Five whys vs fishbone vs fault tree: how to choose a root cause method
Short answer
Use five whys for a narrow problem with one likely chain of causes. Use a fishbone to brainstorm many causes with a team. Use a fault tree when a serious failure can happen several ways. None works if the analysis stops at human error.
Match the method to the shape of the problem
A simple, single-path problem suits five whys. A messy problem with many possible causes suits a fishbone session. A rare, serious failure with several routes suits a fault tree. Many real investigations use two methods in turn.
| Method | Best for | Strength | Limit |
|---|---|---|---|
| Five whys | A narrow problem with one likely chain of causes | Fast, needs no special skill, pushes past the first answer | Follows one path. Easy to stop early. |
| Fishbone (cause and effect) diagram | Early group brainstorming on a problem with many possible causes | Every role contributes; causes group under headings | Lists possible causes, not proven ones. No sequence of events. |
| Fault tree | A serious, often equipment-heavy failure with several ways to happen | Starts from the unwanted event; shows which causes are enough alone (OR) and which must combine (AND) | Needs a well-defined top event, time and skill. Fits equipment better than communication problems. |
Neither the Joint Commission nor CMS requires a tool. The Joint Commission's sentinel event policy says an organization can choose its own process, tools and methods. Both ask for analysis that reaches system causes.
Five whys works for narrow problems
State the problem. Ask why it happened. Then ask why about each answer until you reach something you can change.
Five is a prompt, not a rule. The Joint Commission expects analysis to keep asking why until the systemic causes behind each step are found. IHI says keep asking until the cause is specific enough to test a change and measure its effect.
A 2017 BMJ Quality & Safety article, summarized by AHRQ's PSNet, calls five whys a useful teaching tool. It argues that using it for healthcare root cause analysis may be misguided, because it simplifies complex problems.
Use it for simple problems, or as one step in a fuller analysis. See five whys.
A fishbone puts many causes on the wall early
A fishbone diagram helps early, when a team needs many possible causes on the wall before anyone settles on one. It is also called a cause and effect or Ishikawa diagram.
IHI's toolkit gives five steps. Write the effect in a box at the right. Draw a spine to the left. Choose five or six categories and hang them off the spine as bones. List causes on branch bones and sub-causes below.
Standard categories are Materials, Methods, Equipment, Environment and People.
Its strength is breadth. A pharmacist, an aide and a biomedical technician each add causes the others missed. Its limit: every cause is a hypothesis until checked against records, observation or interviews. See fishbone diagram.
A fault tree shows how failures combine
A fault tree suits serious failures, especially in equipment and engineered processes. It starts with the unwanted event at the top and works down, joining causes with AND and OR gates.
An OR gate means any one cause is enough. An AND gate means every listed cause must be present together. The other two methods do not show that, and it points to the cheapest place to break the chain.
The U.S. Nuclear Regulatory Commission documented the method in its 1981 Fault Tree Handbook, NUREG-0492.
It costs time and skill, and the top event must be precise. Save it for high-consequence events, such as a sterilization failure or a power loss during a procedure, or for a repeat event simpler tools have not explained. See fault tree analysis.
A worked example: twice the morphine dose
This invented surgery center example shows how five whys digs past a first answer. A recovery patient gets twice the ordered morphine dose and needs extra monitoring and a reversal agent. The first-pass answer is "the nurse picked the wrong vial."
- Why did the patient get twice the dose?The syringe came from a 4 mg/mL vial. The order assumed the 2 mg/mL vial.
- Why was the 4 mg/mL vial within reach?Both concentrations sat in one drawer, in boxes with near-identical labels.
- Why were both stocked together?The 4 mg/mL vial was added when the supplier changed. It went into the existing slot.
- Why was the layout never reviewed?The supply change process has no storage check for a new concentration.
- Why is there no such step?The process covers purchasing and billing codes, not look-alike storage. That is the root cause to fix.
