Friday, February 27, 2009

A Call For Congressional Investigation Of The HIT Industry - Is This Industry Trying to Harm Patients? Part 5 of a Series

(Note: Part 1 of this series is here, part 2 is here, part 3 is here, part 4 is here, part 5 is here, part 6 is here, part 7 is here, and part 8 is here. 2011 addendums: a post that can be considered part 9 is here, part 10 is here.)

In this installment of illustrating defects in today's contemporary health IT, I was going to present more sketches (since contractual clauses between HIT vendors and hospitals tend to forbid sharing of actual screens or defects) of the "let's play peek a boo" variety.

I will present them in part 6 instead. These are screens in which clinicians must scroll extensively to, say, match an INR with a coumadin dose, FiO2 with oximetry result, match systolic with diastolic blood pressure, search around nearly empty screens to reveal a single value many rows down, enter a few common diagnoses via the "clickorrhea" of almost fifty clicks, and other patent absurdities from technology touted as "improving medicine."

I am using part 5 to present another issue: A call for investigation of the HIT device industry,
its development, marketing, talent management, contracting, QC, and information sharing practices on bugs and defects.
I have now become aware of organizations with defects/clinician complaint lists for contemporary HIT including CPOE, EHR etc. amounting to well over a hundred pages in one case, and well over a thousand individual items in another, and numerous unquantified but relevant in-betweens.

I recently also received this from a correspondent with expertise in user interaction design:

A funny story [that is actually not funny at all - ed.]

When I managed Operating room systems for a hospital the vendor was [major vendor, name redacted].

The UI was awful. Though the implementation had many problems (prior to my taking the job) circulating nurses were livid. Some actually filed a complaint with the state's department of health claiming the new system endangered patient safety. Alas, the "inspection" ruled it was not endangering patients. [I'm sure the inspectors had significant healthcare IT and informatics domain expertise. Well, no, actually - ed.]

One day I was talking to a representative from [major vendor]. I mentioned their peculiar UI and the guy just didn't get it. He went on to tell me how a single programmer designed and developed the UI all by himself in about a month!!! He thought this was really impressive! I realized in horror, the UI we were seeing every day represented the limited mind of a single technology obsessed programmer who applied his sense of architecture and programming to the patient care routines of operating rooms all across the country.

The electronic chart contained the same fields as old papers forms, but the forms took about 3 minutes to fill while the electronic version took 15 AND while that is somewhat acceptable for long cases [I'd say a doable waste of time, not "acceptable" -ed.], many cases are short, from 15 to 30 minutes and that with the new system the nurse had even less time to take care of the patient - their major responsibility.

This is simply stunning. I might also add "frightening."

The following comes from a person with interest in usability:


I am working at the [name redacted] University Hospital. I asked several [another major HIT vendor] reps about their screen design and why they do not implement better ones. They said there is not a business case to spend the money upfront to do a proper Information Architecture Plan.

I other words, if this is true (and I believe it is from my own experiences with numerous vendors), an HIT company's profits override the issues of a mission hostile user experience and cognitive overload of clinicians. The "business case" means "their profit." Looked at another way, physicians and other clinicians' cognitive energies serve as a bank for HIT vendor deficiencies and unwillingness to develop an appropriate and acceptable user experience.

How many patients and clinicians subject to these devices know about these issues?

I am going to go out on a limb here, similar to 1998 when I became one of the first (to my knowledge) to publicly stick my neck out with what at that time was a controversial website on HIT failures. Its title was "Medical Informatics and leadership of clinical computing", now here under a new title. I had hoped that the HIT field's problems would diminish with the passage of time and with very direct exposure that all was not well.

Instead, it appears the HIT industry and products have gotten worse.

The existence of lists of a thousand+ HIT defects in major healthcare organizations, some resulting in clinician complaints about patient endangerment, and unshareable due to contractual gag clauses common in this industry, is simply intolerable. I am told many of these defects will likely never be fixed. I have only scratched the surface in reporting on these issues, being forced to do generic screen sketches, at my series here .

There is no spin, no excuse, no dialog that can forgive such a situation. One must ask how products with thousands of defects could get into production. The existence of such lists is prima facie evidence that this industry is itself defective and that these devices are experimental - with clinicians and patients as unconsenting participants.
Further, the shielding of the industry from liability on the "learned intermediary" basis is prima facie evidence of misuse of such laws. Even "learned clinicians" are being cognitively taxed and overloaded by these devices; they do not have unlimited cognitive capacity to deal with these defects.
I believe this scenario is likely far worse than any practice I saw in pharma.
It is becoming more clear to me and a number of my like minded colleagues that, as in pharma and the other medical device industries:

