Sunday, July 06, 2008

Mourning becomes Electra

Updated: 28 Jan. 2017 (Fixed broken links)
This blog's title is taken from the play by Eugene O'Neill of the same name, which is an adaptation of the Greek trilogy “Oresteia” by Aeschylus.

The blog was motivated by a device
that automates blood component preparation just licensed by the FDA in the USA. The device's press release is one of the Industry News items in this month's TraQ newsletter. It caught my eye because of the name of the device, the Atreus.

After briefly discussing the names that manufacturers give their products, the blog will examine the pitches used to market automation, the related concept of lean manufacturing, and whether they constitute progress.

First, the product names. Atreus rang a bell because during my youth my Dad used to share his love of Greek mythology with me. Atreus was, to put it mildly, a pretty repellent dude even in a curse-stricken family where each generation gruesomely and tragically murdered family members and others out of some combination of pride, ambition, lust, and revenge.

One version of the Atreus myth (there are several) has Atreus and his brother murdering their father's illegitimate son (their half-brother) to please their mother. Then, after his brother seduces Atreus's wife, Atreus kills his brother's two sons and invites him to a banquet to feed him the flesh of his sons, whom Atreus had boiled. Atreus also saves the children's hands and feet to show to his brother. His brother later rapes his own daughter, who despite being Atreus's niece, becomes Atreus's wife. Atreus raises the child fathered by his brother as his own. Eventually Atreus's wife commits suicide and the child, in loyalty to his uncle (who he realizes is his real father), kills Atreus. Charming family.

Now why would Gambro / CaridianBCT want to name its new device after such a repugnant figure as Atreus? Other products have been named after characters in Greek mythology, for example, Immucor's Echo. In Greek mythology, the beautiful nymph Echo would distract Zeus's wife Hera with stories while Zeus ravished other mountain nymphs. When Hera discovered this, she punished Echo by taking away her voice, except in repetition of another's words. Later Echo fell in love with Narcissus but was rejected. The once talkative nymph eventually changed into rock leaving nothing but her voice to reply to those who shouted out.

It's likely Echo is meant to be a cute takeoff on its role as the little brother (little sister?) of Immucor's Galileo. Galileo contended that the earth rotated around the sun, thus contradicting the orthodoxy of the time (early 17th century) that the earth was the centre of the universe. For this crime he was convicted of heresy and spent his later years under house arrest. Naming products after pioneering scientific figures like Galileo makes sense, but Atreus and Echo....not so much.


Back to the Atreus® and other automated devices and instruments....

Is the Atreus® meant to eliminate "x" FTE staff who currently prepare blood components or just make their work easier and more satisfying?

Manufacturer's typically market automation with claims such as:The instrument/device will
  • do the boring "slug work," freeing staff for more interesting tasks that require more skill
  • prevent repetitive stress injuries
  • increase patient safety by decreasing errors since there are fewer steps for staff to perform
  • increase efficiency by saving time (better TAT with throughputs of "x" tests/hr)
  • increase test reliability by eliminating subjective reading of tests by staff
  • increase test sensitivity and specificity (hmm...can you really do both?)
  • improve process control
  • interface with existing LIS to decrease transcription errors (Oh, oh! Best to take LIS interface claims with a block of salt)
  • ensure positive identification and tracing via improved 2-dimensional bar codes
Notice how some of the standard pitches above emphasize putting staff first - more satisfying work, fewer injuries, easier work (instruments and computers do the "slug work" for you).

To lab managers and directors, marketing pitches include the motherhood issues above but tend to focus more on the bottom line, e.g., The instrument/device will/can

  • save money by eliminating "x" FTE
  • give a return on investment after "x" years
  • be maintained by a rhesus monkey available from our offshore supplier at a discounted price if ordered in volume (Just kidding on this one)
Effects of automation in lab areas such as clinical chemistry have been well documented:
This article in CAP Today sums up the transfusion medicine situation in 2002, as well as the motivations and perspectives of those in charge of the labs:
Naturally, automation fits well with the latest buzzword of the quality movement, "lean," as in "lean manufacturing." Lean is based on the principle that non-value-added activity constitutes waste and should be eliminated.

