the rhetoric of parenting programs: Evidence, theory, and social policy Does not include SW.
(continued)  Barlow et al., 2014 b Group-based parent training programs for improving parental psychosocial health. As a systematic review it only includes some aspects of our PICO. No reference to SW. Furlong et al., 2012 Behavioral and cognitive-behavioral group-based parenting programs for early-onset conduct problems in children aged 3-12 years Does not mention specifically SW. Barlow et al., 2016 Group-based parent training programs for improving emotional and behavioral adjustment in young children.
SW not mentioned.  selection (Shamseer et al., 2015) is included in the appendices of this review (see Figure 1; Flow chart).
Data collection process. One reviewer, for consistency reasons, extracted data from each included study and inserted this into the Excel spreadsheet. Two other reviewers have thoroughly read and commented, and discrepancies were resolved by discussion between all reviewers. Study authors were contacted in relation to any missing information.
Data extraction. We extracted data on the participants, interventions, and outcomes. In addition to that, the extraction sheet included authors, year of study/report, aim/ purpose, type of paper, study population, sample size, study design, and key findings that relate to the systematic review question. One reviewer completed Table 1 Figure 1. Flow chart.
(Characteristics of included studies). Two other reviewers have thoroughly read and commented on the findings, and discrepancies were resolved by discussion between all reviewers.
Critical appraisal. Each of the included studies was appraised using a standardized critical appraisal tool. Critical appraisal forms for mixed methods were tested, such as the Mixed Methods Appraisal Tool (MMAT) Version 2018 (Hong et al., 2018) and Critical Appraisal Skills Program (CASP) tool. Both suggested tools have been standardized and validated and are widely used for systematic review purposes. Each tool was tested independently by three reviewers, with two full-text papers and reviewers agreed on the best tool to work with based on which tool fitted best with the purpose of this review and offered a good selection to cover the types of methodologies used in each of the included studies. It was agreed that the CASP tool, specific to the study design, was the best fit for this review. The twelve included studies were divided between two reviewers with each appraising 20% of the other reviewers' batch. Through the critical appraisal of the included studies, it was found that some studies had gaps in relation to methodological quality and reporting findings but still included contextually rich details that contributed to the overall narrative synthesis and answered our research question.
Risk of bias. Two reviewers independently assessed the risk of bias using the Cochrane risk of bias tool (Higgins et al., 2019), which includes the following domains: random sequence generation, allocation concealment, blinding of outcome assessors, completeness of outcome data, and selective outcome reporting, respectively, and discussed by the authoring team during regular meetings in the data extraction phase (Tables 3  and 4). The results of the risk of bias assessment were incorporated into the findings (the narratives of the review) and limitations sections. Authors of papers were contacted if the information was missing. Significant areas of bias have been identified in the findings section.    Data synthesis. Findings from included studies were synthesized narratively. The "Guidance on the Conduct of Narrative Synthesis in Systematic Reviews" was used to advise the narrative synthesis (Popay et al., 2006). First, a preliminary synthesis was conducted to develop an initial description of the findings, common themes and of included records and to organize them so that patterns across records can be identified. The patterns resulting as well as the common themes were discussed by all three reviewers and agreed. In a second step, thematic analysis was used to analyze the findings. The following five steps of thematic analysis were followed adopting a recursive process (Higgins et al., 2019): 1. Familiarization with the extracted data. 2. Generation of initial codes. 3. Searching for themes. 4. Reviewing themes. 5. Defining and naming themes.
Meta-analysis was not conducted as data were heterogeneous.
Dealing with missing data. We contacted the original investigators for clarifications and to request missing information. If we were unable to obtain this (e.g., two contact attempts over 1-month period), we used data available from the published studies and assessed the risk of bias through the criterion "incomplete outcome data." We did not impute any missing data and discussed all assumptions and subsequent procedures used to deal with missing values in the review.

