The next item of business is a statement on the justice system’s approach to risk assessment by the Cabinet Secretary for Justice and Veterans, Keith Brown. The cabinet secretary will take questions at the end of his statement, so there should be no interventions or interruptions.
15:22
I will update the Parliament on an issue affecting the level of service and case management system, also known as the LS/CMI system.
LS/CMI has been used in Scotland as a paper-based system since 2006, and as an information technology system since 2010. It supports risk assessment and case management for individuals with a history of offending. The LS/CMI system is used by social work and prison staff as one part of a wider set of processes to inform a number of decision points within the criminal justice system, including sentencing decisions, programmes access and prison release decisions.
In 2019, the system was centralised. As of 22 November 2021, all 32 local authorities and the Scottish Prison Service, which had each previously hosted the system locally, were migrated on to a centralised IT system.
The LS/CMI risk assessment tool is not the only risk assessment that is used within the justice system. A number of different tools are used for a variety of different types of offenders. This issue relates only to the LS/CMI risk assessment tool. A number of validated risk assessment tools support and inform professionals in their decision making. Those include risk assessment tools specifically related to violent offending, sexual offending and intimate partner violence and stalking.
In the prison environment, LS/CMI is used as one of a suite of risk assessment tools. The choice of risk assessment tool in that context is partly influenced by the nature of the index offence. Management of the assessed risk is governed by a multidisciplinary team of professionals referred to as the risk management team. The RMT is chaired by a prison senior manager who is supported by a range of professionals including, but not limited to, criminal justice social work, psychologists, health professionals, Police Scotland, local authorities, chaplaincy and third sector agencies. The challenge of assessing and managing risk draws together that diverse range of professions in the shared objective of protecting the public by preventing or minimising harm.
A recurring theme in risk management practice is the need to balance the safety of potential victims with the human rights of the offender. This multidisciplinary approach, which involves a range of professionals, is also often used for managing individuals within the community setting. Risk assessment is dynamic and, as members see, it is holistic in its nature and is never based on one assessment within the system.
I will explain in detail the circumstances relating to two issues that have been identified, and outline the precautionary measures that have been taken at every step to ensure that confidence in Scotland’s public protection arrangements is maintained. I apologise in advance for the somewhat technical explanation around some of the factors at play. Given the importance of the issues, it is right that we take a precautionary approach.
Following a call about a single case, which was raised by a user to the system helpdesk in January of this year, the system issue was explored by the IT managed service provider for LS/CMI and by the Risk Management Authority, which both provide a helpdesk service for the system. That work sought to understand whether the issue affected just that particular user or was more widespread. Following those detailed investigations, it became apparent last week that the issue affected other users of the system, and test scripts were immediately developed to identify affected cases.
The particular systems issue affects the display of information in the risk assessment part of the system. In some particular instances, the numerical risk score value does not match the risk score level that is displayed by the system. A systems issue appears to prevent any subsequent changes being made to that risk level when new information has been entered.
As of this week, there were 103,394 assessments in total on the live system and, of course, individuals can have more than one assessment. There are approximately 24,000 open cases on the live system. An open case is one where there is some on-going management of the individual in the justice system, which requires use of LS/CMI. I am advised that, from the work that was carried out over the weekend, there were 1,317 assessments where the calculated score did not match the final risk need or level. Of those assessments that were affected, 1,032 relate to closed cases and 285 relate to open cases.
The system enables social workers to override the risk level shown on the system and, of the 1,032 closed cases, there are 537 where an override has been applied by social work. That is the professional judgment being applied to a risk assessment. That means there will be 495 closed cases that appear to contain a risk level that is affected by the system error. The 537 cases that have an override applied will need a case-by-case review to determine whether the override superseded any error.
Officials have taken immediate action to review open cases that the justice system is still managing, and I will say more about the review process shortly.
On Friday of last week, my officials issued an immediate update to users of the system—principally, justice social work but also SPS staff—to make them aware of the issue and provide a temporary solution, so that cases could be identified on the system and a form of override applied.
