Public Audit and Post-legislative Scrutiny Committee

Meeting date: Thursday, March 16, 2017

Official Report 570KB pdf

---

Section 23 Reports

---

“i6: a review”

Under agenda item 2, we will take oral evidence on the Auditor General for Scotland’s report entitled “i6: a review”. I welcome to the meeting the Auditor General for Scotland, Caroline Gardner; Mark Roberts, senior manager at Audit Scotland; and Catherine Young, audit manager at Audit Scotland.

Before I invite Caroline Gardner to make an opening statement, I want to put today’s evidence session into context. At our previous meeting, the committee took evidence from the Scottish Police Authority and Police Scotland on the SPA’s 2015-16 accounts. It is clear that challenges remain in ensuring that the expected savings from police reform are actually generated.

The report that we will consider today looks at i6, which is the national information technology system that was intended to replace a number of IT systems that were used by the police. The report states clearly:

“The failure of the i6 programme means that some of the benefits of police reform that should have arisen from implementing it, have been, at best, delayed.”

Given what we know about the Scottish Police Authority and Police Scotland’s finances, that is very concerning.

Today, we want to examine why that programme failed and whether that failure could have been avoided. The report does not cover IT developments in Police Scotland since the termination of the i6 contract, so we will focus on what lessons can be learned from that failure.

I invite the Auditor General to make an opening statement.

In June 2013, the newly established Scottish Police Authority signed a contract for £46.1 million with Accenture to deliver a national information technology system for Police Scotland. The system, known as i6, was intended to replace more than 130 IT and paper-based systems inherited from Police Scotland’s predecessor forces, and to transform how Police Scotland records, manages and analyses information. The system was planned to go live in September 2014, with full national roll-out by August 2015.

In July 2016—three years after signing the contract—the Scottish Police Authority and Accenture mutually agreed to terminate the contract for the i6 programme. The contract was signed by the SPA, but day-to-day management rested with Police Scotland.

The i6 programme failed because of disagreement about the scope of the programme, the interpretation of the contract and the extent to which Police Scotland’s requirements were met by Accenture’s solution. The disagreement arose despite an 18-month procurement process, which we found had followed good practice and included intensive engagement with potential bidders. The disagreement surfaced almost immediately after the signing of the contract in June 2013, and permanently damaged trust, relationships and confidence between Police Scotland and Accenture. The signing of a contract variation agreement in April 2014 reset relationships and put the programme on a more positive footing, but that proved to be temporary.

One of the factors that contributed to Accenture being awarded the contract was its experience of delivering an IT system to the Guardia Civil in Spain. Initially, it was believed that the majority of the i6 system could be based on that system. As time went on, however, it became clear that a significant amount of the system would have to be built from scratch.

The approach to software development that was used in the i6 programme meant that Police Scotland finally received the i6 system for full user testing in June 2015. At that point, Police Scotland identified fundamental problems with the system. In August 2015, Accenture indicated to the i6 programme board that its confidence about meeting the revised go-live date of December 2015 was at 91 per cent. However, following an analysis exercise, Accenture revised its estimated go-live date to April 2018. At that point, the SPA, Police Scotland and Accenture undertook an options appraisal exercise to agree the best way forward.

Following that exercise, the SPA and Accenture entered into negotiations and agreed to terminate the programme in July 2016. They agreed a settlement of £24.65 million. That figure included a refund of all the money that had been paid to Accenture—a total of £11.09 million—plus an additional payment of £13.56 million.

As you said in your opening remarks, convener, the failure of the i6 programme means that some of the benefits of police reform that should have come from it have been, at best, delayed. There are also wider implications for the modernisation of the justice system.

Our audit concentrated on the history of the i6 programme and the factors behind its failure. I have asked the auditors of the SPA to monitor information and communications technology developments as part of the annual audit process. That is particularly important given the emphasis that is placed on the use of technology in the recently published draft policing 2026 strategy. I also put on record my intention to publish in May this year a report on the lessons to be learned from this and other ICT projects that I have reported on.

Alongside me are Mark Roberts and Catherine Young, who have worked on the project. As always, we will do our best to answer the committee’s questions.

Thank you very much. Colin Beattie will open the questioning.

Sadly, we have become accustomed to public sector IT projects that do not deliver. In your report on i6, you say:

“The process for procuring a supplier for the i6 system followed recommended good practice.”

Oddly enough, Police Scotland does not seem to have done anything wrong. It followed the recommended procedure, but at the end of the day that did not work. Is there a problem with the procedure? Is the “recommended good practice” not up to snuff?

I will ask Mark Roberts to comment in a moment, but I think that we can say that in this case Police Scotland did follow good practice and that there were some real benefits from that, although the programme ultimately failed. The quality of the contract that it entered into meant that there was no direct loss to the public purse, and Police Scotland was able to settle with Accenture in a way that recovered the payments that it had made, plus a significant sum in compensation. That reflects the benefit of a thorough procurement process, even though it did not work through in this case.

Is the public purse actually facing an opportunity loss, rather than a financial loss?

Indeed it is. The loss is of the wider benefits that the system was intended to achieve in terms of savings to policing and operational benefits to police officers and staff. We say in the report that there was no single reason for the failure, but it is certainly clear that the procurement approach that Police Scotland took in this case was not one of the reasons for the failure.

Paragraph 8 on page 6 of your report shows

“Deloitte as the external experts on procurement and managing commercial contracts

Eversheds as legal advisers

Exception UK as technical advisors.”

That must have cost quite a bit. Do we know how much those external consultants cost?

I do not have the figure here. My colleagues might, or the committee may want to pursue that with the SPA. Colin Beattie asked earlier about good practice. Good practice indicates that it is important for the client—Police Scotland, in this case—to make sure that it has access to the necessary professional skills. In this case, legal skills, procurement skills and information technology skills were the right ones. The committee might recall that one of the problems with the NHS 24 system was that the contract itself did not stand up to challenge when things went wrong. That did not happen in this case, which is clearly positive.

Even with all that external support, however, the project still did not work; even following good practice, it did not work.

It did not work. As we say in the report, there was no single reason for that. I will ask Mark Roberts to talk the committee through what we think the underlying factors were.

As the Auditor General said, strenuous efforts were made during the procurement stage to bring in external expertise and to learn lessons from experience. There was a large amount of engagement with potential bidders. The process kicked off in 2011 with large-scale meetings with potential bidders that gradually focused down to the preferred bidder. There was also intense discussion with potential bidders; there were around 160 dialogue workshops to discuss the technical requirements of the system.

One of the other factors that we draw out in the report is the waterfall approach to the development of the system. Although it was widely used at that time, we have clearly heard that that approach to software development or system development would not be recommended these days—there would be much more focus on an agile approach to development. There was therefore a methodological issue that meant that it was quite hard for Police Scotland to see the scale of the challenges that were facing the programme until quite late in the overall timeline.

I am still looking at the external support. Paragraph 19 on page 10 talks about the Scottish Government’s gateway team, whose external reviews

“suggested that delivery confidence was either amber or green”

throughout the programme, and its recommendations were adopted by Police Scotland. Again, how could that be? Even at a relatively early stage, there were indications of some problems, although Accenture gave assurances that they were being managed. However, here we are again with an external party doing the necessary box-ticking exercise, and it still did not work.

As you can imagine, we looked at that issue quite closely in this exercise, particularly given our experience of other large IT projects and the problems that they have encountered. Our conclusion in this case was that, because Accenture was giving strong and positive assurances to the programme board about its ability to deliver i6, and because the waterfall approach that Mark Roberts has described meant that the system could not be tested until the whole system was in place, it would have been difficult for the gateway reviews and the other reviews that were taking place to raise stronger concerns than they did. The reviews identified some areas for improvement that were acted on, which is another positive action from Police Scotland, but we feel that the way that the system was designed to be delivered, and the assurances that Accenture gave, made it difficult for the reviews to reach different conclusions.

Are we saying that, no matter what we put in place, if a company such as Accenture in effect lies, there is no protection against that?

09:15  

I cannot speculate about what underlay the assurances that Accenture gave to the programme board. I can tell you—and we say in the report—that the programme board strongly challenged the assurances that it got from Accenture through the process.

The approach to developing such a system is likely to be different in the future. For example, we have talked about the agile approach, which means that we test as we go along rather than have a wholesale development that is handed over to the user at quite a late stage in the process. The agile approach should enable problems to be identified and tackled earlier than happened with the i6 programme. There is clearly an issue with the internal resourcing and delivery capability of any large contract under the waterfall design approach, particularly in the case of IT systems, which tend to be quite a black box, even to people with expertise in technology.

Looking in the round at how things worked, is there an aspect that we can learn from to try to protect ourselves for the future? Is there something that we can do better in order to get an earlier warning that there is a problem?

