Boston hospital experiments with Google Glass in ER | Hacker News Reader

Boston hospital experiments with Google Glass in ER

geekdoctor.blogspot.com

76 pointsilamont73 comments

Now, like all pieces of medical equipment, it would probably good to culture some swabs off them to see what's growing, and how hard they are to clean.

Huh, why would that be good? Why is the sterility of a doctor's eye-wear a concern?

Tabloid scare-stories of "X species of bacteria found living on Y" just feed on the public's ignorance of microbiology; i.e. "wow that many on the average toothbrush!"... well, how many are found in your mouth? In the intestines? How do we define a bacterial species?

Because they touch their eye-wear, and they touch patients and themselves before and after touching their eye-wear.

This probably is the same problem for regular glasses, so it's not a new concern unless the Google Glass gets touched more often.

For sure, that's why hospitals are jam-packed with hand anti-bacterial (fungal etc.) stations. They don't try to sterilise the doctor.

First, those stations are ineffective against some major pathogens (C. difficile and Norovirus come to mind), and alcohol gels aren't particularly well suited for cleaning equipment - they're meant for hands.

We do talk about sterilizing equipment doctors carry - lab coats, ties, stethoscopes, phones, etc. all the time. There's just not a good solution yet, which is one of the reasons one might need to assess:

1. How much of a problem this is for Google Glass in particular?

2. How amenable are the devices to cleaning?

You know, exactly what I said in my first post.

My thought is that if they're using Google Glass as a routine tool, they've increased the number of people wearing eye-wear, and those who are not used to wearing glasses will probably be fiddling with them more.

Carrying pathogens room to room, even when the object doesn't directly touch the patient, is a serious concern:

http://www.nysenate.gov/files/pdfs/DirtyTieFINAL_1.pdf

It's incredibly important. the place you are most likely to acquire an infection / disease and die is the hospital, after adjusting for the fact that you're there because you're already sick. Forgive me for not linking to a source on this

I am a doc and my partner is a medical microbiologist.

Its just about simple bacteria spread - would you want your immunocompromised relative in contact with a doctor that has just been in contact with a contagious infection source without at least trying to minimise the chance of cross infection?

The stethoscope is a similar concern - rarely cleaned properly, often present.

The sterility of a doctor's eye-wear is a concern because of the following simple scenario:

A doctor has been seeing patients all day, and visits one with an active C. difficile infection. They touch the patient's bed (which has been shown to be as contaminated with organism as the skin), and then, before they change their gloves, adjust their glasses - transferring the pathogen to a surface it can likely persist on for some span of time.

Later that day, they adjust their glasses again (seriously, I do this like 20 times a day), and recontaminate their hands before examining another patient.

Glasses - along with shoes, ties, lab coats, cell phones, etc. represent potential sources of cross contamination between patients, and are an active area of scientific and medical research.

I get this I really do, you can invent just-so stories about transmission with anything regardless of whether he's wearing Glass.

Doctor visits patient with <highly transmissable pathogen> touches x, washes hands and but then (!) retouches x and treats an immune-deficient patient etc. My point is x can be forehead, hair, pocket, watch, any variable you can think of and we can't control for any of them.

You seem to think every item a doctor carries needs to be autoclaved at clocking out time, when really the only viable solution is to wash your hands, change gloves etc. — sterilise your contact points, not your person and everything on it.

> we can't control for any of them

Yes, we can. Infection control procedure is exactly that, and we have begin looking at how we can minimize as many of those items as possible. Things like the universal gowning and gloving studies, examining how to clean surfaces and keep them clean, what to do about cell phone and clothing contamination, etc.

There are a massive number of studies showing contamination of things doctors carry, and studies showing heavy correlations between the level of contamination and patient outcomes. It's an active area of research, and commercial development.

Your idea that somehow "Wash your hands and change gloves" is fine, but "Find out if you can use clorhexidine wipes on Google Glass, and if we need to" is somehow Infection Control Gone Mad! really puzzles me.

That's mischaracterising my position, so I don't doubt you're puzzled.

"Now, like all pieces of medical equipment, it would probably good to culture some swabs off them to see what's growing, and how hard they are to clean." leads to you... "You seem to think every item a doctor carries needs to be autoclaved at clocking out time" and claiming tabloid scare stories.

Perhaps it isn't well characterized to begin with?

Let's rephrase: why would glass be any different than a doctor's normal glasses? If it's not, then are you just saying "we should probably treat this like their glasses, i.e. do research but currently do nothing special in a clinical setting"?

Possibly harder to clean than normal glasses - I'm not worried about frying the electronics in my glasses. That's the only difference I can think of, but it's a major one.

Additionally, if they're being used heavily by people not used to wearing glasses, the frequency of touching them is probably higher.

So, essentially no change worth talking about.

Look, I get the idea of "Hey, here's something new, it could change something in a bad way, so lets test it." Fine, I hope someone does that. But the fact that these kind of concerns are the first thing that come to people's mind and make it to the top of this thread means the discussion is bad. Every time someone mentions something revolutionary like a self-driving car, we could spend the discussion talking about that super unusual situation when the self-driving car is more dangerous than a normal one. Or when Amazon starts drone deliveries would could imagine that one time it crash and kills a dog. But this is just a distraction.

> So, essentially no change worth talking about.

Umm...depends on how hard it is to clean. Something we don't know the answer to. I can hose down my glasses with clorhexidine, they're entirely inert. Can I do that with a Google Glass? Does interacting with your glasses as an information gathering tool change usage patterns - for better or ill?

You can't say there's no change worth talking about until you actually do the study. Not even a particularly hard study at that.

> Look, I get the idea of "Hey, here's something new, it could change something in a bad way, so lets test it." Fine, I hope someone does that. But the fact that these kind of concerns are the first thing that come to people's mind and make it to the top of this thread means the discussion is bad. Every time someone mentions something revolutionary like a self-driving car, we could spend the discussion talking about that super unusual situation when the self-driving car is more dangerous than a normal one. Or when Amazon starts drone deliveries would could imagine that one time it crash and kills a dog. But this is just a distraction.

First off, I was like the first or second poster on the thread - that it got a lot of comments is possibly just because it was there.

Second, I'm an infectious disease epidemiologist with a particular interest in hospital infection control, especially surface contamination mediated transmission. Asking questions like this is my job.

> Umm...depends on how hard it is to clean. Something we don't know the answer to.

We don't know whether the new curved cell phone screens will harbor more bacteria either. And we don't know if increased close-range tablet usage by doctors will subtlety make them want to focus less closely on patient sores. And of course we can make strong physics argument about why wifi is non-ionizing and so doesn't cause DNA damage....but couldn't the extra heat in the room increase bacterial growth?

> Second, I'm an infectious disease epidemiologist with a particular interest in hospital infection control, especially surface contamination mediated transmission. Asking questions like this is my job.

This isn't a personal attack, I'm explaining why I don't think this is particularly valuable to the thread.

I'm willing to bet Google Glass has a lot more nooks and crannies for the little bugs to hide than a normal pair of eyeglasses.

Yea, and smartphone have a lot fewer nooks than traditional cell phones since they have no keys. Better do a study to see if we should recommend that all doctors upgrade from their old Nokia!

> Yea, and smartphone have a lot fewer nooks than traditional cell phones since they have no keys. Better do a study to see if we should recommend that all doctors upgrade from their old Nokia!

Um...trying to figure out what to do about smartphones and the contaminants they carry is a very active area of discussion in infection control right now.

Yes, but you can not use a smart phone while treating patients. If you're using Google Glass, you can't really not use it, especially if the reason you're wearing it is to help patients.

I do believe this is a solvable problem (as do the people working on it, I'd imagine), but it hasn't been solved yet.

Perhaps a sealant on the device plus a cleaning regimen will do the trick, but right now I'd put google glass somewhere near neckties (glad that that's been largely dealt with), lab coats and, well, pretty much anything that has places that will readily harbor foreign bacteria as a terrible thing to wear from room to room in a hospital.

So I guess that given the state of things right now (riddled with such harbors), it's perfectly reasonable to give the technical benefits of glass a try. Maybe we'll come up with some good software that will save some lives.

