Don Berwick: improving the safety of patients in England

I was asked by the Prime Minister to produce a set of recommendations for implementation of changes in the National Health Service of England in response to the Francis Report and its sequelae. I chose to do that through a committee, an advisory group. I nominated and we selected about 15 members...15 or 16, I can’t remember the number exactly but it was a wonderful group. Everyone asked agreed. It consisted of scientists, people interested in organisations, organisation theories, safety and quality. Some managers and clinicians from the NHS, although no-one currently in a senior executive role within the National Health Service at the centre, and most importantly patient representatives on the group who were enormously valuable.

We appointed a series of special advisors who we could call on for guidance but who, because of their current roles, would not have been appropriate for membership in the committee. We met virtually mostly about every fortnight for a couple of hours by video. There were seven working groups established and they met in parallel so we had a lot, these people put in a lot of time and I’m very much in their debt. It was an amazing group to lead.

We are indebted to Robert Francis. His report was the initial reason for our being called into existence. Robert himself was enormously generous. He was a special advisor to the group on many occasions. When we needed clarification or guidance or kind of needed to know what went on inside his brain he generously met with us, as did others.

I want to say a special word about Bruce Keogh. Sir Bruce Keogh, you know, produced the Keogh Report on 14 high mortality trusts that came out while we were doing the work that was a completely and separate and independent enterprise. To my knowledge there was no interaction between our group and Bruce, and it is eerie to read Bruce’s report because we really did converge, and I think in terms of confidence about the directionality of recommendations we have two independent findings which point in pretty much the same direction. My hat is off to Bruce for a stunning piece of work.

We began by outlining problems. We don’t want to stay there, this is about the future not the past, but we wanted to say what we thought we were seeing as we examined the mid-Staffordshire story and other stories that emerged during the period of our deliberations. Here are the seven findings that began kind of the problem list for us.

The first is that we just don’t believe mid-Staffs is an isolated case. It is a symptomatic case, a notorious case, but of some problems in the NHS where there are safety issues. There are some outliers like mid-Staffs that have apparently high mortality rates, and Sir Bruce has done a pretty good job beginning to investigate some of those cases. More generally, like any large complex system, the safety and quality is a challenge in the NHS as a whole so we’re dealing not just with mid-Staffs, we’re dealing with the whole system.

The second, there just isn’t anybody in the committee that believes that blame is the right approach with respect to staff. You have 1.4 million people working in the NHS in England. We think they’re in general just as dedicated as any of you would be to making it a great service, to fulfilling, giving, meeting in their own lives, helping relieve suffering, and to start from a platform that somehow everything’s gone wrong and the staff don’t care is just not something we sign on to. We’re working with a different theory. We sense incorrect priorities that did develop. Certainly if you study them in the Staffs case, there was a diversion of leadership attention towards two things really – quantitative targets, a kind of game and a tick the box approach, and second finance and financial diligence – and the more general agenda of making sure the patients get the correct care, the best possible care, the patient at the centre of care, that kind of got lost in the attention to details. They were dealing with trees, not the forest, and we... we scored that as priorities that had gone amiss or had been interpreted to go amiss. Often it was a leader who had articulated something and the staff picked it up as a game, and when they did that it was not what the leader intended.

That things were going wrong could have been known far earlier than it was known in mid-Staffs. There were signals both quantitative, for example the mortality rate metrics which were way off scale. This was not a marginal case of, you know, fine tuning mortality measure, and this had to do with three or four standard deviations of unexpected mortality rates, they were known. And if the quantitative data wasn’t enough there was qualitative data. Staff, patients, carers were speaking of problems and they were not heeded. There was something about what was going on in that setting and which really important information was not being acted on. It was either being explained away or lost entirely.

With respect to outside regulators, one of... one of the phenomenologies we think we observed, and that Robert commented on strongly in his report, is a complexity of the regulatory scheme in NHS England. A lot of agencies responsible for different components of experience and quality and if you add in the responsibility for financial stewardship it even gets more complex. From the field you hear this is tough to deal with. Lots of metrics coming at us, lots of demands, lots of inspections, and regulation instead of becoming an absolutely focused, clear, respectful, simple enterprise has become far too complex and it’s very hard to give care, to manage care, under that kind of environment.

