**Vulnerable Family Meetings: A Way of Promoting Team Working in GPs' Everyday Responses to Child Maltreatment?**

#### **Jenny Woodman, Ruth Gilbert, Danya Glaser, Janice Allister and Marian Brandon**

**Abstract:** This study uses observations of team meetings and interviews with 17 primary care professionals in four GP practices in England to generate hypotheses about how "vulnerable family" team meetings might support responses by GPs to maltreatment-related concerns and joint working with other professionals. These meetings are also called "safeguarding meetings". The study found that vulnerable family meetings were used as a way of monitoring children or young people and their families and supporting risk assessment by information gathering. Four factors facilitated the meetings: meaningful information flow into the meetings from other agencies, systematic ways of identifying cases for discussion, limiting attendance to core members of the primary care team and locating the meeting as part of routine clinical practice. Our results generate hypotheses about a model of care that can be tested for effectiveness in terms of service measures, child and family outcomes, and as a potential mechanism for other professionals to engage and support GPs in their everyday responses to vulnerable and maltreated children. The potential for adverse as well as beneficial effects should be considered from involving professionals outside the core primary care team (e.g., police, children's social care, education and mental health services).

#### **1. Background**

It is now widely acknowledged that scarce resources and high demand mean that children's social care in England is functioning as "an emergency service [1]" which prioritises reactive child protection responses over earlier help and prevention, despite policies which have ambitions to the contrary [1–5]. Qualitative studies report that some families and professionals have encountered a lack of early help or child in need services for families who are below the very high thresholds for child protection services [6–11].

In this context, the spotlight has been thrown on the role of other services which have routine contact with all children and families, such as GPs. Researchers have begun to ask how these services might use their everyday contact with families to respond to vulnerable and maltreated children and their families, especially those not meeting thresholds for services from children's social care [12–15].

There is evidence from qualitative studies that some GPs in England are already using their core skills and opportunities as family doctors to identify and respond to maltreatment-related concerns1 , particularly for children with problems related to neglect and emotional abuse [12]. In these studies, GPs were responding directly to families by "keeping a watchful eye" on them, "standing up and shouting" for families (e.g., advocating for better housing), talking to parents to help them realise "that there was actually a problem with the children" or that "stopping drinking was a good thing", and providing immediate and opportune healthcare for children when parents came to the surgery for other reasons [12,16]. These responses relied on GPs building a therapeutic relationship with parents and/or older children and good links with health visitors [12].

A therapeutic GP-patient relationship might be effective at keeping parents (or other adults important to the child) engaged and motivated to change whilst reducing risk to the child by helping these adults to keep the child "in mind" [17,18]. Without proper team working and professional supervision there is danger that fostering a therapeutic relationship with parents becomes an "accommodative strategy" which affirms patterns of the "bad" behaviour and encourages professionals to focus on parents and overlook the needs of children [19,20]. Good links with health visitors can allow relevant and timely information to flow to the GP to support risk assessment and monitoring of families with preschool aged children [16]. However, with the relocation of health visitors away from GP surgeries, these links may be difficult to maintain and GPs might be working in the context of partial or incomplete information about a child's situation. Systems which encourage team working, mutual peer supervision, information sharing and inter-professional links appear a common sense way of supporting GPs when they are responding directly to families and of minimizing the risks of isolated working.

Regular primary care team meetings for the specific purpose of discussing vulnerable families ('vulnerable family meetings') have the potential to be one such system. Such meetings are recommended as good child safeguarding practice by the Royal College of GPs [21]. However, there is little clarity about the purpose or format of these meetings and we know little about how they work in practice or whether/how they might involve other professionals.

This study used observations and interviews with primary care practitioners to describe the purpose, format and functioning of vulnerable family meetings in four general practices in England. We used the data to generate hypotheses about how these meetings might best work to support GP responses to maltreatment-related concerns and about the role of other professionals in relation to the meeting.