I will ask Mark Roberts to comment in a moment but, as we say in the report, the programme was happening at a time when there was a great deal of political and professional scrutiny of the establishment of Police Scotland. A number of things were also going on around the establishment of the Scottish Police Authority—I have reported on some of the disagreements and tensions.

A previous IT performance management system had failed before the establishment of the i6 programme. There were also some high-profile challenges for policing around, for example, stop and search, armed officers and control centres. All of that probably had the effect of increasing the focus of both the SPA and Police Scotland on ensuring that the i6 system worked, which might have introduced the optimism bias that is often a factor in similar large projects. Mark Roberts can give you a flavour of what he saw from the inside of the audit work.

The programme board rigorously considered the option of terminating the contract as early as October 2013, which was three months after the contract had initially been signed. However, there was an agreement between Police Scotland and Accenture to continue the process and try to make it work. As the Auditor General said, there was a great deal of determination on both sides to deliver the system. That resulted in the signing of the contract variation agreement in early 2014, which was within the first year of the programme. There was willingness to consider the option of terminating the contract, but the ultimate decision was to carry on and to try to make it work. As with any major programme of work, there were always likely to be difficulties and challenges, but it was thought that they could be overcome. However, some of the early problems and their impact on relationships flowed through the course of the whole programme.

Good morning. I will follow on from Colin Beattie’s questions about the start of the process. The report is clear that Police Scotland appears to have followed recommended good practice, and we can accept that Accenture will have gone into the contract not to fail but in good faith. However, I am interested in the role of the external experts who were advising Police Scotland. Things started to go wrong within weeks. Does not that suggest that the advisers—Deloitte, Eversheds LLP and Exception UK—bear some culpability for what happened?

To go back to my response to Mr Beattie’s question, the contract was robust enough to enable Police Scotland, at the end of the process, not only to recover the payments that it had made but to receive a significant sum in compensation. In my experience of similar cases, that is exceptional. Something clearly went badly wrong, and we have identified a number of things that we think contributed to that, but Police Scotland’s contract was as strong as it could have been in protecting the organisation from the direct consequences of what went wrong. There was clearly an opportunity cost in terms of the savings and operational improvements that were intended to be made, but the contract itself was robust.

I accept that the contract was robust in the sense that it allowed Police Scotland to recoup indemnities if something went wrong. However, in terms of its scope and all parties’ understanding of what they were signing up to, it appears to have been significantly deficient. Is that right?

It is certainly true that there was significant disagreement between the two parties about what the contract was intended to deliver. One of the underlying factors involved Accenture’s experience of delivering a system for the Guardia Civil in Spain, which it thought would provide a very strong basis for the development of i6. That was a significant factor in the decision to award Accenture the contract. In practice, however, some of the things that were set out in the contract—this contributed to Police Scotland’s ability to recoup the payments that had been made—differed significantly from the Spanish experience. Mark Roberts might want to add to that.

The crux of the disagreement focused on a very small part of the contract. The contract referred to the user requirements that Police Scotland wanted the system to deliver. Accenture believed that it had presented a solution that would meet those requirements. It quickly emerged that, despite—as the Auditor General mentioned—an 18-month competitive dialogue process, there was a gap between the two perspectives. However, another clause in the contract said that the user requirements took precedence over the contractors’ solution, which enabled Police Scotland to say, “What we say about our user requirements must take precedence, and we want you to deliver that.” Accenture’s position was very much to say, “You have agreed with our proposed solution, and that is what we will deliver.”

Does not that suggest that there was a significant degree of ambiguity in the contract that the parties signed?

We could certainly interpret it in that way.

When I was in professional legal practice, I was paid to make sure that there was no ambiguity in my clients’ contracts. Does the ambiguity in this contract suggest that the failing—or a failing—lay with the professional advisers who had been engaged?

We have not looked into that in detail; the committee would have to discuss it with the SPA and Police Scotland.

To stay on that point, Colin Beattie quite rightly asked how much was paid, and I accept the answer that you gave, Mr Roberts. How much was paid specifically to Accenture? Was that money recouped as part of the penalty that was required from the firm? I suspect that it was not.

We do not have the detail that would allow us to break down the payments that Police Scotland made to the external consultants as part of its client team. As the Auditor General said, the settlement agreement contained an additional payment, over and above the moneys that had been paid directly to Accenture. That reflected some of the additional costs that Police Scotland had incurred, such as estimated staff costs, hardware costs and software licence costs, but I do not know whether it included estimated costs for the third-party advisers.

I wonder whether that should be investigated.

The committee could consider the matter with the SPA.

I will move on in the process timeline. At the start, as we can see, there was apparently an 18-month window in which Police Scotland was getting a great deal of advice—potentially very expensive advice—and it tried, as I suspect all the parties did, to put in place something that worked. There was then a contract variation, to which the same rigour does not seem to have been applied. Would it be fair to say that?

There was not such an in-depth process, but, as it involved a variation of the existing contract, that seems reasonable. I think that, once Police Scotland knew more about where the issues were, it concentrated on where attention needed to be paid. The process involved adjusting the existing contract rather than starting the contract discussions from scratch again. For example, Police Scotland took on responsibility for data migration, which Accenture had previously been going to deliver. They were adjusting the existing contract on a relatively small scale to try to pick up some of the problems that existed—

Who was doing that?

That was being done by Police Scotland, with advice from its external advisers, in discussion with Accenture.

Do you know who those external advisers were?

They were the same advisers who had been involved from the start, as listed in paragraph 8 on page 6 of the report.

Again, the contract variation did not work. Does not that suggest that somebody was poorly advised?

I am not sure that I would follow that interpretation; Mark Roberts will want to come in again on that. The fact that Police Scotland was able to agree a contract variation indicates the strength of the contract that it had entered into. The revised contract embedded the user requirements—as Mark Roberts described—which filled the gap between what Police Scotland had interpreted as the user requirements in the original contract and what Accenture thought that its solution would deliver. The contract variation did not change the cost, although it did change the delivery milestones and the delivery date. It clarified Police Scotland’s user requirements at no additional cost. You are right to say that that did not work, but it was a best-endeavours attempt to enforce Police Scotland’s user requirements in the contract at that point. Again, the issue is the confidence in, and the capability of, Accenture to deliver those requirements within the cost and the broad timescales that had been agreed. Is that a fair comment, Mark?

That is a fair reflection of the situation.

I have a final question. Has any person or group really been held to account for those failures? I accept that there are a number of moving parts but, at a number of points, the ball seems to have been dropped. Who dropped the ball, and what has happened to them?

The formal solution was that the parties agreed to terminate the contract, and Accenture made a significant payment that refunded the £11 million that it had received in milestone payments together with £13.5 million as compensation for staff time and other costs that had been incurred. In a direct sense, therefore, the penalty fell to Accenture.

As the committee knows, there has been a number of personnel changes in Police Scotland and the SPA since they were established in 2013. The broader questions might therefore be better addressed to those bodies.

I will touch on some of the project management issues. You said that there was a waterfall approach that basically involved going through all the phases of the system build and then handing it over to Police Scotland to do the user testing. How many phases did that involve? What was the timescale from the start of the project to when it was handed over to Police Scotland in June 2015?

The project kicked off with a high-level design phase very soon after the contract was signed. In broad terms, there were four further stages. The detailed design, functional design and product testing stages were performed by Accenture, as would be the case with any contractor in a programme management set-up, and the final stage was user acceptance testing. The system was handed over to Police Scotland only at the final stage so that it could see what the system was doing. That happened in the summer of 2015—as you will recall from the Auditor General’s previous answers, the system as a whole was initially expected to go live in September 2014, so Police Scotland did not see it until some nine months after the original go-live date.

During the course of the programme, dates slipped. Waterfall methodology involves one stage following another, but some of the phases that were originally envisaged as sequential began to overlap. The i6 programme board presented that to Accenture as a significant challenge: it asked why the stages had to overlap, given that that would generate additional risk. The overlapping was an attempt by Accenture to keep the programme moving and on track.

It has been a long time since I worked in software development, but I cannot for the life of me understand why someone would sign up to an approach that involves the developer going through the design stages and the build stages—which I assume would involve internal testing—and handing the system over only once that is all completed. Have Police Scotland or its external advisers been questioned on why the waterfall approach was accepted?

09:30  

We asked programme managers who were around at the time why that approach had been adopted. At the time, it was very much regarded as the accepted approach. Accenture provided demonstrations to Police Scotland during the course of the programme, but there was no full handover of the code so that Police Scotland could really get into the nitty-gritty of it.

It is clear—indeed, the chair of the Scottish Police Authority has said publicly—that such a programme will not be undertaken in that way again, and that a much more modular and agile approach to development will always be adopted in future. At that time, however, people were determined to use the waterfall approach.

Did anyone query the approach?