There was no visible support for what we call a system of improvement. We’ll deal with that more in a few minutes, but it means a system by which all staff – clinicians, managers, front line staff – can learn how to improve, they can understand how to improve their own work as part of their work, and be a social system which allows people to learn from each other both within organisations and among organisations. Mid-Staffs was not part, so far as I know, of any collaborative improvement effort that would have kept them in a community of shared effort and there was no centralised resource for that. And from all of that emerges an atmosphere of fear. You can almost feel it as you read the Francis report. As people became frightened of the very information that they could have used as a foundation for their improvement. If you take this list of problems then you can juxtapose solutions. Let me show you the general framework we began to work from as we arrived at our more specific recommendations.

The first is you have to recognise the problem. All improvement begins with aims and a general thought, I think we can speak for the group, is we have to be able to say that things are wrong in order to set them right, and that general idea – that transparency and openness about defect – is absolutely key, but not in a blaming culture. Certainly there are miscreants, and we’ll deal with that in a few minutes, but very, very few. The idea that somehow you can say things are wrong and you’re doing it, or you’re doing it, or this party did it, or that agency did it, is just not mature, it’s not an appropriate reaction to the problems of improvement. We assert the primacy of working with patients and carers.

If mid-Staffs had listened to the patients, listened to the carers or the staff, if someone had said ‘what are they telling us? Put it together’ they would have been on alert, they would have been worried, they would have begun to have some diagnostic work done. Instead the voice of the patient, the voice of the carer, the voice of the staff was muted and eventually more or less ignored.

Quantitative targets were not ignored, they became part of the game - waiting times, particular benchmarks to hit. And it was a ‘tick the box’ mentality that was not favourable to the focus on patients. All of this quelled transparency because as the data became troublesome or inscrutable or worrisome as you don’t hit your target, as people do speak up and you don’t have a way to remedy it, you can’t work on improvement, well you hide, you run away from the data, and that’s what you observed. In a proper improvement culture I think our advisory group is unanimous. Transparency is essential. You have to be able to turn the lights on in order to improve. And with respect to the outside regulatory system simplicity, clarity and channels of responsibility that are easy to understand are crucial to being able to direct the system in response to outside regulation correctly.

If you don’t build skills you don’t get skills. It’s been a long day for me so to wake myself up I thought I’d play a little game with you. Here’s a game, OK? I’m going to say some numbers and you say them back to me. Have you ever played this game? So if I say ’3,6’ you will say (3,6)... OK so now I’m going to give you some numbers and you repeat them back to me. Here they come – 3, 3, 2, 5, 3, 2, 7. (3,3,2,5,3,2,7). Perfect... almost perfect. Are you ready for another one? 6, 0, 3, 7, 2, 6, 5, 1, 2, 3, 4 (laughter). What happened? What happened was we just exceeded the capacity of short term memory. It can store seven items, so you succeed at seven, you fail at nine. If we designed a task that has... in which a worker has to remember has to remember seven items they will likely succeed. If we design a task in which they have to remember nine items they will almost certainly fail. It’s a system’s view of improvement. Now how do I know that? I know it because I studied it. I’ve studied safety, I understand how frail short term memory is as a guide to correct work. Well that means all of us – the managers, the executives, the workforce – they have to know that, they have to understand the relationship between the job design and execution. Blame will help not at all. Design helps, and design depends on knowledge. This is a very small example of what it would be to be an NHS constantly able to use science for proper design and improvement. You get it? We could go on all day long with games like that, but I won’t bother you.

We speak in the report a lot about pride and joy in work. That’s not something that plays well in a lot of crowds. I think it’s essential. You’ve got an enterprise trying to make health care a human right, you have the potential to be a system as excellent as any in the world. People should feel wonderful about being able to work in the NHS. They’re committing their lives to it, their time, their waking hours, their spirit, and there’s no reason why this system can’t aspire to pride in that service. You saw it at the Olympics, you saw a whole world watching you celebrate, you know, a national treasure and now we’re krechting about that treasure. That’s not right, it’s still the same place that you celebrated at the Olympics. Remember it, and tell 1.4 million people that’s true.

Nonetheless, the challenges are big. You’re going to have to grapple with culture. You always do, this isn’t special to the NHS, you’re going to have to think about how people behave, what they think, how they deal with each other, the elements of culture are going to matter. We will talk about regulation, we’ll talk about rules, we’ll talk about measurement, we’ll even talk about punishment, but that’s not the trick. The trick is to develop a culture, and the culture we’re after is a learning culture, and I’m going to try to explain what that looks like. We’re going to use the word ‘quality’ a lot. I’m just harking back to Lord Darcy’s definition of quality in his report. It’s just fine. It’s a combination of properties of the system. We’re after safety effectiveness and total experience of the patient. I want to point out that although mid-Staffs focuses on safety and it gets focused on safety, quality is not divisible, and by working on safety you must open the door to total excellence and keep that mind. Our work is not just about safety, it’s about excellence overall.