<sup>1</sup> We use the term "maltreatment-related concern" to capture the full range of professional concern about child maltreatment, including concern about parental risk factors for maltreatment and compromised parenting or parent-child interaction which is judged to have the potential to become harmful, whether or not the problem currently meets children's social care thresholds for enquiry or action. Therefore, "maltreatment-related concerns" includes children who are judged to be marginally maltreated or at risk of maltreatment as well as cases where practitioners are certain enough to definitively label the child's experience as "maltreatment". The term "maltreatment-related concerns" is consistent with a continuum model of child welfare and in keeping with a public health approach which covers prevention and treatment.

#### **2. Methods**

#### *2.1. Recruiting Data Collection Sites and Participants*

We purposively selected four "best practice" sites with a geographical spread across England from a convenience sample of 11 GP practices who were participating in an audit of child safeguarding led by the Royal College of GPs. The 11 practices have been described elsewhere [22]. We chose four sites where the GP lead for the audit was a child protection "expert" (held a child protection post, was involved in child protection policy or in delivering child protection training) and held regular team meetings to discuss vulnerable families [21]. These characteristics were ascertained via an online survey which the GP leads completed between April and May 2010. The lead GPs acted as gatekeepers by introducing the study to their colleagues and arranging for the researcher to come to the vulnerable families meetings. Before the vulnerable family meetings started, the researcher (JW) explained the study and collected the emails of GPs, practice nurses and health visitors who were willing to be interviewed. The researcher emailed each of these professionals and where they agreed, arranged to interview them. The four GP leads were also interviewed. A total of 21 professionals expressed interest in the study but four later declined due to lack of time. We conducted a total of 17 interviews.

#### *2.2. Interview Participants*

The participants tended to be experienced health professionals and to have worked for an extended period with their current team: almost two thirds (*N* = 11) reported being qualified for 20 years or more (range 1–40 years and 58% (*N* = 10) had worked in their current team for at least 10 years (range 6 months to 2 years) Eight of the interviewees (47.1%) were "experts", defined as described above). Just over half the respondents were female (*N* = 10/17; 60%). There were five participants each from Practices 1 and 2, four from Practice 3 and three from Practice 4. The two health visitors worked in different sites from each other, as did the two practice nurses.

#### *2.3. Conducting Interviews*

One researcher (JW) conducted the interviews with 14 GPs, two practice nurses and two health visitors from the four sites. In the interviews, the researcher elicited narratives by asking the participants to choose two or three "children, young people or families who had prompted maltreatment-related concerns" and describe their concerns and involvement. In keeping with the aim of allowing participants to tell their stories and control the content, the interviews were free ranging with minimal steering from the researcher. Interviews were face to face, conducted between November 2010 and September 2011, lasted an average of 50 min and were audio recorded and later transcribed. In total, we collected 837 min of interview data from 17 participants (602 min from the 14 GP participants).

#### *2.4. Observing Team Meetings*

The same researcher (JW) observed one vulnerable family meeting at each of the four data collection sites between January and June 2011. During the meetings, the researcher was a complete observer, taking no part in the meeting and sitting apart from the professionals. Thoughts and details were recorded as hand-written structured notes (the researcher noted the apparent purpose of the meetings, the content and the 'tone' of interaction between attendees as well as any additional thoughts).

#### *2.5. Analysing Data*

We used a thematic approach to analysis as we were interested in identifying recurrent and common themes about responses to maltreatment-related concerns within a defined group (GPs). Our approach was inductive (data driven) and interpretive (rather than simply descriptive) [23]. Although the inferior status of thematic analysis persists, a robust and in-depth thematic analysis can be as insightful and skilful as other "branded" analytical approaches [23]. The analyses generated a detailed description of seven GP responses (monitoring, coaching, advocating, opportune healthcare, recording concerns, referring to other professionals and joint-working with other professionals) as well as theory about the types of families for whom they were used and the contexts which facilitated them. These results are published elsewhere [16].

At the start of the study, we conceived of vulnerable family meetings as one possible 'response' to maltreatment-related concerns in general practice. However, as the analysis progressed, we saw that they were perceived by GPs and other members of the primary healthcare team as a facilitator of other responses (rather than a response in their own right) and were being used in this way. As the analysis continued, we asked detailed questions of the data regarding purpose and form of the meetings and it is these results which are presented here. To generate hypotheses about the most useful form and structure of these meetings, we asked ourselves: 'Are there particular characteristics of meetings that seem to support the stated or implicit purpose of the meetings?' The resulting hypotheses are presented in the discussion of this paper.