Findings
Critical appraisal. The search initially identified 423 papers, once duplicates were removed, and these were screened by three of the authors resulting in 386 being excluded because they did not meet the inclusion criteria of the review (see Figure 1). In total, 37 papers were assessed through critical appraisal. This resulted in 12 studies meeting the inclusion criteria for this systematic review. Descriptive analysis can be found in the characteristics of included studies (Table 1). Eight of the papers used a qualitative methodology, however, one of these, Barratt (2012) was a project rather than a research study, observing parent-child interactions during a parenting assessment. This was included in the review because the findings were deemed relevant to the research question. The remaining four papers were quantitative, using a randomized controlled trial (RCT) design. All of the studies were undertaken in the United Kingdom.
Common themes. Developing relationships. The relationship formed between the parents and the social work team was a significant finding in three of the studies. Whittaker et al. (2014) used the concept of surface and depth to explore parents' experiences of using family support services. The approach adopted by staff was identified as a key theme in this paper. Parents felt that they were being monitored during their interactions with social services which they perceived as unhelpful. When parents received praise from the workers, they felt that they were able to change, which they expressed as feeling trusted by social workers, thereby enhancing the relationship. Barratt (2012) argued that parents who actively engaged with the parenting assessment process and developed a positive relationship with staff improved the outcomes for their children. Positive relationships with center staff resulted in more children remaining with the family. Social workers also reported that engaging in the multifamily group program enabled them to observe the family's interaction with their children and other parents over a longer period of time. Workers noted sustained changes in parental behavior and identified the skills learned from other families, which they felt enhanced the quality of decision making in their final reports (Barratt, 2012). Killan et al. (2017) found that social workers perceived the relationship with parents who exhibited threatening behavior as weak. Interestingly, social workers who had received training in motivational interviewing rated the relationship with families less positively than those who had not engaged in this training. In the systematic review of motivational interviewing (Boyle et al., 2019) a statistically significant difference was found in the MI skills staff demonstrated while delivering MI irrespective of the training they had received so this is perhaps not a surprising finding.
Symonds (2019) undertook a qualitative study that aimed to investigate how both parents engaged with parenting services. Initial telephone conversations undertaken by social workers to parents who had been referred for parenting support were recorded, with parental consent. The recordings were transcribed and analyzed using conversation analysis. Social workers attempted to invite the other parent in 50% of the conversations (14 out of 28) and on six occasions this invitation was accepted. The author concluded that this form of invitation may have a correlation between specific questioning techniques and the recruitment of both parents to a parenting program. This is particularly relevant for fathers who are unrepresented in these programs (Symonds, 2020).
Impact of parenting programs on behavior. Two studies in this review focused on the impact of parenting programs on the child's behavior. Sonuga-Barke et al. (2018) undertook an RCT to compare the efficacy of the individualized New Forest Parenting Program (NFPP) with the group based Incredible Years (IY) program in preschool children with Attention Deficit Hyperactivity Disorder (ADHD). Parents found no difference in the child's symptoms from either of these programs. However, children who engaged with the NFPP revealed statistically significant differences in conduct problems when compared with treatment as usual (TAU). It is important to recognize that parental ratings of their child's behavior may overinflate the treatment effect due to bias as they are invested in the process. NICE guidelines support group-based programs over individualized programs because they are cheaper. In this study, the NFPP was found to be cheaper to implement and the team concluded that the guidelines should be reviewed based on this evidence. Beckett et al. (2012) conducted an RCT comparing two programs in children aged 5-7 years with disruptive behavior. The study focused on improvements in reading ability comparing the IY program with Supporting Parents on Kids Education in Schools (SPOKES). There were four arms to this study; IY, SPOKES, combined IY and SPOKES and TAU. Children in the intervention group had significantly fewer behavioral problems than the control group. Parents reported feeling more confident in dealing with their children. There was a significant improvement in reading ability with IY compared with the control group and positive parenting behaviors increased significantly. Both groups showed improved child behavior but only the IY group improved the children's literacy. The external validity of this study was strengthened by the fact that the two different study sites included participants with very different ethnic profiles and yet the results were similar across both groups thereby strengthening the generalizability of the findings. Barlow et al. (2018) conducted an RCT that focused on the emotional behavior of parents. The study compared the parents under pressure (PuP) program with a control group who received TAU to determine if this program would reduce the potential for child abuse. The PuP program supported skills development by providing children with a nurturing environment. Parents were supported with workbooks, including exercises for the parents to work through with their children, mindfulness techniques and video feedback. Measurement tools identified that parents in the PuP had a reduced risk of child abuse as well as clinical improvement in the levels of observed abuse. There were statistically significant improvements in emotional management, psychological well-being and depression in the PuP group. However, neither group showed any improvement in parenting stress. There was also no difference in the social and emotional well-being of the children between the two groups. The PuP parents felt that mindfulness was helpful in managing their emotions which was confirmed by social workers. The researchers concluded that mindfulness strategies impact positively on parental functioning and child abuse risk and should be rolled out to other families where children are perceived to be at risk.