Work is on-going to identity the specific affected cases and, on Tuesday, my officials issued details of the open cases and locations of those cases to users of the system, and asked them to specifically review risk levels and scores and take any necessary actions.
The action in those cases will have been for users of the system to determine whether the risk level that was shown for the cases that they manage was correct, and to apply an override if not. Users of the system were also asked to involve partners, if there was any impact on the on-going management of that individual within the justice system. My officials are assembling returns from users of the system as they do so. To date, 150 returns have been received, and no users of the system—social worker or SPS—have advised the Scottish Government of any public protection risk as a result of that systems issue.
Justice social workers, whether community or prison based, are trained professionals and will always apply professional judgment to every individual that they manage. The nature of risk assessment is holistic and wide ranging. It is not mechanistic and is never solely based on the LS/CMI risk assessment tool.
The initial error has been investigated as I have described. Further to that, and as part of that investigation, previous change logs and helpdesk calls have been reviewed in parallel, to explore any issues that interact with the systems issue. It seemed only sensible to do that. From those investigations, it appears that there might be another area of risk scoring, in relation to alcohol or drug use, that creates an error. Although the extent of that is not known, it is clear that it might affect the risk score. As presently reported, the initial evidence on the system indicates that the score is likely to be higher rather than lower, so it overstates risk rather than understates it.
Given that this second issue has been identified as potentially affecting cases, I have—again as a precautionary measure and to ensure that we take no risks with public protection issues—agreed the following actions.
All social workers have been asked to review all open cases on the following priority basis: cases that are due for imminent consideration of release from prison, or for entry into multi-agency public protection arrangements, or to move on to licence.
Justice social workers have been asked to move, with immediate effect, to the paper-based system that is part of the agreed contingency plan in the event of any system failure. Extensive support from Community Justice Scotland has been put in place should there be any immediate training implications. Other risk assessment tools in the system are paper based, however, so the system that those professionals move to will already be familiar to many of them.
We are working with the IT company that manages the system and, if necessary, additional expert IT capacity will be deployed to assist with the rigorous assurance process of every element of the system that we now need to carry out.
I have convened a risk review group, which will be led by the Risk Management Authority, to work through, as a matter of priority, the open and closed cases to assess whether the errors have had an impact on how those cases were managed in the system. I reiterate that this work is being done to provide further assurance, and not because we are aware of any issues around the management of offenders due to the issue. I am keen that this work concludes swiftly, and I am, of course, willing and eager to report back to the Parliament on its outcome as soon as that group has concluded its work or has initial findings that it is appropriate to share.
In terms of further actions, officials have written to other justice partners with an interest in these matters, including Police Scotland, the Parole Board for Scotland, the Scottish Prison Service and the Scottish Children’s Reporter Administration. Officials have also written to victims organisations to ensure that they are sighted on the issue and so that officials can offer necessary reassurance around the work in hand. If any of the victims organisations would find it reassuring to meet with me so that I can reassure them about the actions that are being taken, I am more than happy to make that offer, as well as to involve them in these processes.
I have also written to the Criminal Justice Committee today, to make it aware of the issue. I will continue to update the committee as more information becomes known.
It is important to reiterate that this issue has not resulted in any concerns being expressed by social workers, in the community or in prisons, around any offenders who are in the justice system. Our approach will always be precautionary and evidence based.
I am updating the Parliament on the issue to ensure openness and transparency, but I hope that the steps that have already been taken, and the openness with which we are dealing with the matter, provide members with some reassurance.
I have confidence in the professionalism of our justice and health professionals who every day manage changing and evolving risk across a range of offenders. As has been explained, LS/CMI will never be the sole determinant of how the justice system deals with the risk that is associated with an individual. There is much more by way of judgment and process involved, and determination will often involve a multidisciplinary range of professionals, who are never just following what is displayed on the system. I am very grateful to them all for their continued support in ensuring that we retain confidence in how we protect the public from offending behaviour as we move forward.
I will continue to update the Parliament on the matter, as appropriate. I am happy to take any questions.
The cabinet secretary will now take questions on the issues that have been raised in his statement. I intend to allow around 20 minutes for questions, after which we will move on to the next item of business.
