Errors pervade all health systems. Health care in the United States may cost more, have more resources, and be more customer friendly than that delivered by the United Kingdom National Health Service (NHS), but the epidemiology of errors is probably much the same.^1,2 Even the French system, recently declared the “best” in the world,³ has during this summer’s soaring temperatures publicly failed many of its older population when they desperately needed help.⁴

Poor quality and unsafe care, we have come to understand, are caused by faulty systems and not by faulty individuals and no single group is to blame; “every system is perfectly designed to give precisely the results that it gets”. Even though collated figures about poor quality or unsafe care may be alarming—it is estimated that 5000 people may die each year as the result of hospital acquired infections and that for a further 15 000 deaths hospital acquired infections are a “substantial contributory factor”⁵—the effects of the faults in the design of health care are insidious. For every one person who is harmed by the system of care, many more are unwittingly put at risk, not offered available appropriate options and choices, or simply left bewildered by a system that seems to be overwhelming. These people do not present discrete groups but are scattered among the many who receive good, error free care. Although we know much more about the extent of the problems and something about the causes of poor quality care and the sources of errors, managing to change health systems so that patients not only consistently receive better and safer care but also so we can identify those who do not, is proving a huge challenge.

The papers in this supplement to QSHC are published to coincide with a Nuffield Trust–BMJ Group conference, Working differently, for better safer care, that aims to explore some of the changes needed to working practices if health care is reliably to deliver better, safer care.

Health care is not the only industry to have to face the need to improve safety. Hudson (see pp i7–12 this issue),⁶ describes the changes made over the years in the airline and oil industries and considers the lessons applicable to health care. With vestigial reporting systems and a culture of safety that can only be described as pathological or reactive, health care has much to learn from these safety conscious industries—despite the differences. The relationship between pilots in the cockpit is recognised as central to safety management—something that any team that has worked with an awkward member should recognise. Good working relationships, trust, and understanding are crucial for safe delivery of health care. Edwards (see pp i21–4 this issue) argues that better understanding between doctors and managers is vital if health care is to change enough to ensure safer, better care.⁷

Patients receive care from health professionals, and the roles of doctors and nurses and other health professionals are the usual focus for discussion about the quality and safety of care. But hospitals and surgeries depend on all those who work in them—those providing the infrastructure and facilities as much as anyone. Many people work, often in difficult conditions and during unsocial hours, to do essential cleaning and carrying, and caring. Toynbee’s (see pp i13–5 this issue) recent experience as a porter, cleaner, and health care assistant in London uncovered a separate world operating within health care with its own rules and culture, in which work is sub-contracted out and links between the workers and hospital management are tenuous.⁸ The people who do these jobs have direct contact with patients. They help care for patients. Unless they too are properly valued and allowed to be part of a team then any quality improvement initiative will be incomplete.

Berwick (see pp i2–6 this issue) writes that accelerating healthcare improvement will require large shifts in attitudes and strategies for developing the workforce.⁹ In short, working practices will need to change, for some perhaps out of all recognition. Barber and colleagues (see pp i29–32 this issue) suggest, for example, radical changes to prescribing: doctors will become “directors of therapy” and pharmacists and nurses working in partnership with patients will prescribe drugs.¹⁰

How long will it take before health care can boast a culture of safety that is proactive or generative?⁶ A key factor in industries that demonstrate through their working practices that they take safety seriously, is recognising that what they do is potentially dangerous; its time that health care recognised this too. As Chantler has said, “Medicine used to be simple ineffective and relatively safe. Now it is complex, effective and potentially dangerous.”¹¹ We are still operating in a system that evolved in that safer world.