Although it would be useful for the reader to know which interview participants came from which practice, we judge such a level of detail to potentially compromise participant anonymity. Unless otherwise stated, we report themes that emerged from at least two of the practices. Where there are multiple quotations to support a single point, they come from participants who worked at different practices.

#### *2.6. Ethics and Research Governance*

Ethical approval for the interviews and observations was given by Central London 1 NHS Research Ethics Committee on the 8 October 2010 (Reference 10/H0718/6). For each of the four sites, approval was given by the relevant Research and Development Unit of the Primary Care Trust (PCT) who issued a Letter of Access permitting the research to take place.

#### **3. Results**

#### *3.1. Format of Vulnerable Families Meetings*

Table 1 shows that the meetings in each of the four practice were highly variable in their frequency, number and range of attendees, "tone" of interaction between professionals and selection of children and families for discussion.

One practice (practice B; Table 1) allocated an hour every fortnight to discuss problems in adults (e.g., cancer patient or palliative care patients) and, for the second half of the meeting, discussed children and young people:

It could be anything, from a child that's constipated to a child who's got a serious medical diagnosis to a child who is a cause for concern, and we'll also discuss all the children who have child in need plans or child protection plan (GP, interview data).

In the observed meeting in practice B, just over half of the children and young people were discussed due to maltreatment-related concerns (*N* = 7/13, see Table 1). The other three practices had meetings bimonthly (practices A and D, Table 1) or quarterly (practice C; Table 1) for 45 minutes to an hour, which were specifically dedicated to discussing children and young people who had prompted maltreatment-related concerns and their families. In three practices (practice B-D, Table 1), GPs, health visitors and practice nurses attended the meetings, with general practice (practice D) or health visiting service managers (practice C). In practice A, there were a wider range of professionals from education and health (Table 1).

In two practices, the "tone" of the meeting was relaxed and chatty, seemingly on account of the meetings being well-established and all professionals knowing each other well (practices B and D; Table 1). In one of these practices, there was an air of efficiency with a list of patients to discuss and wider information being readily to hand whilst the other meeting appeared more disorganized. In contrast, the other two meetings were characterized by tentative and guarded interaction between professionals, seemingly on account of the professionals not all knowing each other (practices A and C; Table 1) and the meetings being very newly established (practice A). In one of these practice, a GP subsequently commented that the presence of the researcher had made people more cautious and self-conscious about what they said (practice C). The guarded nature of meetings appeared to be a defense against the uncertainty and "not knowing" inherent within child protection work, which manifested itself as a preoccupation with the thresholds of concern for information sharing (practice A) or a reluctance to admit that the professionals did not know the answers to questions being asked. (practice C).

Only one practice had a clear method of selecting children and young people for discussion (practice B, see Table 1 for reported selection criteria). In this practice, the health visitor also brought children for discussion via a list of children she knew to have a child protection plan and be registered at the practice. In the other practices, selection of children and young people relied on GPs and practice nurses remembering their concerns and bringing them to the meeting.

### *3.2. Purpose of Meetings*

Fourteen (82%) of the interview participants (including both practice nurses and both health visitors) spontaneously mentioned the vulnerable families meetings and did so in the context of them being "good", "pivotal" and "important" (GP, health visitor, GP participants, respectively).




3.2.1. Monitoring, Review and Follow-Up

In all four practices, the meetings were seen as important for monitoring children and families. This came across implicitly in the observed meetings (Table 1) and more explicitly in the interviews:

Well, we would, um, um...I'll probably discuss it at the primary health care team meeting next week […] to those families we...you know...we always talk in our meetings They'll be discussed again, um, so I wouldn't...you know, that's the way we sort of monitor these sorts of families (GP, interview data).

So, yeah, you know, the meetings that we had, we kind of discussed it, that case, and, you know, you obviously hear other information from the other members around the table, "Oh, well, so-and-so and so-and-so..." and they know all about the families and things (GP, interview data).