We are not aware from looking at the board papers that there was any challenge from people asking whether there was an alternative development methodology. Catherine Young might have more detail on that.

In the latter part of the programme round, there was some discussion about whether to develop some of the modules in an agile fashion. However, as Mark Roberts said, there is no evidence that such an approach was fully considered, given that both parties had already agreed to the waterfall approach. The modular approach was growing in popularity, but at the outset the waterfall approach was used.

Was it accepted that Accenture would set out the approach that it would adopt and Police Scotland would sign up to that?

There was an agreement between Accenture and Police Scotland—or rather, the organisations that preceded Police Scotland, because that process was going on prior to its establishment. The original business case was agreed by what was then the Association of Chief Police Officers in Scotland; the entire procurement phase predated the establishment of the SPA and Police Scotland.

Was there any visibility of, or transparency around, Accenture’s internal testing before the handover to Police Scotland?

As I said, there were demonstrations of what the system was doing. Police Scotland’s programme management team visited Accenture’s development centres in Newcastle and Mumbai. However, as I said, there was no detailed handover of the code.

Did Police Scotland know whether Accenture was doing proper testing?

It knew that Accenture was doing testing. There was quite a lot of challenge from the programme board about Accenture’s methodology for testing, which the board thought did not reflect industry standards. Accenture was saying, “This is our testing methodology, which we have used elsewhere in many cases, and that is what we will use,” but there was tension around the approach.

With the exception of the system demonstrations, which basically involved getting people in a room like this one and firing stuff up on a screen, was there no visibility of the actual results of the testing?

There was not, as far as we are aware.

A number of flaws were identified at the product test stage, and there was a lot of disagreement between Accenture and Police Scotland about how critical the errors were. There was a lot of challenge from board members about the errors. When the programme moved into the user acceptance stage, some of the errors that had been found at the product test stage were unresolved.

There was also a lot of challenge about the methodology not being standard industry practice. A new programme manager for the Police Scotland side had a significant IT background and was able to get in and look at the coding. Board minutes show a lot of discussion about how the detailed coding did not meet standard industry practice.

That does not sound very satisfactory.

It is clear that the waterfall methodology has serious flaws. How confident are we that the waterfall approach is not being used for other IT projects in the public sector, and that a more robust approach is being taken? That would involve the products being handed over in phases to allow users to test them more robustly as each phase is completed, which would give them more confidence in the products overall.

As I said in my opening remarks, I am planning to publish a report in May that looks across the range of IT projects on which I have reported over the past few years. It is certainly true that an approach such as the waterfall method, which was widely used five to seven years ago, is now much less commonly used, and we are seeing much more use of the agile approach. However, in some of the reports that I have produced, we have raised concerns that the agile approach is being used by people who do not fully understand it or are not properly trained in it. I will pull all that together in my May report, which should give the committee some assurance about the current position.

Thank you.

I want to follow the same line of questioning. According to the report, there was a year and a half’s worth of dialogue but even so, at the end of the day, the functionality of Accenture’s programme never met Police Scotland’s requirements. That is a fairly fundamental issue, and I am trying to understand exactly how it came about. Was Police Scotland clear from the outset about its technical requirements for the system?

Yes. I will ask Mark Roberts to come in, but the user requirements specified in the contract were the basis of Police Scotland’s ability to achieve the settlement that it did achieve when the plug was pulled.

Police Scotland was—and probably still is—confident that its user requirements set out exactly what it wanted the IT to do, and Accenture responded with a solution that it felt met those requirements, with some qualifications, and believed that Police Scotland had agreed to that. Accenture felt that going beyond its solution there had been what it described as a continual elaboration of requirements—in other words, the scope was getting bigger and bigger. At the same time, however, Police Scotland maintained that it had not changed its user requirements.

As I said in answer to Mr Kerr’s question, there was a precedence clause in the contract that said that the user requirements were paramount. That is what Police Scotland continually referred back to, and that is what was at the heart of the difference between Police Scotland’s feeling that it was clear about what it required and Accenture’s feeling that there was continual expansion of the scope of the programme.

That is very helpful. So Police Scotland had been clear about its requirements and expectations from the outset.

On page 8 of the report, you talk through the procurement process and the early design stage. It says in paragraph 14 that, in November 2012, as part of that long process—we know that procurement can take time—

“Accenture scored highest against the technical and implementation criteria”.

However, it did not score the highest on cost. Was there a failure in due diligence on the procurement side?

When we spoke to people, it was made very clear to us that Accenture’s experience of delivering a system to the Spanish police force, as the Auditor General has described, was a very important criterion, albeit that that was a simpler system. That was reflected in the weighting assigned in the procurement scoring exercise. At the same time, it was made very clear that, although cost was an important factor, the quality of the system was what Police Scotland really wanted. That is why the technical and implementation criteria were very important in Accenture’s securing the bid.

On the question whether there was a failure in due diligence, as we have said in the report—and as the Auditor General has said—we think that Police Scotland followed best practice even though that did not manifest in a successful outcome, as we have discussed. Nevertheless, the procurement process was robust.

Following on from that answer and Colin Beattie’s earlier point—and, indeed, going back to the point that Police Scotland had been clear about its requirements—I wonder whether it would be fair to suggest that during the process Accenture had in any way been misleading with regard to its offer and what it was advising it could produce for Police Scotland. Given that, at the end of the day, the user requirements and what Police Scotland wanted were significantly different to what it got, is it fair to suggest that Accenture misled Police Scotland?

A fair reflection might be that Accenture had underestimated the complexity of the system. A repeated theme throughout the challenge presented by the programme board was the extent of knowledge of policing and the specific requirements of the Scottish policing environment of the people who were getting involved from the Accenture side. That issue was raised repeatedly by the programme board.

In response to Colin Beattie’s questioning on this subject, the Auditor General said that the SPA board provided challenge. What kind of challenge did the board present to Accenture at the time, and how robust was it?

It was the i6 programme board that provided the challenge. I will ask Catherine Young to talk you through it, as she is the one who has spent an awful lot of time reading the board minutes and talking to people about what happened at the meetings.

At the initial stage when a gap between the requirements and the contractor solution was identified, there was a lot of discussion about the fact that neither party had picked that up in any of the 160 dialogue sessions. It all came back to the interpretation of the requirements.

The search function was quite a significant issue at that stage. The Accenture solution said that it would do one thing, but the requirement was different, and that was picked up only at the very early design stages. Because the requirements obviously took precedence in the contract, Police Scotland’s standpoint was that the solution must fulfil those requirements.

Throughout the other stages of the programme, the challenge was around the quality of information that the board was receiving from Accenture, and there were questions around the test strategy and the methodology that were being employed. I should also point out another layer of governance, which related to day-to-day contract management. I believe that, at that point, there were more discussions around the detail of those issues.

A lot of concern was also expressed about the timeliness of documentation, because an awful lot of material was being provided to the programme board at the very last minute. As Catherine Young has said, the quality of the documentation was another focus of challenge.

Finally, for completeness, I should point out that, after the quite early deterioration in relationships, the programme board asked Accenture to express its confidence in delivery as a percentage instead of simply using red, amber and green flags, and those percentage ratings were challenged quite strongly by the programme board.

I have another question on that point. As the report highlights, there was a breakdown in trust and confidence at a very early stage. If there was no trust and no confidence in Accenture’s ability to deliver at such an early stage, why did we continue? Why was something not done earlier? If it was never a workable relationship, how could it ever have been made productive?

That is a very good question, and one that we touched on in our earlier answers. A great deal was riding on the programme both for the police in Scotland and for Accenture. You are right that problems emerged very soon after the contracts had been signed, but Accenture was very concerned to deliver the system, which was seen as important for it globally. For the police, there was real concern to demonstrate that it could deliver an IT system that fulfilled its requirements in order to boost confidence in the new policing arrangements in Scotland and because of its importance in achieving the financial savings and the operational benefits that had been foreseen in the business case.

That is one of the areas where we think that lessons might be learned about not allowing such pressure—such optimism—to colour the decisions that are being taken when problems start to arise. That does not mean that we think that the actions that Police Scotland took to try to bring the contract back on track were not appropriate, but there is a question whether the plug could have been pulled sooner.

That takes me neatly to my last question. Paragraph 26 on page 11 of the report mentions the

“failure of a previous police ICT project in 2012”

You would have thought that lessons had been learned. However, that paragraph finishes by saying that there might have been

“misplaced optimism about the prospects of success”.

It sounds to me like there has been misplaced optimism throughout the whole process. Is it fair to suggest that a lot of this was done on a wing and a prayer, with eyes closed and people hoping that it was going to get better, only it never did?

09:45  

I do not think that that is fair. There was an element of optimism—and you are right to point to that paragraph and sentence in the report—but it was not naive optimism. Police Scotland and the programme team did what was required of them and did it thoroughly, against a backdrop, as Mark Roberts has said, of a complex but real disagreement with their contractor about what the system was going to deliver.

