Tuesday, March 1, 2011

A Japanese Starbucks Model for Healthcare Quality

I'm flying back from Japan to the US today and on the way I purchased a vegan (soy) Green Tea Latte from Starbucks.   I relish the Japanese version of this beverage because it is unsweetened (even a bit bitter), since the Japanese typically do not like sugary foods and drinks.   As I ordered the beverage, the clerk behind the counter handed me a special quality control token to ensure I would not receive any dairy products.

Although my choice of soy is related to a lifestyle choice and not a medical condition such as a milk allergy, Starbucks in Japan wants to ensure "zero defects" in the delivery of soy-based drinks to eliminate accidental allergic reactions.

When my beverage was ready, I handed the soy "safety token" (pictured above) to the Starbucks barista who then double checked my receipt to identify that the beverage ordered was the beverage delivered.

Imagine how this Six Sigma/LEAN approach could be applied in hospitals.   Upon admission, patients could be given a token (a bar coded wrist band, a laminated card, an upload to their smartphone etc) that is handed to every caregiver before a medication is administered.   The patient and caregiver verify accuracy together so that right drug is given to the right patient at the right time.

I realize that many hospitals including BIDMC have implemented or are planning bedside medication verification (bar code the patient, bar code the medication, bar code the caregiver and scan/scan/scan before a medication is given) and electronic medication administration records (close the loop between orders and actual administration by charting medications administered electronically), so we're headed in the right direction.  Meaningful Use Stage 2 and 3 will likely include several new medication safety workflow requirements.

However, in the meantime, if Starbucks can use laminated cards to prevent allergies and ensure zero defects in my Green Latte, hospitals should feel inspired to examine their medication workflows and think about simple solutions to do the same for our care delivery processes.


Mike LaChapelle said...

Great example, John. I used to travel with a colleague who was diabetic. He would order a diet soda and then use a test strip to determine if it was really diet or not. I was regular soda a significant percentage of the time.

Sunny said...

That is the idea as a newborn getting a tag in the feet with parent's name and everytime you check in and check out the baby they verify it with the tag of the parent. Also for cardiac patients who are also diabetic, there is a special wrist band that the food delivery guy checks when dispensing meals. I have seen these in hospitals everywhere...

I have a feeling Starbucks borrowed the idea from hosptials rather than the other way round. :)

Suki Tsui said...

RFID & RTLS are huge at HIMSS this year. There are even showcase, symposium and education session. It is definitely one of the hot topics at the conference.

However, HIE (and interoperability) still has my vote as the hottest topic. :)

Mark Neuenschwander said...

Thanks John for your good work. Sure hope bar coding makes it into MU 2 and 3.

geekgoalie said...

These sorts of process controls are an improvement. What sorts of engineering controls could we put in place to further reduce the potential for error.

Pie said...

I think that this is a matter of making it more widespread. When I was in the hospital recently for the birth of child #3, reading the numbers on my wife's wristband and matching it against our daughter's ankle band was standard.

There were no drug allergies involved, so no other checks for that system. Sounds like we need to just make this more widespread.


Patrick Pichette said...

Great post John, I love this type of progressive thinking, and how small ideas can make a significant impact (good job Starbucks). In Canada, our government has spent billions on a national EHR project over the last decade, which has yet to be launched. Yet, less than 20% of physician offices use an EMR (difficult to feed an EHR if GPs continue to use paper charts). Lets rapidly solve a series of small issues before trying to launch a national EHR...which is putting the cart before the horse.

Mark Graban said...

Wow, great discovery. Thanks for sharing.

With a process-based mistake proofing method like this, I think it's interesting to think how the process could fail or be intentionally circumvented. Let's say where was a language barrier and you didn't know what that token was for and you discard it? Could baristas somehow intentionally circumvent this to suboptimize their own efficiency?

Any discussion of hospital technology, like barcoding, should include a discussion like that. Can clinicians circumvent the bar coding by printing extra wristbands, etc.? Why are they being pressured into cutting corners (ie why is there too much waste that's taking up their time and creating pressure?

I know one hospital where I.S. built in a "check digit" that was only printed at certain authorized bar code printers, this prevented the nurses from being able to use any extra bar codes that might have been printed.

Final thought is that I'm not blaming nurses or others for doing this - we need to look at the whole system to understand why it is happening.