A fishbone on the same event would put look-alike labels under Materials, restocking into the old slot under Methods, no barcode check at the draw station under Equipment, shift-change noise under Environment, and float staff unfamiliar with the stock under People.
A fault tree would put "twice the dose" at the top, with an AND gate: wrong concentration drawn AND nothing catches it.
| Possible action | Strength | Why |
|---|---|---|
| Remind staff to read the label twice | Weaker | Relies on memory. |
| Add a second-nurse double check | Weaker | Depends on someone noticing every time. |
| Stock only one morphine concentration in recovery | Stronger | Removes this look-alike mix-up. |
| Add a storage review to the supply change process | Closes the root cause | Stops the next look-alike arriving unnoticed. Needs an owner and an effectiveness check. |
Treat "human error" as a question, not an answer
Every person who made a mistake worked inside a system that made it possible. Find the conditions and fix those. The VA's guide gives five rules for cause statements. The Joint Commission points to the same guidelines.
- Show cause and effect between each statement and the event.
- Use specific, accurate words, not vague or negative ones.
- A human error needs a preceding cause.
- A procedure violation is not a root cause and needs a preceding cause.
- Failure to act is causal only if there was a duty to act.
Regulators ask the same thing. CMS tells ASC surveyors to ask whether the center stops at the immediate cause, such as staff error, or probes for root causes.
OSHA's recommended practices add that a good investigation asks whether the worker had the right tools and time, enough training and proper supervision.
Where these methods fit in IncidentKit
IncidentKit's investigations record contributing factors, five whys and a disposition. Lauren drafts the investigation from the report, and a person reviews, edits and signs. Human-authored RCA templates are rolling out.
Run a fishbone or fault tree with your team. Record the causes that survive scrutiny as contributing factors. Each can lead to a corrective action with an owner, due date, evidence and effectiveness check. See the root cause analysis and CAPA guide.
Frequently asked questions
Is five whys enough for a root cause analysis?
Usually not for a serious event, because it follows one chain. The Joint Commission asks for analysis that keeps asking why until systemic causes are found for each step. Use five whys inside a fuller analysis with a timeline, cause statements and an action plan with owners and measures.
What categories go on a fishbone diagram?
IHI's toolkit lists Materials, Methods, Equipment, Environment and People, and suggests five or six in all. Adjust them to the event. Make each cause specific enough that you could test a change and measure whether it worked.
When is a fault tree worth the effort?
For high-consequence or equipment-heavy failures where several things must combine, such as a sterilizer failure or a power loss during a procedure. If one clear chain explains the event, five whys is faster. If the cause is unclear, start with a timeline and a fishbone.
Does the Joint Commission or CMS require a specific method?
No. The Joint Commission says the organization can choose its own process, tools and methods, but the analysis should focus on systems. CMS guidance for ASCs names no method. It asks whether the center stops at the immediate cause or probes for root causes.
How many whys should I ask?
As many as it takes to reach a cause you can change at the system level. Five is a prompt, not a rule. Stop when the next why would only point at a person or something outside your control.
Sources
- VA National Center for Patient Safety: Guide to Performing a Root Cause Analysis (2021)
- IHI QI Essentials Toolkit: Cause and Effect Diagram
- AHRQ PSNet: The problem with “5 whys” (Card, BMJ Quality & Safety, 2017)
- U.S. NRC: NUREG-0492, Fault Tree Handbook (1981)
- The Joint Commission: Sentinel Event Policy (CAMH Update 1, July 2026)
- CMS State Operations Manual, Appendix L: ambulatory surgical centers
- OSHA 3886: Recommended Practices for Safety & Health Programs in Construction (incident investigations)
- AHRQ PSNet primer: Strategies and approaches for investigating patient safety events
Reviewed against the sources above on Oct 5, 2026. Rules change: confirm current requirements with the issuing body or your counsel before relying on any summary.
Start with one incident.
Create your kit in about ten minutes and report the first incident the same day. Free to start, no card.