It is unusual in the public sector for the client not only to recover what they have paid but also a sum in compensation, and that compensation reflects the fact that Police Scotland followed good practice in procurement. The project itself was undermined by the difference in view about what the system should deliver.

This has been going on for four years now. The original plan was for the project to save about
£200 million or thereabouts over a period of time. Does the compensation element include compensation for not realising all or part of the £200 million savings that would have been realised had the project been handled properly and competently?

First of all, Mr Neil, it is good to see you back.

Thank you.

Secondly, we are talking about £200 million of financial savings plus other operational benefits across a 10-year period, the phasing of which is set out in appendix 1 of the report.

The information that we have been able to obtain from the SPA is that the compensation payment was not for the benefits forgone but for staff time and other direct costs such as software licences and hardware maintenance that had been incurred, and it was the result of a negotiation between the SPA and Accenture rather than a precise figure. In short, therefore, the information that we have is that it covers those direct costs rather than benefits forgone.

Should it have included benefits forgone?

That is a question that you will have to ask the SPA.

As Auditor General, do you think that it should have included that element to protect public money?

That is clearly my main concern about the programme. As I have said in the report and in my opening remarks this morning, the project was a central part of police reform with regard to the savings that were intended to be achieved and, much more significantly over time, the improvements in the way that police officers and police staff carry out their work and the way in which the whole criminal justice system can operate.

Those benefits have been delayed at best, and there is currently no plan for when and how they will be achieved. I am not in a position to say whether it was ever feasible for the SPA to negotiate compensation for that opportunity cost from Accenture, but the question now is how we go about achieving those benefits in future.

Were there any restrictions in the contract that would have prevented compensation being claimed for savings forgone when things broke down?

I would be surprised if there were restrictions of that sort. The contract was about what would be delivered at what price rather than the compensation arrangements if it were not delivered.

Do we know whether the SPA asked for any element of compensation to cover the savings forgone, or did it just forget about it, ignore it or not raise it? Did Accenture say, “No way”?

As it was a commercial negotiation, you would need to ask the SPA about the content of it.

When do you think that Police Scotland is going to be in a position to put a plan in place to recover the situation and get a computer system that realises those savings?

I have said in my report and on the record that it is urgent that it does so. It has forgone having those benefits in place now, based on the original delivery dates.

The policing 2026 strategy that was published a couple of weeks ago is centred on much better and more flexible use of IT to help police officers be out and about doing their work on behalf of all of us across Scotland. Currently, there is no plan and no date for achieving those benefits. It is something that the committee might want to explore with the SPA and Police Scotland.

It is definitely something that we will want to follow up.

Finally, you said that you would publish the lessons to be learned from this episode. When are you going to do that? Can you give us any indication this morning of what the three or four main lessons are likely to be?

We are working towards a planned publication date of May. As I said in response to Colin Beattie, that report is intended to pull together lessons learned from all the IT projects on which I have reported during my time as Auditor General. I would prefer to hold off on commenting until that report is published, but many of the examples will be familiar to members of this committee.

What was the Scottish Government’s role in all this?

Mark Roberts might want to pick that up.

As we have discussed, the Scottish Government provided us, at various points before the awarding of the contract and in the early stages of the post-contract timeline, with gateway reviews of certain aspects of the programme. For example, it looked at governance arrangements and made recommendations in that area.

In the first year of the programme, after the awarding of the contract, the then deputy chief information officer in the Scottish Government joined the programme board to provide additional support. That individual has a commercial and software development background, and he supplemented the existing programme board by providing external challenge and expertise.

We have seen the same thing happen in the national health service: the Scottish Government sends in troubleshooters after problems arise, which is clearly the right thing to do. However, did the Scottish Government have any formal responsibilities in this area, and were those understood by both Police Scotland and Accenture?

The contract was signed between the Scottish Police Authority and Accenture. Although I am sure the Government was aware of what was going on, it did not, over and above its membership of the programme board, have any direct involvement in the day-to-day management or the scrutiny of what was happening.

You have outlined the political context and say that the Government had no day-to-day role in the scrutiny of what was happening, but at that point there was a lot of pressure on Government to make a success of the legislation, which had passed through Parliament not long before. Was there any indication from the research that you did in your audit that the Scottish Government could have taken a more proactive role in trying to facilitate a better relationship or a more profitable contract?

We have not seen any evidence to suggest that that was explored in any great detail. It was very much the case that the set-up was going to be maintained and day-to-day programme management handled within Police Scotland, with ultimate oversight through the Scottish Police Authority and—even more indirectly—through ministers.

Do you know how often reports went to ministers?

I am sorry, but I do not.

Paragraph 20 of your report states:

“The i6 programme had difficulties almost immediately”.

We have all been in a situation in which, looking back on a project’s lack of success, we have thought, “Actually, there were teething troubles right at the start. Maybe we should have called a halt to it or looked for an alternative at that point.” Was there any indication in your audit that that option had been considered?

As I said in response to an earlier question, there was consideration as early as October 2013—in other words, within three months of the contract being awarded—of whether to terminate the contract at that point. As we describe in our report, and as we have said today, there was genuine determination on both sides to try to make the project work. It was recognised that teething problems, to which you have alluded, were inevitable, and the parties involved were going to try to resolve those issues. However, it is clear with hindsight that, ultimately, that did not happen. Whether those involved could have pulled the plug earlier than they ultimately did is a moot point, given the context of the system being a very important element of police reform that was going to improve information management, sharing and analysis significantly across the national force.

Was there any ministerial involvement in the consideration of whether to terminate the project?

Not as far as I am aware. I think that that was done at SPA or Police Scotland level.

You have said already that the report makes it clear that the project was of global significance for Accenture. It is a big company and it was a huge contract. Do other countries have a public sector body, such as we might be lacking here in Scotland, that can guarantee good practice in IT?

We have said in evidence to the committee before that the United Kingdom Government digital service is doing some interesting things. It is too soon for the National Audit Office to be able to give an assurance that the digital service is avoiding things going wrong—no approach will ever give you a blanket guarantee that things will not go wrong—but it has been investing in a centre of excellence that has up-to-date high-level digital skills, in training people more widely across public services, and in ensuring that the most up-to-date approaches are being used.

Some of that thinking is being picked up in the Scottish Government’s new approach to information systems. The chief information officer can tell you more about the approach that it is taking, but both approaches are an attempt to tackle the conundrum that Mr Neil has identified in the past, which is that it can be hard for the Government and for public sector bodies to attract the scarce and highly paid skills that are needed to make projects of that nature work well. The UK Government’s approach is one example that we have seen, and the report that we will publish in May will give other pointers to international good practice.

You have described what the UK Government is doing. Is any similar work happening in Scotland?

You have heard from us and in evidence from the permanent secretary, the chief information officer and others about the approach that they are taking. We have said that we think that the approach that is being taken here in Scotland is definitely a positive move, although it is too soon for us to say that it is having the desired effect and is operating effectively and at the scale that is needed—especially given the range of opportunities for using digital technology to transform the way public services are delivered and reduce costs. There are moves being made, but it is too early to say whether they are fully effective.

To go back to the Scottish Government’s role, do you know the last point at which the Government provided public assurance on the project?

I would prefer to check when the final gateway review took place, and to put that information to you in writing. I would say that it was in 2014, but I will need to confirm that.

At various points, the Scottish Government said that delivery confidence was either amber or green. Given all the background disputes that are outlined in your audit evidence, why did it say that?

We answered a question earlier about the value of the gateway reviews and the other external checks that were done. As we say in paragraphs 18 and 19 of our report, the Government identified some improvements but was broadly giving amber and green ratings for delivery confidence. That comes down to two things, the first of which is the waterfall approach that was being taken, which made it impossible to see flaws until the system was handed over for user testing much later. The second thing is the levels of assurance that were being given by Accenture at that point, which is a common theme all the way through: Accenture was expressing very high confidence. The approach to producing the system meant that it could not be tested by the users independently of Accenture until later on, so at that point the reviews were focusing on the processes that were in place and relying on the assurances that were coming from Accenture.

Would it be fair to say that the amber or green warnings were more about the Government’s confidence in the two parties trying to sort out a problem, rather than being ratings for actual delivery of the project?

One of the earlier health checks—at the contract variation stage in April 2014—comments on the relationship challenges between the two parties, but that review focused very much on the programme management and contract management side of things. There was some discussion about whether Police Scotland or the SPA would oversee the contract management; the report recommended that it should be done by Police Scotland, and that an extra layer of governance should be built in with an IT scrutiny forum. That was one of the health checks that had an amber rating. At the next point, Police Scotland had taken on board that recommendation and put in place an IT scrutiny forum that included representatives from the SPA and Police Scotland. It moved to green at that stage.