So our reports are organised on these categories and I’m going to walk you through the recommendations and then dive in a little more deeply. Each of these categories had a working party more or less associated with it and I commend my colleagues for the hard work they put in in those working parties. OK, the over-arching goal...this is just the unifying goal for the whole system. We are recommending that continual and neverending improvement of the wellbeing of the patient and the particular case here, reduction of harm. The Prime Minister, I think quite dramatically and courageously, set out zero harm, and I’m glad he did, that’s a start, it’s a guiding start, it’s a north star, zero harm. The actual agenda day-to-day is continual reduction of harm, that’s the way to think about it. That should be intended, monitored and embraced throughout the system. The reflection in the leadership arena is that this then becomes the job of the leader. If the leader doesn’t embrace this in the daily leadership activity focusing on quality and patient safety, at the top of the priorities list look at that list – investment, enquiry, daily enquiry by the leader, improvement, regular reporting, encouragement support – don’t expect the staff to invest. They’re going to follow the signals of the leader no matter how much they care about their work.

The third area of recommendation is about patients and carers and here we are talking about complete empowerment of patients and carers way beyond tokenism, way beyond occasional representation of focus groups. We’re talking about the presence and involvement of patients and carers at all levels in the system at all times.

The fourth set of recommendations are around staff. This has turned out to be the lightening rod recommendation in some ways today. One of them we believe, as any normal caring person would be, there should be sufficient staff now and in the future, and it is a job of the entire system to assure that. The question will arise as to whether we’re going to set numerical targets for that, but the answer is ‘no we’re not, not at the moment’ and I’ll deal with that in a few minutes.

In terms of building staff capacity I would put this at the top of the list of recommendations. It is to have a service that invests, I think, from the centre as well as throughout in the constant improvement of the capabilities of the workforce to invest in their own learning and improvement, and that is at two levels – the individual learning more and more how to improve their own work, and to do that in a collaborative environment in which people can share the results of their improving efforts, so all teach, all learn, there’s learning going on everywhere, and that’s what we call a learning organisation, and the NHS invested in the processes of learning, not just within, but among organisations. An asset in that is transparency. You remember in midStaffordshire account transparency was a threat – it produced fear. If the numbers appeared I might be in trouble, I would be judged a failure, something bad will happen to me or I will feel helpless and can do nothing about it. Well that’s not what we want, but we don't want to get out of fear by avoiding transparency. The the idea of total commitment to revealing and sharing knowledge about what’s going on is a very important recommendation for our committee, and in that measurement process - the transparency process - knowing what the patients and carers are saying and thinking is crucial and we just can’t say strongly enough the importance of listening to the patient and care voice as an essential asset in monitoring the state of the system.

What about the hard edge? Not just the learning edge but the supervision and enforcement edge? Well, we understand the need for regulation, and in particular we think that there needs to be an apparatus in the country that looks at performance in the NHS. We think the current apparatus is too complex. We hear that in the Francis report, we’ve seen it in our own investigations. There a diffusion of responsibility because too many agencies are responsible for different parts of excellence and if you add in responsibility for financial stewardship it gets even harder, and so we need something like simplification and clarification of the regulatory regime.

With respect to the hard edge of regulation enforcement piece, which frankly Robert Francis was strong on - stronger than I think we feel - we recognise the need for some strong regulation, but we’re... we’re taking a hierarchal approach to regulation which the final level of regulation, the introduction of criminal sanctions or serious sanctions is the port of last call. It’s the last place to go. What we need is a tiered system of regulation and oversight which is much more responsive to actually what’s going on on the ground, and maybe other committee members will want to speak up before I’m finished.