  • A congressional investigation of the HIT industry's development, marketing, talent management, contracting, QC, and information sharing practices (on bugs, UI problems and other defects) is warranted;
  • These defect lists need to be public record so the public can evaluate whether their consent to be subjects of these devices is warranted;
  • Vendors should not be immune to liability for defects caused by cavalier IT development, testing and talent management practices; the article "Failure to Provide Clinicians Useful IT Systems: Opportunities to Leapfrog Current Technologies", Methods Inf Med 2008; 47: 4–7, by Ball et al. (PDF here) spells out some of these issues;
  • A societal process needs to be set up in parallel to those other two biomedical industries, pharma and [physical] medical devices, of staged clinical trials, validation, and regulation of HIT [virtual] devices by impartial bodies with appropriate, unconflicted expertise on board. "Finding a Cure: The Case for Regulation And Oversight of Electronic Health Records Systems", Harvard Journal of Law & Technology 2008 vol. 22, No. 1, by Hoffman and Podgurski (PDF here) presents a case for oversight;
  • The evaluation needs to go far beyond the CCHIT features and functionality check, an almost useless evaluation criteria in my mind, directly to usability and effectiveness. For a good summary of CCHIT's deficits see p. 132-134 (p. 30-32 of the PDF) in the Harvard Journal of Law & Technology article above. The major problem in HIT seems not absence of features, but a mission hostile user experience.

From the latter
Harvard Journal of Law & Technology article:

The benefits of EHR systems will outweigh their risks only if these systems are developed and maintained with rigorous adherence to the best software engineering and medical informatics practices and if the various EHR systems can easily share information with each other. Regulatory intervention is needed to ensure that these goals are achieved. Once EHR systems are fully implemented, they become essential to proper patient care, and their failure is likely to endanger patient welfare.

Indeed.

Mine may sound like a radical position, but it is not. The existence of such secretive lists and lack of transparency is a highly significant problem. There is no other industry I know of that can push devices with hundreds or thousands of defects, some safety critical and others enhancing risk, onto the public, while preventing dissemination of defect lists.

Imagine an aircraft industry where planes in the sky or air control devices have thousands of defects, some potentially leading to collisions or crashes. Pharma is being pushed to release negative trials results, for example.
These issues need to be addressed in Washington before we dump $20 billion+ into a defective industry, instead of providing those funds for needed patient care of the poor and uninsured.

-- SS

4 comments:

Cetamua said...

Just a question: Don't hospitals request (demand) to have a pilot program to evaluate how well this stuff actually WORKS before ordering and implementing those EHR systems? Don't they happen to have a lawyer READ the proposed contracts?

Hell! Any John and Jane Q Public would take, at least a car ride before buying one. Not to mention go to Edmunds.com or cars.com. Talk to friends and a trusted mechanic; all these things known as due diligence.

Since when working in a health care environment absolve managers from applying common sense?

I guess I'm being way too down to earth here.

Cetamua said...

I forgot to ask a question to Dr. Silverstein:

From your post, am I correct to infer that physicians working in an Institution have very little say in the decision-making process of choosing a particular EHR?

The key passage of your post that triggers this question was:

"I asked several [another major HIT vendor] reps about their screen design and why they do not implement better ones. They said there is not a business case to spend the money upfront to do a proper Information Architecture Plan."

In other words, clients do not demand it, otherwise, vendors would be *forced* to do it. Which lead me to conclude that physicians are removed from the equation a priori.

Correct? Or way off base?

InformaticsMD said...

Don't hospitals request (demand) to have a pilot program to evaluate how well this stuff actually WORKS before ordering and implementing those EHR systems?

No. Many hospitals are run by - well, let's just say, the same type of people who put the world into a bit of an economic crisis. Not the brightest bulbs. See this PDF, for example.

Don't they happen to have a lawyer READ the proposed contracts?

See above.

all these things known as due diligence.

I agree, but couldn't stop a CIO from putting desktop computers on the ceilings of ICU's for his convenience, putting patients at risk. See the first case at my HIT website here. What can I say?

From your post, am I correct to infer that physicians working in an Institution have very little say in the decision-making process of choosing a particular EHR?

That is usually the case, whether they are contractors, or even worse, owned practices (employees).

In other words, clients do not demand it, otherwise, vendors would be *forced* to do it. Which lead me to conclude that physicians are removed from the equation a priori

Correct, and worse.

They are forbidden from speaking about problems to other hospitals, or of showing screens. Plus, the doctors provide a liability shield to the vendors - the "learned intermediary" between the computer and the patient.

Doctors take all the risk.

Thus, their diligence and ingenuity in working around system defects serves as the vendor's bank.

My website on HIT failure is nearly unqiue, after TEN YEARS. Ans not for lack of interest - see the poster in the "other materials" tab. Think about that, and what it might mean about an industry that muffles dissent.

It makes pharma look like kids in a sandbox.

-- SS

Anonymous said...

"The major problem in HIT seems not absence of features , but a mission hostile user experience"

Medscribbler is all about "the user" "the user" "the user" In its development, as chief architect, I continually send ipod toting, warcraft playing programmers interface results back for them to "fix" We have regular "interface" meetings. Our reward:

Doctors who have been so indoctrinated that they demand a 'certified" solution that automatically generations a note according to some terminology engine.

Our return:
We are adding those "features" but only as we solve the user interface problem.

Cultural Anthropolgy was my major area of study - and how to relate one cultural group to another. First I know of to coin the idea of the concrete/relational to abstract/linear compendium of thinking.