Introduction
Greetings!
The Department of Veterans Affairs National Center for Patient Safety (NCPS) supports and leads the patient safety activities for all the VA medical centers. Since 1999, NCPS has developed tools, training and software to facilitate patient safety and Root Cause Analysis (RCA) investigations. This guide functions as a cognitive aid to help teams in developing a chronological event flow diagram (an understanding of what occurred) along with a cause and effect diagram (why the event occurred). RCA teams have found this book an effective aid with these sometimes nettlesome activities. We welcome your comments, suggestions, and feedback.
James P. Bagian, M.D., P.E.
Lead Developers
Joe DeRosier, P.E, C.S.P.
Erik Stalhandske, M.P.P., M.H.S.A.
Additional Contributions
Jean Alzubaydi, John Gosbee, Amelia Landesman, Jeff Lee, Janine Purcell
Other tools developed by the Department of Veterans Affair National Center for Patient Safety:
- Patient Personal Freedoms and Security : Escape and Elopement Management (Version 10/01)
- Patient Personal Freedom and Security : Fall Prevention and Management (Version 10/01)
- NCPS Triage Cards for Root Cause Analysis (Version 10/01)
- The Healthcare Failure Mode Effect Analysis Process (Version 8/02)
Sources Consulted
- Veterans Affairs NCPS Patient Safety Improvement Handbook, January 30, 2002.
- Marx, David. Maintenance Error Causation. A technical report prepared for the Federal Aviation Administration; June 9, 1999.
- Gano, Dean L. Apollo Root Cause Analysis. Yakima, Washington: Apollonian Publications, 1992.
How To Use This Aid
First, go to the Flow Diagramming tab to develop the chronological sequence of the series of events leading up to the incident.
- Follow the sequence through Initial, Intermediate, and Final Diagramming.
Second, go to the Cause and Effect Diagramming tab to identify the conditions that resulted in the adverse event or close call.
- Follow the sequence from steps one through three.
Third, go to the Causal Statements tab and develop root cause/contributing factor statements, actions and outcomes.
Detailed Explanation of How to Use This Aid
The first step in an effective investigation is uncovering what occurred. The Event Flow Diagramming steps (the first set of tabs) accomplish this task. For effective problem resolutions, teams also need to uncover why something occurred. Use the Cause and Effect Diagramming steps in the second set of tabs to meet this goal.
Definitions
Throughout this guide there may be unfamiliar terminology. Below is a list of terms and definitions. These definitions are also highlighted in red. You can see the definition by leaving your mouse cursor over the red text -- Try it here.
- Cause and Effect Action:
- Something momentary and fleeting (e.g. stepping on wet floor, turning on machine).
- Cause and Effect Condition:
- Something existing over a period of time (e.g. temperature of room, competency of staff, waxed floor).
- Cause and Effect Diagram:
- A systematic method of determining the causal conditions underlying an event or close call.
- Event Flow Diagram:
- A chronological diagram of the series of events leading up to an adverse incident or close call.
- Hindsight Bias:
- Failure to recognize factors and the myriad of decisions made by an individual prior to the event or act which resulted in the adverse event occurring due to being prejudiced by the final result.
- Root Cause/Contributing Factor:
- Any event in the chain of causes that, when acted upon by a solution, prevents the problem from recurring.
Next Section: Event Flow Diagramming
Diagramming
Introduction: Event Flow Diagramming
Event Flow Diagramming helps the Root Cause Analysis Team identify what happened in the event. Three steps depict the diagramming process:
To See: | Go To: | |
|
Initial Diagram | |
|
Intermediate Diagram | |
|
Final Diagram |
Why are we presenting Event Flow Diagramming?
- Event flow diagramming is one of the critical tools for many process improvement activities including root cause analysis investigation.
- This initial step gives all members of the team the same understanding of what occurred, and helps avoid differing interpretations of the same event.
Helpful Hints -- Initial Flow Diagramming
- Visit the site of the event and observe activities with the RCA team to confirm your understanding of the process.
- Work with the process owners to confirm your flow diagram.
- Remember! This chronological flow diagram should reflect what actually occurred.
Diagram
Initial Event Flow Diagramming
An Initial Event Flow Diagram is an outline of the story that progresses chronologically from the first known fact, includes the actual event being reviewed, and concludes with the final known event.