Let me go a little more deeply into the recommendations as outlined. The over-arching goal, I say to you, is that the system becomes oriented around continual improvement as its central aim. All leaders are invested in quality of care and safety as a priority, and by leaders we mean all leaders from the Prime Minister’s office right down to the head of the ward. This is a comprehensive investment by leaders in the patient at the centre of care. For patients in public involvement we’re arguing for a rather bold level of engagement of patients in the processes of design, regulation and scrutiny of the system, not just activation of patients in the individual care/patient relationship. We’re really talking about quite a revolutionary role for patients much more centrally in the design and conduct of the system. For the staff’s efficiency piece we are recommending that NICE – the National Institute – be chartered, as I believe it already has been now, to come up for a formula, an algorithm that could be used by managers of the system to adjust in real time to the needs of staff at the sharp end. Several members of the committee would have preferred that we come out with a particular staffing ratio. As a whole we backed off about that a bit. What we’re saying is if you run a hospital you should know what the science is saying about appropriate staffing ratios, you should adapt that to your local context and hold yourself accountable for that, your Board should hold you accountable for that, it should be transparent, but at the moment we’re deferring to NICE to come up with the algorithms that would help you know what would be appropriate locally in terms of response to patient acuity, demographics and particular stresses.

The training capacity building issue is key and we are really urging a very serious step forward for the NHS, even more than now into the mastery of quality in patient safety sciences as part of the preparation, and ongoing education of really everyone in the service, and we very strongly wish that they would emerge as an ambitious programme of collaborative learning among organisations in the NHS. It has worked in this service in the past, it will work in the future. You know, the press have been picking up a lot of what I’m telling you today and I kind of wish this was their headline – it’s not, but this is probably the most important actionable recommendation. We’re saying a system of collaborative learning and improvement will pay off handsomely for you as it has in the past.

We’ve put in the report one model for the improvement skill piece, what it is people should be equipped to understand and know. This comes from Kaiser Perminente, it’s adapted to the NHS, and you can’t read this of course, but in the report if you want to see more about the skill building piece that we’re talking about, it is there. Our position on transparency is unequivocal and complete, which is we’re arguing for a norm in the NHS that all non-personal data and quality and safety, whether that data are assembled by Government, by organisations, by professional societies, by anyone, they should be shared in a timely fashion with anyone who wants it, including an accessible form, the public. Again this would be a world setting leadership act, because I don’t know of a system that could be at a transparency at the level that our group is urging...and among the things to be transparent about is what patients and carers are saying about their care.

For structures we have a bit of a tightrope here. I must say in our committee, in our group we probably, if we had complete. If we just felt we could do anything at all or urge the NHS to do anything at all we would simplify the regulatory scheme now. It’s just too complicated, there are too many players. And this, possibly even the separation of financial from clinical regulation, is not particulalry wise, and neither is the separation of enforcement from assessment, but there’s... right now you’ve been through a lot of restructuring and we’re kind of saying ‘well OK, take a deep breath, you know, another restructuring may not be useful right now, but you better keep your eye on this one because if this doesn't get simpler, it’s going to be very hard to have regulation be part of the overall improvement agenda’. Indeed we recommend and review, a prompt review, of the degree to which co-operation is emerging among the regulators which is key.

With respect to enforcement we did introduce one new recommendation for a criminal sanction. We did this, I would say, hesitantly, but we feel it’s appropriate, that there are occasionally very, very rare instances - and I want to emphasise how rare we’re talking - about in which there is completely negligent and neglectful, wilful, reckless behaviour in those cases. As in any industry, we believe there should be recourse to sanctions that currently, we think, would require a new Statute.

Overall my elevator speech about what we’re recommending is tough. We produced more recommendations than I had hoped, but there are four basic principles to what you’re watching as you see this play out, and I want to make sure these are clear, because this is what should be acted on.

Number one is put the experience of the patient first – the patient comes first, no matter who you are in the system. Second, hear the patient, hear the carer, empower the voice of the people we’re trying to help. They have more information almost than anyone else in the system. Third, invest in the growth and development of the capabilities of staff, their ability to improve what they do and their ability to work together to improve what they do both within and across organisations, and that will require an investment. And finally, take a big leap toward transparency that is absolute and complete.

We believe that the potential’s enormous. I think my group emerges from our work with optimism, that the sky’s the limit for the NHS of England. It can set the global pace for safety if you choose. After all what other nation could have had a mid-Staffs occur and have the complete mobilisation of intent interest, all sectors involved, except one that has a National Health Service like you do, that the very fact that we’re here today fretting about what happened at mid-Staffs is the strength of the system because it means you can respond, you can get together and decide ‘wait a minute, this is not what we intended’ and as a total community take action.

