Root-Cause Analysis: Correlational vs. Causative Findings
It has been mentioned in the first three parts of this series that root-cause analysis is an important process to use in order to discover and find solutions to vexing issues that arise in long-term care. However, at the end of the third series, it was also mentioned that root-cause analysis may not be the best name for this investigatory process. In fact, it often gives the misconception that all problems, no matter how complex, can utilize root-cause analysis and will always determine the exact cause that led to these problems. This is far from the truth and it is the reason I wanted to follow up with this final thought on this topic.
In many cases root-causes can be found. However, as this author has tried to impress through the first three installments on root-cause analysis, it is often a complex process in which cursory and simple approaches can fail, and sound root-cause analysis entails individuals that are well-practiced in their craft. Furthermore, in the real world of health care, especially long-term care, problems are often not unitary, but multimodality issues. It is quite nice when a specific, single, unitary cause can be isolated, leading to the emergence of the problem.
However, in reality, problems are often quite complex and multifaceted. Therefore, using approaches that seek to examine something quite linearly when a lack of linearity exists will often not address the multifaceted nature of the problem. For instance, Toyoda's "five whys" is a simple problem solving tool based on asking the question why until one achieves no further ability to ask this question. It is quite linear, looking to ultimately isolate, in a reductionistic manner, the particular problem. When a problem is simple and unitary in its dimensions it can be used. However, more often than not, many higher level problems have a number of reasons for their existence. Moreover, those who have been involved in statistically attempting to isolate and analyze the reasons for a problem's emergence find that at times the best that can be found is correlation and not cause.
Many problems, due to their complexity, defy simple single root cause solutions. A resident, who has an unplanned hospitalization and has end-stage cardiovascular and kidney disease, along with metastatic cancer to the liver and also suffers from late stage Parkinson's disease, is hospitalized with severe hypertension and mental status changes. Is it due to any one of these conditions, two of the four, three of the four, or all of them combined?
Or possibly you have a problem with staff turnover, which does not only seem to be traced to pay, but to morale, competing facilities in your geographic location offering better fringe benefits, and the turnover being found more exclusively among men in your facility than women. Certainly you may be able to isolate a single issue here, but quite often you may only be able to find a correlation.
Finally, how many readers have attempted to isolate a single cause behind a person with dementia and many behavioral manifestations that arise in these residents? You think you have found the cause to their behavior, only a few days later to see further exacerbation in their behavior after you thought you determined the cause without exception. Also, as more variables are involved, the geometric progression for ascertaining any true isolated cause increases. More often than not, issues often work in a combinatorial manner that lead to correlational, rather than causative, factors being found.
This stated, correlational findings, where relationships are found rather than a single cause, does not need to be viewed as a weakness. In fact, understanding that some problems will only demonstrate trends, patterns, associations, and relationships and that many problems exist with a level of probability and not absolute certainty, helps to strengthen your problem-solving ability and investigatory process. Finding correlations in your search for problems still can lead to strong and sound solutions to problems, and by failing to delude yourself by thinking that you can always find cause, the strength in your ability as a problem solver is enhanced. However, it must be realized as well that correlational explanations are always based on a level of probability that cannot be mistaken for certainty. Therefore, the astute problem-solver will not confuse correlation with cause. Correlation is not equal to cause and should not be interpreted as such. However, correlational results are frequently what are found in the real world of long-term care where problems demonstrate a complex and multifarious nature.
As was mentioned in this overview, the name of root-cause analysis can be misleading. It often leads many to believe that a cause for all issues can ultimately be found. However, as was mentioned here, not all problems can be clearly solved with a discovery of root-cause. In this article the reader was introduced to the concept of correlational understandings, in which relationships between certain events may be found without ever determining the ultimate cause. This is not a short-coming and in fact, realizing this is actually a strength in your pursuit toward finding solutions to your long-term care problems. Unfortunately for many in long-term care, when they are introduced to root-cause analysis, correlation often fails to be presented.
Yet, our lives, the events in long-term care, and the complexity of issues that we face are typically not as linear and easily discoverable as is commonly perceived by those who are introduced into root-cause analysis. In fact, as was stated, root-cause analysis as a name for a multitude of processes that attempt to solve a multitude of issues may be best understood as just simply problem solving strategies. This at least would not minimize the complexity that problem solving often entails. As most statisticians will tell you the events in our world, and long-term care is no different, are probabilistically based, and in many cases absolute causes can never be found.
It is my hope, that after a four-part series on root-cause analysis, or long-term care problem solving, one comes to appreciate the complexity and the difficulty that exists in finding and solving problems. However, this does not mean that it should not be attempted or that one has to specialize in health care problem solving to be successful. What I hope I have conveyed, and what I hope was successful in this series and that the reader is able to take with them, is a greater appreciation of long-term care problem solving. Furthermore it is my hope that the reader has come to realize that given the appropriate practice and diligence, individuals can learn to be very successful problem solvers in long-term care facilities. The benefit for having a facility that employs individuals who are successful problem solvers is ultimately immeasurable. I welcome any comments and feedback you may have on this issue.