It is absolutely amazing to watch the unfolding saga the moment a real patient enters real data into Google Health from his hospital’s medical records. The way the marketing folks tell us, this is a seamless exercise that gets you up and running on personal health records (PHRs) like Google Health instantly.
The reality, as recounted here first on April 1 by our own e-Patient Dave, shows otherwise. The emperor has no clothes, and it apparently took a single real patient to try it out to show how such a seemingly simple idea — pulling in medical data from your medical records to help you track your health — is really far more complex than even Google realized.
But the fun really started watching the media coverage unfold once The Boston Globe published this article about e-Patient Dave’s experience with Google Health and his attempt to import data from his hospital’s electronic medical record.
The first person to hit the spin cycle was none other than John D. Halamka MD, Chief Information Officer (CIO) of Beth Israel. He’s the guy responsible for helping implement the Google Health PHR connection from the hospital side of things (Dave’s hospital, that held his electronic medical records, the Beth Israel Deaconess Medical Center (BIDMC)). He strangely suggested:
[The problem the article described] is really unrelated to Google, Microsoft, or any Personal Health Record (PHR) provider.
Well, okay, except that it is directly related to those providers and BIDMC insomuch as it affects what the patient sees in Google Health. You really have to ask yourself who did the quality control on this product from either side of the fence — Google Health or the hospital — before saying, “Yup, this baby is good to go!” It’s not like it’s any great secret that insurance billing codes rarely match medical procedures or diagnoses. Anyone who has worked in healthcare for more than a week gets up to speed on that issue very quickly. So why would anyone think sending billing codes as bits of useful and accurate data is a good idea? And how, in any universe other than this one, would the two be “unrelated”?
But it’s okay. Dr. Halamka (who really is more on top of these things than most hospital CIOs) assures us the fix is in. You just might have to wait a decade or more before you see useful results, because all of the procedures and languages hospitals currently use to describe things in electronic medical records aren’t changing any time soon. Meanwhile, his blog entry mentions other ways they’ll try and fix things to make Google Health actually usable from a medical records perspective. To have Halamka respond so quickly and transparently to the problem is nothing short of revolutionary — if healthcare IT is going to change, it may very well be because of people like Dr. Halamka.
Next up, you have good old iHealthBeat, a California Healthcare Foundation electronic publication targeted at those interested in “technology’s impact on health care.” You’d think any mention of The Boston Globe article or this serious issue might at least mention in passing who discovered the flaws in this data exchange system between Google Health and BIDMC. Most of this “article” is simply a quick rewrite of the Globe article, taking entire sentences from the original article and publishing it as though it were original reporting. However, unlike the Globe article, the spin is definitely on the technology and physicians, with no mention of how the flaw was discovered — not by a CIO or Google Health, but by a lowly patient trying to use the tools provided him. You might think that would garner at least one sentence.
At least eWeek is up-front that it’s taking its quotes and article sourcing directly from The Boston Globe, in their article entitled Google Health Accused of Inaccuracy in Electronic Medical Records. I’m not sure the headline is correct or accurate though, as nobody accused anyone of anything. e-Patient Dave simply described the problems he had in importing his hospital data into Google Health. That’s hardly an “accusation.” And at least they credit him with finding the problem in the first place. In the European edition of eWeek, the article morphs into Doctors Criticise Google Health Data. If anyone was criticizing Google Health data, it was the patient who was trying to actually use his PHR as intended.
The Industry Standard also sources The Boston Globe’s article, and credits Dave deBronkart’s experiences up-front. This blog entry focuses on what a patient can do to find and correct such mistakes, quoting Halamka about reviewing your hospital’s medical record on you to ensure it’s accurate. Unfortunately what is not mentioned is that for most patients, such a review would be next to useless (unless you have a degree in medical information science). Medical records weren’t made to be reviewed by patients, and so browsing through it would be as useful to most of us as browsing through a 14th century Latin text.
Dr. Jay Parkinson writes over at The Business Insider, Google Health Will Belly Flop (GOOG). The full entry is well worth a read, but here’s a key quote:
Here’s the problem. Each and every piece of multi, multi-million dollar bloated piece of crap healthcare “electronic medical record” (sounds about as dated as electronic mail, doesn’t it?”), is simply a billing engine to communicate a medical diagnosis to insurance companies with the hopes of maximizing how much doctors are paid. Each diagnosis and procedure has these numerical codes. They are a ridiculously robust antiquated language, like the code written to power the Commodore 64. There are people who speak this language – the 100 or so medical billers who are holed up in the basement of every hospital. They are the people you never see when you visit the hospital, but they’re the ones speaking the language that maximizes how much your hospital gets from your insurance company.
And now Google wants you to have access to this coded language that you, nor your doctors, speak. The assumption is that this will empower consumers and drive advertising revenue to Google based on these medical codes. And then you log in to Google Health and see a language that’s impossible to interpret and, most of the time, simply inaccurate. But these are your medical records found in one of the highest ranking hospitals in America.
He suggests Google should just give up on Google Health while it’s ahead (and maybe Steve Case who tried to foment a revolution in health care with his Revolution Health might agree). I do agree that anyone who tries to whitewash the extent and complexity of the problem or suggests it might be easily solved is living in a dream world. Dr. Parkinson hits it out of the park with his observations on this one.
There are a few other media mentions of the piece, and I’m sure more will be on the way. At the end of the day, I think Dave’s experience demonstrated that this piece of the Google Health really wasn’t ready for primetime, and hospitals shouldn’t have gone along with them and agreed to allow the import of their data into Google Health. As I said earlier, how a hospital electronic medical record works is no mystery to the management who lives and breathes these records every day of the week. How anyone imagined this was going to result in useful, beneficial and most of all, accurate health data for patients to see and use is beyond me.
Yes Google should just give up its business in EMR. We don’t need a 800 pound business Guerrila to keep people’s EMR.
The more I digg into this story the more I am, very strangely, becoming a big fan of Google Health and Microsoft HealthVault. These may be our only possibility to grab our health data and do somethinh constructive with it.
I am becoming convinced that we got blindsinded by the privacy issues (yes they are real) and never questioned the data quality in existing EHRs. So what if these are fully protected by HIPAA, if they transmit useless or worse, erroneous data.
We assume that Dave’s problems were specific to Google but we don’t now if transfer of data from BIDMC to another institution would be of a higher caliver or not. Lots of unanswered questions!
Here’s another user experience – I uploaded my allergy list, medication list, and immunizations onto HealthVault, and they appeared as exact replicas of what was in my record. There isn’t much I can do with it there except pass it along, but at least it came across accurately.
Do you have conditions and procedures you could upload?
Was it medication history or just current ones?
(Nosy, ain’t I? Well, you know, you don’t have to answer anything you don’t want.)
Current meds, allergies and imms. No conditions or procedures available yet, from my EHR system. I can change/edit the data if I want, which changes the signature to mine.
The problem is that most of these systems are too complicated and requires having to sort through all sorts of irrelevant information. That’s why we developed an easy to use, HIPAA compliant, on line site, http://www.housedoc.us, that lets patients list their medications, and include only the important medical information. Its also lets doctors and patients communicate with each other by email.