I would be grateful if members who wish to ask a question would press their request-to-speak button now or type R in the chat if they are using BlueJeans.
We now know why this mysterious, last-minute statement on justice was kept quiet until it was released to us. Wow! Clouded in jargon, it all sounds like a simple technical error off the back of another bungled IT centralisation project. The reality, though, could not be more stark—or, actually, shocking. These are vital systems that are used to score the risk to the public of criminals before they are released early. The admission today that there are potentially hundreds of cases in which the assessment of that danger was wrong will be—and should be—a source of grave concern to us all. The fact that there are 495 of them is shocking.
The justice secretary reassures us, of course, that officials have checked 150 of those cases so far, and he tells us that
“no users of the system ... have advised the Scottish Government of any public protection risk”
whatsoever as a result of the so-called “systems issue”. The key word that he missed out there is “yet”. They have not found any public protection risks yet, because they do not know the full picture yet.
Trying to disguise this in technical jargon is one thing, but let me try to get some immediate clarity for the public. First, how many prisoners were released early or wrongly when they should not have been? Secondly, when will the public find out whether they have been put at risk, in any way whatsoever, by any of this gross incompetence? Thirdly, will the justice secretary categorically rule out—right now—the early release of any prisoner whose assessed risk to the public is, or even might be, wrong?
The member makes reference to a centralisation process—I forget how he described it—in a derogatory way. We do not know this yet, but it may have been the centralisation, which happened in 2019, that was the means by which the technical glitch, which is what it was, was discovered. However people may like to describe it, there was a technical problem with the program. Centralisation meant that all the councils and the Scottish Prison Service were on the same system, which may well have helped us to find the issue in the first place.
I mentioned that the work is on-going. We were advised of the situation last Friday. The work that has been done over the weekend and right up to this point has led us to the statement that I have made and some of the facts that I have given, but there is more to be done. I have been very clear about that. There is no way around the technical information. I would have been slated had I not given the technical information behind the issue.
It is true to say, and I am perfectly willing to admit, that I would have liked to come to the chamber with all the facts in front of me, but there was also pressure to tell the Parliament as soon as possible. I have conceded that more information will come out, and I will be happy to report to the Parliament in the future.
The member asked whether anybody has been released. I have given the facts as I know them, which are that all the returns that have come back so far indicate that there is no risk of somebody having been released early. On the second point, which was to do with the alcohol factor, it may well be that the risk was overstated. However, we will, again, have to wait until more facts come forward.
Therefore, there is no evidence—as yet, as the member rightly said—of any risk to the public from anybody having been released when they should not have been. As I have said, more information will emerge. It is not the 495 cases that the member talked about but the 200-plus that I mentioned, 150 of which have come back so far. However, it is also true to say that there is a much bigger piece of work needed to go back through the history of the situation right to 2012, to make sure that we got it right in relation to the closed cases as well.
I am trying to be as open as possible, and I am happy to come back to the Parliament and answer more questions in the future, when we have more information.
I thank the cabinet secretary for bringing this urgent matter to the attention of the Parliament. A pre-brief would have been helpful to provide some explanation of the system and the terminology in the statement. I recognise that the social workers and health professionals and their risk management teams will be working very hard to resolve the issue. It is not at all clear to me from the statement what the real risk is to communities from what seems to be quite a significant error. That is my sense, having had first sight of the statement only an hour ago.
I realise that social workers and multi-agency partners will be taking a holistic approach to all cases. Will the cabinet secretary give an example of how the issue may be detrimental to offenders, given that he said in his statement that the risk might be overstated? More importantly, where it might put communities at risk, what is the cabinet secretary potentially concerned about? I think he said that there are 1,032 cases to be reviewed in total. Will he confirm whether I have understood that correctly and what resource will be needed to review what seems to be a high number of cases in a system that relied so heavily on an IT model?
I will answer as many of the questions as I was able to take in. On Pauline McNeill’s first point, she is right that it is technical information and the statement has lots of that information, including the figures that I mentioned. She asked for something like a glossary or explanation of some of the terms. There is more information in the letter that I have sent to the Criminal Justice Committee, because it is easier to put the information in a letter.
