Friday 31 January 2014

Human Factors in Child Protection (2): Decision-making

Decision-making is central to child protection. Is this a child protection case? Is this child at risk of significant harm? Is emergency action required to protect this child? Is a child protection plan necessary? Should we start care proceedings? Each of these questions represents a step requiring someone or some group to make a decision, often a weighty one.

Wrong decisions can have momentous consequences. A case may be closed prematurely. A child may be returned home to suffer further abuse. A professional may wrongly decide that her or his concerns do not amount to a concern of abuse or neglect. As a result a child may be harmed, or worse.

So how do we take decisions? Psychologists [2] distinguish between different types of decision making. Below I look at each of these in turn.

Rule-based decision-making involves the decision-maker following procedures which determine the outcome. These often take the form "If X do Y". They may look straightforward, but problems often arise in determining how to interpret the procedure. For example Lord Laming recommended a procedure following the Victoria ClimbiƩ tragedy along the following lines: "If a child's first language is not English, an interpreter must be used ..." [1]. But what do we mean by a "first language" (first language learnt, most fluent language, language spoken at home, language spoken at school etc.) and who can be an interpreter (qualified professional interpreter, fluent speaker, person with some knowledge of the language, member of child's community, member of child's family)? And what happens where the child prefers to speak in (say) English, but is more fluent in another language, or where it is very difficult to find a qualified interpreter in a language not much spoken locally?

Clearly having procedures is a good idea in simple situations, where ambiguity and uncertainty do not apply. Procedures are actually useful in clear, simple and predictable situations. However, for complex situations procedures rapidly become too elaborate and eventually become unworkable. Highly involved procedures can result in unacceptable delays in action. Where procedures are too complex they can be difficult to understand and steps may be missed. Sometimes the wrong procedure for a given situation is applied. And sometimes the procedures themselves contain 'error traps' because the person who drafted the procedure did not foresee some particular set of circumstances. Last, but by no means least, complex procedures which make the task more difficult often result in people developing informal 'work arounds' to get the job done. These can have disastrous consequences, not least because managers may naively think that a procedure is being followed, when it isn't.

In contrast intuitive decision-making relies on the expertise of individuals. An experienced practitioner can spot patterns in situations and react rapidly and very often appropriately. This type of intuitive decision-making, or 'pattern-matching', is necessary where there is insufficient time to generate lists of options or to carry out any kind of detailed analysis. It is often essential in emergency situations. Often the response is 'workable' but not ideal and it is usually quick. This type of decision-making is suitable for 'first responders' in most types of emergency situations, but experienced practitioners are required. It is by no means infallible. Pattern 'mismatches' are always a danger and some very serious accidents have involved very experienced people making the wrong intuitive decision. Intuitive decision-making is not suitable for dealing with new circumstances or situations which are not well understood. It is not suitable for use by inexperienced people, especially novices.

Analytical decision-making involves identifying options, analysing each and selecting the best. Typically the decision-maker thinks through the consequences of each option, exploring the various scenarios, possibly identifying positives and negatives of each and comparing likely outcomes. As much relevant information as can be marshalled to inform the decision is the ideal. It is also important to identify and analyse the 'do nothing' option.

Clearly analytical decision-making is likely to produce optimal solutions more often than other methods and it is well adapted for complex situations. However, it requires a lot of time and effort and places a heavy burden on staff. It is not appropriate for emergency situations and difficult working environments, especially where a response is required quickly.

Creative decision-making involves coming-up with new and untried solutions to unfamiliar situations. It requires inventiveness and high levels of resourcefulness. It is very time consuming and makes very heavy demands on those who undertake it. It is also inherently risky. However, it is sometimes unavoidable when decision-makers are faced with completely new and novel situations about which little is known.

Implications

In child protection we need to use rule-based decision-making sparingly, only in situations where it is appropriate. Procedures will only work for simpler processes. In the past there has been a naive assumption that procedures could be extended more or less indefinitely to cover a wide variety of complex situations. This has resulted in large complex procedural manuals, not better child protection decisions.

In child protection work we should rely on intuitive decision-making only where it is absolutely necessary to provide a quick response, for example when considering emergency action to protect a child. This type of decision-making is suitable only for very experienced staff, but in Britain there has been an unfortunate tradition of providing first contact child protection services using recently qualified or agency social workers. No matter how experienced the decision-maker, intuitive decision-making requires careful review at the earliest opportunity, when time permits.

We need to recognise that much decision-making in child protection will be analytic. This requires good working conditions and staff have to be given the space, time and resources to analyse and assess options and to come to an optimal conclusion.

Creative decision-making in child protection should be avoided unless it is absolutely necessary to deal with new and challenging phenomena. It needs to be recognised that creative decisions in child protection will always be risky and should be avoided if possible.

No approach to decision-making is without a downside. Many decisions in child protection are inherently risky. Decisions should be reviewed where possible using the analytic approach, but where this is not possible the risks need to be realised. Reviewing decisions is time consuming and hard work, but it has to be recognised that it is frequently necessary.

