Learning together to safeguard children: developing a multi-agency systems approach for case reviews

Putting it into practice - Preparation

Identifying a case for review

  • A review should be initiated to answer a particular question or questions.
  • Those questions should not be restricted to understanding why harm has been caused to a child and how it could be avoided in future.
  • Curiosity can usefully be focused on a whole range of practice issues.
  • There are good reasons to focus on routine practice, practice that practitioners and/or families are happy with and innovations that seem to be working well.

There needs to be a reason for conducting an inquiry or case review regardless of the method of learning used – some curiosity to answer some question. However, the reason does not need to be a specific adverse event happening to a child. It can just as well be:

A review triggered by a case considered to represent routine or normal practice can give a deeper picture of how the system is operating to support frontline workers. One featuring new working practices or innovations, for example addressing parental mental health, can contribute towards an evaluation of their effectiveness. Given the ‘deep negativity’ that surrounds the social work profession in particular, ‘whereby few have a good word to say publicly about it’ (Jones et al, 2007: 1) there are also good reasons to highlight cases involving multi-agency working that professionals and families feel positive about. To what extent do these indicate robust systems or involve chance elements?

Selecting the review team

  • A systems review requires a team not just one person.
  • Knowledge of the key professions involved can be beneficial.
  • Outsider status can help workers engage openly in the process.

The systems case review should be carried out by a team of people. The amount of work involved is likely to be too much for any one individual and the critical dialogue between team members is vital to the quality of the analysis and learning. The team should reflect the key professions involved in the case under review. In our pilots, workers’ active and open participation seems to have been aided by our independence from the organisations whose practice was being reviewed.

SCIE’s model is not premised on the two-part process stipulated for SCRs whereby individual management reviews (IMRs) are undertaken by each relevant service and subsequently brought together by an independent person commissioned in an overview report.

Identifying who should be involved

  • People involved in the case include both workers and family members.
  • Without family members key perspectives will be missed.
  • Identifying the professionals whose roles and contributions were most significant may only be obvious over time.
  • It is important to include managers, supervisors etc., not just those who had direct contact with the family.

Ideally, all personnel involved in the case, or part of the case, under review should be involved in the review process. This includes both workers and the members of the family themselves. Research suggests involving the family is possible but this is under-developed in the present model.

Ideally, all personnel from whatever sector and/or agency and at all levels within organisations should be involved in the review. However, as the majority of cases run over a significant period of time, this will often not be realistic. Consequently, judgement is required as to whose roles and contributions were most significant. This is not necessarily self-evident at the beginning of the review, but instead can emerge gradually over time. It is important to try to identify staff who were seen as key by members of the family as well as by professionals. It is useful to involve staff and family members themselves in this process.

Given their management roles and responsibilities related, for example, to supervision, budgets and performance indicators, it is important to include significant first-line managers and not only the staff who had worked direct contact with the family.

Preparing participants

  • Participants need a detailed introduction to the approach.
  • A face-to-face meeting is recommended.
  • The requirements for confidentiality must be made clear.

It is vital that participants are given a thorough introduction to a systems approach before the case review begins. Otherwise it would be difficult for them to participate actively. In the pilot case reviews we initiated contact with participants using an introductory letter – see Appendix 1.

Subsequently, an introductory meeting in which participants can meet the review team face-to-face is recommended. The aim of this is to ensure that they understand the aims of the approach, what it entails and the part they are being asked to play. It also serves to demonstrate in a very tangible fashion the nature of the relationships and dialogue with participants that the review team wants to develop. It can also serve to foster the beginning of a group identity and, therefore, the possibility of joint ownership, across agencies, of the review process and findings.


It is crucial at an early stage that the review team clarify and reassure participants about the priority given to learning over blaming in the systems approach. Organisational backing for this stance also needs to be concretely stated and details about confidentiality clarified.

As a collaborative approach involving a multi-agency group of workers the review team cannot guarantee to keep everything that all individuals tell them confidential. Interim and draft final reports, for example, will draw on the content of individual conversations and need to be shared and discussed with the group. It is important, therefore, that all draft reports remain confidential to participants in the review team and are not, for example, shared with other staff or managers from the participating agencies.

In final reports that might be made public, geographic identifiers should be removed, professionals referred to only by their role and the family by pseudonyms.

Next in this section: Data collection