Root Cause Analysis as a Process, Part 3
In the first two parts of this series, I expounded upon the need to develop a sound problem statement and to understand and not be misled by terms such as facts, validity and truth often used superficially in examining issues. In the third part of this series, a discussion on the process of root-cause analysis will follow. The key term here is process. Root-cause analysis is not a static or compartmentalized element, it is a dynamic process that looks to define the problem, search the data, analyze the data and come up with solutions to deal with a current issue and hopefully forestall future issues. Therefore, as was mentioned earlier in the series, it is a process that needs to be practiced toward developing sound analytical skills, which aid in narrowing down and specifically targeting the problem.
Anderson and Fagerhaug write that the process of root-cause analysis should focus on seven major areas: Problem understanding, problem brainstorming, data collection, data analysis, identification of root cause, eliminating the problem and implementing a solution. Notice that they also agree that one needs to start off with a well-developed understanding of the problem. Also notice that root-cause analysis is typically based on the group, brainstorming ideas, many of which may go unnoticed if a single person is involved exclusively in the problem solving approach. This is important because on a sociological and social-psychological level, studies have demonstrated that bringing people together often helps to generate ideas that they would not have developed in exclusivity.
Following the brainstorming period data collection, data analysis, and root-cause identification should ensue. However, it must be emphasized again that appropriate data collection, sound data analysis, and subsequently a targeted root cause is only as good as the problem statement that has been formulated. If you start off with the wrong problem, no matter how good your data collection, analysis and identification is, it will be all for naught. As was mentioned in the first series, a clearly identified problem drives everything that follows. Therefore, it will determine what data you collect, the analysis that you use, and the identification that follows. If you determined that problem X is the issue, when it is really problems Y or Z, the data, analysis and identification will become futile procedural elements that have not aided your endeavor in finding the truth, and ultimately the cause, associated with a particular issue.
Finally, as the process continues, root-cause analysis looks at more than just identifying the cause, but also eliminating the problem and instituting a well developed solution to 1) deal with the current issue and 2) prevent future issues of a similar nature. The process also should not be viewed as a perfectly fluid process with no glitches. Because it is analytical and deals with individuals as a group looking to add ideas toward finding cause and ameliorating the particular issue, there are often many hills to climb among the group members.
A question often arises about what type of analysis should be employed. The answer to this question is that there is no one correct type. There are two major types of analyses that are often employed: quantitative analyses based on numbers, statistics and mathematical equations and qualitative analyses based on non-numerical data analysis. Neither is necessarily superior to the other, and skilled practitioners will often determine the analytical tool based on the problem and the type of data collected. There are literally hundreds of techniques for both qualitative and quantitative data analysis that can be used. However, due to the brevity of this article, none will be covered. This needs to be reserved for a more thorough discussion of data analysis. But, it should help to illustrate the complexity of root-cause analysis; a process that due to its complexity cannot be entered in a hasty and unpracticed manner.
In the third part of this series I covered many of the steps that are involved in the process of root-cause analysis. It is an iterative process that is shaped by those involved in the process and by the type of problems that are being addressed. Moreover, it is an analytical process that is highly useful to those in long-term care, which can help target and address problems in a constructive manner. However, it must be stated that many problems are quite multi-dimensional and for some of these problems finding the cause is not always feasible. Therefore the term root-cause analysis in itself can be misleading, with the reader thinking that if they followed the appropriate algorithm, a cause will always be ascertained. It is because of this that a final installation on root-cause analysis will be written to explain why correlation, and not causation, is at times the best we can hope for in solving very complex issues that arise in long-term care. I look forward to any comments you may have.
Reference
Anderson, B. & Fagerhaug, T. (2006). Root cause analysis-Simplified tools and Techniques. Milwaukee, ASQ Quality Press.