Briefing and de-briefing can be used to support decision-making in child protection. De-briefing the team after an important piece of work can be an ideal opportunity to review the success or otherwise of the decisions taken. What went right and what went wrong? How can we do it better next time?

In a briefing we can identify and plan for likely outcomes before a significant piece of work is undertaken. In this way a well organised briefing can reduce the pressure to produce rapid decisions in a new and challenging situation.

Much decision-making in child protection is undertaken in groups, rather than by individuals. Planning or strategy meetings, reviews and child protection conferences all involve several people, often representing a range of agencies. Decision-making in groups has some advantages but also carries risks. I will be writing a further post of this subject in February.  

Notes

[1] Lord Laming, The Victoria Climbie Inquiry Report  CM 5739, HMSO, Norwich, Jan 2003. Recommendation 18, Paragraph 6.251

[2] See for example Flinn, R et al Safety at the Sharp End, Ashgate, 2003, Chapter 3.

Sunday 26 January 2014

All at sea ... four preoccupations of child protection policy

Since Christmas my Internet access has not been great, hence the dearth of posts on this blog. It's not helped by being on a slow boat to Brazil. We are now on our way back, making our way down the Amazon from Manaus. It's the West Indies next...
Being cut off from the news, especially the more trivial news, has its upside. With less day-to-day clutter around I have been able to reflect on some historical themes. Watching a large wave dissipate its energy as it slapped up against the ship's side the other day, it suddenly occurred to me that we can look at the history of child protection policy in Britain as falling under four headings - the four preoccupations of child protection policy. The similarity of each to that spent wave will become apparent.

The first preoccupation I see as being dominant in the time between the Maria Caldwell report in the early 1970s and about 1990. I call this the procedural preoccupation. The emphasis of child protection policy was on requiring the development of both agency and interagency procedures with the aim of standardising working practices and making it easier for different agencies to "work together". Of course an important tool for implementing this policy was the development in England of the government publication Working Together and the equivalent documents in other UK countries.

The procedural influence lasted quite a long-time because developing procedures for an activity as complex as child protection has no end. Indeed inertia carried over with the development of new procedures extending into present times. About half of Lord Laming's 108 recommendations in his 2003 report into the death of Victoria ClimbiƩ were of a procedural kind.

But the striking thing is that at the time there were few who saw the important limitations of this approach, which are many. The central one is that while it may be helpful to outline a procedure for a relatively simple process, such as calling an interagency meeting or deciding who should attend, it is not usually sensible to create procedures for complex tasks in which there is a lot of variability - such as interviewing a child and family or making a decision about whether the child is at risk of significant harm. Attempts to create procedures for more complex tasks - which are the ones in which critical errors often occur - resulted in ever lengthening rule-books and complex procedures which were difficult to understand and follow.

The second policy preoccupation I see as concerning the collection and dissemination of information. Whereas the procedural preoccupation places emphasis on following rules, the information preoccupation focuses on having what some people see as the 'right' material to do the job.

During the 1990s there seemed to be evidence from both academic studies and from inspections in England that the needs of children and young people were being overlooked in child protection inquiries, which were seen as focusing too narrowly on the forensic question of whether or not maltreatment had occurred.

The re-focusing debate sought to re-direct practice towards gathering information on children's needs. Central to this approach was the development by the Department of Health of a range of assessment tools, designed to help practitioners reveal the complex needs of abused and neglected children, by collecting lots and lots of information about them. An early example was the so-called 'orange book' which was used as the basis of conducting long in-depth assessments where there was already a concern of abuse and neglect. But the longest lived and arguably best developed tool was the Framework for the Assessment of Children in Need, which was developed by an 'expert' group at the Department of Health.

The problem was that you can quickly have too much of a 'good' thing. "The framework" may have appeared to many as an important advance, but there was much about it that was arbitrary and poorly justified. Why was asking questions about some things vitally important while other things were ignored without explanation? The policy-makers were not interested in those sorts of questions. Rather they tried to impose the 'framework' - which had originally been put forward only as recommended practice - as a mandatory tool. The initial and core assessments which became part of the Working Together child protection procedures were the first steps. Then the framework was adapted into the 'exemplars' of the Integrated Children's System, a blueprint for a computer system that seemed to become more and more bizarre as time went by. Not only had the numerous questions of the assessment framework become reified as database 'fields' but the completion of these fields was to be policed by 'workflow' which required data to be added within predetermined timescales in order to prevent the record and its user being escalated for management action.

There had been a rapid movement from 'professional tool' to bureaucratic nightmare.

An emphasis on 'prevention' led to calls for the scope of collection and dissemination of information to be widened. In the wake of the Laming report, the Common Assessment Framework was developed - a 'Children in Need Framework Light', perhaps. This was widely focused on any children about whom concerns were unresolved, but while the benefits of so doing were 'hyped' there was no research evidence to show the effectiveness of such an approach.

The accompanying initiative was the ContactPoint computer system which had the aim of creating a computerised record for every child in England (probably in Wales, but bizarrely not in Northern Ireland and not in Scotland!). This is not the right place to re-examine ContactPoint, but it seems to me that the further we get from its much welcomed demise in 2010, the odder it looks.