The meeting allowed GPs to monitor children and their families in three ways. First, the team could act proactively by anticipating important or stressful points in a family's life or changes which could impact on parenting (e.g., the birth of a new baby):

when we met regularly as a whole team, the whole practice, we were [..] recognising for example that the mother was nearing term [in her pregnancy], that the parents were complying with [methadone] treatment and all was going well (GP, interview data).

Secondly, participants gathered and shared information with other professionals at the meetings, usually the health visitor, who was pivotal for supporting GPs' monitoring and review of vulnerable children and their families:

Of course [we ask health visitors what they know] at the child protection meetings we have on a regular basis at the practice, you might just say, "Is anyone worried about this family?" (GP, interview data).

In practice C, the cases that were brought for discussion at the meeting seemed to be ones where a professional felt that they had lost sight of the family or not achieved follow-up (Table 1). The professionals turned to the meeting as a way of catching up on what had happened since their initial concern, although as described above this was hindered by a reluctance to admit professional anxiety, uncertainty and incomplete knowledge around the case.

Health visitors were viewed as a conduit of information between children's social care and the police and general practice:

We found out about this [a very serious domestic violence incident] because the police alerted social services and social services obviously relayed the information to the health visitor. And in such instances where we have reports from social services, they're always brought up in our primary care team meetings so that everybody is aware of what is happening in that family at that particular time (practice nurse, interview data).

Thirdly, concerns by practice nurses and GPs prompted proactive information gathering via the health visitor:

And I brought her up at the meeting where the health visitors were present as well, […] And, again, the health visitors said they would go out and visit. And, in fact, when they got out to visit, the grandmother was also concerned that the mum, the child's mum, had been going out late at night and not coming back. And, of course, they had years and years of watching their daughter with drug use, and they thought she was using again (practice nurse, interview data).

But [the meeting] makes you think, and... and there have been occasions where people [health visitors] have said well perhaps I'll go around and see them again next week, just to see how they're getting on...(GP, interview data).

#### 3.2.2. Risk Assessment

From the GPs' and practice nurses' perspective, the main purpose of monitoring, review and follow-up was to aid risk assessment and make timely decisions about when a concern warranted being shared with children's social care. In this way, information sharing at the meeting was seen to contribute to a "bigger picture" and facilitate more accurate risk assessment:

*Interviewer:* How did having the vulnerable family meeting influence your view of her and...and the family?

*Respondent:* Um, I suppose it...it concerned me more in that it was, you know, it was part of the bigger picture and... it was more, oh, right, I didn't realize all of those people were kind of in the same household. So it was helpful from a...from a bigger picture, you know, thinking about all of the children rather than just her.

#### 3.2.3. Sharing Risk

One participant felt that the vulnerable family meeting allowed her to live with her decision not to refer a family to children's social care after her concern was stepped down following information from a health visitor:

I can live with it [the risk] and actually it's shared because I have told the health visitor, I have told the other doctor and I will mention it again when I go to the practice meeting and that is as much as we can do at the moment and I don't believe she is in immediate danger (GP, interview data).

3.2.4. Encouraging Inter-Professional Trust and Inter-Agency Working

One practice (practice A, Table 1) aimed to build inter-agency trust and joint working between general practice, local schools and mental health care services (CAMHS; Child and Adolescent Mental Health Services). This approach was newly conceived and it was apparent it would take a long time to build these links. The attendance of professionals from outside general practice resulted in a preoccupation with confidentiality and the ethics of information sharing, so that this was the primary topic of discussion and only one family was discussed.

#### 3.2.5. Learning and Peer Supervision

In one meeting, (practice C, Table 1) the stated aim was to learn from general principles rather than support the management of individual cases. However, there was no evidence of this happening in the meeting which was observed. There was no other indication in the observation or interview data that the professionals saw the meetings as a way of learning or accessing peer supervision in this difficult area.

#### 3.2.6. Joint Decision Making

In practice A (Table 1), the GP leading the meeting explicitly stated that the meetings were not a forum for decision making. There was no other mention of joint-decision making as a purpose of the meetings.

