Add root-cause analysis to your PI armament
April 1, 1998 8 minutes read
Add root-cause analysis to your PI armament
Find out what sets the stage
By Patrice Spath, ART
Brown-Spath & Associates
Forest Grove, OR
Root-cause analysis (RCA) is a systematic approach to identifying the underlying reason for an undesirable condition or problem. Once causal factors of the event are identified, effective corrective actions can be formulated.
It's important to distinguish between the apparent cause of a problem and the underlying reason. For example, suppose a patient falls and injures himself. The apparent cause of the problem might be lack of supervision, a slippery floor, failure to answer the patient's call-light, or any number of probable causes. The apparent cause, however, is most likely a manifestation of an underlying root cause such as inadequate staffing, fall risk assessments not routinely done for all patients, or haphazard patient education.
RCA complements rather than replaces other performance improvement activities of your organization. Performance improvement involves four activities: quality planning, measurement, assessment, and improvement. Within each activity, many different tools are used, and RCA is just one more process improvement technique that can be used to identify the cause of problems.
Mere identification of the root cause does not automatically assure its correction. The output must be linked to other components of the performance improvement process, such as effective project management, error correction, and trend analysis. Although RCA is a vital element in any performance improvement program, it is most effective when used in conjunction with other program components.
The standards of the Joint Commission on Accreditation of Healthcare Organizations in Oakbrook Terrace, IL, require that an RCA be performed as part of the intensive analysis of a sentinel event occurrence.1 However, RCA techniques can be applied to any problem investigation. A systematic analysis of the events that are causing an undesirable situation would be a beneficial component of any performance improvement project.
Another use of RCA is prevention. While not addressed in the Joint Commission's standards, RCA can be used in a proactive mode to examine current operations. Identifying and improving high-risk areas or activities can prevent an undesirable event from occurring. To obtain full benefit, health care organizations should use RCA for both reactive and proactive purposes.
When an undesirable patient care incident occurs, it is important that physicians, managers, and staff use a logical thought process to arrive at the event's most probable cause. The steps of this logical process - an RCA - are listed below. These steps should be incorporated into your organization's sentinel event policies/procedures to guide all problem inquiry activities.
1. Determine what happened. Have a designated person within the organization undertake this preliminary investigation. This may be the quality management director, the risk manager, or another person or department selected by the leaders to be the principal sentinel event investigator. The investigator collects the facts surrounding the event. This will involve discussions with those directly involved, collection of relevant policies/ procedures, and gathering of other evidence. It's important to quickly assemble all the facts while they are still fresh.
Flow chart is useful tool
After gathering information, the principal investigator can use a common QI tool, a flow chart, to illustrate the pre-event process. A type of flow chart specifically designed for RCA work is called an events and causal factors (E&CF) chart. This flow chart depicts in a logical sequence the events and causal factors for an incident occurrence. E&CF charts were originally developed by the National Transportation Safety Board to use in accident investigations, but are now being used by many industries in their RCA activities. Unlike a common flow chart, the E&CF chart helps investigators see the direct link between each step of a process and root-level or systemic causes of undesirable performance.
An example of a simple E&CF chart is illustrated on p. 81. The sentinel event being investigated with this chart involves a situation in which a patient's subdural hematoma was not diagnosed by the emergency physician, and the delay in treatment contributed to the patient's death. The events described in the rectangles represent the process steps that led up to the event. The causal factors that appeared to influence each process step are shown in the ovals. The investigation team will greatly expand the E&CF chart as it gathers more information about the sequence of events and identify causal factors.
2. Select an investigation team. The RCA becomes multidisciplinary at this step. People familiar with the processes underlying the event are brought together to conduct the formal investigation. The information gathered by the principal investigator during the last step ought to reveal who should serve on this team. Using information gathered by the principal investigator, the team begins its analysis of the event. Additional information will continue to be gathered as necessary.
3. Thoroughly study the event. The goal of the RCA is to discover all cause-effect relationships from which practical corrective actions can be derived to improve total performance. By answering questions such as these, the investigation team can narrow the list of causes and begin to formulate solutions:
· What was different this time? Sentinel events are rare occurrences. The investigation team should determine what was different in the process that allowed the event to occur. Compare a flow chart of the process as it should occur to the events leading up to the event; you may discover differences. By acknowledging what was different in the process at the time of the event, the team may acquire insight into why the event happened, particularly with the first occurrence of a problem.
· Did people do what they were supposed to do? The answer here might be difficult to obtain if the organizational climate is not supportive. However, it can point to problems in systems, training, or performance appraisals. It's important to keep the investigation team focused on the performance of the system and not just on how individual people performed.
