The trusted source for
healthcare information and
Fault trees uncover complex causes
Understand what the date are telling you
By Patrice Spath, RHIT
Brown-Spath & Associates
Forest Grove, OR
Root cause analysis (RCA) is a technique used during an incident investigation to find the fundamental system deficiencies that caused the event. It is also a tool used during a failure model and effects analysis (FMEA) to determine the most likely causes of process failures so that corrective actions can be taken before an adverse event actually occurs. There are many ways of discerning root causes — the most common being a series of "why" questions that may eventually lead to the discovery of fundamental causes.
Even when RCAs or FMEAs are performed flawlessly, it can be difficult to recognize the inter-relationship between causal factors and root causes. Often the fundamental system deficiencies are not single point failures but rather a combination of factors. For example, a data entry error in the computerized medication ordering system may not by itself result in a adverse drug event.
But when this error is combined with other mistakes, such as failure of the double-check system in pharmacy, the likelihood of an adverse drug event goes up. When looking for root causes, whether after an occurrence of a significant adverse event or for failure prevention purposes during a FMEA, it is important to understand the various relationships within the complex systems of health care delivery. If these relationships are not considered, the actions taken to improve patient safety will be less effective.
A tool used in other industries to evaluate complex system relationships is fault tree analysis (FTA). This analysis would be useful in health care to identify the fundamental root causes of unsafe situations, which often result from a complex interaction of multiple failures. FTA is useful for several purposes:
To conduct an FTA, a fault tree diagram is developed. A fault tree is a graphical method for systematically listing various sequential or parallel events or combinations of faults that must occur for a particular undesired event to occur. The fault tree diagram is then analyzed to identify where several things must fail together to cause another failure or where only one of a number of possible problems need to occur to cause a significant process failure. Once the relationship between failures in the process is clear, appropriate corrective actions can be taken.
Fault tree analysis can be integrated with a root cause analysis by creating a diagram for each significant contributing cause that contributed to the event. It can also be integrated with FMEA by developing a diagram for each high-risk failure. Construction of the fault tree involves working downward from the top event (the contributing cause or failure) asking, what could cause this event to occur? This question continues to be asked until the end of your universe is reached — the point at which your organization can no longer control the causes. At this point, you've found the root causes. Once the diagram is completed, it serves as the basis for developing appropriate actions to reduce the risk of critical failures and safety hazards.
Suppose, for example, your organization conducts an FMEA on the process of insulin infusion therapy. One of the potential high-risk failures is that the nurse doesn't change the patient's insulin concentration as ordered by the physician. This failure mode becomes the top event in the fault tree diagram.
To build the diagram for this failure mode, the FMEA team brainstorms the "whys" to the top event — why would the nurse not change the patient's insulin concentration? There could be several immediate causes that would create this problem:
One of the unique characteristics of FTA is the ability to examine the relationship between immediate causes and root causes. This is done by determining if the causes, as identified by the FMEA team, happen independently or if two or more must happen together to cause the top event to occur. The relationship between causes is illustrated on the fault tree through the use of special symbols called "gates." The OR gate is used when causes are independent of one another. The AND gate is used when two or more causes must occur together to create the previous failure.
In this example, the FMEA team determines that any of the three immediate causes — order illegible, order ambiguous, or order missed — could by themselves produce a situation in which the nurse might not change the patient's insulin therapy concentration. Thus, the OR gate is used to connect these immediate causes to the top event on the fault tree. These causes are illustrated on the first row beneath the top event on the fault tree diagram.
The FMEA team continues to build the diagram by identifying the immediate or contributing causes and finally the root causes that exist or co-exist to create the failure depicted as the top event. The "why" questioning ends when the team gets to a point where the causes are out of the organization's control (e.g., patient variables) or when the team runs into a dead end when asked, "What would cause this to occur?" An example of a completed fault tree diagram for the event — nurse doesn't change the insulin concentration — is shown here.
Once the fault tree diagram is completed, analysis and action planning begins. In this example, it is apparent that illegible orders result from a combination of two root causes: poor handwriting skills and lack of accountability.
The team identified handwriting practices and accountability as being dependent; meaning that one cause cannot be fixed without also addressing the other cause. Notice in the diagram that an AND gate was used to connect these two root causes to the immediate cause — order illegible. There are several independent root causes (as depicted by the OR gates in the diagram). For example, understaffing is an independent root cause of missed orders as well as staff not having adequate rest time between work shifts. For action planning purposes, this means that understaffing can be dealt with independent of the rest time concerns. However, changes in staffing practices could only reduce the likelihood of missed orders. The issue of adequate rest time between work shifts would also need to be addressed if the goal were to eliminate missed orders.
Using fault tree analysis during an RCA or FMEA can help the team identify the inter-relationships between various complex factors. By clearly recognizing the risk points in a process, the team can determine where action plans should be directed to achieve maximum benefits.