I am sure that the committee will want to discuss the issue in the future, and I am more than happy to provide further explanation and briefing on it if the member wishes. A great deal of work is being done by the social workers and the review panel that I have set up. That will involve a substantial amount of work for the people involved, which will be done as a priority.
Pauline McNeill mentioned the issue of risk, which I mentioned in relation to the alcohol aggravator. Bearing in mind that the final judgment on a case is a professional judgment by the people involved in the system, the risk scoring, which seems to have been the issue, was giving that aggravator too high a bearing. Sometimes, when alcohol was no longer deemed to be a risk factor, the risk scoring might not have been showing a reduction in risk. That is one of the things that is being investigated.
I am happy to provide more information to the member. Either she can write to me or, since both she and Mr Greene are on the committee and the convener is here too, I am happy to provide information through that route and to come back to Parliament as well with more information when we have it.
In his statement, the cabinet secretary mentioned that, in the prison environment, the LS/CMI tool is used as one of a range of risk assessment tools. Will the cabinet secretary expand on what other risk assessment processes are working alongside that system?
The member is correct in saying that it is never just the LS/CMI assessment that determines, for example, prison release. A multidisciplinary team of professionals, which is referred to as the RMT, governs the management of assessed risk in prison. A prison senior manager chairs the RMT, supported by a range of professionals including, but not limited to, criminal justice social workers, psychologists, health professionals, Police Scotland, local authorities and third sector agencies. A range of other specialist tools for assessing the risk of sexual and violent offending are also often used alongside the LS/CMI—it is not, to use the old phrase, the only tool in the box. It is probably not even the most important one, considering the professional judgment of the list of professionals that I have just given.
Yet again, the Scottish National Party’s justice system fails law-abiding Scots. On the day that the Parliament debates international women’s day, we do not know how many women that incompetence has put at risk. The cabinet secretary has to approve the first grant of temporary release applications for murderers and rapists who are serving life sentences. I have a specific question: are any of his approvals among the botched cases? If so, how many?
Notwithstanding the further work that we have to do, I am confident that the answer will be one that reassures the member—he will perhaps be reassured if I write to him with the exact details of the matter.
The member is making an outright political attack when the matter is quite serious and the public are possibly concerned. Honestly, we are just getting sick of this script—“The SNP Government this, the SNP Government that.” Perhaps the member should treat the issue as seriously as it should be treated. The issue is serious and can cause people concern, so it deserves a serious discussion rather the political rhetoric that we have just had from Mr Findlay once again.
What lessons have already been learned? Will the cabinet secretary reiterate the details of the next steps that the Scottish Government is taking to resolve the issue?
Throughout this situation, we have sought to take a precautionary approach, to ensure that all potentially affected cases are reviewed and to revert temporarily to the LS/CMI paper-based system while we assess the impact of making the required changes to the IT function. The member will have picked up that we want to do that with the open cases first and then move on to previously closed cases.
I have mentioned that we are also asking the risk management authority to urgently convene the review group, which would draw in other justice partners as needed. The work of constituting that group began today. Once the immediate and on-going review of live cases is concluded, the review group will examine the overall impact of the two errors in the IT system and I will report back to Parliament at that stage—Parliament permitting, of course.
Has the cabinet secretary been provided with a breakdown of the kind of offences that those who are involved have been convicted of? For example, does he know how many sexual offenders are involved in the cases, and has he been advised of the potential implications of the errors in those cases?
It sounds as though the work will be resource intensive at a time when the legal system is already under massive pressure. What information has the cabinet secretary been provided with by the various agencies about the work that will likely be involved?
On the first point, as far as I am aware, it would not be possible to give a breakdown of the offences, because, as I mentioned, many different tools are used in relation to different offences, often in conjunction with other tools, and all those tools are used in conjunction with professional judgment. However, I will look into the matter and see whether I can provide more information to the member.