(This may seem like a retraction, but I was really just comparing Google Glass vs regular glasses in my original reply).

[Edit] this was intended as a reply to the person the parent was quoting.

Wow, I just noticed your nick and wish I could give you more upvotes for that alone!

It looks like you have a background in epidemiology, so not sure if you are seriously concerned or posting for the pure perfection of deploying your nick. But it is an interesting question and assumedly one they thought through.

The alternative inputs they mentioned (head tilts) may have been designed to minimize needing to touch Glass. Certainly touching glass after having ones hands on some patients would be a concern -- and for the health provider given that it's rather close to the eyes.

Providers do touch smartphones, tablets, pens, etc. when they are with patients. I wonder whether Glass can be chem wiped. Many of the chemicals used in the wipes are reportedly weak carcinogens as well (at least where I trained), so personally I would be concerned about putting something on my face that had been wiped repeatedly during a 16 hour shift.

However, if it's not touched, then it's no worse than eyeglasses and likely to be less of a fomite than a white coat or a necktie.[1]

[1] http://online.wsj.com/news/articles/SB125859205137154753?mg=... (Plenty of peer reviewed studies, but this is publicly accessible.)

grins

It was a perfect use of my nick, but more importantly I'm a computational epidemiologist who works on hospital acquired infections. It's mostly a curiosity, to see if the gesture and command based Glass system results in less carriage than some other tools. It likely cuts down on some things, but I touch my own glasses all the damned time to adjust them, and I'm used to them - I've been wearing them all my life.

As someone who trained recently (after hand washing became a widespread practice), I can say that I developed a consciousness I didn't have previously. Though I don't practice now, I still get this strong sense that my hands are unclean and need to be washed when I blow my nose, touch something dirty, am around someone who is ill, etc. But not touching my face (or glasses) was a much harder habit to break and something I still do subconsciously. So I'd also be curious how often docs using Glass do this.

Also, hospitals already sterilize surgical instruments, but I did a quick search and it looks like some manufacturers are designing devices meant to be deployed in nursing stations in response to mobile.[1] So maybe more use of these devices will actually drive increased use of rapid sanitizers within the clinic, as a secondary benefit.

http://www.infectioncontroltoday.com/news/2013/09/steridesig...

Yeah, I'm waiting for "dump your stuff in here" boxes at nursing stations that at least give everything a UV once-over or the like.

What is significant about the GP's nick?

http://www.letmewikipediathatforyou.com/?q=fomite

I get it now. I didn't know if he was someone notable.

> I didn't know if he was someone notable.

Not really, no.

Interesting article, but 2 things stood out:

1. "Google Glass immediately recognizes the room and then the ED Dashboard sends information about the patient in that room to the glasses"

This sounds like a redundant level of indirection. Why not just have a patient-unique QR code on the bed?

2. Very OT, but the use of "emergently" made me bristle. "Emergent" has a very well-established meaning that is not "very urgent". Sorry, end of rant.

1. Or QR on the patient's wrist identifier.

2) Agreed OT, but "emergent" to roughly mean "urgent" is pretty established in medicine. Here's a more thorough discussion:

http://blog.amamanualofstyle.com/2013/01/23/emergency-emerge...

An apparently deep discussion, but the author cherry picks definitions to match his argument. For example, he states two definitions for Merriam-webster, but they are only definitions 1a and 1b. There are 4 more.

Stedman's dictionary requires an account, so I can not verify at this time.

Well, here's Stedman (a standard medical dictionary) on 'emergent' from...a hundred years ago:

emergent 1. Arising suddenly and unexpectedly and calling for quick judgment and prompt action.

http://books.google.com/books?id=OG0XAQAAMAAJ&pg=PA290&dq=st...

QR code on the wall is probably better in this case because it's easier to spot from a head mounted display. The barcodes on wristbands are used mainly for patient ID when administering medications, where the user is close to the patient. Being able to walk into a room and look at the same place each time has the advantage here.

Maybe. But the ERs I work in rarely have one patient rooms, and the patients have a tendency to wander.

