Gary Oftedahl: Beyond RCA, we need more than a hammer

September 30, 2009 at 1:22 pm 2 comments

In today’s increasingly complex health care  environment, with a rapidly proliferating capacity to provide effective, but complicated care, there is an associated increased risk to the patients we serve.  Patient care, in almost any setting, is an inherently hazardous endeavor.  Since the sentinel IOM report, To Err Is Human, was published, there has been an increasing focus on addressing safety.

The need to develop systems which both provide effective care, but also minimize the risk to our patients of an untoward effect is obvious.  However, even under the conditions of vigorous efforts at risk assessment, the actual outcomes of complex systems will often surprise their designers,  often with tragic consequences.  There is increased pressure on health care systems to develop effective safety programs which can assess and learn from both the consequences of unexpected events, but also from the near misses often seen in health care.  In my previous days as a practicing physician, we viewed the near misses with a sense of relief and moved on to another day, accepting the inevitable likelihood that these near misses and unexpected accidents were an accepted part of medicine.  After all, we’re dealing with people, from a patient, health care worker, and physician perspective, and errors will occur.

Since the IOM report, we’ve seen an increased focus on avoiding “never” events, occurrences which should never happen, under any circumstances.  Indeed, in Minnesota, there is a list of 28 “never” events developed by the Minnesota Department of Health which require submission of a report of their occurrence, and the actions taken to evaluate the cause of said episodes, linked to a “corrective action”  (

I would suggest that our ability to assess and address the issues leading to these “never” events is limited by the tools we use in our evaluation.  Our capacity to identify methods and processes which deal with the increased risk of unexpected human and system errors is exceeded by the rapidity of development of new and complex opportunities for error.  An old adage notes that “If the only tool you have is a hammer, every problem looks like a nail.”  In health care, we have emphasized the use of RCA (root cause analysis), with a recent increased focus on FMEA (Failure Mode and Effects Analysis) in investigating the cause of both the “never” events and in some more progressive institutions, the  “near misses” which may eventually lead to a more  serious eventual incident.

While valuable at a certain level, we in health care still tend to understand errors in terms of breaches of personal and professional accountability.  Few of us have received or understand the types of systems-based issues, human factors engineering, and human behavioral science which would enrich our ability to delve deeper into the complexity of systems and personal issues which led to an unfortunate never event.

In looking at other high reliability organizations (described by Karl Wieck and Kathleen Sutcliffe in their sentinel book, Managing the Unexpected) we see the need for immediacy, completeness, and commitment to a process which is transparent, part of the culture, and cognizant of the limitations of human behavior. In health care, we still see surgeons believing that they’re capable of functioning in a fully alert manner after 36 hours of uninterrupted calls, multiple major surgeries, and emerging fatigue.

There are opportunities to move forward in identifying tools other than the “hammers” we see in RCA and FMEA.  I’m not belittling their role, and their value at a certain degree, but we need to move beyond accepting them as the gold standard in risk assessment.  In fact, while bemoaning their limitations, we rarely see these tools applied in a systematic fashion in many institutions.  Rather than focusing on a single root cause, we need to further our understanding of the complexities of human behavior, the intricacies and convolutions of our processes, and the need to move from solely personal and professional accountability as a driver of our efforts.

At the Institute for Clinical Systems Improvement (ICSI) we’ve been involved in working on safety and learning more about what tools might be valuable in expanding our ability to address it.  While understanding the value of RCA and FMEA, we need to consider the possibility that there  are other aspects of evaluation which may be necessary to add to our “toolbox.”  In light of this, we’re delighted to have Mike Silver, a human factors expert from the Utah/Nevada QIO (HealthInsight) leading a workshop on November 11, 2009 ( providing us information on human factors applications, fault tree and causal tree analysis, and applied behavioral analysis, as potential levers in increasing our capacity to address the multiple causes which may be leading to the unexpected, and dreaded never events.   It’s only a start, but I really don’t want  to be stuck with only having a hammer, when the world requires a more elegant set of tools.

The focus on health care reform is essential, but we need to continue to focus on how we can ensure that our patients are not only receiving the most appropriate and valuable care, but also the safest.  Going to a clinic or being admitted to a hospital should not be considered risky behavior.  Unfortunately at our present level of understanding of errors, and addressing them, it is the norm not the exception to have that  perception.  It is our challenge and responsibility to ensure our toolkit is equipped with the best and most modern “tools” in creating the safe  environment our patients expect.

Entry filed under: General Info, Health Care Redesign. Tags: .

Kent Bottles: Communities and Knowing Yourself Kent Bottles: The Problems That Health Care Reform Must Address


  • 1. Robert Latino  |  October 1, 2009 at 9:09 am

    I would like to address Gary’s specific comment:

    “Rather than focusing on a single root cause, we need to further our understanding of the complexities of human behavior, the intricacies and convolutions of our processes, and the need to move from solely personal and professional accountability as a driver of our efforts.”

    As an author, educator, consultant, software developer and more importantly practitioner of RCA and FMEA for the past 25 years, I both agree and disagree with Gary’s comments.

    Based on his statement above, this tells me that what people are calling RCA, is NOT RCA at all.

    If you do an analysis and find one (1) “root cause”, you are not doing RCA.

    If you do an analysis and do not delve into “the intricacies and convolutions of our processes” and other systemic issues, you are not doing RCA.

    I spent the first 15 years of my career in the engineering discipline in heavy manufacturing. I spent the last 10 years doing RCA and FMEA in healthcare. I can tell you that healthcare in general is not doing RCA near the depth that industry does.

    Healthcare’s definition of success of an RCA is typically to meet the minimum requirements of the Joint Commission so they can pass an audit and get their Federal monies.

    For this reason there is no published correlation that shows a direct link between accredited hospitals and patient safety. As a matter of fact, since the IOM report that Gary cited from 1999, there has been an increase in the number of adverse events in healthcare. The RCA and FMEA guidelines from the Joint Commission have been in place since 1996 and most hospitals are accredited. How can that be? How can patient safety decrease when preventive efforts like RCA and FMEA have been in place for 13 years?

    Either RCA and FMEA do not work as Gary is surmising; OR RCA and FMEA are improperly applied in healthcare. I submit the latter is the case based on 25 years in this niche business.

    It is not the tools, it is the application of the tools and the propensity of the receiving agency or management to accept less than acceptable analysis results.

    Robert J. (Bob) Latino
    Reliability Center, Inc.

  • 2. Cory Boisoneau  |  October 1, 2009 at 10:45 am

    Hi Gary,

    Great post and very informative. As someone who has been working with more and more health care professionals lately, your perspective was very helpful. Do most health care professionals view RCA as a simple tool that is used just to find the “root cause”? At Apollo, we teach a robust method of RCA that seeks to identify and understand all of the causes of a given problem, with the goal of finding effective solutions that prevent problem recurrence.

    From reading your article, it seems the RCA methods you may be referring to are the old, standard methods such as 5-Why’s or Fishbone analysis. While these tools can be helpful and do have a place, they are becoming outdated and are simply just not advanced enough to examine all causes and show how they relate to one another. Also, neither require you to back up causes with evidence, leaving open the possibility of guessing causes.

    Once again, thank you for the insightful post.


    Cory Boisoneau

ICSIorg Twitter




Get every new post delivered to your Inbox.

%d bloggers like this: