Hacker News Reader

Comment by davycro | Hacker News Reader

Emergency doctor and hacker here. The closed systems and lack of support for interoperability between EHRs upsets me the most because it leads to patient harm.

I just had a patient transferred from an outside hospital for abdominal pain and somehow their CT scan was lost in transit. Because it was an emergency we had to CT scan them again which doubled their radiation exposure and their risk for a kidney injury from the iv contrast. It was midnight so it would have taken until business hours to obtain the scans from the hospital.

Things like this happen almost daily.

View on HN

Just-graduated MD/MPH and former EMT and ED admin here. I empathize strongly, since the first time I watched a 16-year girl who'd been in a MVC and first transported to Duke, who needed sub-specialty consults at UNC get re-scanned.

I'd read about CT dosing and realized that we had just increased this girls lifetime likelihood of ovarian or endometrial cancer by perhaps about 1/1000. For absolutely no reason other than the fact we couldn't get the images 8 miles down the road from the Duke ED, and the attending wanted to 'just be sure'.

I asked the resident why they couldn't send them digitally and they just laughed. That was 2009.

It's 8 years and a few hundred billion of national EMR spend later - and you know what? We still can't send an image between the two EDs. UNC and Duke were the first two nodes/servers on Usenet back in 1980[1], and 37 years later we can barely exchange medical data using our combined 1.3 billion dollars of Epic EMR implementations.

This problem harms people needlessly every day. Please use your voice as a provider to remind people whenever possible.

[1]https://en.wikipedia.org/wiki/Usenet

EHRs are bought by people who don't understand what they are getting, used by people who don't get to decide on what they are getting, and the effects are noticed after millions were poured into integration.

EHR companies have bigger incentives on selling than on making shit better. Worse, if I was to make an EHR startup with amazing UI and all around amazeballs features, I'd never get into a single hospital convincing their CTO to replace the current EHR system and re-train doctors. Never gonna happen.

Welcome to the problems of Enterprise tools, we've been having these for decades and nobody has solved it well yet.

The only real solution is for a hospital to take say 5 doctors of various tech and age levels, and create a mini EHR (regardless of cost) and test it. Then switch to another. The doctors would be EXPECTED to be inefficient, but they will be the scouts of the hospital. But hospital CTOs don't think in this way. Rarely are they technical people.

The closed systems and lack of support for interoperability between EHRs upsets me the most because it leads to patient harm.

As a patient here's what gets me: UNC has one EMR system, my GP has another, and Duke has yet another. I've been a patient at all 3 at various times. I also use Strava, Fitbit, and the like. Now you would think that it would be trivially easy to use an API and export and aggregate all of my lab results, for things like blood pressure, cholesterol, etc., so I can graph, say, my blood pressure and my Strava activity together. But nooooo... all of these EMR systems either have no export functionality / API, or if they do, it's something byzantine and just-short-of-impossible-to-use.

Of course not every patient is a data geek who's going to use machine learning, statistical analysis and visualizations on their own data. But for those of us who want to, the roadblocks to doing so are infuriating.

The doubly sad part is that this is almost 100% a policy / business issue and not a technical one. We've known how to share data for a long time. Heck, IIRC, a scenario much like the one listed above was used as an illustrative example for justifying the WS-Federation[1] protocol way back when.

Edit: Yep, this document[2] explaining WS-Federation actually uses an emergency room scenario to justify the need for WS-Federation. Not the exact same scenario as above, but the point stands. We've had protocols and technologies for doing this stuff for a long time.

[1]: https://en.wikipedia.org/wiki/WS-Federation

[2]: http://download.boulder.ibm.com/ibmdl/pub/software/dw/specs/...

Well, they are all required by law to support import/export of a CCDA, and Duke's EHR does. The problem is that it's poorly supported in the workflow, and may require using something like a HIE.

Also, Fitbit data is medically useless beyond establishing that you do or don't exercise.

I use a Fit bit brand scale, and I would argue that my weight data is useful.

They weigh you at any doctor's office, on a more accurate scale. Weight trends are important, but at a granularity of a handful of measurements per year.

Unless a patient some specific chronic illness like diabetes or asthma, most patient recorded data isn't medically useful.

most patient recorded data isn't medically useful.

If you're talking about diagnosing a particular disease, you're probably right. That's not what I'm talking about. I'm talking about having an overall holistic view of my health and fitness related data, and being able to combine it so I can look for interesting correlations / patterns to suit my own interests.