Establishing an accurate chronological depiction of the series of events preceding the adverse event or close call is a critical step towards discovering the root causes/contributing factors and determining the corrective actions to implement. Include only those key events that are crucial to understanding what transpired.
Note: It is not necessary to include the amount of time that elapsed between events, but if the information is available it may provide valuable insights in later problem solving steps.
Event Reported:
A patient in a locked ward was found lying on the floor in his room with 3rd degree burns to his chest and arms. The patient was last seen requesting a cigarette. A partially burned posey was still attached to the patients wheelchair.
While the events preceding the discovery of the burned patient are not explicitly defined, it is possible to determine the series of events which led up to the injury by referring to clues given in the narrative above. Using this information, an initial Event Flow Diagram for the situation might look like this:
? | → | Patient is wearing posey in wheelchair | → | Posey ignites, burns, and breaks | → | Patient falls out of his wheelchair | → | Patient found burned, lying on the floor |
Diagram
Intermediate Event Flow Diagramming
The rule is simple: when addressing each event in the flow diagram, ask why each event occurred until there are either no more questions or no more answers. If the answer results in blaming an individual or group of individuals ask another why question. Also remember to get answers to the Why X 3 (why times three) questions what happened, why did it happen, what are you going to do about it. At this stage, use the Triage questions to help identify system and process issues reference related questions under the event.
Beware of hindsight bias (a.k.a. Monday Morning Quarterback). Teams often jump to conclusions thinking they know the cause of the adverse event without doing any investigation. The natural tendency is to think in terms of a straight line from the adverse event to the preceding action. In reality, multiple decision points are encountered and must be dealt with. These environmental factors and decision points must be understood in order to identify the root cause or contributing factors leading to the adverse event. Try to think why it would make sense to you to do that activity or make that decision.
The team uses the intermediate diagram to develop and answer questions to get to the final understanding of the event. Follow the sequence looking for what caused each event, and add to the sequence as needed. This often reveals any gaps that may exist between events.
*These reference specific questions from the VA NCPS Triage questions for Root Cause Analysis.
Diagram
Final Event Flow Diagramming
After the why questions and the Triage questions have been answered, interviews conducted and references consulted, the team constructs the final event flow diagram.
Last, ask for the relevance or significance of each event (the so what? question), and capture answers under the event. This will help the team identify potential root cause/contributing factors.
The Final Event Flow Diagram should resemble the following model (the actual number of events will vary):
Statements
Event Flow Diagram Causal Statements Examples
Potential root cause/contributing factors (Causal Statements):
- The lack of staff training on the medical center restraint policy resulted in the patient being tied to his wheelchair with a posey, increasing the patient's agitation, and increasing the likelihood of injury by fire.
- Not having restraint alternative devices available resulted in the patient being tied to the wheelchair, increasing the patient's agitation and the likelihood of injury by fire.
- The policy of providing lighters to ignite cigarettes gave the patient the ignition source needed to injure himself and/or others by fire, which occurred.
- Inadequate staffing level on the floor due to an employee illness resulted in the smoking room being left unsupervised. This created the opportunity for the patient to return to his room with smoking materials, leading to the injury by fire.
- The highly combustible nature of the restraint device increased the likelihood of injury by fire, which occurred.
Next Section: Cause & Effect Diagramming
Diagramming
Cause & Effect Diagramming
With the assistance of a Cause and Effect Diagram, an RCA Team can uncover chains of causal links that will lead them to the root causes/contributing factors of the problem and allow them to implement effective barriers and solutions.
Teams report that this diagramming helps achieve some of the magic of transitioning from focusing on what occurred to a deeper investigation of why an event occurred. Also, this approach encourages examining system and process vulnerabilities instead of asking who is responsible. A Cause and Effect Diagram is constructed along a four-step process:
For Step: | Go to: | |
|
ID Problem | |
|
Brainstorm Causes | |
|
Complete Diagram | |
|
Causal Statements |
Why are we presenting Cause & Effect Diagramming?
After review of hundreds of RCA investigations, we have seen that teams are able to successfully determine what occurred and create an accurate chronological flow diagram. However, teams may struggle with adequately developing why the event occurred. They have more difficulty creating a causal sequencing of events. This tool will help teams progress from what happened to why it happened.
Problem
Identify Problem Statement
Event Reported:
A slip and fall accident resulted in an elderly patient fracturing his hip. The man, wearing only his hospital gown and slippers, had last been seen walking down the dimly-lit hallway that leads to the outdoor smoking shelter.
