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

Putting it into practice - Data collection

There are two important sources of data relevant to a systems investigation – the written records of different agencies and conversations with key staff, service users and carers. Reviewers need continually to be comparing the data from these different sources, so that each helps to make sense of the other – critically appraising documentation in light of participants’ narratives as well as further questioning staff about their narratives in light of information the documentary sources reveal.

Selecting documentation

  • Documentation forms the formal record, but access may be restricted.
  • Records provide checks on accuracy but also insights into cultures of communication and how tools are shaping practice.
  • The order in which formal records and one-to-one conversations are accessed is arbitrary; each brings its own bias.

Records provide the formal account of professional involvement. In an SCR, access to these documents is legally permitted. In other contexts, access may be restricted, with a consequent limiting effect on the analysis of practice. These written documents provide essential details but are necessarily and intentionally selective and, therefore, incomplete.

Documentation provides a vital check on the accuracy of the basic factual details of the case. People’s individual accounts are likely to be influenced both by lapses in memory and in being remembered through the filter of knowing what happened later in the case. Separate agency sources also provide a check on accuracy of any one, thus identifying gaps or mistakes in understanding that need to clarified.

Documentation can give significant insights into the cultures of communication both within and between sectors. It can highlight what is included and what becomes written out of the formal record, and to what effect. It can give an indication of how tools are actively shaping practice through the ease or difficulty review team members have in making sense of the information contained (c.f. White et al, 2008: 12).

Reviewers can choose whether to examine the multi-agency documentation before conducting conversations with participants or vice versa. Each will bring its own biases because what you see as significant depends on what you have already found out. New information will continually come to light against which you have to rework your developing overview and analysis. You may realise that you have omitted an important data source, be it document or person, or that you have incomplete information from a particular data source because certain questions and issues have only just become apparent and therefore could not have been explored earlier. Consequently, there will often be the need to return to both participants and documentation in order to follow up. For some individuals, a second conversation may be necessary.

One-to-one conversations

  • Conversations provide the essential viewpoints of the people involved.
  • We have developed a structure for the conversations, but this can be used flexibly to guide the discussion.
  • The style of engagement should be relaxed and the conversation conducted with genuine curiosity and respect.
  • We found it advantageous that the same two members of the review team conduct all conversations.
  • A written record of the conversation is essential.

One-to-one conversations are essential because they provide the data that allows us to build a picture of how things looked to the people involved, at the time they were involved. For this reason, the conversation begins with a narrative account of the person’s involvement, unstructured by the interviewers. Participants are then asked to identify key practice episodes which they believed influenced the way the case developed. Referring to the list of ‘contributory factors’ from various aspects of the wider system, described earlier, the person is then encouraged to consider why they acted as they did.

It is particularly important that the style in which conversations are facilitated should be relaxed and conversational and demonstrate genuine curiosity, openness and respect. If we are asking participants to trust us enough to speak to us in detail about the intricacies of their involvement, we need to respond in such a way that shows we are indeed worthy of such trust.

We found that it is better not to give rules as to how participants should prepare for these conversations. This allows people to bring their own approach and professional or personal norms, which become a further data source, throwing light on both individual and sometimes wider team cultures e.g. relating to the value of paperwork. An excerpt from the letter we sent to participants concerning the conversations forms Appendix 2.

Conversation structure summary

  1. Introduction
  2. Hearing their story/narrative
  3. Identifying turning points or ‘key practice episodes’
  4. Clarifying their ‘local rationality’
  5. Discussing contributory factors
  6. Highlighting things that went well
  7. Their ideas about useful changes
  8. Summing up
  9. Reflections on conversation process

A more detailed version of the conversation structure can be found in Appendix 3.

Two members of the review team should take part in the conversations. This allows one to take the lead in listening and taking notes, recording ‘subtle points that may otherwise be overlooked’ (Taylor-Adams and Vincent, 2004: 11), with the other taking the lead in responding and asking questions to get the participant to elaborate or to prompt their thinking.

We also learnt that there are significant benefits to the same two people facilitating all the conversations. This allows for the overview of the case to be developed more quickly in the course of successive conversations and, consequently, overlaps and discrepancies to be pursued in the course of conversations, thereby minimising (thought not eradicating) the need for follow-up later.

Some form of written record or transcript of the conversation is essential. We learnt from the pilots that shortcuts, such as filling in a data extraction form straight after a conversation, are likely to be too distorting because they will reflect our picture of the case at the time so omit what might be crucial counter-evidence.

Next in this section: Organising and analysing data