Impact of social factors on families. A number of studies in this review identified complex social factors which contributed to the challenges parents experienced when providing effective parenting. Frequently participants were single parents, facing financial challenges, in low paid work or unemployed and some had a criminal record (Barlow et al., 2018;Sonuga-Barke et al., 2018). Many were victims of domestic violence, associated with drugs or alcohol misuse. Parents facing these complex problems frequently revealed a history of childhood abuse themselves resulting in many of them being brought up in care (Barratt, 2012;Tarleton & Porter, 2012). Poor parental role modeling and growing up in an environment where violence was perceived as normal behavior, contributed to this behavior being modeled in their own parenting (NICE, 2017) and with children often becoming neglected or looked after by the local authority (Barratt, 2012;Sen at el., 2014). Over a third of the parents with learning disabilities (LDs) were themselves brought up in care and 41% of these parents had their own children removed from their care (Tarleton & Porter, 2012).
Domestic violence was frequently seen in conjunction with drug and alcohol abuse and either contributed to or resulted in mental health issues. Tarleton and Porter (2012), evaluated the effectiveness of the valuing parents support service (VPSS) in parents with LDs, and found that 28% of children lived in families where domestic violence was a feature. The families experiencing VPSS were socially and physically disadvantaged, with 63% of participants having poor social skills. Despite this, both parents and workers valued the trusting and honest relationships they developed and appreciated the support provided during care proceedings. The authors concluded that VPSS had a positive impact on family life and argued that it could be extended to other areas of children's social work (Tarleton & Porter, 2012). Mental illness is also frequently seen in conjunction with domestic violence. Domoney et al. (2019) are evaluating a parenting intervention that focuses on the whole family. The families recruited for this study were experiencing domestic abuse at recruitment. Measurements and frequency of current domestic abuse, levels of mental health disorders and drug and alcohol issues were also assessed. This evaluation focuses on the whole family and commences during pregnancy and extends through the first two years of the child's life. To date, the researchers have identified that families value the service and the whole family involvement and appreciate not feeling judged. Ghaffar et al. (2012) undertook a qualitative study of 42 families in the North of England, exploring their experience of child protection systems. Parents with drug or alcohol issues identified that the support provided by the Child Protection Plan improved the care their children received. But <25% of parents who experienced domestic violence felt that the system provided them or their children with appropriate support. Women in abusive relationships felt that workers did not appreciate the strength of purpose required to leave an abusive relationship. Families in this study appreciated the clarity and honesty of social workers but felt social workers lacked empathy and that the counseling support available to their children was inadequate. Ghaffar et al. (2012) recognized a limitation of their study was that parents from ethnic minority groups and fathers were underrepresented. Killan et al. (2017) however, identified that when there were concerns over parent's alcohol or drug usage or domestic violence that the social workers assessment of the Working Alliance Inventory was not associated with a reduction in scores compared with the parents and observer's assessment.