First, create an Event Flow Diagram (see Final Event Flow Diagramming) to clearly define the events which took place:
Patient is wearing hospital issued slippers | → | Patient leaves bed to smoke a cigarette | → | Patient walks outside through dimly lit hall | → | Patient slips, loses balance, and falls | → | Patient fractures his hip |
Second, develop the problem statement based on the understanding of the event:
Fractured Hip (Problem Statement)
Hint: Teams may have difficulty in agreeing on a problem statement -- it should reflect what you are trying to avoid. Keep it simple (noun-verb if possible).
Step 1: Hints
- Effective problem resolution begins with agreeing upon a definition of the problem to be fixed.
- By forcing an explicit discussion of the problem, consensus is reached at the outset regarding what we are trying to modify, or correct, and the whole team understands why they have been assembled.
- Begin the cause and effect diagram after you have collected your information from interviews and research and have developed a chronological flow diagram. Completing this advance work gives the whole team a basic knowledge of what occurred, and will help make for a successful cause and effect diagram.
Causes
Brainstorm Primary Causes
For the second step, have the leader prompt the group by asking caused by statements repeatedly. In this example he would ask, The fractured hip was caused by
Hints:
- Have leader repeatedly ask the caused by question
- Use post-it notes to record suggested causes and evaluate afterwards
- When searching for possible causes remember the rules of brainstorming:
- Dont judge suggestions
- Aim for quantity
- Get everyone involved
When all suggestions have been exhausted, review the list of causes and choose only those that are most relevant to the event. These primary causes should be so directly relevant that, if any were removed, the event most likely would not have occurred.
Remember to have at least one action and several condition statements.
*Deleted; do not lend themselves to corrective actions
Definitions:
- Action
- Momentary and fleeting (e.g., stepping on wet floor, turning on machine)
- Condition
- Exists over time (e.g., temperature of room, competency of staff, waxed floor)
Hints:
- Some groups become concerned that a particular causal branch could fit a couple of different areas --- relax! Put it in either place.
- When should you stop building a causal branch for your diagram? Keep working until you run out of information or decide there is no merit in further review.
Step 2: Hints
- This activity may at first feel contrived as the leader asks the caused by question --- However, it is an effective prompt.
- You may find that the group is initially not clear on the distinction between action and conditions --- revisit the definitions for them.
- The group may argue about what is the best choice for the specific action --- dont get sidetracked by this detail.
- Remember the goal is to explore several possible causes to generate better solutions.
- There is no absolute rule about the correct number of conditions. Logistically, working with three to five seems to be effective.
Diagram
Complete Causal Chain
Now that you have identified an action and conditions, the next step is to complete the causal chain. Take each of these one at a time and have the leader ask the group caused by until you have completed a particular branch of the diagram. Use post-it notes to capture responses and run them out horizontally. For example, consider the action, fell in hallway, which directly preceded the fractured hip
- Facilitator asks group Fell in hallway caused by
- Capture the responses by the group. Here the response is going outside for a smoke.
- Facilitator asks group going outside for a smoke caused by
- Capture responses
Once the group has exhausted a particular leg of the diagram they move to the next primary cause.
Each of the primary causes is addressed resulting in the following completed
Cause and Effect Diagram
Root cause/contributing factor statements can be taken from the cause and effect diagram.
Next Section: Causal Statements
Statements
Cause & Effect Diagramming Causal Statements
Once the cause and effect diagram has been filled out, the team needs to develop root cause/contributing factor statements. Based upon our example, we have developed the following:
- Due to the VA smoke free policy and nicotine addiction, our patients will use exterior smoking shelters; this increases the risk of falling and injury.
- Due to the age, lack of exercise, and comorbidities, patients are likely to have a compromised sense of balance, increasing the risk of falls.
- Due to the shortage of resources, there is reduced staff; this increases the likelihood of not recognizing and correcting maintenance issues, resulting in a fall risk due to inadequate lighting.
- The purchase of slippers intended for a non-ambulatory population created an increased risk of slipping hazards by patients walking with these slippers.
Root Cause/Contributing Factor Statements
After you have developed an understanding of what occurred (event flow diagram) and why it occurred (cause and effect diagram) the next step is developing root cause/contributing factor statements.
