Five Whys

There are four similar but important tools that organizations, including hospitals, use to learn from failure and improve their key processes and procedures following accidents and near-misses (I've talked about these in two previous posts, "The Failure of Foresight" and "Hotwash") - they are Root Cause Analysis, Apparent Cause Analysis, "After Action Review", and "Hotwash".  

One of the techniques that is common to all four of these techniques is the "Five whys technique" developed in the 1930's by Sakichi Toyoda, the founder of Toyota.  It's relatively straightforward and simple to use.  When a problem occurs, you can drill down to the root cause by asking "Why?" at least five times.  The advantage of asking "Why?" at least five times is that we avoid making incorrect assumptions based on past biases and other traps.  

The "Five whys technique" has increased in popularity in health care in the last several years, and it is now recommended by the World Health Organization, the National Health Service in England, the Institute for Healthcare Improvement, the Joint Commission, and a number of other organizations focused on health care quality and safety.  According to Taiichi Ohno (whose Toyota Production System inspired Lean manufacturing), "The basis of Toyota's scientific approach is to ask why five times whenever we find a problem … By repeating why five times, the nature of the problem as well as its solution becomes clear. The solution, or the how-to, is designated as ‘1H.’ Thus, ‘Five whys equal one how’ (5W=1H)."

Here is one common example (from Wikipedia):

Problem: Our vehicle doesn't start in the morning for work.  

First Why?  The battery is dead.  

Second Why? The alternator is not functioning.  

Third Why? The alternator belt has broken.  

Fourth Why? The alternator belt was beyond its useful service life and wasn't replaced.  

Fifth Why?  We did not follow the manufacturer's recommended service schedule.

  

Root cause: We did not follow the manufacturer's recommended service schedule.

Solution: Follow the manufacturer's recommended service schedule.

By asking "Why?" five times, we didn't stop our analysis too early.  In this case, if we would have stopped after the second "Why?", we would have just taken the car in to replace the alternator, when the better long-term solution was to follow the manufacturer's recommended preventive maintenance schedule, which would have prevented the problem in the first place.

There's a great article in the journal BMJ Quality and Safety that calls into question the "Five whys technique" (see "The Problem with 5 Why's" by Alan Card).  Card talks about one of the most famous case studies of the "Five whys technique" which focused on the problem of the Washington Monument slowly deteriorating.  I've seen a few different versions (some involving the Washington Monument, others involving the Lincoln Monument), and I suspect that the story isn't completely true.  However, it's still interesting to review and illustrates a potential problem with the "Five whys technique":

Problem: Washington Monument is deteriorating.

First Why? Harsh chemicals are being used to clean the monument.

Second Why? The monument is covered in pigeon droppings.

Third Why? Pigeons are attracted by the large number of spiders on the monument.

Fourth Why? Spiders are attracted by the large number of midges on the monument.

Fifth Why? Midges are attracted by the fact that the monument is lit up at night.

Root cause: Lighting up the monument at night attracts midges.

Solution: Stop lighting up the monument at night.

As you hopefully can see, in this particular case, the "Five whys technique" oversimplifies the problem, and the solution is neither realistic or perhaps appropriate.  Now, I would first state that the "Five whys technique" is just a tool, and tools are only good if they are applied correctly and in the right situation.  Card argues that it shouldn't be used at all because it forces us into a direct cause-and-effect relationship to a single root cause, which may not be appropriate (and certainly wasn't appropriate in the Washington Monument case).  I think that is perhaps overstated, but it's clear that you have to use all of the tools in the proverbial quality improvement toolbox.  I wouldn't abandon the "Five whys technique" just yet.

I think part of the issue is also in how the problem is stated.  We need content and/or subject matter experts to help us and make sure that we are solving the correct problem.  For example, let's look at the famous case of the RMS Titanic, which sank in the North Atlantic after striking an iceberg on April 15, 1912.  Two different problem statements result in two different solutions - let's take a look:

Problem: The Titanic sank.

First Why? It ran into an iceberg.

Second Why? The crew didn't see the iceberg in time.

Third Why? The lookouts didn't have binoculars.

Fourth Why? One of the ship's officers left the shift in Liverpool with the key to where the binoculars were stored.

Fifth Why? He forgot.

Root cause: The ship officer forgot to unlock the binocular locker prior to leaving the ship in port.

Solution: All officers should follow a checklist prior to leaving the ship.

I am not making this story up (for more details, see the Wikipedia post on David Blair) - the fact that the lookouts didn't have binoculars is one commonly cited reason that contributed to the loss of the Titanic.  But would anyone be satisfied with the proposed solution here?  Let's re-phrase our problem statement in a different way and see if our solution is better:

Problem: 1,500 people died when the Titanic sank.

First Why? Not enough passengers got into lifeboats.

Second Why? There weren't enough lifeboats.

Third Why? White Star Line chose to meet only the minimum requirements (at the time) for the number of lifeboats.

Fourth Why? Too many lifeboats clutter the decks.

Fifth Why? Lifeboats are large and unwieldy.

Root cause: Lifeboats are large and unwieldy.

Solution: Increase the number of collapsible lifeboats.

Here we are attacking a different problem and proposing a completely different solution.  But again, would anyone be satisfied with the solution proposed this second time?  Probably not.  I would argue that if you gathered a multidisciplinary team of subject matter experts, you would likely come up with more than one root cause and, hence, more than one solution.  Incidentally, the BMJ Quality and Safety journal published another great article ("The problem with root cause analysis") on some of the drawbacks to the Root Cause Analysis technique, one of which trying to focus on just one single root cause.

In summary, I want to conclude by emphatically stating that I do not think that these techniques should be completely abandoned.  Both the "Five whys technique" and Root Cause Analysis technique have not only helpful in my experience.  We just need to be careful about placing all of our eggs in one basket and apply the right technique to the right context and at the right time.