Across different healthcare systems worldwide, published reports of inquiries into serious incidents urge that the lessons that emerge must be learnt and that such incidents should not happen again. But they do. The lessons, it seems, are rarely learnt and the perception is that health services are not trying hard enough—if at all.^1,2 While it might appear straightforward, in reality not only are the reasons for the incident usually multifactorial, but implementing change is fraught with difficulties. This should not be taken as an excuse, but as a stimulus to understand why it is difficult to learn lessons and how the experience of incidents can be used by others to prevent future harm.

One of the fundamental barriers to learning lessons is the lack of a safety culture in health care. A difficult but essential aspect of health care is the need to accept that people, processes, and equipment in highly complex systems will fail. While the great majority of treatment is carried out effectively and safely, all organisations will at one time or another experience a serious incident. By accepting this, people and organisations can then focus on learning and making changes which seek to design out opportunity for error or failure and develop defences and contingency plans to cope with these failures.³

However, the all too common response to a serious incident is to focus on individual frontline workers who may then face disciplinary measures, suspension from work, and subsequent professional censure.⁴ A good safety culture, therefore, is one where an organisation promotes active awareness, where staff are encouraged to speak up and identify conditions and practices that might lead to an incident, and where staff—when they do speak up—are treated fairly.

Furthermore, there is a tendency to concentrate on immediate causes such as human error rather than on organisational causes which are prevalent in many incidents.⁵ While the immediate causes of an accident are important, and putting these right may prevent the same incident from happening again, making changes to the underlying or latent causes is more likely to prevent future incidents.⁶ By ensuring that incidents are seen as systems failures it is possible for departments—and even organisations—other than those involved in the incident to learn far more than would first appear possible.⁵ This could be achieved by investigating incidents using a consistent methodology across healthcare settings and sharing the relevant lessons to enable wider learning.

Lessons learnt following incidents are all too often confined to those directly involved rather than disseminated throughout the health service.⁵ Most serious incidents repeat past events, but the learning from previous incidents is usually at an individual or, at best, department level and individuals then leave the organisation taking their knowledge with them, depleting the organisation of its memory.⁷ A further challenge to spreading change across health services is that, unless an individual or team is directly involved in an incident, they may feel it could not happen to them and therefore that any change or learning required does not apply. In addition, even when people feel they can speak up, they may not believe that this will make a difference to the system within which they work.⁸ Organisations often fail to show the benefits of reporting and to demonstrate how staff can influence the organisational levels of risk or change the system.

Another problem is that learning is not sustained: an organisation may focus intensively on a problem for a short time but is distracted when new priorities emerge or staff move on. Organisations need to ensure that they set up knowledge and risk management systems for sharing lessons on an ongoing basis. To retain a corporate memory all incident investigation records must be retained and the recommendations made supported by implementation plans. A final report should clearly describe the incident, the causal analysis, and proposed recommendations for change. Many investigators make the mistake of producing an unfeasibly large number of recommendations—itself a barrier to learning—or they provide recommendations which staff can do little about and lack clarity on who is responsible for implementation. Reports should have key recommendations which are prioritised, realistic, and sustainable. These should be translated into implementation plans which are monitored until completion and evaluated for impact. These reports are then used to disseminate the learning to help others.

It is also important to note that serious incidents should not be the only catalyst for change. Organisations should take as much notice of incidents that were prevented or which caused little or no harm. These provide vital information on where a system is failing or likely to fail. If tackled early, the changes made could prevent a more serious incident occurring.⁹ Consequently, incident reporting systems need to be quick and easy to use, and enable staff to report all levels of incidents in sufficient detail to enable appropriate analysis. There is a need to prioritise incidents for further detailed investigation and, if the analysis demonstrates significant themes and clusters of incidents in relation to specific factors, resources should be targeted at the areas that require appropriate changes. Feedback to all staff to demonstrate these changes is essential to re-enforce the benefits of reporting.

The effects of serious incidents for patients, their relatives, and healthcare staff are long lasting. We therefore owe it to our patients and our staff to have an effective system that ensures that the lessons learned are long lasting. If we put into place the processes suggested, we have a realistic and sustainable chance of ensuring that experience of one incident can prevent future harm.