Like its cousins quality systems and six sigma, there is an entire industry built around lean. It helps to develop a business around a concept if you can give insiders the jargon that makes them "in the know" and keeps the non-enlightened out.
And lean is now huge in health care. Here's a lean hospital in the UK - for patients arriving in the emergency department, no more seeing doctors - docs do not add value to the process:
Hospitals, laboratories, and blood centres send staff to lean conferences to learn the principles and jargon of lean (muda, kaizan blitz, kanban, JIT, poka-yoke, etc.). Seems that everyone is making a buck from lean.
Try googling lean AND health care - on 6 July 2008 the search gave >7 million hits. Most seemed to be websites for consultants selling lean advice.
So is automation progress? Is lean progress? Progress can be defined many ways but it generally means improvement or growth, whether for individuals, organizations, societies, or humanity.

Are highly automated or lean labs progress for

  • staff who no longer have lab jobs? Have they gone on to bigger and more rewarding careers and lives?
  • remaining staff who load the instruments and press the buttons, letting the equipment do the analyses. Is their job satisfaction improved?
  • senior staff who do less but more interesting hands-on lab work and have more time to read computer printouts, attend meetings, and program spreadsheets with error management data and the like.
  • lab managers and directors who can more easily meet their goals with fewer resources now that costs are down?
  • patients, who constitute the true bottom line? Is their health care improved and safer?
Many would say, of course, automation and lean constitute progress - it's the current conventional thinking (orthodoxy). Even questioning automation and lean is heresy. However, if questioning of automation and lean principles bugs you, you may be experiencing cognitive dissonance.

On the other hand, if you are certain that automation and lean are progress for mankind, read on because shangri-la approaches:

Add robotics to lean hospitals and soon we'll have gotten rid of all the non-value-added waste in the health system, as well as most of the health professionals.

I wonder if there is a way to automate administering transfusions as this seems to be a area where deadly errors continue to occur? FromTraQ's July newsletter - International news: UK and USA:

Transfusion nurses could have improved and more satisfying work by investigating only unusual events that actually require their skill set. (I write with tongue firmly planted in cheek.)

I would call the automated transfusion device Electra after Atreus's granddaughter, who wanted her brother to avenge the death of their father by killing their mother, leading to the concept of the Electra complex.

I'm not altogether sure whether automation constitutes progress from a big picture perspective (benefits humankind), but it's irrelevant. Automation is here to stay and will only get more and more pervasive in health care.

Take that, Atreus! The curse goes on....
Comments are most welcome BUT, due to excessive spam,  please e-mail me personally or use the address in the newsletter notice. 

 Addition (3 Sept.2008)
Thanks to those who have commented.

In these blogs I'm really just asking questions, not providing definitive answers. Readers should decide for themselves based on their experience and reading the literature. Responses may vary widely depending what automation has meant for individuals personally.

Few of us have extensively studied the issue of automation over time. And what studies have been done get their results based on the questions asked, using measurable outcomes like reduced number of FTE staff, money saved according to defined algorithms and assumptions, and self-reporting by remaining staff about job satisfaction, etc. The internal validity of some of these studies is open to question and their generalizabilty is not universal.

As noted, "Many would say, of course, automation and lean constitute progress - it's the current conventional thinking (orthodoxy). Even questioning automation and lean is heresy."

My view is not so certain:
"I'm not altogether sure whether automation constitutes progress from a big picture perspective (benefits humankind)." It may or it may not.

Taken to its logical conclusion, and using a reductio ad adsurdum argument, if automation constitutes progress (a better life) for humans, then we should automate everything, leaving a few of us to do the routine work (load instruments, push buttons, etc.), while the rest of us sit around and problem solve until there are no problems left to solve because the automation has been perfected. Now, what to so with all this leisure time? Create a new Renaissance? Die of boredom?

I've reduced the proposition to an absurd conclusion, which could be a fallacious argument, just to show where my abstract and theoretical thinking about automation leads.

About inadequate staffing causing errors, I'm not sure that's been proven in the cases cited by the authors of "Canada's pathology" in CMAJ. Although it's logical that being overworked could be a contributing factor in causing errors, it would be risky to assume that it's a significant cause in individual cases without evidence. Many of the pathology errors were caused by the repetitive actions of a single pathologist such as Dr. Charles Smith, whose autopsy decisions were not always based on the evidence.

Lab systems are supposed to be designed to prevent and detect and correct any errors. Clearly the overall system failed in the case of the serious pathology lab errors that occurred.