Wednesday, June 30, 2010

Electronic Medical Records Follies

One of those oddities—I’m being kind here—of the health care system is that a doctor, after doing a treatment, does not actually know when, if or how much he or she will be paid for said treatment. It’s all up in the air, thanks to that wonderfully efficient system called America’s health insurance industry. Moreover, the inefficiencies caused by that goofiness are more than just the first-order oh-you-have-to-wait problems: the insurance company will frequently try to lowball doctors on payment, which leads to negotiation which leads to an epic of lost time and lost man-hours. It’s no wonder that about $300 billion is spent on administrative tasks in health care.

Now, the NYT article linked to above attributes the problems to electronic medical records and their relative disuse among doctors (and, when used, the tendency to update them at the end of the day rather than at real-time which slows down the whole claims process). And apparently a strong number of Americans (42%) don’t even realize whether their doctors use electronic medical records or not (I'm one of them!) And there are a disturbing number of studies that show that the electronic medical records don’t do what they promise: improve care while slashing costs.

The theoretical case for electronic medical records remains more than strong; why hasn’t the practice fit the theory?

I’m afraid that the execution of electronic medical records has been more than lacking. It’s that lack of quality execution, by the way, that represents a very lucrative opportunity to whatever business figures it out.

The first problem is who’s to organize it. Part of the problem is that electronic medical records are often pushed by insurers, who (naturally) design them to meet the needs of insurers:
"When you're trying to read the notes of your colleague [in an electronic record], it's almost impossible to figure out what happened to the patient," says Rushika Fernandopulle, an internist, instructor at Harvard Medical School and co-founder of consultant Renaissance Health. "You have to read through two pages of all this junk that's put in to increase billing."

Dr. Fernandopulle gives an example of a note a doctor might write on a paper chart after seeing a patient with a sore throat: "Patient has a sore throat, no fever. I think this is viral pharyngitis. No need for treatment. I reassured the patient." But if a note for that visit had been generated by an electronic system designed to maximize billing, it could be pages long. "You would have to wade through an awful lot of stuff about whether they had stomach pain, diarrhea, weakness in their muscles—almost any question imaginable, most of which wouldn't be relevant to the problem at hand," Dr. Fernandopulle says.
Anecdotally, I have it on good authority (Hi Mom!) that some electronic medical records do not allow you to create graphs. My guide to life is that if Microsoft has figured out how to do it (Excel!) than it’s a pretty good bet that a common-sense solution exists; furthermore, if your functionality is worse than Microsoft’s relevant product, you have huge problems. Let’s dwell a little more on the graphs problem: the reason has to do with the fact that the relevant graphs are for a little-used corner of medicine (i.e. it’s not like tracking blood pressure over time). Now, you—being perfectly sensible—have a very good question: well why can’t you just input whatever variables you’d like? As I said, it’s irrational and dumb.

Probably there’s room for a third-party—whether it’s a startup or someone else—to do this right and design a proper medical record.

The second problem is cultural. Older doctors are rather famously technology adverse (do you realize how many jokes there are about doctor’s handwriting?), meaning that the records have to have that Apple-like quality of being immediately understandable and intuitive. Moreover, (as mentioned above) doctors have to commit to using the records immediately on the spot so as to save the redundancy of administrative upkeep later.

The third problem is universality and sharing. The highest ideal of electronic medical records would be for everyone to have one, and every relevant medical personnel to be able to call it up when relevant. This suggests cloud computing to me. Even more so than, say, Facebook, the privacy concerns are pretty much ubiquitous. The key is to be able to give and manage the requisite permissions (you could, for instance, allow all emergency room types to access your records with a code? I’m not sure here.), which should be very doable. What's more, the best electronic medical records system would allow you access to your own medical records at will--the fact that 42% of Americans don't know whether their doctors use electronic medical records is one suggestion of how one-sided the relationship with health care system is.

Beyond the obvious effects of properly designing and propagating a standard medical record—more productivity! less costs! fewer errors!—there’s the less obvious stuff. One example that comes to my mind is the absolute usefulness to academic researchers: users could opt-in to allow academic researchers some access to their data (who could, say, search for white males in their sixties to recruit them to examine, uh, cholesterol and its relationship with cancer or something.) And I’m sure there are even more effects that I just haven’t thought of yet.

As with so much in health care—hell, the entire economy—it’s something of a wonder that we do so much so incompetently when we could do so much more so much better. The gap from here to there represents a permanent, massive upgrade to both our national wealth and health. Let’s figure these things out.

No comments:

Post a Comment