Three of the studies in this review identified the impact of unemployment and single parenting, which was in some studies associated with criminality (Barlow et al., 2018;Sonuga-Barke et al., 2018;Whittaker et al., 2014). Only 5% of the 57 participants in the study by Whittaker et al. (2014) were employed and almost half were single parents. Attendance at the family center provided parents with practical help with key skills and social support. However, parents in this study did not feel that the social workers understood the challenges of being a single parent and frequently felt that they were told want to do rather than being understood within their social context. Parents found it helpful if social workers asked parents what future changes they would like to see. Whittaker et al. (2014) concluded that adopting relationship-based approaches made parents feel valued and respected which was more likely to result in a change in behavior. Sheppard et al. (2010) compared two different assessment tools, the Parenting Concerns Questionnaire (PCQ) and the Parenting Stress Index (PSI) at four children's centers in Southern England. In this study, almost 20% of attendees demonstrated clinical levels of stress on the PSI. However, the PCQ identified severe family and environmental problems at a higher level than that identified by the PSI. The authors argued that these factors could adversely affect child development and parenting capacity and concluded that the PCQ appeared more sensitive to variations in vulnerability factors than the PSI. The key factors were single parenting, family size of three or more children and being in receipt of income support. Whilst this study was relatively large, with over 300 hundred participants, it was undertaken in a rural area of predominantly white parents and therefore lacked generalizability.
Health and psychological concerns. Mental health issues were frequently experienced by parents being supported by social services. Killan et al. (2017) found that social workers reported significantly lower scores on the Working Alliance Inventory in families suffering from depression. These findings need to be interpreted with caution as of the 610 cases in this study only 131 participants completed the final interview and questionnaire suggestive of sampling bias. Ghaffar et al. (2012) found that 40% of parents in their study with mental health issues felt they received good support. This compared more favorably than the parents who experienced domestic violence, whose views were more mixed regarding the level of support for themselves and their children. In the study by Sonuga-Barke et al. (2018) the General Health Questionnaire screened for mood disorders such as anxiety or depression. Scores of 11 or more were considered to reveal probable mental health problems. In this study of the 235 participants, the threshold was identified in between 73% and 79% of the population from the three trial arms. This is perhaps not surprising given the fact that their children had ADHD but is still a very relevant finding.

Discussion
This paper reported on the findings from a systematic review of parenting interventions used by social workers to support vulnerable children in the United Kingdom. The research question was: What is the impact of current parenting interventions used in UK social work with vulnerable families with children (0-11 years)? We identified four themes encapsulating the impact of parenting interventions in social work: 1. Developing relationships. 2. Impact of parenting programs on behavior. 3. Impact of social factors on families. 4. Health and psychological concerns.
This review identifies evidence to support that parenting interventions can have a positive impact in some areas. However, there are limitations that are explored in this discussion.
The importance of relationships was generally seen as central to parental interventions whether between families or professionals. Parents who developed positive relationships with professionals were more likely to achieve positive outcomes (Barratt, 2012). But positive outcomes were broadly limited to improved communication between parents and professionals rather than improved outcomes for children. However, the author did identify that there was some evidence of improvements for some children illustrated by the fact that more children were returned to the care of their families by the end of the project.
Identifying relationships as important was on one level self-evident. The importance of relationship building is a core social work value. Arguably, the importance of the relationship is self-evident in social work practice (Ruch et al., 2010). What is less obvious is how and to what extent relationship building improves outcomes for children during parental interventions (Killan et al., 2017). Even when the research studies focus on the importance of "relationship-based" practice (e.g., Ruch et al., 2010) positive outcomes are not generally clearly identified (Ruch & Julkunen, 2016).
However, this finding of the limitations of social work knowledge is nothing new. As long ago as 1971 Olive Stevenson, the first editor of The British Journal of Social Work, wrote that there is "little evidence about the effectiveness and appropriateness of social work interventions" (Stevenson, 1971, p. 235). Whether times have changed in social work knowledge and interventions goes beyond the purview of this study. However, as recently as 2014 Barlow et al., argued that "there is currently limited evidence about the effectiveness of the available tools in the field of child protection, and further piloting of these instruments and research is now needed." (Barlow et al., 2014b, p. 10). As a whole, the interventions tended to focus on the relationship between the social workers and the parents. There were some interventions involving the children but the main discussion on relationships concerned mainly improved/good parent and professional relationships. To some extent as parenting interventions parents are bound to have an important role. However, there appeared to be a danger that the children were not sufficiently at the center of the interventions (Munro, 2011). As Munro (2011) highlights the child protection system should be child centered. The focus may frequently involve engaging and helping parents. However, it is essential to keep assessing whether parenting interventions (as well as other types of interventions) are providing sufficient improvement in the capacity of the parents to respond to their children's needs.