I think that one of the final checks was in August. I think, too, that a technical assurance check was done very late on when the system was already going to fail. I do not have that in my chronology, but we can double check it.

10:00  

It is probably worth stressing the point that the Scottish Government’s various reviews and health checks were technical processes. They considered the content of the contract and how it was delivered rather than being policy oversight of the establishment of the SPA and Police Scotland.

John Foley of the Scottish Police Authority said that there was no detriment to the public purse as a result of the i6 programme. Is that an accurate statement?

It is accurate in relation to the direct cash costs to the SPA and Police Scotland. They recovered the payments that they had made to Accenture and a significant sum larger than those payments to compensate for staff time, software licences and hardware maintenance. In those terms, and bearing in mind the fact that the negotiated sum was a negotiation and an estimate, there has been no loss to the public purse. However, it is also true to say that the i6 programme was to be central to achieving many of the benefits of police reform—the savings that would help to close the funding gap on which I have reported to the committee previously and the operational benefits to enable police officers and staff to carry out their work more effectively.

As Mr Neil said, the issue is the loss of the benefits.

Yes.

Good morning, Auditor General. Accenture said that the user requirements were increasing but Police Scotland said that they were staying the same. Who was telling the truth?

Mark Roberts has given you a clear picture of the user requirements, as set out in the contract, being the element that took precedence within the contract. The contract variation embedded that in what Accenture agreed to. There were changes to the programme that went through the contract variation process, as we would expect in any area, but the core of the problem was not Police Scotland continually revising its requirements.

There were changes, and the original contract included a change-control process that allowed mutually agreed changes to be made. Accenture argued to the programme board that there was an elaboration of requirements and that Police Scotland was making the project more complicated than it had initially said it would be, whereas Police Scotland said that it had, in its user requirements, been clear about what it expected the system to do. As we have described, that gap was never filled.

Given those differences of opinion, have Accenture and Police Scotland both accepted the report’s findings?

Yes. As the committee knows, I agree the factual content of my reports with the people on whom I am reporting. In formal terms, that relates to public bodies—the Government, Police Scotland and the SPA in this case—but, as a matter of courtesy, I sent a copy to Accenture as well and asked for its comments on the report’s factual accuracy.

We touched on the fact that there has been no detriment to the public purse and that the operational benefits have suffered as a result of the failure. Has it had an adverse effect on the staff?

There is no question but that it has done so in terms of opportunity cost. I say in the report that police officers and police staff still have to use 130 manual or IT processes, many of which come from the previous forces before Police Scotland was established. There are problems with officers and staff having to enter the same data into multiple systems and to search multiple systems to get information back out again. Time is wasted doing that and there is a risk of errors being introduced when data is entered again. All that makes the jobs of police officers and staff more difficult and reduces the time that they have available to spend providing policing services.

Have there been any improvements to the IT system at all?

There certainly have. Police Scotland has invested in a new national network and in replacing outdated hardware, including desktop computers. Other projects have been under way to deal with matters such as managing custody, which is an important part of police business. All that is good and is progress, but it is not a substitute for the benefits that i6 was intended to achieve.

You mentioned the policing 2026 strategy, which is quite ambitious on use of technology. Do you think that a single national IT system for policing is still viable?

That is a question that you should ask the SPA and Police Scotland, but their response to my report has been that they would not go about the process in the same way again. It has been quite a salutary experience, with regard to how to procure the necessary IT and how to plan the process. Having learned from that experience and from how things are done in the technology world more widely these days, they would take a more incremental approach.

You said that there is currently no timescale for an IT system. This might also be a question for Police Scotland and the SPA. Are there any indications that they are even thinking that far ahead at the moment?

As the Auditor General said in her opening comments, we will follow up on that through our annual audit process. We heard about the development of smaller components that built on systems that had been used by other police forces elsewhere in the UK or by the previous Scottish forces, but it had been done on a piecemeal basis. We are talking about the development and roll-out of a very small part of the system.

We are waiting to see what the underlying implementation plans are once the policing 2026 strategy is finally agreed. The police’s ICT strategy will be key to its achieving the vision that has been outlined for 2026.

Do you think that the police and the SPA will have that ready when the policing 2026 plan is finalised?

I think that that is a question for the SPA and Police Scotland.

Exhibit 1 on page 7 of your report shows the six areas that represent 80 per cent of policing activity. Given that the i6 system was designed to deal with those six areas, if it had worked and was now up and running, would it already—in the light of the policing 2026 strategy—have been obsolete?

I am sorry to be boring but, again, that is a question for the SPA. What I can say with confidence is that i6 was intended to cover around 80 per cent of the information flows around policing. Many of the case studies that are used in the 2026 strategy, such as those that involve police officers gathering information about a missing person or someone who is acting strangely, would fall within the categories that are shown in exhibit 1, and that information would have been available to other people in the wider criminal justice system. Inevitably, the system would have needed to continue evolving, as our phones and tablets do, but I think that the core of it was absolutely aligned with what the policing 2026 strategy is about.

I see that Mark Roberts wants to add to that.

The Auditor General got in before me. I was merely going to say that it would have been hard in 2010 or 2011 to envisage officers being based in police cars and using tablets as a way of recording and sharing information at the origins of the i6 programme, but the fundamental principles of information sharing and being able to do searches across the whole system remain valid.

I want to ask a question that the public would like to know the answer to. Accenture seems to have held its hands up and accepted a significant degree of responsibility, but that begs the question whether Accenture should still be able to tender for public contracts and, if so, whether it still has a prospect of winning them.

I am sure that Accenture is tendering for public contracts. We are not aware of whether it is actively involved in tendering for any in Scotland or whether it holds any in Scotland; I am afraid that we do not know that.

Would you be able to find that out and write to the committee about it? Is that information commercially confidential?

I suspect that information about potential bids would certainly be commercially confidential. Maybe that is a question for Accenture.

We can explore that in Parliament.

Did you identify any operational errors resulting from the failure of the i6 programme?

I think that it would be impossible to link any operational failings directly to the failure of the programme. That said, we are all aware of things that have gone wrong with call handling and other aspects of operational policing. The i6 system was intended to make those things more reliable, more efficient and more effective for police officers, staff and—more importantly—for all of us in Scotland.

Thank you very much for your evidence.

I suspend the meeting until 10.15.

10:10 Meeting suspended.  

10:15 On resuming—  

---

“Changing models of health and social care”

We move on to consider the Auditor General’s report entitled “Changing models of health and social care”. We did not have time to consider the report at our meeting on 9 February. I apologise for that and thank the panel for coming before us again today.

I welcome our witnesses from the Scottish Government. Paul Gray is director general of health and social care and chief executive of NHS Scotland, Shirley Rogers is director of health workforce and strategic change, Christine McLaughlin is director of health finance and Fiona McQueen is chief nursing officer.

Paul Gray does not wish to make an opening statement, so we will move straight to questions, the first of which is from Alex Neil.

As this is the first public meeting that I have been at since my recent incident, I put on the record that I received excellent care from the Scottish Ambulance Service and the cardiac unit at Edinburgh royal infirmary. It was second to none. I think it is worth while to record that, given the bad publicity that the national health service gets—very unfairly—on a day-to-day basis.

I will begin with a simple question. We know that health inequalities are one of the major reasons why we have such a substantial health challenge in Scotland, but the “Health and Social Care Delivery Plan” seems not to deal with health inequalities. Indeed, it barely mentions them. Why is that?

First, I am pleased that we were able to look after you, Mr Neil. I am glad to see you back—despite your question. [Laughter.]

The arrangements for the governance of the implementation of the delivery plan include a strand on population health improvement—I will ask Shirley Rogers to say a bit more about that—because we regard the improvement of the health of the population and the work that we do on prevention as being absolutely essential to the longer-term development of an engagement with the population that helps everyone to understand what they can do to develop and improve their own wellbeing as well as what the services, and not just the health service, can do.

I ask Shirley Rogers to say something about how we plan to integrate that with the overall delivery arrangements that are already in place.

It is very nice to see you looking so well, Mr Neil.

Thank you.

Public health and population health improvement is one of a very small number of key themes of the delivery plan, and you will see quite a lot of attention around primary care and all the initiatives around public health improvement and through that lens. We are also proposing in the plan to bring together a public health resource for Scotland to try to make sure that we have a cohesive and strategic approach to public health and population health.

We are taking forward a number of strands of work around alcohol, obesity and all the social inclusion areas around health. We have spent a lot of time looking at investment in health visitors, for example, to try to put that process in place. We also have the early years collaborative work, which I know you are familiar with and which is about giving children the best possible start.

The programme takes all those pieces of work into account and tries to bring them together into something that is comprehensive and can give us measured deliverables, and that is really where we will be targeting the firepower of the delivery plan over the next four or five years. That involves trying to make sure that we have comprehensive plans in place across the piece and proper programme management to make sure that they are delivering the sorts of outcomes that we need.