#### **4. Discussion**

#### *4.1. Key Findings*

The format of the vulnerable families meeting varied greatly across four "best practice" general practice settings in England. The meetings were primarily used to monitor children, young people and families, aid good risk assessment and support decisions about when to refer to children's social care. These uses of the meetings in this study go beyond those previously described: in feedback about the meetings in one Clinical Commissioning Group area, participants saw their primary purpose as information sharing to support early identification of problems and their secondary purpose as building relationships between professionals [24]. There was no evidence in our study that vulnerable family meetings were used for peer supervision and learning in this difficult area or to support joint decision making about how GPs and the primary care team should manage the family in their everyday contact with them.

Other qualitative studies of multi-disciplinary team meetings to discuss vulnerable families in social work [25] and patients with dementia in primary care [26] have reported that participants saw joint decision making as one purpose of the meetings. Both these studies are set in the US. Effective team meetings in other patient groups in primary care are characterized by clear goals [27] and it is likely that this would also improve the functioning of vulnerable family meetings.

#### *4.2. Facilitators of a Fit-for-Purpose Meeting*

From the results, we would suggest four key facilitators of a fit-for purpose vulnerable family meeting (based on the data, a fit-for-purpose meeting is one that promotes monitoring and follow-up of children and their families, good risk assessment, shared risk, joint-working, learning, peer supervision and joint decision making about referral to children's social care and/or continued management within primary care services).

#### 4.2.1. Flow of Meaningful Information into the Meeting

In the meetings we observed, GPs relied on health visitors to act as a conduit for information from children's social care and/or the police and to bring to the meeting detailed information that they had collected during home-visits. There are several challenges to using health visitors to underpin the monitoring, risk assessment and risk sharing functions of the vulnerable family meeting in this way. Health visitors only work with families with preschool-aged children (though they may extend their work to older siblings [28]), their caseloads may not overlap with the geographical area covered by GP practices (*i.e.*, they only work with some of the families registered at the practice); and they may not have time, energy or incentives to attend GPs' vulnerable family meetings. Health visitors may not themselves receive important information from other agencies: in their overview of a series of studies about safeguarding children in the UK, Ward and Davies reported that heath visitors were just as frustrated as GPs at the lack of feedback from children's social care ([9], p. 119). Wider interview data from our study also revealed that health visitors may not pass on all relevant information to the GP even if they are aware of it and even if they attend the vulnerable family meeting [16]. Assuming that health visitors know and share relevant information from other agencies might leave GPs erroneously thinking they have the bigger picture and impede good risk assessment and monitoring. It might also place an unfeasibly large burden on the overstretched health visiting service.

One alternative would be to create a system for regulated information exchange such as automatic notifications of referrals to police or children's social care or an electronic database that could be shared across agencies. However, this type of "techno-rational" solution [29] has been criticized for prioritizing administrative work and formulaic agency responses over thoughtful practice and therapeutic work with children and families [30–33]. As Hall argues in his analysis of inter-professional communication in public enquiries into child death, we should not assume that it is easy or simple to transfer "information" from one professional to another [32]. Professionals from different agencies will draw on the rules, beliefs and habits of their profession to make (different) interpretations of the same piece of "information" [34]. Even when information is shared, the meaning of that information can easily get lost as it changes hands, especially in the context of child protection work which engenders professional anxiety, uncertainty and vulnerability to blame. "Lost" information might be exaggerated with automatic systems, which also threaten to deluge general practice in the sheer volume of information they might receive about their patients, for example domestic violence incidents via the police.

Another alternative would be to resource another professional whose job description included collecting ensuring information flow into the meeting from social care, education and health visitors and feedback to these agencies. This professional could be located either within primary care or within children's social care. A similar role already exists in some hospitals in the UK in the form of liaison health visitors and paediatric liaison nurses. Collecting information prior the meeting might improve its function: a qualitative study of team working for patients in Belgium primary care reported that well-planned meetings worked best [35]. Any member of liaison staff would have to be sufficiently skilled to elicit and transfer the *meaning* of information, which might require interagency training [9] and to ask questions around the information exchange rather than passively accepting what they hear [32]. There are obvious resource implications of using a trained and skilled professional to fulfill this role. One possibility is for the meetings (and staff time) to be funded by the Clinical Commissioning Group as a Local Enhanced Service [24].