A patient unique QR code would need to be kept with the patient at all times in an easily visible place. Maybe a good solution for the long term but would require a lot more changes/training to ensure the codes stayed with the patient. On the flip side most hospitals already have a pretty reliable system for tracking which patients are assigned to which room so this approach can leverage the existing infrastructure and process.

1. A QR code on the bed would still be a layer of indirection and if it's patient specific, something else to print and get lost. Patients are already given bracelets with bar codes. But there are many reasons why they may not have used bracelets with their system: It may have felt too intrusive a way to greet patients. The backend system already maps patients to rooms and providers are used to thinking this way. Providers are also accustomed to confirming ID verbally with patient and always using the bracelet to be absolutely certain when dispensing meds or starting a procedure. Finally, in an emergent situation the bed can be surrounded by providers, and asking everyone to step back to scan a bracelet just to look something up in Glass would not be okay.

2. If this use of "emergent" makes you bristle, then stay away from hospitals because all healthcare providers and especially ED providers use the term this way. Every specialized domain has its own lingo and I doubt they care to change over a layman's concerns. Imagine how ridiculous it would sound if a non-technical English major tried to correct our use of all the technical lingo we use. (But this is seriously OT. So if someone is interested in exploring this, it might make an interesting separate blog/HN post.)

3. FYI, Halamka is a rockstar techie doc who built his hospital's EMR, serves as CIO, helps draft federal standards, sits on multiple boards, still sees patients, and runs an organic farm on the side.

Edit: Added more info and dialed down the frustration that on second reading might have sounded too much like snark.

>1. A QR code on the bed would still be a layer of indirection and if it's patient specific, something else to print and get lost. Patients are already given bracelets with bar codes. But there are many reasons why they may not have used bracelets with their system: It may have felt too intrusive a way to greet patients. The backend system already maps patients to rooms and providers are used to thinking this way. Providers are also accustomed to confirming ID verbally with patient and always using the bracelet to be absolutely certain when dispensing meds or starting a procedure. Finally, in an emergent situation the bed can be surrounded by providers, and asking everyone to step back to scan a bracelet just to look something up in Glass would not be okay.

Yes, when I thought about it a bit more, I could see advantages to having this indirection. An d if that is how the "system" is already set up, then all the better. I can imagine though that there may be cases of room transfers without a system update which may cause problems at times, but they will do anyway, not just for supporting this technology.

> emergently.

As a self-styled knitpicker and guardian of language, I had to look it up. The references I found [1] [2] indicated emergent can mean calling for immediate action or urgent, however, it's only in the weakest use of the word. And I find the adverb form of the word another step weaker.

I believe the definition you're highlighting is related to the dictionary definitions referring to "appearing casually or unexpectedly," such as emergent phenomena, which are, arguably slow. These near opposite meanings of the word are unfortunate, b/c people don't really know what sentences using the word are supposed to mean: Is an emergent page one that appeared gradually over time or one that must be attended to immediately?

I find the author's repeated use of the word clumsy and non-descriptive. Exactly the kind of trendy abuse of a word to signal self-entitled and undeserved superiority. To get paged emergently doesn't describe the process of getting paged for an urgent issue, since the action of paging is different from the issue (how was the page button pressed?). To "emergently reverse" blood thinners could just as easily have been stated as "quickly, immediately, or urgently."

Therefore, I would follow your suggestion and drop use of the word as a high-falootin' synonym of "urgently."

[1] http://www.thefreedictionary.com/emergent

[2] http://www.merriam-webster.com/dictionary/emergent

'These near opposite meanings of the word are unfortunate, b/c people don't really know what sentences using the word are supposed to mean'

By some strange coincidence, there is a discussion on such antonymous words under the current front-page article on mispronunciation of words: https://news.ycombinator.com/item?id=7384406

Guardian of language? That's a bit of a silly position to take, isn't it?

I would say that almost everyone who studies language agress that "guarding" language is bad.

http://en.wikipedia.org/wiki/Linguistic_prescription

Did you mean you are a self-styled 'nitpicker', and not "knitpicker"?

...wink...

'Emergent' is well-established in medicine to mean exactly how it was used in the linked article. There is nothing 'trendy' about that usage (see link to 1914 Stedman's Medical Dictionary definition below).