As you are well aware, some of those ambitions go way beyond the lifetime of any Government and into the next 20 or 30 years to try to improve population health in those ways. In this, the first of the delivery plans, we set out the principles that we will take forward with the view that, over the next few years, we will be able to learn, develop and put forward additional proposals that will improve the situation.

I agree with all of that and would not object to any of it, but there are matters that I do not see in the plan. For example, it is clear that the deep-end practices have a major role to play in reducing inequalities. We need to get away from the flat funding of general practitioners so that we positively discriminate in favour of those who need the additional resources to reduce inequalities in order to prevent some of the health problems that we face. However, there is nothing in the plan to tell me, for example, what additional resources deep-end practices will get, the anticipated impact of additional resources and when and where the practices will get them. That is the kind of information that I would be looking for in a delivery plan. The inverse care law has been with us since about 1971, but we have never really tackled it. It seems to me that, as the evidence mounts, we have to tackle it.

On a wider point, the Government is introducing a child poverty bill to set targets for 2034 to reduce child poverty. There is nothing in the plan about what contribution health and social care will make to achieving those targets. I would have thought that that kind of strategy would need to be in a delivery plan.

Just to be clear, the issue of public health improvement is covered in pages 16 to 20 of the plan, so it is not just mentioned in passing. As the committee will know, we are putting additional resource into primary care. Negotiations are going on at the moment on the contractual basis of the engagement with general practice, and we are building on our commitment to having multidisciplinary teams and a hub-based approach.

I take Mr Neil’s point about deep-end practices. As the delivery board, which Shirley Rogers chairs, develops the detail of its proposals, certain matters will inevitably come to the surface. However, additional funding is already in place. A £128 million transformation fund was set up, and I will bring Christine McLaughlin in soon to say a bit more about the components of that and the additional investment in primary care. I do not think that we are disagreeing, Mr Neil, about the importance of primary care and its centrality to the shift in the balance of care. That underpins a great deal of what we do and will help us to move more activity out of the acute sector and into primary and community settings, bearing in mind that we are speaking in the context of health and social care integration. If the committee would find it helpful, the finance director can say more about the detail of the funding.

I think that we are all familiar with the overall funding figures, but the issue in a delivery plan such as this is the distribution of the funding. If we just continue with the previous distribution of funding, the “worried well”, as Duncan McNeil used to call them, will continue to get massively unfair access to the health resource and social care resource compared to the access that the people who need it more and most get. Obviously, things such as dealing with the use of tobacco and—I hope—eventually implementing minimum unit pricing for alcohol will help with public health and help reduce inequalities. However, there is an issue about the NHS and social care resource, because I do not see a strategy in the plan for dealing with the inverse care law that is any different from how we have unsuccessfully dealt with it over the past 30 years.

I will bring in Christine McLaughlin shortly, but I point out the work that we are doing on the alcohol framework, on delivering the maternal and infant nutrition framework, on the mental health strategy and on supporting mental health in primary care. All that work is intended to tackle the clear position that, if someone lives in a more deprived area, they will find services harder to access and will be more likely to be unwell. We are not disputing those facts at all. The components of what we are doing in the plan are central to tackling that issue, and the investment that we are making in physical activity and the work that we are doing on building success in schools among other things are intended to tackle that issue.

I take your point about the importance of specificity in what we do about things such as deep-end practices. Perhaps Christine McLaughlin can say a little more on that.

The funding is implicit in everything that we are trying to do, rather than being a ring-fenced component for tackling inequalities. In the spend on primary care, there is an overall metric about shifting £250 million into primary care and shifting £500 million overall from acute to community services. As you know, the deep-end practices are a very specific area and our approach is more about trying to understand how we can enable things like GP clusters to have the impact that they need to have in the parts of the country where they need to have it. We are also using data on variation to understand why there is a higher prevalence of certain conditions in particular parts of the country and how we can deal with those.

It is about having a whole-system data-driven approach. We are trying to approach the funding by being more flexible and less mandatory at the centre about how that money is best applied locally. The finance is there to go where it needs to be directed to get the most benefit.

I do not dispute any of that—all of it is right. However, when you read the document, it looks like a mixture of a statement of intent and a wish list, rather than a delivery plan. The English business plan for the national health service, which is also done on a national basis, is much more detailed in that it gives likely impact assessments and sets out where resources will be distributed and how they will be delivered.

Let me give you an example from early on in the plan. It says on page 8—I do not disagree with this aim—that

“By 2018, we aim to: Reduce unscheduled bed-days in hospital care by up to 10 percent ... by reducing delayed discharges, avoidable admissions and inappropriately long stays in hospital.”

We have been doing that for years, but how are you going to do it so that you achieve the target? The plan tells me what the target is and what you want to do, but it does not tell me what you will do differently in order to achieve that target. The plan is riddled with that problem. It is not a delivery plan; rather, it is a statement of intent and a wish list, with which I would not disagree. It does not tell me how it will be delivered, what the impact assessment of the various measures is or what the resource allocation is. I can see those things in the English plan, but I cannot see them in the Scottish plan. I am happy to park that and give you some time to think about it.

The best thing that I can do is to give you three pieces of assurance and then offer the committee something in writing if that would be helpful.

First, the plan relates to the budget that we have—the sum of money given to health and the money available in other budgets, including local government—and is therefore a funded, rather than an unfunded plan. Secondly, the level of detail that Mr Neil is asking for is precisely what we are working through with the delivery board. That is why we have a delivery board. It is not simply aspirational. We are confident that the sums of money that are available are sufficient to deliver what we have been asked to do. Thirdly, Mr Neil spoke about risks to delivery. Of course there are risks to delivery—things change over time—but we have programme management and governance arrangements in place to deal with that. That is why we have governance and oversight.

If the committee would find it helpful, I would be more than happy to provide a more detailed breakdown of the components of the plan, how it will be delivered and at what cost.

As an example, in a delivery plan, I would expect to see figures relating to where, when and why money is being spent and what you are doing differently that will make a difference. I would also expect to see some estimate of the impact that that will have on health inequalities and achieving the targets and why that will be different to what has been done before. We need to wait and see what Harry Burns’s review of targets says, but the media concentrate all the time on when we miss our targets, even though we are doing excessively better than the health service down south. You do not get credit for that; you only get abuse because of missing the targets.

The question that we have to ask is: how will you achieve the targets for accident and emergency and all the rest of it in future? The treatment time guarantee is a good example. The plan does not tell me how, between now and the end of the plan period, you will achieve the treatment time guarantee. That is what I mean. You have to bring all this together so that somebody who comes new to the matter independently and neutrally can see the total picture. I do not see it as a plan; I see it as a worthwhile document, but in essence it is a statement of intent and a wish list, not a plan as I would understand one.

10:30  

Do you have a question?

No—I think that Paul Gray has offered to come back to us.

I am happy to go further on the matter, convener, or to take other questions—I will be guided by you—but I have committed to writing to the committee with more detail, as Mr Neil has asked, and I am happy to do that.

If you could, please. Do you wish to comment on Mr Neil’s last point, or do you just wish to write to us?

What we have is a plan for delivery. We have given specific dates by which we intend to do things, and I have committed to writing in more detail about the underpinning finances and where they will go. I believe that it is a robust plan, and it is underpinned by data, evidence and clinical advice. We are happy to share all of that with the committee if it would find that helpful. I would describe the plan as a firm proposition, not as Mr Neil has described it.

One area that really concerns me is integration joint boards’ integration of care at very local level. The Auditor General’s report highlights the poor progress on that. It says:

“Evidence suggests that the new partnerships with statutory responsibilities to coordinate integrated health and social care services, integration authorities, will not be in a position to make a major impact during 2016/17.”

That is a bit of a worry.

Paragraph 30 of the report says that

“A common issue with many of the new care models being introduced across Scotland is a lack of evidence about the impact, implementation costs, efficiency gains or cash savings, and outcomes for service users.”

That seems to be a big grey area, but it is such an important area, and you highlight it in your delivery plan as being fairly essential. Can you give me a feel for what is happening in that regard?

The “Changing models of health and social care” report was prepared by Audit Scotland in March 2016, before the first full year of operation of integration, which we are now in. A number of developments have taken place in the course of the first full year of integration that give me confidence that we are heading in the right direction.

Not every partnership is at the same point on the trajectory—there is a degree of unevenness. However, I have 10 or 15 examples here—I promise I will not go through them all—of areas where integration is working well, which can give us confidence. I will share those examples with the committee in the written response that I have already agreed to give to Mr Neil.

