May 29, 2017

Work automation without comprehension

Lara Goiten in a review of the book Let me heal: The Opportunity to Preserve Excellence in American Medicine, authored by Kenneth Ludmerer, writes:

Moreover, my receptiveness to teaching by senior physicians plummeted. There was very little time for such instruction, and when it did occur I was unable to concentrate because I was constantly reviewing in my mind the work that needed to be done, or being called away to tend to my many patients’ sudden needs. It seemed an unaffordable luxury to learn about the mechanisms or finer points of disease. Whereas previous generations had emphasized searching for and learning from what was new or unusual—what stretched and challenged current understanding—now training emphasized classification of patients into categories and algorithms in order to cope safely with the pace.

For example, many teaching hospitals provide their residents with “protocols” (often in the form of flow diagrams, or prewritten orders for tests and treatments) for common problems such as chest pain, pneumonia, heart failure, and stroke. While protocols may make residents more efficient and provide a basic safety check, they also devalue innovation and individual initiative, and discourage thoughtful consideration of unusual or unique features of individual patients. As Ludmerer points out, while standardization may impose a floor on performance, it may also impose a ceiling.

With so little time to think about patients, we would order batteries of tests roughly corresponding to whatever anatomic area was brought to our attention, sometimes before we’d even seen the patient. This was the only way to make sure (we hoped) that we wouldn’t “miss anything.” The tendency to overtest began as a survival technique, but by the end of residency it was ingrained as a style of practice—and this excessive use of tests is one driver of health costs.

Lara Goiten, Training Young Doctors: The Current Crisis, June, 2015

Dr. Wolpaw, a resident physician writes:

It is easy for residents like me to take care of patients with diabetic ketoacidosis. All we have to do is type “DKA” into the order entry system and choose the preassembled order set. This will instruct the nurse to begin a continuous infusion of insulin, specify how and when to titrate this insulin based on serial measurements of blood sugar, and provide algorithmic protocols for the repletion of fluids and electrolytes. There are multiple phases, dictating a series of coordinated changes that occur as the blood sugar lowers and the body recovers.

Order sets are best suited for conditions like DKA where the diagnosis is known and the treatment is standardized. When used properly, they can increase efficiency and decrease the chance of oversight or error. But as this type of automation proliferates, inevitably physicians will become dependent on it. I personally do not always know the detailed steps of the care I deliver, placing my signature on orders that I may not completely understand, trusting that the system I work in is appropriately designed and updated.

This is a concept that is inherently disturbing to many physicians, and I would expect for many patients as well. What if the computer system goes down or is hacked? What if a doctor who was trained in a technologically advanced environment goes to work in a setting with significantly less electronic support? Shouldn’t doctors know, inside and out, what they are doing?

Benjamin Wolpaw, Learning in the era of automation, Academic Medicine: May 2017 - Volume 92 - Issue 5 - p 578

Dr. Wolpaw has gained insight, as pointed out by Goiten and Ludmerer, that doing the work doesn't imply having an understanding or comprehension of what's being done. Understanding and comprehension are processes not directly observable, but they take a lot of effort for complex knowledge such as the practice of medicine.

No comments:

Post a Comment

1. You should attempt to re-express your target’s position so clearly, vividly, and fairly that your target says, “Thanks, I wish I’d thought of putting it that way.
2. You should list any points of agreement (especially if they are not matters of general or widespread agreement).
3. You should mention anything you have learned from your target.
4. Only then are you permitted to say so much as a word of rebuttal or criticism.
Daniel Dennett, Intuition pumps and other tools for thinking.

Valid criticism is doing you a favor. - Carl Sagan