They weigh you at any doctor's office, on a more accurate scale. Weight trends are important, but at a granularity of a handful of measurements per year.

The "more accurate scale" part is irrelevant given that your bodyweight can fluctuate 3 or 4 pounds from one day to the next, based on hydration levels, undigested food in your stomach, etc. Trends over time, even if measured on slightly less accurate scales, are really more meaningful.

Anyway, I'm not really sure why we're even having this discussion. Are you arguing that EMR systems shouldn't make their data easily available to the end user? Because otherwise, this whole thing about Fitbit data is a red-herring. What I do with the data, what other data I mash it up with, etc. is really not what's at issue here.

No im just arguing that EHRs shouldn't import fitness tracker data. There's literally no evidence that it's medically useful. Every doctor and researcher I've ever talked to thinks it's a pointless distraction.

The problem is not necessarily closed systems or lack of interoperability support.

In your situation with CT scan, it sounds like there was no electronic way to share the digital version of the CT scan. Furthermore, if there was an electronic way to share it, your ER where the patient presented would need to be connected to any and all other facilities they could be transferred to.

To further complicate matters, and electronic CT scan isn't really EHR functionality, it goes straight from the CT device to a PACS (picture archiving and communication system). There may be integration with the EHR, and definitely with the RIS (radiology information system), but outside of those two, you're talking about transferring 100s of MBs to GBs of data to the other hospital.

If I ran your hospital IT team (or rather the hospital you're referring people to) I would prioritize integrating the two. Duplicate scans cost them money, and if you have choice over where you refer patients - the integrated hospital will always be preferred right?

The problem is that IT and regional monopolization shouldn't drive patient outcome. The technology is a cancer here -- it's literally made everything other than prescribing (Note that pharmacies do interoperate) measurably worse. Paper & fax was better.

A family member just was hospitalized for a serious issue that requires rehab, and the default from the hospital is to get you into an affiliated facility (a nursing home). The better facility (a rehab hospital) for this person's condition is in a different network. We handled data transfer by putting the records in my hands and delivering them to the physician hands at the rehab facility.

If you're not aware of these differences and pushy, you'll get sent to the default and likely have a worse outcome. So my family member gets stuck with a worse quality of life, and the taxpayer via Medicare gets to spend exponentially more on providing care.

it sounds like there was no electronic way to share the digital version of the CT scan

Seems unlikely. A few years ago I was given a CD-ROM of the scan of my jaw and was able to download a free DICOM viewer to look at it. These things seem to be standardized.

Wasn't really excited to take that much dose to my head but at least I have the souvenir of a 3D view of my jaw...

By electronic I meant over a network - a burnt CD may have well been the media that got lost in OP's situation.

The problem is that the digital scan is isolated to his ER's network, and without a business relationship between the two parties, a CD is the only way to transmit the data.

This is pretty key and why HIEs exist. They're a hackish solution though. Some sort of national data exchange would be useful.

I wrote a DICOM fetch and view app (mid 2000s). Image viewer was the easy part. Network interop was a mess. Worse than any ASN.1, SNMP, or XML-based interop work I've done. I can't imagine it's gotten any better since.

It's even worse in some circumstances.

My GP migrated to a pretty good EHR in 2007 as an efficiency measure when he was a partner in a ~<10 doctor practice. That was awesome because it reduced friction and improved continuity if you had to see a NP or another doctor. That solution wasn't fully Medicare compliant, so they moved to another system that definately sucked, but wasn't much different from a patient perspective.

Now they got gobbled up by some mushrooming health network centered on one of the hospitals. They just migrated them to yet another EHR that is cloud hosted (Athena?) -- except they forgot about the "R" part of EHR. They literally migrated nothing and the medical staff needs to re-enter all information from billing to health history. It's just irresponsible as people don't necessarily remember all of their health history.

You can't even leave as like 75% of GPs near me are in this network and affected by the same thing. The other network in the area concentrates on a different geographic region.

With the EHR gaps (interoperability, communication, 100% adoption), what do you do in the ER when a patient is unresponsive or otherwise doesn't know the medications they are taking?

It seems the most common answer is attempt to call the patients pharmacy or rely on the best info at the time. I'm working to solve this problem at least for Medicare chronic care patients - we basically instantaneously provide physicians a pharmacy created medication history and 6 months RX claims data - and while we are focusing on primary care, ACOs, and transitional nursing care...maybe this might be of value in the ER.

Has shipping patients with the required information been considered ?

Reply on news.ycombinator.com