The Institute for Healthcare Improvement is aiming to prevent 5 million medical injuries through an improvement campaign in the US. Here Don Berwick says safety requires a passionate determination by managers

The first step toward systemic improvement is to intend it. That takes courage, because it involves staring reality in the face - the reality of defect. The problem of gaps in the quality of care is immense in all nations.

The quality gap is much, much bigger than most people think. That poses healthcare leaders, especially leaders in systems of political accountability for care, with a psychological as well as a strategic challenge. Leaders need to address both the felt needs of their voting public - which are usually to do with service characteristics, especially waits and delays - as they are real and urgent.

But that is not enough. Leaders will actually need to add to public concerns by surfacing and addressing gaps the public is far less aware of, such as safety hazards, unreliable care, waste and clinical outcomes of care that fall short of the scientific potential.

A very helpful scheme for classifying and studying opportunities for improvement was suggested by the great Japanese quality scholar Professor Noriaki Kano. He said there are three types of improvements: first, those that reduce defects experienced by the people we are serving (such as avoidable infections or insulting interactions); second, improvements that reduce costs, while maintaining or improving the experience of the person served (such as not losing a laboratory test and having to repeat it); and, third, improvements that represent new features or enhancements, some of which add costs - such as finding a cure for a disease previously thought not treatable. In general, healthcare has not been bad at the third type of improvement - but it has not done too well with the first two.

It will help us think clearly if we can talk, not just about improvement as an abstract concept, but about a specific improvement - patient safety.

As we have begun to understand the ways that healthcare fails, patient safety has in many nations surged to near the top of the agenda, right behind cost. That is not because safety is the only dimension of care that matters. But it is in many ways the most understandable: it is the easiest bridge to building shared will between the public and the system. You do not have to be a doctor or nurse to know what it means to be harmed by the care that is supposed to help you.

Secrecy or openness

In the Institute for Healthcare Improvement, we use a repeated mantra to describe the essential elements for large-scale change: will, ideas, and execution. You have to intend to get better, you have to have ideas about alternatives to the status quo that can get you there, and then you have to make it real - execute changes.

Mobilisation for change starts with will; creating will is the first duty of leaders of change. Hurting patients needs to become not just conscious, but unconscionable.

The political and policy agenda to improve patient safety should be informed by as deep an understanding as possible of the causes of healthcare injuries. Because, to make care safe, we are going to have to redesign care thoroughly and that is not a simple matter at all.

The first answer that many people reach for is simple but not scientific: blame somebody. Blame the incompetent doctor, the careless nurse, the stingy executive, the saboteur. Blame the government. If all else fails, blame the patients.

That cannot work. It can never work. People who use blame to pursue safety are showing that they do not understand the scientific foundations of safety.

It is incorrect because of the fundamental premise of modern improvement science, which I first heard from my colleague Paul Batalden: 'Every system is perfectly designed to achieve exactly the results it gets.'

Safety will get better through the redesign of healthcare, and in no other way. That is a tough message. It is a lot tougher than just blaming someone when things go wrong. It is especially tough on leaders, because it plants responsibility for safety firmly on their shoulders.

The workforce cannot tame healthcare. The workforce is trapped in the defects. The young doctor or nurse who finds himself or herself trapped in an error feels guilt, demoralisation, horror, loss of confidence. What they do not feel is powerful, able to change anything.

Unsophisticated systems - untutored leaders - perpetuate the hazard. The reason is that they encourage secrecy. They do that by blaming people. When people are scared of what will happen to them if they tell the truth about errors, they hide the hazard, instead of curing it. No modern safety system of any merit relies on blame and reprisal as a mainstay of improvement.

Naming the enemy

As healthcare leaders, we should create and nurture an environment that accepts human frailties, that discloses hazards and injuries, and that encourages and celebrates openness, honesty, and ambition about those injuries. Otherwise, our patients and communities will continue to pay a high price in hidden healthcare injuries. We cannot have it both ways. We have to choose either blame, secrecy, and injury, or, openness, forgiveness, and safety.

