Friday, April 16, 2010

Health Information Technology Basics From Calif. Nurses Association and National Nurses Organizing Committee

I have never before seen a document like the one entitled "Health Information Technology Basics", by the Institute for Health & Socio-Economic Policy, California Nurses Association and the National Nurses Organizing Committee. It is available at this link (PDF).

It is long but contains rather interesting views on the issue of health IT, management and clinicians (nurses).

I find the following passages of particular interest as they reflect views we express on this blog, and attack the notion of cybernetic miracles being wrought:

Skill is the ability, drawn from education and experience, to do something expertly. It can also be defined as the effective exercise of professional judgment in non-routine situations.

Following prescribed rules, as a machine would, makes an employee competent to perform tasks, but it doesn’t make the employee skilled. They can do their job as long as there are no surprises. But when something unexpected happens, the rules break down, and caring for patients means facing the unexpected every day. Only skilled health professionals can cope with the unexpected. To know what to do, they have to rely on their own judgment. The exercise of judgment is the essence of skill.

... Much health information technology is skill-degrading. As the work of health professionals becomes increasingly automated, they lose the ability to do their jobs without HIT. To make matters worse, they’re expected to keep pace with machines. They serve the machines rather than doing the more gratifying work of patient care, and ultimately they’re compensated less well.

And this:

Displacement [pf people by machines] is hard to spot because it’s unlikely to appear as a one-to one correspondence; that is, you probably won’t find a robot sitting in your colleague’s chair tomorrow. It’s more likely to happen piecemeal, over an extended period, and through attrition.

  • The job of patient care will be redefined, privileging technical over clinical skills.
  • The hospital will begin to hire more HIT specialists and fewer RNs.
  • Functions performed in the past by health professionals will be fragmented and reallocated between machines and less-skilled employees.
  • Increased technical efficiency will enable the hospital or HMO to expand without expanding its workforce.

And this:

Use of any hospital technology must be consistent with safe, therapeutic, and effective patient care. Health information technology is a complete unknown in this regard. It’s an enormous social experiment designed by computer scientists and implemented by hospital administrators. HIT hasn’t grown organically from the needs of patients but has been imported from other industries. Known as enterprise resource planning, it’s adapted from similar technology designed to manage business operations on a massive scale and already being used to run the world’s largest corporations.

Caring for patients isn’t business. It requires compassion, judgment, and advocacy. Because RNs have the moral right and legal duty to advocate for patients, they have to be able to override the automated decision-making of HIT designed to serve business interests.

RNs have to work collectively to control health information technology rather than trying to fix it. It’s important to recognize that tinkering can’t fix HIT because its primary purpose is to mechanize, or routinize, patient care. It’s designed to quantify the unquantifiable, to replace the patient with an imaginary statistical norm. High-quality healthcare can’t be mechanized because it depends on people—on patients and caregivers—and people are infinitely more complex and capable than computers can ever be.

Amen to that.

Read the entire document for a union-oriented view of health IT. It contains many truisms regarding HIT irrational exuberance and the control issues of the healthcare management class.

-- SS


Anonymous said...

Eureka! This analytical report provides insight in to the January 2010 death at Harvard's flagship hospital of world class reputation.

It explains why the nurses and professional staff missed bradycardia for at least 20 minutes on a heart ward with the patient on a monitor as reported here:
Was the etiology really "alarm fatigue" or something that the report describes, or the thought disruption from the user unfriendly patient management computerization systems?

By the way, Dr. Blumenthal of ONCHIT czar hails from this hospital. Ever wonder what he has to say about it while he serves up kool aid?

The writer of this blog should post the cartoon depicting the MGH patient asking for the nurse (as he is dying).

Anonymous said...

This document points to another interesting change in the medical field; the growing importance of nurses at all levels of care.

Often we as patients have relationships with nurses; we do not have the same relationships with doctors. Time and money have forced the doctor into a supervisory role. Entering a room they make a pronouncement with out ever speaking with the patient and then turn and leave with the parting words; the nurse will explain everything.

My personal feelings are that in today’s world doctors diagnose while nurse perform the actual care function. HIT eliminates the doctors’ choices, while trying to limit the nursing function to that of simply following orders.

Doctors have been skeptical of the rise of public’s acceptance of the NP/PA. Doctors are great at finding zebras, but often fail in the interaction department of the run of the mill office visit. My experience is that most NP/PA’s are older and have a great deal of practical experience, having seen all of the problems with HIT and the rush, rush of modern medicine.

Doctors are trained to make quick decisions. Here are the labs; here is the prescription, next patient please. HIT feeds this need for speed, and rewards those who embrace this model with higher patient counts and a higher income. As pointed out many times on this blog doctors are also facing the subtle, and not so subtle, pressures of trying to manage a business model that is coming under increasing pressure.

The nurses have this one right. I respect doctors, but see forces such as HIT driving them away from patient care and into a role as patient monitor. A role that will diminish their standing in the public’s eye, and change the way we do medicine in this country.

Steve Lucas

Anonymous said...

Steve may need to rethink this statement: "HIT feeds this need for speed, and rewards those who embrace this model with higher patient counts and a higher income." HIT does none of this.

Most reports are that it slows the function and thought speed of effective clinicians and nurses. Why? The data is not in hand. The data needs to be searched from silos without efficient screen displays to enable astute connecting of the dots.

There are a sufficient number of misidentification and other mistakes, especially with electrical order entry, such that the output and records should not be trusted as being accurate.

No doubt, medical care is being radically altered, and not for the better, as the literature portends.

Joseph Arpaia, MD said...

I agree with this post completely. I use 95% of my training for the 5% of the time when something unexpected happens. It is my experience that tells me when the situation is in that 5%.

The problem with using algorithms in HIT is has nothing to do with the speed of the computers or the storage of information. The problem is that a complex perceptual problem has a solution space that is measure 0 in the space of solutions that will fit the data. Human perceptual processes somehow get this right almost all the time. And we have no idea how.

An example of this is vision. We experience a 3D world from a 2D retinal image. There are an infinite number of 3D perceptions which would fit the 2D image, but we see only one. Optical "illusions" are the rare situations where our visual system is fooled. For more examples see Visual Intelligence by Donald Hoffman, Ph.D.

So, we can't give the computer the same data that the clinician is receiving, because you can't input all the subtleties of a clinical history and physical into a computer. But even if we could, we still don't know how to model the human perceptual processes. For a mathematical formulation of this see Observer Theory, by Hoffman, Bennett, and Prakash (which may be out of print).

I have been away from the machine vision field for awhile, but the last I knew we were nowhere close to being able to replicate the human visual system's ability to determine structure from motion. A clinical decision is far more perceptually complex, and we don't know the mathematics of how to replicate that. There is no way we can create algorithms for processes we cannot model accurately with mathematics.

Anonymous said...

These systems are making money for the vendors concomitantly with depreciating the care provided to patients. There should be outrage in the medical community.

Why are those entrusted with the care of patients putting up with this invasion?

It is suggested that they raise a ruckus.

Anonymous said...


I stand corrected.

One of the great strengths of this blog is the fair and open discussion that takes place.

Steve Lucas