The second question was about the resources that are required to carry out the work. It is the nature of the job that they will have to do it—it is very important, so the work will begin right away. As the member knows very well, it is in the nature of the work that is done by criminal justice social workers and staff in the Prison Service. Justice professionals will be familiar with the paper-based system that will be used while the review is undertaken, not least because some of the other systems that they currently use are still paper based and because some staff will have used the LS/CMI paper-based system prior to the IT system being adopted. It will not be new to those members of staff, but I recognise that the review will take time. Given the on-going, resource-intensive nature of the work, I am happy to report back to the committee and the Parliament on it.
I refer members to my entry in the register of members’ interests.
I was working in the sector when the LS/CMI was introduced, and I recall the fairly robust training programme that we had to undertake. Although the override option, as it was called, involved robust protocols, including several levels of management approval—I hope that that is reassuring for my Conservative colleagues—it was nonetheless always by far the most contentious part of the LS/CMI. It is therefore perhaps not surprising that that aspect is receiving more Government scrutiny. That is surely a good thing, because it should help us to make the system more effective and consistent.
On behalf of my former colleagues and other criminal justice social workers working across Scotland, will the cabinet secretary provide more detail about the support that is being provided to them as they move, with immediate effect, to using a totally paper-based system?
They will be helped by the review process itself. In any event, it is a good idea to have the review, and it was good that what was picked up, in the first instance, by somebody in the Prison Service was not accepted as a glitch. It could have been perceived as such, but it was not written off in that way. There were thorough checks to see whether there was a wider issue, which there was, and that led to further checks and a second issue in the system being found.
The system was introduced in 2006, I think, but it moved to being IT based in 2012 and has now been centralised. To go back to the very first point that Jamie Greene made, I am not aware that there was anything in the centralising process that caused the issues; we know that that is not the case in relation to the two technical problems that have emerged. Indeed, it might be that centralisation, which is supported by all the different justice agencies, is the means by which the problems have been found.
As Fulton MacGregor asks, we will continue to provide support to all the professionals who are involved. I recognise the points that were made by him and Katy Clark that the matter is providing additional work in a pressurised area. We will make resources available to ensure that the professionals can get through this without it affecting their other work.
I thank the cabinet secretary for early sight of his statement, but I echo Pauline McNeill’s concerns about some of the technical language that was used.
It is clear that a key component for assessing the risk of convicted people—a component that feeds into how the justice system deals with them—has been displaying information wrongly. That has potentially been happening for years, and it has potentially impacted decisions on release, sentencing and who should be subject to MAPPA monitoring in relation to sexual offenders. Will the cabinet secretary confirm whether the IT system has been overstating or understating the risk not just in alcohol and drugs cases but in general? When does he expect the urgent review of the remaining cases that is now under way to be completed?
On Liam McArthur’s first point, I cannot provide confirmation. We have received information on 150 cases—the number was 150 first thing this morning, but it will be higher now; that is the pace at which the situation is being worked through—and those cases have not yet thrown up any public protection issues, which, in relation to the idea of overstating or wrongly stating, does not yet concern the professionals who are involved. I might be able to say more about that as more information comes back to us and as we go through the rest of the functions.
I apologise, but I have forgotten the second part of Liam McArthur’s question.
I have mentioned the means by which we are trying to push through the review. We will have significantly more information in the next fortnight, but I do not want to commit to when the review will be finished, because we must be sure that we have had a complete check. As well as finding out about the issues and concerns, we want to ensure that the entire system is working.
I cannot give Liam McArthur the assurance that he asked for just now, but we should have much more clarity on the major issues, many of which have been raised by members today, within a fortnight.
Three more members are seeking to ask questions, so I ask for succinct questions and answers.
Will the cabinet secretary provide a commitment that the Scottish Government will continue to keep Parliament and the Criminal Justice Committee updated as the situation develops? Will he outline how he will do so?
Yes. As I have said, I am happy to do that. I think that I am due to appear before the committee for about two and a half hours, on different issues, next week. It will be up to the committee—[Interruption.]. I apologise—I did not hear that. I will be happy to answer any questions that are asked. If the committee wants to change the nature of its questions as a result of today’s statement, the Government will be responsive to that.