We will need knowledge, learning, and action. All of those depend on leaders. The leader's task is to establish a context that encourages curiosity about hazards and injuries, and that supports continual change to reduce those defects. To get that done, leaders have those three primary tasks: to ensure will, to nurture ideas, and to make improvement daily work.

Will comes first. We have got to stare the problem of patient injuries in the face. We have to be absolutely honest about the modern burden of injury to our patients, and we have to refuse to accept that burden as inevitable. We cannot conquer an enemy we do not name. The exact same principle will apply to any dimension of care you intend to make better, be it safety, effectiveness, patient experience, timeliness, or efficiency.

After will, we need ideas. If every system is perfectly designed to produce the results it gets, then asking the status quo system to do better makes no sense at all. All improvement is change; though not all change is improvement. Leaders have a duty to arrange for the continual inflow and discovery of alternatives to the status quo.

What is the other way; the new way? First, we must all give up outmoded beliefs and theories, like the reliance on blame, or the hope that measurement and reports systems are enough.

Many nascent safety systems invest excessively in measurement and reporting. They invest an enormous amount of currency - both financial and political - to mount massive projects on safety reporting, or on reducing delays, or becoming more patient centered, or reducing costs, without anything close to the proper investment in redesign, research, and, most crucially, technical support to the field to redesign processes and work.

We do need measurement and reporting, but the point, in the end, is not to measure, it is to change.

Most healthcare experts will tell you that hospital-acquired infections are, to some degree, inevitable. Some people would not even classify them as patient injuries - certainly not as errors. We disagree.

In 2005, the Institute for Healthcare Improvement in the US began a national effort that we called the 100,000 Lives Campaign. We asked hospitals to band together to save 100,000 lives in 18 months, by adopting a few simple changes in their care that can reduce needless deaths.

Over 3,100 hospitals joined, and they were supposed to make six changes in all. Three bear on infections that people get in hospitals - ventilator-acquired pneumonia, surgical site infections, and bacterial infections of indwelling central venous catheters. For each of these hazards, we asked hospitals to adopt new systems of care that ensure, absolutely, that every eligible patient gets every single one of the simple, evidence-based procedures that are known
to reduce the chances of the infection.

Inspirational results

What is new is the commitment to absolute reliability, developing standards, computer supports, reminders and teamwork processes that make the right things the easiest things to do - the defaults.

The results have been inspiring. Dominican Santa Cruz Hospital in California celebrated one full year without a single ventilator-associated pneumonia. Mercy Health System in Oklahoma City had over 1,000 consecutive surgery cases without a single infection. And, Dr Peter Pronovost from Johns Hopkins led a consortium of 70 hospitals to reduce central venous line infections by over 50 per cent, saving 1,578 lives and 81,020 hospital days. He also helped them save over $165m in costs averted.

Based on the experiences of the 100,000 Lives Campaign, the IHI has just launched a new and expanded effort, the 5 Million Lives Campaign, which will try to help 4,000 US hospitals prevent 5 million medical injuries in the next two years, beginning with investment in the original six changes of the 100,000 Lives Campaign, plus six new ones.

The campaign hospitals are not just showing strong will and good ideas. They are also tackling effectively the third, and hardest, job of safety leaders: execution. In the US, healthcare is highly distributed - care is being given in thousands, even hundreds of thousands of places by millions of talented, well-meaning workers, to tens of millions of patients.

To make care safe, changes have to spread that far, that wide. It is a matter of day-to-day management. The same leaders and managers who today create and manage budgets, schedules, supply chains, training, supervision, hiring, and clinical care itself, need to have safety on their screens every day, every moment. Like quality, overall, safety can be no more a 'programme' for a healthcare organisation than breathing can be a programme for a human being.

Don Berwick is chief executive of the Institute for Healthcare Improvement in the US. This article is based on his speech during an IHI visit to Cardiff last month. The full version is available atwww.ihi.org