· Did people know what to do? The answer to this question helps to point out holes in present systems that prevent people from knowing what to do. Although staff can't be trained or procedures provided for every possible situation, the team can determine what critical thinking skills were lacking and how they can be reinforced through coaching or education.
· Who reported this event? Often this is important and may give clues to problems with the reporting system itself. Was the person who reported the undesirable situation the one who should have first noticed it? If not, make sure to talk with the person who should have first noticed the event and find out why he or she didn't report it. The goal of this discussion is not to place blame, but to find and resolve the underlying system issues that need correcting.
· Did any event/action prevent an even worse situation? Answering this question will help the investigation team think through solutions to the problem and identify current problems in procedures and training.
· Has this happened before? Perhaps this should be the first question! The answer gives vital information regarding the nature of the problem and the efficacy of any previous corrective or preventive actions taken.
· What was done before to fix it, if anything? If nothing was done, it's time to do something now. If previous solutions didn't work, it's time to redesign corrective actions.
4. Identify the event's causal factors. Here's where the investigation team begins to answer the "why" question. The team uses structured analytic techniques to discover all influencing factors that contributed to the event. There are two major kinds of causal factors: direct and contributing. Direct causes represent the apparent reason for the event. A direct cause may turn out to be the root cause, but until this is confirmed by analysis, the investigation team should not make this assumption. Contributing causes represent the process elements and underlying system issues that created the environment in which a sentinel event was more likely to occur. These, too, may become root causes once the investigation is complete.
Some questions the investigation team can ask when identifying the causes of an event:
· How could this event have been prevented? By asking how the event could have been prevented, the investigation team can identify processes or underlying systems that were inadequate.
· What might prevent this event from happening again? The answers to this question will help the team focus its attention on all possible corrective actions. However, don't jump too quickly to the action planning phase. Wait until all root causes and solutions are identified and analyzed.
5. Select root causes. Through a systematic analysis of the direct and contributing causal factors, the investigation team arrives at the root causes - the underlying issues that, if corrected, would reduce the likelihood of future similar events. Root causes can include system deficiencies, management failures, unqualified staff, performance errors, omissions, nonadherence to procedures, and inadequate health care team and/or organizational communication. The team should find one to three root causes for a particular event. If they come up with more than that, the team should dig deeper into their answers about "why" the event occurred.
To ensure the investigation team has analyzed the sentinel event fully and correctly, check its proposed root cause(s) against questions on the checklist inserted in this issue. For each identified root cause, the team should be able to answer Yes to all the questions. If the results of this evaluation are not conclusive, continue working on root-cause identification until the tests are passed.
6. Develop corrective actions and a follow-up plan. After selecting the root cause(s) of the sentinel event, the investigation team can begin work on the second phase of the investigation: problem solving. In this step the team will decide on workable solutions for the root causes. If the event resulted in a serious patient injury or a death, the team will probably already have implemented preliminary corrective actions to minimize the risk of another undesirable event. During this step, the team should revisit these knee-jerk solutions to be sure they are the best choices for achieving lasting change.
The corrective actions will vary according to the root causes identified. An interactive process is the preferred approach for generating solutions. That is, the team members should work together to review each root cause and then begin generating a list of possible solutions. These solutions should be linked directly to the root causes derived from facts and analyses.
Problems that are "fixed" without a plan for structured evaluation tend to resurface later. That's why it is important for the investigation team to identify qualitative or quantitative measurement data that will be collected during and following implementation. Because sentinel events are rare, don't rely on sentinel event incidence data to measure the success of your solutions. Define measures that evaluate the impact of your solutions on the root causes of the event. The investigation team also should determine when measurement data will be aggregated and reported, and what group is responsible for analyzing the data to determine the effectiveness of the solutions.
7. Prepare a RCA report. Upon completion of the project, the team should prepare a report that clearly and concisely conveys the results of the investigation in a manner that will help the reader understand what happened (the event description and chronology), why it happened (the causal factors and the root cause), and what can be done to prevent a recurrence (the proposed corrective actions). A sample report format is inserted in this issue. If the RCA report must be communicated to the Joint Commission, a more comprehensive report using the required format must be prepared. In this report, be sure to thoroughly document all investigations and the team's rationale for choosing the root causes it did. The more complete your documentation, the more likely your investigation will be deemed acceptable to the Joint Commission.
Many of the investigative techniques used in a reactive RCA can also help prevent incidents before they occur. In next month's issue of Hospital Peer Review, you'll find out why proactive RCA should be an important component of your organization's performance improvement strategy.
Reference
1. Joint Commission on Accreditation of Healthcare Organizations. Comprehensive Accreditation Manual for Hospitals: The Official Handbook. Oakbrook Terrace, IL: 1998, PI.4.3.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content