#### 4.2.2. Systematic Ways of Identifying Cases for Discussion

In the observed meetings, there was a clear need for systematic ways of selecting children and families for discussion. This might be achieved, for example, by identifying children from codes in their electronic primary care records and cross referencing concerns with children and parents known to be vulnerable by other professionals (as in practice B, Table 1) or by selecting families where parental risk factors are known to the GP practice, such as domestic violence, drug and alcohol abuse and suicide attempts. This latter approach is already used throughout all Dutch ED departments [36–38] and in some English EDs [39] to identify at risk children. Any protocolised selection of children for discussion might also require a filter of professional concern about the child, to make the meeting directly useful to practice [35] and to make numbers manageable.

Having a systematic way of identifying children and families for discussion at the meeting might minimize disruption from the absence of key staff (as seen in practice D, Table 1) and avoid reliance on professionals' memory, confidence and motivation to bring cases to the meeting. Due to disincentives to record maltreatment-related concerns, and particular disincentives to use maltreatment-related codes in the patient's notes, these codes might not identify all concerns known to the practice. However, the Royal College of GPs has recommended a simple intervention to improve coding of maltreatment-related concerns [22], which increased coding by about 30% in 10 practices in England (personal communication). Giving codes a specific and visible purpose (using them to "case-find" for the vulnerable family meetings) might promote their use.

#### 4.2.3. Limiting Attendance to Core Members of the Primary Care Team

There is a high level of anxiety and professional vulnerability in child safeguarding work, which can muddle and overwhelm professionals [30,31,40,41]. Qualitative studies have described how defenses to anxiety and vulnerability creep into team work and joint-working practices [31,41]. The pre-occupation with confidentiality and the reluctance to admit that "the ball had been dropped" that we observed in two meetings might be exactly this type of defense and have been descried as such elsewhere [31,41]. A paralyzing pre-occupation with confidentiality is commonly described in joint child safeguarding work between social works and health professionals [11,25].

Keeping the meetings small (limited to the core primary care team) might help create a "safe space" which allows for the uncertainty and "not knowing" in concerns about child abuse and neglect and has been described as essential for thoughtful practice and peer support in child safeguarding work within children's social care [31]. As opportunities for informal professional reflection, peer review, team building, and moral support are squeezed by a target-based primary care system [42], a formal "safe space" might be increasingly important.

The evidence-base does not clearly tell us whether it is better to have small homogenous groups of primary care professionals in vulnerable family meetings or to have larger interdisciplinary meetings. A literature review of team work for complex patients in primary care concluded that occupational diversity in a team promoted positive impact on patient care [27]. However, other qualitative studies of interdisciplinary teams for late-stage dementia in primary care [26] and child safeguarding work [25] report that true collaborative discussion and joint-working rarely took place, and could even be absent in the context of co-location of professionals [25]. Co-locating social workers in paediatric primary care settings was one component of a multi-component intervention that improved health outcomes of vulnerable and maltreated children [43,44]. On the other hand, our data suggest that inviting social workers to vulnerable families meetings will only work if there is complete professional trust and ease between the social workers and primary healthcare professionals, taking an unfeasibly long time to build and rebuild each time a social worker changed post.

In an ideal world, professionals from other agencies could come to a vulnerable family meeting and engage in uninhibited and profitable discussion and joint decision making. However, it seems that this is not so easy to achieve in practice. Limiting attendance at the meetings to the core primary healthcare team might be the type of "street level" solution that Hood describes as being a practical solution to the inherent messiness and ambiguity ("the swampy lowlands") of everyday practice [33].

If vulnerable families meetings were limited to the core primary healthcare team, there would be need of structures to challenge assumed and dominant ways of working and to promote inter-agency working outside the meeting. Such structure would guard against meetings which reinforced "routine dysfunction practice" [45] promote inter-agency working to avoid "silo working" [46]. Inter-agency training might be a one way of doing this, although the same review that reported inter-agency training to be a positive and valued way of promoting inter-agency trust also reported that GPs rarely attended [9]. Given the monitoring role of LCSBs and statutory requirement for GP led clinical commissioning groups to sit on these boards, LSCBs might be able to help promote interagency working around children discussed in vulnerable family meetings. However, LSCBs face significant challenges in term of participation, leadership and resources [9].

