The implications for action from the growing body of research on nursing and patient safety are straightforward: hospitals seeking to improve safety outcomes should put a premium on adequate nurse staffing, a high proportion of registered nurses, a well educated nurse workforce, positive nurse-physician relations, and responsiveness of management to addressing problems in patient care identified by nurses at the bedside.

A critical mass of research confirms an association between hospital nursing capacity and patient outcomes, both within and across countries with differently organized and financed health care. Recent studies undertaken in the United States, Canada, England, Switzerland, New Zealand, the Russian Federation, and Armenia^1–6 all show that the adequacy of nurse staffing and the quality of the nurse working environment are associated with the quality of patient care. In hospitals with poor nurse work environments, patients tend to be at a heightened risk for adverse outcomes—including mortality.

The National Quality Forum (NQF) (www.nqf.org), an organization representing a wide array of stakeholders in health care in the United States, has recommended that all hospitals should monitor the quality of the nurse work environment as part of their efforts to improve the quality and safety of health care. The International Hospital Outcomes Study has demonstrated the applicability of the instrument recommended by the NQF—the Nursing Work Index^7,8—in a wide range of countries. Our findings confirm that low levels of hospital nurse staffing and deficiencies in the nurse working environment are associated with poor patient outcomes including excess deaths in a broad array of countries.^1–6

If there is so much evidence for a link between nursing and patient safety, why hasn’t nursing been a higher priority on national safety agendas? Much of the attention in patient safety research and practice has been on preventing errors, while safety experts point out that reducing latent conditions that increase the risk of error will result in safer care more quickly.⁹ Nurse understaffing, fatigue, burnout, inadequate education, and a poor practice environment are classic examples of latent conditions that predispose individuals to make mistakes, and make it difficult for others in the environment to identify a mistake before the consequences are serious. These latent conditions have received far less attention than the factors directly connected to the errors themselves. Reducing latent errors requires culture changes which, while not necessarily expensive, demand a commitment by top management to a different style of decision making with greater devolution of authority to nurses at the bedside commensurate with their high level of responsibility for the welfare of patients.

In this issue of QSHC Tourangeau et al¹⁰ review the determinants of mortality for patients who have experienced acute care hospitalization. To the evidence-based recommendations which they make, we would add two more. The first is that clinicians and leaders must realize that improving safety in health care is not only a matter of implementing “new and improved” procedures and equipment (such as computerized provider order entry), but is also about fundamentally rethinking the environments in which care is delivered. Even though technological solutions and scrutiny and re-engineering of discrete clinical judgments and actions can and should work hand in hand with organizational approaches to safety problems, these two approaches are still often in competition with each other. There is still a tendency to view organizational thinking as too ethereal and distant from the day to day concerns of clinicians and patients to be of much use. That is changing. Researchers need to continue to make organizational context “real” to policymakers, administrators, and clinicians by describing as concretely as possible what good organizations and good conditions in organizations look like, and clearly demonstrating how important outcomes differ under various scenarios. Only then can we achieve the right balance of thought and action for the prevention of active and latent errors.

Our second recommendation is that stakeholders must realize that positive organizational features in hospitals and other healthcare settings (such as staffing numbers and positive nurse-physician relations) do not appear spontaneously, nor do they usually exist in isolation. They are put in place and maintained over time by skilful managers and executives who operate from a vision of patient care driven by an understanding of patient needs. Researchers also need to focus more attention on the interrelatedness and precursors of organizational factors in safety—namely, strong and consistent leadership. On a practical level, managers and executives must be carefully selected, developed in their roles, and listened to and supported in their decision making. Authority and resources must be invested in clinical leaders, both at the front line of organizations and at the executive level, and meaningful and positive interdisciplinary relations must be fostered by any institution hoping for improvements in both organizational qualities and safety, let alone one aiming for excellence. Well functioning interdisciplinary teams, which often need the intervention of top management and role modeling by senior physicians for implementation, are also required. A careful reading of the evidence base linking better work environments to patient safety shows that a culture change could improve the productivity of nurses and other personnel at the bedside, thus making it possible to achieve more with the resources in hand.

Not surprisingly, the review by Tourangeau and colleagues¹⁰ concludes with a plea for more research on organizational correlates of safety, which we heartily support. The enormous progress made in a short period of time reflects the investments of researchers and funders internationally. There is a clear need for more study of the mechanisms responsible for organizational effects on mortality and other untoward patient outcomes. The nature of the relationship between staffing and outcomes needs further clarification, including the circumstances under which it operates the most strongly. There is also a need to connect work on the properties of organizations (including entire organizations and their subunits and broad summary measures of quality) with safety research and practice investigating specific aspects of the delivery of clinical care such as medication errors and failures of team communication.

There is consensus that the goal proposed by the Institute of Medicine to halve the rate of medical errors within 5 years has not yet been achieved.¹¹ In their paper Tourangeau and colleagues¹⁰ show how much work lies ahead to connect organizational components and patient outcomes. More importantly, however, they give us a tantalizing glimpse of how much might be achieved in terms of reducing errors and poor patient outcomes by the investment of resources by hospitals in their work environments, especially in nursing and the management of nursing services.