I will give the committee two examples now, however. I met representatives of Ayrshire and Arran NHS Board and the integration partnerships a week past Friday, and I saw how that health board is managing patients with chronic obstructive pulmonary disease heart failure and diabetes much more effectively, which has resulted in a 49 per cent reduction in emergency admissions and a 36 per cent reduction in bed days in the first six months, for that case load. I can provide more detail on that.

The Dumfries and Galloway integrated hub has developed a single point of access for community health and social care services, which is bringing together health and social care services called STARS—short-term augmented response services—and the third sector for people who are registered with two general practices in Dumfries. I can also provide more detail on that. It has reduced the referral time from GP to care input by up to 15 days, there are quicker interservice handovers and there is increased capacity. It has also improved communication, staff morale and team working. There is evidence that those measures are working.

Has there been an effective transfer of budgets to make that happen? That seemed to be the key thing that was delaying integration—everybody had their own little fiefdom, which they would not give up. How effective has budget transfer been and how has it been achieved?

The first year of any operational delivery is clunky, to use a non-technical word, and we learned a lot in the first year on budget setting. There was a great deal of to-ing and fro-ing, and a great deal of debate, discussion and negotiation about budget transfers. I will not pretend otherwise. We have been through that process and the evidence suggests that it will be much smoother this year. I will bring Christine McLaughlin in on that, in a second. We are learning from what we have done—as people always do—and improving as we go along, which is the right process. I am not claiming that such a transformational change as the integration of health and social care will go from nothing to perfection in two years—nothing of that scale will, or would. However, I have clear evidence of progress in local areas.

Issues of governance, leadership and trust have been worked through because those were essential precursors to effectiveness. Delivery on the ground is now making a difference. If the committee would find it helpful, the finance director will say a bit about the budget process.

I would be interested to hear that.

Paul Gray is right: for the first year, there was due diligence in advance and, through the year, people understood what their resources were able to buy and where there were levers or pressures.

As partnerships plan for the 2017-18 financial year, there is a much greater level of engagement all round between integration authorities, NHS boards and local authorities. That does not mean that everybody agrees all the time, and there are still issues in some parts of the country about signing off on the resources to deliver the plans that are in place, but there is much more sharing of information, and integration authorities and chief officers are part of the discussions about how resources are used on a more day-to-day basis.

We are in some places starting to get into shared risk and shared ownership in relation to delayed discharge performance and how we share investment in new services that we are looking to plan, for example. There is definitely evidence of more mature relationships in respect of budgets. It will not be until we get into the bigger issue of how that impacts on the acute sector, for example, that we will see some of the traction and some of the metrics. However, on delayed discharges, we are starting to see good work and progress in respect of the number of people whose discharge is delayed.

On page 9 of the delivery plan, under “Shifting resources to the community”, you say that by 2021

“spending on primary care increases to 11 percent of the frontline NHS Scotland budget.”

Is that an arbitrary figure? What is it based on?

The figure is based on our assessment of the correct relative share to get the shift that the evidence suggests we need, so it is not an arbitrary figure. It is also part of a negotiation between the Government and the representative bodies, as all those things are the subject of negotiation. There is clear evidence that putting extra capacity into out-of-hospital care provides an opportunity to transform services and deliver better outcomes.

Good morning, panellists. Paul Gray talked about governance, leadership and trust. We are seeing a massive change in the NHS, with a shift from acute to community care. As we know, and as we have heard in previous evidence sessions, members of the public are attached to buildings—they see hospitals as the main places to provide care. In a lot of cases, they are suspicious and slightly wary of change when it happens on such a big scale. What is the Scottish Government doing to engage with the public to raise awareness of the policy to shift the balance of care? What discussions, consultations and public engagement are taking place on why services need to change?

I use the term “shifting the balance of care” as shorthand, but I suppose that the public would not engage particularly well with that phrase. Concrete examples of what would be different and better for them are helpful to the public. For example, the rapid elderly assessment care team in West Lothian visits people in their own homes and provides multidisciplinary care. People learn about and come to appreciate that service by experiencing it. That is hugely important. Abstract statements of what might happen are hard for people to grasp.

We are working with the health boards and integration boards to ensure that there is the appropriate level of engagement, as defined by the Scottish health council. The great benefit of the integration boards is, of course, that they engage local authorities’ elected members. I believe—I say “I believe” because I have seen it—that local authorities have well-developed approaches to engagement with the public, which we benefit from and learn from. Through the processes that have been established, we are ensuring that there is appropriate engagement. The issue is to make that engagement specific and—I was going to say “concrete”, but that might bring us back to buildings—real, rather than making generalised propositions about “shifting the balance of care”.

We should not start the conversation in the wrong place. We should not say, “We’ve got this building and we’re going to knock it down” because a sentence like that always gets people nervous—and rightly so. Instead, we should start the conversation by saying, “Here are the developments in healthcare that would allow us to provide you with a better service. Here is how that could be delivered. Here is how you can be engaged in deciding and agreeing what your care should be.” That would result in be a much better conversation.

Engagement does not just come through the forum of health boards or integration partnerships directly. For example, the conversations that are going on in the context of the national clinical strategy, particularly on realistic medicine, are between clinicians and individuals about what care would be best for them. Most often, when people are engaged in decisions about their care, the outcomes are better for them. Most people prefer not to be overtreated.

The chief nursing officer could say more about that, if that would be helpful to the committee.

As Paul Gray and Gail Ross have said, people are anxious about what happens to buildings. The relationship between clinicians and the public is incredibly important. We know that the public trust clinicians’ decision making. We need to help people to understand that they can have better outcomes at home or in more local settings from treatment for which they would in the past have needed to go to hospital. I suspect the care that Mr Neil received is a slightly more acute version of that situation, with a rapid transfer to a hospital—and perhaps not to the nearest hospital—and a much shorter length of stay there before going back home. The community always appreciates good outcomes when its experiences them. We will have that dialogue with people through the whole strand of realistic medicine and in progressing the national clinical strategy.

We are all involved, including local authorities and NHS boards, in reassuring members of the public that they will receive the best possible care. What is the best way for NHS boards to engage with the public?

10:45  

That is a big question. The short answer is that it depends what the issue is. I can give examples of methods that are not good ways of engaging with the public. One is letters, because people cannot ask a letter a question. Engagement with the public must involve people engaging with people. If I was to give one fundamental answer to your question, that would be it.

Giving people something to see works: if you are describing a facility, you should, as early as possible, give folk an idea of what it will look like and where it will be located. If you are describing a new care pathway—in other words, a new way of treating people—you should get someone to talk through it and perhaps use a case study with a patient and a clinician having a conversation, rather than produce a dry document that says, “Here’s the new care pathway” for a particular condition, which in a sense is asking people to work out what the change would mean for them. I much prefer to give people something to look at or to hear and—above all—the opportunity to ask questions. We should not be afraid of being challenged. The public have some really good ideas about what would be best for them.

When massive change is taking place in a big organisation, the best advocates for that change are often the staff themselves. Are your staff fully informed about and on board with the changes? Are they able to talk to the general public about what those changes are and what they will mean?

I will bring in Shirley Rogers and Fiona McQueen on that point, because it is hugely important. I do not think that we always get that element right. We have partnership arrangements that work well and are well regarded, but if we are talking about investing in talking to people, we need to treat our staff and the public well. That is essential.

The best part of my job is going out and talking to people, staff and patients, because that is where I learn most. I believe that the leaders in health and care systems ought to devote a significant proportion of their time to speaking to the people who deliver the services and the people who receive them. That is how they will learn whether the system is working. Shirley and Fiona will both have things to say about that.

I will pick up a couple of themes and address the specific point that was made. To go back to Alex Neil’s observations about what is different, the planning and delivery processes that are signalled in the plan give huge—and, I think, very welcome—prominence to national, regional, local, community and individual aspects. If we lose sight of those aspects, we will miss a trick. The prioritisation that Alex Neil was talking about, whether that relates to general practices or to particular needs, is addressed specifically by our not determining things at national level and instead allowing local communities and IJBs to prioritise actions.

That approach is important for a number of reasons—not least in relation to Gail Ross’s point about communication. It is important because it means not only that communities get what they need but that communities themselves are advocating for the changes that they want. That is really important.

We are creating a national narrative, but we are also making sure that we are taking the approach that I have set out. If I think about what I have been doing in the past three months since we published the plan—

If you could make your response a bit tighter, that would be great. Thank you.

I will do my best.

We have spent a lot of time talking about health to all the communities. We have talked not only to chief executives and chairs, but to staff—nurses, human resources people, organisational development people and all the rest—so that they can do precisely what Gail Ross described. The focus of this year’s NHS event is on people making change happen. We will have not just 10 or 12 people but 156,000 people talking to their communities about the changes that are proposed.

My question follows on from Alex Neil’s questions on the recommendations in the Audit Scotland report on financial flexibility for NHS boards. I appreciate the answer that you gave about that aspiration, but I want to drill down into the detail of what steps have been taken to follow up on those recommendations and what financial plans are being prepared.