We have in our report then, taken these recommendations and passed them out as as requests, I guess would be the way to say it, to different stakeholders, and I want to just walk you through some of these top line requests. For the leadership – the top leadership – for the Government, the Prime Minister, the Secretary of State, the NHS leadership, you’ve got to talk the safety and quality talk and that is not going to go away. That’s every day, all the time. It’s relentless, and never-ending. It will never not be part of your job to assert that the NHS is about quality and the patient experience. And part of your duty is to assure, as Bruce Keogh did in his brilliant report, that when warning signs ring, response happens. As my friend Lucian Leape ‘when something happens, something happens’, and that confidence the public has that when there is an alarm it will go off, as it did not go off, it was not responded to at mid-Staffs. It will not happen. It will require leadership right up to the top to invest in the improvement capabilities of the NHS and use the theory of improvement. Understand how much pride and potential joy there is in the workforce if you, as a leader, can call upon it, and I think we’re asking the senior leaders to address the complexity of the regulatory system. At the moment if there’s no taste for restructuring then absolutely insist on unprecedented levels of co-operation among the players. We’re talking here about CQC and Monitor and Trust Development Authority and others, they must co-operate. If they do not co-operate in an unprecedented level we recommend restructuring. We have specific guidelines in there about the restructuring that could be useful. If you run a trust or a clinical unit, put patients in the centre and listen to them, not just in their individual care but in terms of design and conduct and leadership of organisational processes at every step in care and every step in governance. Monitor quality and safety constantly. If you don't understand the data, if you’re not getting the data, you won’t be able to respond to the warning signals or to the day-to-day agenda of improvement.

Several members of the group brought to our attention and eventually won over the entire group on the particular issue of safety alerts. This is what Liam Donaldson used to call ‘The Orange Wire Problem’. I’m going to fly home tonight on a 777. If there’s a burned orange wire when it arrives in Boston at 5 O'Clock, by midnight every 777 in the world will know it – you just don’t ignore it. That’s a safety alert and I think we now have converged as a group on the importance of really getting the safety alert system right. There should be no compromise on that.

We think of will it be important for leaders of organisations to embrace complete transparency, you’ll have to swallow hard to do it, but it’s time to go there, and again I think if you’re going to set an international standard. You’ll have to be part of the training of staff, as we talked about earlier, and we would recommend that every single organisation in the NHS will be part of least one multi-organisational collaborative from now on and forever. There’s nothing like a network to learn about how to make things better in an area you’re currently working on. And then for staffing levels we have a particular staffing ratio at the moment, even though we aren’t completely unanimous on that decision that’s where we’ve come down, but we do think it’s your duty to use evidence based tools to assure adequate staff levels, and soon if things go well you will have guidance on that from Monitor.

For regulators you’ve heard our request. Simplify your work, clarify your work, align your requests and demands. If they’re going to be through regulators making an inspection at Salford Royal in April do it together, not separately. Don’t drive the leadership crazy by having three separate inspections that are not co-ordinated. Reduce the waste, co-operate fully and seamlessly with each other and that, I must say, is asking for above and beyond the current levels of co-operation.

For educators and professional regulators, those who are setting standards for professions, you will need them to assure the capacity and involvement of professionals in quality, as team mates and leaders in continual improvement. And a big challenge to the professional organisation, as the Royal College is for example, is to embrace transparency. We believe it should become the norm for Royal Colleges that when you do audits, when you do studies of what’s going on in your profession, turn the lights on and show everybody what you’ve discovered, that it’s time to get beyond the other belief system.

For staff and clinicians you have a duty which is to be participating in improvement of systems of care, the way my colleague, says it is ‘you always have two jobs: You have the job and you have the job of improving your job’, and that involves skills and commitment and you have... you should be able to rely on leaders making it possible for you to do that, but that becomes a duty which means you have to learn how to do that, so for all of us that may mean going back to school a bit. If you see something go wrong, say so, and expect of your leaders that you can say so. It is not acceptable for a leader to silence a worker who is seeing a safety problem. That’s not acceptable and that’s one of the duties we’re now putting on leadership value, and embrace openness and especially when people are speaking about things going wrong, and remember the patients are in a new position. Given the vision in our report is that their powerful, they’re there, they’re with you, invite them and train them to participate as coproducers. And for the patient’s side, if you’re able, if you’re willing - this is not a compulsion it’s an invitation - be an active partner in your own care and be an active partner in the design and redesign of the system that’s giving you care, and please speak up. Again not a compulsion, it’s an invitation, but you have a right to speak up and you have a right to an expectation that you will be heard.