#### 4.2.4. Locating the Meeting as Part of Routine Clinical Practice

Lastly, locating the vulnerable families meeting as a part of the routine work of general practice might facilitate motivation of the team (by labelling safeguarding as core clinical work) and frequency of meetings. This might be done by including discussion of vulnerable children in regular meetings about all problems in children, which could also be nested within a meeting about problems about adult patients as in practice D. Locating safeguarding work as core 'medical' practice was identified as one of the drivers of GP engagement with safeguarding work in the previous analyses of the data from this study [16].

These suggested facilitators need to be tested in practice.

#### *4.3. Strengths and Limitations of Our Study*

This was an in-depth qualitative study that achieved a level of candour from participants, as described in previous publications [16] and used researcher observations as well as professional accounts to generate hypotheses. As an in-depth qualitative study of "best practice" sites, we cannot assume that our results can be generalised to all general practice settings in England. Indeed, we found high variation even between a small number of "best practice" sites. Variation is likely to be greater still over the tens of thousands of GP practices in the UK (there were more than 8000 GP practices in England in 2013) [47]. There are currently no data about the proportion of GP practices in the UK which hold vulnerable family meetings, how frequently they do so or with what perceived purpose.

As described above, other qualitative studies above have generated hypotheses about effective team-working and interagency work about other patients in healthcare settings or about vulnerable children in social work or multiagency teams. Currently, there are no epidemiological data on the effectiveness of vulnerable family meetings on outcomes for children, young people and their families in primary care and this was beyond the scope of this study.

#### **5. Conclusions**

Vulnerable family meetings were being used for monitoring children, young people and families in some practices in England. We suggest an opportunity was being missed for peer supervision in this difficult area and joint decision making about how to manage everyday contact with specific vulnerable children, young people and their families in primary care. Using the meetings for peer supervision and joint decision making might help GPs manage the "not knowing" inherent in concerns about abuse and neglect and minimize the potential for isolated working.

Our study suggests that there is likely to be some way to go before team meetings in primary care reach their potential for safeguarding children and young people. To maximize the potential of the meetings, practices are likely to need properly resourced and skilled support.

Other professionals should recognize that the meetings are a potential mechanism for engaging and supporting GPs in their everyday responses to vulnerable and maltreated children. The police, children's social care, education and mental health services should feed information into these meetings and provide support when requested, although attendance by outside professionals might cause more problems than it solves.

A large scale survey of a representative sample of GP practices should be undertaken to test the hypotheses generated by our study. We need a study conducting a series of small and linked evaluations to test the most effective way of running a vulnerable family meeting in primary care. These should include epidemiological data from patient records as well as qualitative data and should include a comparison group and, if possible, a randomized design.

#### **Acknowledgments**

We would like to thank the four practices and participants who gave up their valuable time to make this study possible. This study was funded by an MRC/ESRC interdisciplinary studentship awarded to J.W. (grant number G0800112).

#### **Author Contributions**

J.W. designed the study with R.G. and M.B. J.W. conducted the interviews, analysed and interpreted data. She is the guarantor. R.G. and M.B. designed the study, supervised the analyses and contributed to the writing of the article. D.G. and J.A. contributed to the design of the study, interpretation of findings and contributed to the writing of the article. JW, RG and MB took responsibility for the integrity of the data and the accuracy of the data analysis.

#### **Conflicts of Interest**

RG is supported by awards establishing the Farr Institute of Health Informatics Research at UCLP Partners from the MRC, in partnership with Arthritis Research UK, the British Heart Foundation, Cancer Research UK, the Economic and Social Research Council, the Engineering and Physical Sciences Research Council, the National Institute of Health Research, the National Institute for Social Care and Health Research (Welsh Assembly Government), the Chief Scientist Office (Scottish Government Health Directorates) and the Wellcome Trust (MR/K006584/1).

#### **Reference**


Help for Children and Young People." Available online: http://www.childrenscommissioner.gov. uk/content/publications/content\_747 (accessed on 31 July 2014).