There is, perhaps, a narrow point to be made about how boards achieve financial balance in individual years as well as a wider point about generally living within your means and having sustainability.

On the narrower point about flexibility, I hope that our position is quite clear: we provide services that give boards a level of flexibility from one year to another. That is, by and large, something that we can plan for; with, say, a major new project such as a new hospital, we can plan for that over a number of years. You will know that a number of boards are facing financial sustainability issues and, as you will recall from our discussions on NHS Tayside, we provide financial assistance that we look to have repaid.

Just now, we are working with boards on their five-year plans, which will underpin the delivery plan; at the end of the month, boards will be giving us their local delivery plans, which will focus in detail on 2017-18. We will therefore have plans for individual boards and, as part of the delivery plan, we are now asking them to start to work on a regional basis. I expect to move to a position where we would look to provide regional resource envelopes, which would give another level of flexibility—a regional level, if you like, sitting one layer down—in the deployment of resources. If money were to become a barrier to moving or sharing a service, we would want to take such barriers away.

The portfolio—and, indeed, the Scottish Government—needs to be able to balance overall, but within those parameters we are doing everything that we can to provide flexibility. In return, we ask boards to look far enough ahead to understand the longer-term position, to plan as much as they can for the knowns and the unknowns and to be really clear in the risk assessment that they carry out.

That leads nicely to my next question. What work are you doing with boards to drill down at a more local level and ensure that there is sufficient resource and investment in the community to meet the objective of shifting the balance of care from hospital to the community?

As part of next financial year’s funding settlement, we have put in place some very clear criteria, including our expectation that the total sum of money being spent on non-acute areas will be maintained. That might be only one sentence in a letter, but it marks a real shift in clarifying the direction that we want to go in with the resources that are spent in the NHS.

In that respect, we have made it very clear that we expect spend on integration authorities, which amounts to about £8.2 billion, to be maintained, and spend on primary care and mental health to be maintained, too, as a minimum. The additional investment that we are putting in, such as the £70 million for primary care and the additional money for mental health, should be over and above that core recurring funding, and we have asked for assurance from boards that that is being delivered. It is quite a crude measure, but it means that we expect best use of the resources that are being maintained in that area. We will see in the local delivery plans that we receive at the end of the month the boards’ assurance and their descriptions of how they are doing that work.

On how to ensure that funding levels are maintained—this touches on a question that Colin Beattie asked—I note that the delivery plan states that you will

“Ensure Health and Social Care Partnerships increase spending on primary care services, so that spending on primary care increases to 11 percent of the frontline NHS Scotland budget.”

Given that, according to a previous answer, that has been subject to negotiation, and given your comment that you are looking for assurances, what can you do to ensure that that funding level is met? Where are the details and the plan to back all of that up and to allow us to monitor whether it is being achieved?

I will answer very briefly, and I would like to ask the convener a question, if I may. When I said that it is subject to negotiation, I was not suggesting that the 11 per cent is subject to negotiation.

Convener, I know that you have other business today. Can you give me guidance on how long you would like to allow for this session? We will then try to tailor our answers to an appropriate length.

We have time, Mr Gray. I am not as concerned as I was a few minutes ago, so please say what you need to say.

Thank you very much. I will hand back to Christine McLaughlin, but first I want to make the point that I am not suggesting that we will negotiate the 11 per cent in a different direction, but what happens to it must necessarily be the subject of negotiation.

We have written back to the Health and Sport Committee in some detail to give clarity on that point. There are a number of measures. There is the 11 per cent, but I think that the bigger point is the shift of the £500 million by the end of this parliamentary session. The modelling that we have done suggests that that should be a realistic projection over that timeframe. In fact, if that level of sum cannot be achieved, there will be no shift in services on the ground.

The assurance comes not so much from the money and the overall conditions that we set, but from the ability to see a service that was provided in an acute hospital being provided in a different way in the community.

We now need to build up the detail of those individual changes on the ground to understand how they all come together. I can give you that detail in our assumptions. By the end of this parliamentary session, we should have more than £14 billion in healthcare. The additional cash investment needs to get directed to the areas that we need it to get directed to. We have a level of planning assumptions about health funding in cash terms over that period, which allows us to have a bit more certainty about how, if the position continues, funding will be directed. On that basis, getting to that £500 million shift does not seem unreasonable to me.

Mr Gray, based on the evidence, what is better when it comes to social care and the public purse—private providers or carers who are directly employed by a local authority?

Generally speaking, continuity of care is important, so bringing in providers from time to time is much less likely to lead to a good outcome for people. It is not that there cannot be a good outcome, but you would want to be sure that you were using an approach that provided continuity of care.

The same quality standards would be expected, were we occasionally to use the private sector—there would be no difference. I believe that having care provided by the national health service, which I speak for specifically, is the right thing to do, because, as the leader of the national health service, I am able to put in place a set of standards, a set of behaviours and an expectation of integrity, and I ensure that those things are followed by having in place systems and processes such as iMatter and engagement with partnerships. Therefore, I believe that that approach is most likely to yield good care.

You are talking about having carers who are directly employed by the IJBs.

I am talking about carers who are directly employed by the NHS. Similarly, local authorities have directly employed staff who are subject to local government standards and guidelines. Again, I focus my answer on continuity of care and the ethical standards of the public sector.

My next question might be more for the Auditor General. We know that it costs the health service three times as much to employ an agency nurse as a contracted nurse—we have discussed that before. Do you know what the cost difference is between social care workers who go into people’s homes and are employed directly by the NHS or the IJB, and those who are employed by private companies such as Avenue and other care providers?

I do not have that level of detail. The information could easily be got for you, but the problem is that it will vary from area to area. I am perfectly happy to provide it, to the extent that it is available.

Could you, please? That would be useful. I will make the same request of the Auditor General.

Yes, I can do that.

11:00  

Page 18 of the delivery plan states the intention to

“Consult on a new strategy on diet and obesity.”

Am I right in thinking that an obesity strategy is due out from the Government any day now? Does the delivery plan refer to the same thing?

As far as I am aware, yes. Our approach is to have a consultation on an area that we believe needs to be given further impetus. That is why we are doing it.

If the obesity plan is due out any day, why consult on a new strategy?

We are putting out a plan for consultation.

I see. That clears that up.

Paragraph 36 on page 21, which is about getting it right for every child, states:

“In addition to actions included in the main components of work above, we will drive this agenda through: continued implementation of Children and Young People (Scotland) Act 2014, in particular, the Named Person and the Child’s Plan”.

My reading of that is that the implementation of the named person and the child’s plan is continuing. Has that changed since John Swinney’s statement a couple of weeks ago?

I am genuinely not entirely sure what question you are putting to me.

On page 21, paragraph 36, which is on GIRFEC, states quite clearly that the actions recommended by the plan include continued implementation of the named person scheme. Is it continuing to be implemented, or has its implementation halted since the Supreme Court ruling?

The civil service will proceed in line with what the Deputy First Minister has said. That would be my answer to your question.

My interpretation of what he said was that there was to be a pause. I take it that your civil servants will be pausing the implementation of the scheme. Is that right?

The delivery plan was published in December 2016, when the situation was as it was then. If ministers subsequently introduce a pause, we will operate in line with what the Government and the Parliament agree.

Do you know that civil servants and NHS boards have paused implementation of the named person scheme?

I know that we react to what ministers say exactly as they would expect us to. The CNO can confirm that.

As part of the delivery plan, there is a commitment to 500 additional health visitors. Part of the work that we did on supporting young people through early intervention to reduce inequalities involved introducing what we call a universal health visiting pathway, to standardise how our health visitors support families from pre-birth to school age. Some of that can, at times, be conflated with the named person scheme. The named person has duties and responsibilities, which have been paused and are being reviewed, but we are continuing with getting it right for every child and with implementing the universal health visiting pathway, which means additional home visits for families by a qualified health visitor.

We are continuing with that element, which would support the implementation of the named person scheme in terms of data sharing and responsibility. As Paul Gray said, one element of the process has been paused, but the benefits of other elements, such as having additional health visitors to support families in a more comprehensive way, are continuing. However, we are cognisant of the legislative issues.

I would like to ask a supplementary to that, because it is now three years since we decided to increase the number of health visitors by 500. How many of those 500 have now been appointed?

I can send you the specific numbers. We have trained more than 398, and on top of that we will be training an additional 163 health visitors in 2016-17. We are moving to be on target for implementation of the target for the number of health visitors in training, if not qualified. We are making good, steady progress.

When do you reckon we will have 500 additional health visitors actually working?

That is a moveable feast. I can give more accurate projections and trajectories in writing.

I thank everyone for their evidence this morning.

11:05 Meeting continued in private until 11:18.