Relationships are important in every type of social work practice (Department of Education, 2014; Munro, 2011). Relationships with parents characterized by honesty about what needs to change and why, sensitivity and a willingness to listen to parents' points of view, may be better able to help parents become motivated and engage with services (Dumbrill, 2006;Forrester et al., 2008;Laming, 2003). This relationship needs to be maintained after an intervention to make sure parents are provided with ongoing support to reduce risks of the return of problematic behavior (Department of Education, 2014). One of the papers reviewed suggested there were two types of interventions: surface and depth (Whittaker et al., 2014). The former was skills based and focused on practical support. However, the workers felt the parents could benefit more from a depth approach. It is this depth approach that supports the rationale for sustained support once the formal intervention is ended.
A recent study by Forrester et al. (2019) explored the connection between communication skills and outcomes with families. Where social workers visited families more often there was a stronger connection between skills and outcomes, but the connection was limited, and further research was needed.
Sometimes parents felt interventions were not helpful or supportive (Ghaffar et al., 2012). There was some evidence that parents with different issues might have more positive responses to interventions. Parents with drug and alcohol misuse felt they were positively supported to improve their care for their children. On the other hand, parents with mental health problems and those experiencing domestic violence did not feel as well supported. They felt the process was stressful and overwhelming. Victims of domestic violence also felt judged. These families appreciated clarity and honesty but did not always find that the social workers were empathetic and there was a lack of counseling support for children who needed it (Ghaffar et al., 2012). Indeed, generally, there was limited focus on children and their relationship with professionals and families.
Some studies compared programs whereas others evaluated individual interventions. As we have seen there was evidence of the interventions bringing about improvements. However, parents had mixed feelings about the role and empathetic approach of their interventions. Some of the programs were evaluated using an RCT, which gave rigor to the evaluations. However, where evaluations compared programs the results tended to suggest the impact of the tools was similar (Barlow et al., 2018;Sonuga-Barke et al., 2018). This could have contended as an example of "dodo bird effect" (Lundahl et al., 2010). The dodo bird effect is the suggestion that no one intervention model or theory is clearly superior (Prochaska & Norcross, 2007). Perhaps the most significant feature of the parenting interventions was the variety of interventions. They ranged from groupwork to individual involvement. There were a number of audit tools. A common factor of all parents and families was the level of deprivation, which was understandable given parents were vulnerable and required social work interventions. From our review of the studies, it was clear that there was limited evidence to support one intervention being more effective than another. Given the complexity of the families (Sonuga-Barke et al., 2018) we wondered whether reliance solely on audits could capture the wide range and challenges in these families. Hood (2019) makes this point when he argues that the strengths and limitations of different tools and measurements are dependent on the theoretical frameworks in which they are used, and these frameworks reflect the contested nature of child protection and its political and organizational structures (see also Boyle et al., 2019).

Limitations
This review identifies evidence to support that parenting intervention has a positive impact. However, there are limitations to this study. The papers had a variety of approaches and outcomes which prevented robust comparisons. The internal limitations of each of the studies preclude generalizability. The small number of research papers also limited the extent to which there could be wider generalizations. Further research is needed to reduce limitations and provide opportunities to evaluate studies longitudinally 2-3 years after the intervention has been completed. An intervention may well be successful at the end of the evaluation. The real question is: is this intervention still in use or perhaps more importantly, is the impact still sustained? Finally, the effectiveness and limitations of any intervention are linked to the conceptual system underlying the intervention (Higgins, 2013). These tools themselves are the products of the debatable nature of child protection and its organizational setting (Hood, 2019).