Clear and specific root cause and contributing factor statements will lead the team to a more accurate depiction of the events with a focus on system-level vulnerabilities. These, in turn, will prompt the development of better actions and outcome measures.
Do root cause/contributing factors meet the following criteria:
- If corrected, reduces likelihood of adverse event.
- Meets all five of the rules of causation (see the following page for explanation).
of Causation
The Five Rules of Causation
Rule 1: | Clearly show the cause and effect relationship. | |
---|---|---|
(i.e. If you eliminate or control this root cause/contributing factor will you prevent or minimize future events?) | ||
Wrong: | A resident was fatigued. | |
Correct: | Residents are routinely scheduled for 80 hour work weeks; as a result, the fatigued residents are more likely to misread instructions, which led to an incorrect tube insertion. |
Rule 2: | Use specific and accurate descriptors for what occurred, rather than negative and vague words. | |
---|---|---|
(i.e. Avoid words such as poorly, inadequately, haphazardly, improperly, carelessness, complacently) | ||
Wrong: | Poorly written manual | |
Correct: | The training manual was not indexed, used a font that was difficult to read, and did not include any technical illustrations; as a result the manual was rarely used and did not improve performance by the equipment operators. |
Rule 3: | Identify the preceding cause(s), not the human error. | |
---|---|---|
Wrong: | The resident manager made a dosage error. | |
Correct: | Due to no automated software to check the dosage limits and no cognitive aids on dosing, there was a likelihood of this dosing error, which resulted in three times the appropriate level of insulin being ordered and administered. |
Rule 4: | Identify the preceding cause(s) of procedure violations. | |
---|---|---|
Wrong: | The techs did not follow the procedure for CT scans. | |
Correct: | Noise and confusion in the prep area and production pressures to quickly complete CT scans increased the probability of missing steps in the CT scan protocol; this resulted in an air embolism by inadvertently using an empty syringe. |
Rule 5: | Failure to act is only causal when there is a pre-existing duty to act. | |
---|---|---|
Wrong: | The nurse did not check the STAT orders every half hour. | |
Correct: | The absence of an established procedure for nurses to check the STAT orders on the printer created the vulnerability that urgent orders would not be administered; this resulted in the BOLUS of antibiotics not being administered. |
Next Section: Actions & Outcomes
Outcomes
Actions & Outcomes
Actions
Actions are the critical component of the RCA -- and present challenges for the teams. Strong and well-crafted actions have a clear link to the vulnerabilities and are readily understood. The table below presents some categories and types of actions that might be considered. Stronger actions are viewed as those that are more likely to be successful in accomplishing the desired changes, rendering greater utility for the effort expended. Note: you may need multiple actions to address a single root cause/contributing factor.
Do actions meet the following criteria:
- Address the root cause/contributing factor
- Are specific and concrete
- A cold reader understands and can implement
- Will be tested or simulated prior to full implementation (when feasible)
- Process owners were consulted
Recommended Hierarchy of actions
Stronger actions
- Architectural/physical plant changes
- New device with usability testing before purchasing
- Engineering control or interlock (forcing functions)
- Simplify the process and remove unnecessary steps
- Standardize on equipment or process or caremaps
- Tangible involvement and action by leadership in support of patient safety
Intermediate Actions
- Increase in staffing/decrease in workload
- Software enhancements/modifications
- Eliminate/reduce distractions (sterile medical environment)
- Checklist/cognitive aid
- Eliminate look and sound alikes
- Read back
- Enhanced documentation/communication
- Redundancy
Weaker Actions
- Double checks
- Warnings and labels
- New procedure/memorandum/policy
- Training
- Additional study/analysis
Outcomes
Outcome measures provide us a confirmation that what we hoped to accomplish did in fact occur. A well-designed measure will document the effectiveness of our change, and serve as a powerful bargaining chip. With this information, those that are responsible for process changes can be shown that they have made a difference. Likewise, managers and leaders that have invested resources in patient safety can be assured that this is a prudent use with demonstrable impact.
Do outcome measures meet the following criteria:
- Measures effectiveness of the action not completion of the action (e.g. measure that falls assessment occurs for x% of new patients admitted, NOT measure the training of staff on falls assessment)
- Quantifiable with defined numerator and denominator (if appropriate)
- Define sampling strategy and the timeframe for the measurement (e.g. random sampling of 15 charts per quarter)
- Set realistic performance threshold (e.g. dont say 100% compliance unless it will be met)