Yet More Musings on Muda

mri.JPG
Michel Baudin’s blog article More Musings on Muda, meant to end fruitless debate on the definitions and categorizations of the types of waste, in fact led to some brief but interesting exchanges on Twitter. The discussion centered around the claim by lean healthcare practitioners that in healthcare there is a lot of grey area between overproduction or overprocessing. I disagree with this notion. Two questions to support this idea were:


Is an unneeded MRI overproduction or overprocessing?
What waste is drawing five tubes of blood when only one is needed?

The answers seemed straightforward to me. I will purposely not give my answer here, to encourage thought.
These may be a bit harder:


Performing a Cesarean section at physician discretion rather patient medical need
Writing a prescription for 12 days’ supply of medication when 9 days’ supply will suffice

With the notable exception of cataract surgeries performed in India, patient care fundamentally does not resemble a production process; it is a service process. Any production steps are ancillary to care delivery. For example the preparation of medicines is production, but the process of care delivery involves diagnosing, prescribing, and having the patient follow a regimen of taking these meds. If too many meds were prepared, that would be overproduction. If unnecessary meds were prescribed, that would be a waste of processing. The excess in the former case can be reduced to match the patient needs, while the latter can be eliminated totally.
Masaaki Imai explains the 7 types of waste in his book Gemba Kaizen:

Muda of overproduction is a function of the mentality of the area supervisor, who is worried about such problems as machine failures, rejects, and absenteeism and who feels compelled to produce more than necessary just to be on the safe side. This type of muda results from getting ahead of the production schedule. When an expensive machine is involved, the requirement for the number of products is often disregard ed in favor of efficient utilization of the machine.

A good test of whether something is overproduction or overprocessing is to question the reason why the waste exists. Overproduction is often a conscious act of producing more than necessary in order to hedge against future failures, maximize the output or utilization of a resource, to keep busy when there is no other work to be done, or just because you had a pot that served 6 when you were making dinner for 4 people.
The underlying causes of the waste of overprocessing, a.k.a. the waste of processing or the waste of “processing itself” as it is called in works attributed to Taiichi Ohno, are somewhat more varied and less obvious:

Sometimes inadequate technology or design leads to muda in the processing work itself. An unduly long approach or over run for machine processing, unproductive striking of the press, and deburring are all examples of processing muda that can be avoided. At every step in which a workpiece or piece of information is worked on, value is added and sent to the next process. Processing here refers to modifying such a workpiece or piece of information. Elimination of muda in processing frequently can be achieved with a commonsense, low -cost technique. Some wasteful processing can be avoid ed by combining operations. For instance, at a plant where telephones are produced, the receiver and the body are assembled on separate lines and later put together on the assembly line. To protect the surfaces of the receivers from scratches as they are being transported to the final assembly line, each receiver is wrapped in a plastic bag. By connecting the receiver assembly line and the final assembly line, how ever, the company can eliminate the plastic-wrapping operation.

Processing waste is performed because of vague standards or poor understanding of the needs of internal or external customer. Examples from my experience include:
Painting the interior of a body panel: The painted body panel was a pair with a hatch on the other side of the vehicle, which was opened and visible, needing to look good. However the body panel side had no need for the paint, only the hatch. The painting step was removed, along with various inspection and touch up processes that guaranteed the quality of this unnecessary process.
Cleaning a table top: Sawdust and grime was removed from the part at every step along the batch process. Linking the process in a flow eliminated the need to clean between steps because the part did not rest for long enough to accumulate sawdust or grime.
Another way to keep overproduction and processing clear is to understand whether the waste being observed is an input or an output. If the activity is an input, it cannot be overproduction. Likewise, if the activity is an output it not the waste of processing.
The types of waste which are observable as inputs are:

  • Motion – moving arms, eyes, torso as part of the process
  • Transportation / Conveyance – moving things using legs, wheels or conveyors
  • Inventory – raw materials or in-process goods waiting to be consumed at the process
  • Correction / rework – this is a close relative of both Processing and Defect wastes, typically bundled into the latter
  • Waiting – people not able to work because something is missing or wrong
  • Processing – any process, containing some, all or none of the above, that is itself not necessary

The types of waste which are observable as outputs are:

  • Overproduction – excess output of a process
  • Defects – nonconforming product or service
  • Inventory – work in process, finished goods

I believe Mr. Timothy U. Wood bears some responsibility for the confusion between overprocessing and overproduction. The “over” that Tim attached to the waste of processing itself is itself a waste of processing. The four-letter word adds no value, does not further clarify definition and in fact causes people to associate it with overproduction. It is, as we sometimes say, overly redundant.
When observing waste, we need to ask three questions:

  1. What does it cost?
  2. What causes it?
  3. What are the countermeasures?

The costs, causes and countermeasures to the various wastes tend to be rather specific to the type of waste. Remembering the names, or even that there are 7 types of waste, is far less important than being able to identify them all accurately. If there is in fact an 8th or 9th type of waste to unique to healthcare that lies in the grey area between processing and overproduction, by all means let’s capture and categorize this muda so we can all benefit from a quicker and more accurate of diagnosis.

2 Comments

  1. Lech Ambrzykowski

    April 1, 2012 - 11:47 pm

    Thanks for the food for thought.
    I have one issue with the idea of discussing waste in medical services – with production, safety and security should be embedded in the design of the product per se (we can meddle with the production process to some extent), with medical services, however, medical staff sometimes adds cost / time for additional procedures to decrease risk, increase *likelihood* of a better treatment. In a given case. Even with mature evidence-based medicine I assume this will make it a crucial difference compared to production or even other types of services.
    When glancing through your post I realized we start with medicine, but soon enough the examples draw closer to manufacturing. Makes me wonder about the general reference points (or even role models) we choose.

  2. Nick Jenkins

    April 2, 2012 - 5:32 pm

    I think the experimental/scientific methodology is well embedded in medical thinking so the answer should be obvious – if the odds of a beneficial outcome increase with an extra MRI or drawing five tubes of blood, then it’s not waste, because (on average) it’s producing a better outcome.
    Waste and value are related to the outcome – the customer’s in production and the patient’s in healthcare. The problem you have in healthcare is that (assumption here) the variability is greater than in production so you need a larger, historical data set to demonstrate the value (or waste) inherent in a process step.
    Eliminating that step then becomes an experiment at the gemba with recorded outcomes and conclusions.