In January, on a beautifully clear summer day, I was at Milford Sound , the stunning fiord on the west coast of New Zealand’s south island.
As we sailed up the sound, the captain of our sightseeing ship told us that Milford Sound airport is one of the busiest in New Zealand, although it can only be used by small light aircraft. Every few minutes one such passed overhead, making its winding passage up the sound and then swinging sharply right between the towering sides of the fiord. Before the final approach to the tiny runway a tight one hundred and eighty degree turn had to be executed. Flying into Milford Sound is not for the faint-hearted!
Watching these small planes landing at Milford Sound, and knowing that several hundred do so safely every day, made me reflect on why an apparently unsafe form of travel – the airplane – has now become so safe. It is not just better technology and more experienced people, but an attitude of mind that creates a safety culture. Part of that is an insatiable curiosity about what can go wrong and how it can be avoided. If you have to fly into Milford Sound everyday you can’t afford to close your eyes and hope for the best. You have to anticipate the worst and prepare for it.
All of which brings me to the issue of Critical Incident Reporting. Back in the 1940s an American air force colonel called Flanagan  came up with an apparently simple idea. To improve aviation safety, he argued, it was not good enough just to understand what caused particular accidents. Rather we need to know about situations in which accidents might have occurred, but didn’t. We need data about the errors that are made in normal practice, which do not result in a fatal outcome and which often do not come to light. In short we need to study near misses.
Professor James Reason articulates the need for such an approach rather well: ‘Without a detailed analysis of mishaps, incidents, near misses, and “free lessons,” we have no way of uncovering recurrent error traps or of knowing where the “edge” is until we fall over it.’ 
Critical Incident Reporting is one means by which we can obtain data for studies of near misses. Professionals – pilots, ships’ officers, train drivers, doctors, social workers, or indeed any workers whose tasks are safety critical – are provided with a simple means of reporting a critical incident or near miss. The reports are submitted confidentially and the results aggregated and reported in such a way that nobody can tell who in particular was involved. Then the original report is destroyed to ensure continuing confidentiality. That way people will tell the truth and will report incidents which otherwise may never have come to light.
Critical incident reporting has been long established in Britain in aviation and shipping  and is strongly advocated in medicine, particularly anaesthesia and intensive care  . Back in 1990 I co-authored an article  on the applicability of the technique to child protection social work. We argued that reports into child abuse disasters reveal that factors contributing to the death of a child commonly occur in normal practice. These include poor communication, professional disagreements, vacillation, uncertainty in response to aggressive families, pressure of work and burdens being inappropriately placed on inexperienced staff. We concluded that it was only possible to understand the causes and effects of such malfunctions by carefully documenting how they occurred in normal practice.
If properly implemented the benefits of Critical Incident Reporting are obvious and profound. Sadly it is all too easy to get the implementation wrong. A crucial mistake is to be careless about the arrangements to ensure confidentiality; if people believe that they may be identified by a report they will not wish to participate.
An example of poor implementation appears to have occurred in the NHS in Scotland where copies of reports were apparently retained in a filing system. Although heavily redacted these reports had to be made available to the media under the Freedom of Information Act, raising the possibility that individuals or particular units could be identified.
In contrast CHIRP, the aviation and maritime critical incident system, ensures confidentiality by being an independent charity rather than a branch of government or of a particular airline or shipping company. And reports of incidents are quickly destroyed so that no-one can trace the identities of those involved.
The CHIRP website promises:
“CHIRP always protects the identity of our reporters. We are a confidential programme and, as such, we only keep reporters’ personal details for as long as we need to keep in contact with them. When a report is closed off, all original correspondence is sent back to the reporter and all notes are shredded. The reporter’s personal information never gets input in to our database.”
I believe that a Critical Incident programme for child protection could work well in Britain. It would need to be set up as an independent charity, but could receive donations from public bodies as well as from the general public. Independence and confidentiality are crucial but should not be difficult to achieve. And it is likely that such an initiative would not only result in greatly improved safety but might also result in reduced costs as agencies learn more about how to avoid costly and unnecessary errors.
 Flanagan J. C. “The critical incident technique.” Psychological Bulletin 1954; 51: 327–58
 Reason, J. “Human error: models and management.” British Medical Journal 2000;320:768-770 (18 March)
 Cooper JB, Newbower RS, Long CD, McPeek B. “Preventable anesthesia mishaps: a study of human factors.” Anesthesiology 1978; 49: 399–406
 Mahajan R. P. “Critical incident reporting and learning”
British Journal of Anaesthesia 105 (1): 69–75 (2010)
 Mills, C. and Vine, P. “Critical Incident Reporting – an Approach to Reviewing the Investigation and Management of Child Abuse.”
British Journal of Social Work (1990) 20, 215-220