Skip to content


Pre-disasters and Halloween Ghosts of Mistakes Past

John Irving’s “The World According to Garp” is a great book with twists and turns through all kinds of adult scenarios.  T.S. Garp (T.S. stands for “technical sergeant” which is what his comatose father was known as to his nurse mother) grows up as the pages unfold and eventually becomes a parent.  The story keeps coming back to the theme of his anxiety for the safety of his children.  He shares the same desires of all parents, to keep his kids safe from the dangers of the world.  So Garp goes out looking for a house to buy but can’t seem to find the right one with all the safeguards he seeks until one day he’s looking at a house just as it is struck by an airplane falling mysteriously from the sky.  Garp stood in awe as the events unfold and finally concludes, “We’re safe here, it’s pre-disastered.”  In Garp’s world, he figures that just because disaster already struck his new house it can’t happen again.  Wrong – dead wrong.  The fact is that things we think can never happen, or could certainly never happen again, actually happen (again) all the time.  And they keep on happening until we uncover and eliminate the causes and thus prevent their recurrence.

Perhaps the most notorious example of repeating disasters is historical.  Sir Winston Churchill popularized the saying, “those who fail to learn from history are doomed to relive it” (but it was actually first said in 1905 by George Santayana).  Churchill was referring to Hitler’s insane strategic blunder to invade Russia in late June, 1941 only to flounder twelve weeks later in the brutal Russian winter.  This was the same fatal mistake Napoleon made 120 years earlier; both lost their armies and both met defeat.  In “All Hell Let Loose” Max Hastings quotes a German officer watching his troops freezing and dying on the outskirts of Moscow, “the ghosts of Napoleon’s army hover ever more strongly above us like malignant spirits”.  For practitioners of the high reliability mindset, this statement is a poignant reminder that similar mistakes can and will happen and those who don’t learn from them will be haunted by the malignant ghosts of previous errors.

Failure to understand this simple concept was a critical contributing factor to two fatal outcomes that I recently reviewed.  The two unfortunate incidents were virtually identical; except they happened six months apart.  Both were procedures that were undertaken after hours and could certainly have waited for more experienced people and for a time when more help was around in case of problems.  Without going into details, the two incidents involved the same procedure that should have been routine in the ICU.  Each had uncommon but not unexpected complications that couldn’t be dealt with because they were done at off hours, late one evening in one and on a weekend in the other case when experienced surgeons were not around to help manage the problems that occurred.

As students and advocates of the high reliability mindset we know that the only way to prevent an error from recurring is to collect honest and complete incident reports.  Then openly discuss them in a non-blaming manner, study, and understand their causes down to the last detail.  Furthermore, we know that the major value of incident and near miss reporting is to use these reports to determine the root cause of the problem with thorough analysis.  A retrospective approach to error analysis, called root cause analysis (RCA), is the most efficient way to investigate incidents and errors.  RCA has its foundations in industrial psychology and human factors engineering and for some time the Joint Commission has mandated the use of RCA in the investigation of sentinel events in accredited hospitals.  RCA provides a structured and process-focused framework with which to approach sentinel event analysis.  One of the guiding principles of RCA is to avoid the pervasive and counterproductive culture of individual blame. RCA must be used to “shine light, not heat” on errors, near misses and bad outcomes otherwise it defeats the objective that all participants contribute honest information.

The first of the two key aspects of RCA is data collection which is the establishment of what happened through structured interviews, document review, and/or direct observation.  This is used to generate a sequence or timeline of events preceding and following the event that take into account ALL aspects, ALL participants and ALL actions.  The second element of RCA is data analysis that is a detailed process to examine the participants and sequence of events generated above with the goals of determining the underlying contributing factors – the “who did what, how and why”.  The end result of this investigation is to ask and answer 5 “why questions” after the “what happened” has been determined.  By asking “why?” after the “what” the true root cause can be arrived at.  Statistically, after answering 5 questions that start with “why” there is a 95% confidence level that the true root cause will be answered.  In this way, individual, systems and organizational issues can be identified and addressed, and active errors are acknowledged and corrected.

Just as the cases I mentioned above, a systematic application of RCA would have uncovered common root causes that linked these two accidents.  Careful analysis should lead to system changes designed to prevent future incidents by limiting invasive and potentially dangerous SICU procedures to “the light of day” unless there is a dire emergency.  When there is an urgent need to do a procedure in the evening, all team members should be informed, so that they can mobilize in a hurry if things go awry.  Failure to study and learn from past mistakes dooms all of us to be stalked by their ghosts – Halloween or not.

Share on Facebook

Posted in High Reliability Mindset, Uncategorized.

Tagged with .


2 Responses

Stay in touch with the conversation, subscribe to the RSS feed for comments on this post.

  1. Tony Kern says

    Ken, Well put with great historical examples. RCA drives understanding, but FAR TOO OFTEN we can’t get anyone to ask a single “why?” Even with truly adverse outcomes, we see defensive posturing instead of learning. RCA is a deep dive but simple debriefings can drive improvement with an open mind towards improvement.

  2. skulljockey says

    I would take things even farther .There are many interesting points ( but perhaps disjointed) to make about this blog.
    Firstly , your stories of nocturnal misadventures beautifully demonstrate that is not long shift that place patient at risk but lack of supervision. This has been well supported in the medical education and patient safety literature.

    Since this has an Halloween theme, let me be the devil ( advocate). The process of RCA is not without controversy . I favour the position of Charles Vincent and others that the term RCA is misleading. There is rarely a single cause of an event. Although of course
    we need to find out the “why” to be able to let the family and patient know what happen but what is more important is to look forward and understand the system failure. so it is the retrospective look at an incident that I ( and others) find problematic without a deliberate look at the future. That is why even after extensive “processing” of RCA, driven by our organizations, the same incidents re-occur. So why does the same incident occur six months apart?

    Being of Irish descent, I love the condensation of Halloween and your play on the light of day and “shine light not heat” examples . Of course the Celtic celebration of Samhian or the final harvest is is the root of Halloween. Traditionally all lights were extinguished and then relit from the central Druids representing the beginning of the Celtic New Year. A new beginning..a new understanding…The analogies are obvious.

    Thanks for your provocative blog.



Some HTML is OK

or, reply to this post via trackback. All comments are moderated and may not appear immediately.