While the procedural preoccupation resulted in a proliferation of procedures, so the information preoccupation resulted in a proliferation information systems. As with procedures, so with information - the limit is endless, and there is always an argument for having more, but often little to show in terms of concrete benefits for all the effort.

That brings me to the third preoccupation, structure. For the last thirty years of the twentieth century there was little change in the structure of the organisations which delivered child protection services in Britain. In the new century, post-Laming, things began to look different. Laming himself had something of a structural orgy in his report, suggesting a variety of new organisations and structures to provide what looked like a national child protection service. The government of the day didn't go that far, preferring to leave the primary role to local authorities, but they did start to mess with structures in quite significant ways.

Overall responsibilities for child protection, in central government, were transferred from health to education, which subsequently became the Department for Children, Schools and Families. At local level education and children's social services were merged to form 'children's departments'. Not only that but 'children's trusts' were set up to co-ordinate children's services across all relevant agencies and Area Child Protection Committees were retitled Local Safeguarding Children Boards (LSCBs) and given beefed up membership and new responsibilities. New roles such as Lead (elected) Member (for children's services) and Director of Children's Services were created. If progress could be measured by the number of new signs on doors it would have been rapid.

Sadly as with so many structural reforms the waterbed effect has been substantial. It is not just the fact that the new government post 2010 has subtly undermined some of the 'reforms'. Rather the changes have slowly unravelled and there is little evidence of things being any more effective as a result of the changes. LSCBs are a good case in point - a new name, new members, new responsibilities but pretty much the same lack of effect! Children's Trusts, which were much heralded as key co-ordinating structures, have now all but disappeared.

That brings me to the fourth policy preoccupation, which I call the managerial preoccupation. This is the crude idea that by setting central targets for children's services, and enforcing them through rewards and punishments, quality and safety of services can be improved. This was very much on the agenda of the last government, up to 2010, but the present government has also reverted to target setting, despite rhetoric against it, for example in setting specific targets for local authorities regarding adoption. Hopefully the thoroughgoing critique of target setting in the Munro Review will now give central government policy makers pause for thought. But what is difficult to eradicate is the target setting mentality at local level. Time will tell whether the message of Munro is absorbed and acted on or whether organisations revert to type.

I think each of these four preoccupations has been generally unhelpful. They have resulted in complex rules and regulations, poorly focused and poorly designed information systems, wasted resources in re-structuring organisations and wasted effort in trying to achieve arbitrary objectives, which often have unforeseen and unintended consequences.

That's why I think we continue to be "all at sea" - just like I have been since the start of January.

The Munro report seemed to me to offer the possibility of a safe port in heavy seas - perhaps even an alternative preoccupation. It stresses the need for a 'systems approach' to understanding how failings and errors in child protection occur and how to secure improvements in safety and quality. The idea is that it is by understanding the weaknesses in relevant systems, rather than simply blaming the idiosyncratic weaknesses of individuals or teams, learning takes place. When significant failings occur there must not be a knee-jerk resort to more procedures and more blame, to setting targets and enforcing compliance. Rather there must be a genuine attempt to understand how and why systems fail.

Sadly I do not detect much evidence of the widespread impact of the Munro reforms. I suspect that many people do not fully understand Munro's analysis or the implications of her arguments. It is, after all, unfamiliar territory.  While there is widespread acknowledgement of the quality of her report, I do not see much evidence of practical implementation at local level.

I would like to see two new preoccupations develop in the future: quality and safety. These I believe offer the prospect of more tangible benefits.

In the 1990s there was a nasty little initiative focused on looked after children called Quality Protects. The title may have been good, but the initiative was based on silly little targets and performance indicators that were an insult to all concerned. Quality is not about daft ideas like this. It is about gaining a deep understanding of the needs of those who receive the service and inventing innovative and creative ways of meeting them.

A great deal is known about quality. A vast literature exists, little of which little has been systematically applied to child protection. In particular we know that service quality has to be measured by reference to the satisfaction of those who receive the service, which means asking them what they think. So quality in child protection has to be child-centred. There will never be a quality service unless we listen to what abused and neglected children have to tell us.

Secondly quality is not just about conforming to specifications. It is about a constant and never-ending search for improvement. Small, frequent improvements are particularly good because they are often easy to implement and cumulatively over time they can have a dramatic effect. Often those who know best how to improve the service are those who directly deliver it - the people at the front line.

It is a bit similar with safety. We cannot improve safety unless we understand what mistakes occur and how they are caused. We need information systems that give us good data about the things that go wrong. I've been banging on about confidential near miss or critical incident systems since 1990 and I don't intend to stop. We need to create a reporting culture, where people are not punished for reporting their mistakes, but encouraged to do so. Would you want pilots to hush up near air misses? Why should abused and neglected children expect less?

Understanding mistakes is all about having the right conceptual framework to understand human factors. Psychologists have given us good theories and concepts. All we need to do is to apply them to child protection.

Back to the sun lounger on deck eleven ....