Thursday, May 31, 2007

Badly-designed EHR forces mother to decide on sick child's gender

As a medical informaticist who studies the gaps between business computing and clinical computing, and laments the problems caused by the leadership of the latter by representatives of the former, this story speaks for itself:


This story, related by a pediatric specialist, exemplifies the gap between the expertise and thinking of business computing specialists and healthcare informatics specialists:

‘Mother referred for delivery and management of a fetus in utero with diagnosis of hypoplastic left heart – baby born, middle of night, terribly sick, unexpectedly more wrong than expected, and with truly ambiguous genitalia. Can’t get an xray, prostaglandin-E (a lifesaving medication used in babies born with functionally absent, or obstructed, blood flow out of the heart), or any medical test, until there is a “Medical Record Number” in the computer. Can’t get one of those until we tell the computer whether the baby is a boy or a girl. And there is no way to bypass that and get on with delivering emergency care. (Almost unimaginable - ed.)

Not the way one wants to break the news of an intersex (gender not definite) problem to a new young mom, of a very sick baby…. So I ran back to Labor & Delivery, sat down face to face with the mom, still groggy from meds and caesarian section surgery – and said something like, “Your baby is very sick, and we don’t know precisely why. And we can’t tell for sure right now whether your baby is a boy or a girl. In order to get x-rays and medications, we need to tell our computer whether your baby is a boy or a girl. Which do you want the baby to be?

The 18 year old new mom looked me right in the eye, nodded her understanding, and said, “I want a girl.” So I said, “OK, she’s a girl. I’ll come back and talk to you as soon as we get her stabilized.”'

Postscript: We did get her stabilized, and she was a genetic girl, but she died waiting for a heart transplant that never came. And that mom spend nearly every waking minute with the baby for the whole 3 months we all waited for the heart transplant that never came. Truly an amazing person, that mother. All children should be so lucky.

Never since have I ever designed a database with gender as a binary part of the primary key. Out in the business world, people are still doing that. And I’m sure some of those people are still trying to sell their business stuff to healthcare. Babies aren’t the only ones who might need a bypass (intersex is actually quite common), and consider the roadside bomb victims in Iraq – or any other explosive, burning trauma – is gender your first concern?


(My response is that anyone with even a smidgen of pediatric healthcare knowledge knows that gender can be one of “M”, “F”, or “unknown.” Apparently, the designers of the system did not envision such knowledge as important for informing system design .)

-- SS

6 comments:

Anonymous said...

Even more surprisingly, they seem to have utterly ignored both the US and international standards. For over 15 years the standards have clearly stated that administrative sex can be male, female, unknown, or other. I find this story so bizarre that I wonder whether it is true. A system like this could not issue an MRN for a trauma victim, because they are usually unknown sex, name, etc. for the early stages of treatment. It seems odd that you would have a pediatrics specialty like this in a facility that cannot handle trauma patients.

InformaticsMD said...

Yes, hard to believe, but it's also hard to believe that, say, a CPOE system could be as flawed in 2005 as it was, say, in 1992, when I was involved in an implementation of its predecessor.

I report, you decide.

InformaticsMD said...

For more on the wonderful state of health IT and how it uniformly abides by the best clinical and engineering practice, and goes in like a snap as long as "all the right processes are followed", see this recent article.

Anonymous said...

A system we have here has fields for genotypic sex and phenotypic sex. Regardless, and as much as I hate having the untidy 'Other', it's pretty much a requirement in medicine since nature is not always so clean cut. BTW, does it bother anyone else as much as it does me when people insist you use the term Gender when they really mean Sex? I'm just sayin'.

Anonymous said...

Regardless, and as much as I hate having the untidy 'Other', it's pretty much a requirement in medicine since nature is not always so clean cut.

I think nature is the easy part. Now we have LGBT etc. etc.

Anonymous said...

I'm with you, noodle. Sex and gender are not the same thing and if the question of sex can be so uncertain, the question of gender is exponentially more so. Using the term "gender" to mean "sex" severely restricts the kind of conversations one can have about sexual diversity and gender constructs in society. Don't be afraid to say "sex" when you mean sex!