I have already committed to come back to Parliament on the issue. There is a judgment to be made about when it would be appropriate for me to do that. It is a judgment on which we cannot win. I want to make sure that Parliament is informed as soon as possible. That has been the injunction from the Presiding Officer. I have sought to inform Parliament as soon as possible. It is six days since we were made aware of the issue. Much work has been done to make sure that we can get as much information as possible to members. As and when we have more information to provide to members, I will, of course, come back to the committee and the Parliament. In the meantime, I will also be happy—this relates to Pauline McNeill’s questions—to correspond with individual members.
I thank the cabinet secretary for providing early sight of his statement and for his stated intention to keep Parliament informed and to be as transparent as possible, which I take in good faith.
Although there is clearly a place for risk assessment systems such as the LS/CMI, does the issue that we are discussing underline the case that victims, survivors and people who are convicted of crime are, first and foremost, people with individual needs and, therefore, highlight the importance of having sufficient capacity and resources to treat individuals with respect and care? Will the cabinet secretary outline what additional support he is providing to ensure that that is the case?
In relation to the specific issues that we face, I have mentioned the additional resources for the review group and the Government’s willingness to provide additional resources to help the professionals who are involved in that exercise.
However, I think that Maggie Chapman’s question is a wider one about resources more generally. Our recently produced vision for justice will give her a clue as to how we intend to best use the resources that we have. She mentioned victims. As I mentioned in my statement, we will contact victims organisations—I think that they will have been contacted by now—and will involve them in the process through membership of the risk review group, should they wish to take that up. We will work with victims organisations rather than directly with victims, for reasons of which Maggie Chapman is, I am sure, well aware.
I was a little surprised by the response that the cabinet secretary gave to my colleague Russell Findlay. As a very experienced minister, the cabinet secretary knows that, when he comes to this place, he will be scrutinised by members on all sides.
I have a specific question about this catastrophic failure. As of the moment the cabinet secretary entered the chamber, how many offenders have been released, possibly wrongly, and are still out on the streets, as it were, unchecked? He said that the problem was first identified by a user in January, but it took until last week for the Government to realise that it affected hundreds of cases. When was he first made aware of the scale of the problem? Why did it take him from some time last week until today to come to Parliament? I welcome his coming to Parliament—it makes a change. [Interruption.]
Mr Kerr, could you please finish the question, because we are running out of time, as I have already indicated? You have had quite a long run-in. Could you please conclude your question?
I am trying to finish my question, but it does not help when the Deputy First Minister shouts abuse.
Please just conclude your question.
Well, I have asked my questions. When was he first made aware of the situation? Why did it take him from last week until today to come to Parliament?
Mr Kerr makes a fair point about my response to Mr Findlay. However, if a question comes with all sorts of political rhetoric added to it, I will respond to that. I am happy to respond to factual questions, as it is a serious issue, and that is what I have tried to do.
I will respond directly to Mr Kerr’s question. On 24 January, a member of the Scottish Prison Service who was using the system found an issue with it. They contacted the help desk that is provided by those who are there to support the system. They had to be certain that it was not an individual user issue, and it took time to do that. They ran tests in parallel with the system. That took until 23 February. [Keith Brown has corrected this contribution. See end of report.]
The Government was advised that there was a wider issue with the system on Friday afternoon last week. That is when we were told. Over the weekend and since then, work has been going on non stop in an effort to get a resolution for the system and the workaround of a paper-based system, and to gather more information so that I could make as full a statement to Parliament as possible. I think that we have acted pretty quickly. Of course, we are open to criticism, but I think that that was the right way to do it.
I have tried to answer the question of how many people have been released who should not have been released by saying that, of the 265 open cases that have to be looked at, more than 150 have come back with no public protection issues whatsoever. That number was from this morning, so it might be higher now. Of course, if there is any change to it, I will update members through the processes that I have already mentioned.
That concludes the ministerial statement.
Before we move to the next item of business, I remind members of the Covid-related measures that are in place, and that face coverings should be worn when moving around the chamber and across the Holyrood campus.
Air adhart
Public Service Broadcasting