"Emergently" used in this sense is extremely common medical lingo and has nothing to do with the author. It sounds like you're judging this physician based on your own ignorance of the parlance of his field. There's no need to check how Merriam-Webster thinks the general population uses the term.

> This sounds like a redundant level of indirection. Why not just have a patient-unique QR code on the bed?

This necessitates entering the patient's room. A QR code on the door as well allows for accessing information without disturbing their sleep, for example.

"We replaced all the Google components on the devices so that no data travels over Google servers. All data stays within the BIDMC firewall."

I'd love to know more about that.

It's just a small Android strapped to the side of your head. Google 'AOSP'.

Thanks for the hint but I think that was obvious. My point is: the quote makes it seem easier than it is to do such thing and still be 100% confident and sure you will never ever have a confidentiality or privacy breach with patients data involved. Google Glass is fantastic but also very scary for a lot of people, I can't even imagine how that would turn out inside a hospital.

"still be 100% confident and sure you will never ever have a confidentiality or privacy breach with patients data involved"

That is never true, ever, in any circumstance, with any system, ever.

Stop wishing for unicorns and come back to the real world.

The question is, is it reasonably safe, and does it help patient outcomes? Keep in mind, the more likely it is to improve patient outcomes, the more risk is considered acceptable.

There's a certain niggling worry I have about googling "how to hide my data from google"

https://duckduckgo.com/

Presumably the thing is on a network without internet access. Boom, problem solved.

It doesn't mean patient data can't be leaked to the world (or even to wrong people inside the place).

Glass without an Internet connection doesn't normally work, so they must have done quite a lot.

The normal and promoted programming model is quite similar to Chromecast, where you push a subset of html pages to it and it displays them. That push occurs via your system sending a message to Google's server that pushes the update on the Glass.

That all changed last fall[1]. They released an SDK, so now it's pretty much just like making an Android app.

[1] http://arstechnica.com/gadgets/2013/11/google-launches-glass...

Apparently, it's all open source[1] and available to be hacked away at[2][3], including cutting out calls back to Google.

1: http://arstechnica.com/information-technology/2013/04/google... 2: https://code.google.com/p/google-glass-kernel-source/ 3: https://developers.google.com/glass/tools-downloads/system

The example in the article of the physician being able to respond quickly to a brain bleed and an allergic reaction because of information available on Glass was very compelling.

While it may not be practical right now for a lot of hospital use cases, being able to react quickly with data at your fingertips for emergency medicine seems like a great use of the tech. I'm sure the utility will only expand as the technology and the software integration improves.

Was the post deleted? It is still cached

http://webcache.googleusercontent.com/search?q=cache:xB8946F...

I won't be surprised to see products like this grow into a huge industry. Just imagine... surgeons wearing Glass or a Glass-like device which can alert the surgeon to abnormalities while performing surgery. Or have a HUD for the surgeon to monitor all sorts of data while performing surgery.

What if a "normal" doctor could pull up a patient's information while chatting with them, and have the current chat (which usually goes something like: "Any new problems?" "No." "Started smoking recently?" "No.") transcribed to the patient's file. Way less clunky than reading through the manila envelope of charts.

The scenario about being in ER and needing the patient's allergy medication, how did the doctor access the info on Glass? Was it just always on screen? Or voice activated? Could the same effect be achieved with a "Star Trek" style computer? The doctor would say the same command. I guess I can see how it might be more convenient in some situations to have it as visual info instead of a voice response.

I'm curious as to why Glass would bring about changes in medical care. For instance, I hear someone was interested in using it in surgery to record things. But certainly the basic technology of a head-mounted spycam has been around for quite some time. I was under the impression that Glass's innovation was making a hip, consumer-level product. Not that Glass brought new camera technology, for instance.

Ok, so hospital is one of the place that are so far behind in term of technology and someone had the idea of using google glass in ER.

Priorities seem to be a hard thing to do in hospitals.

1. Hospitals are not monolithic. They are a mixture of tried and true procedures and desperate, new-wave experiments.

2. Why would you say hospitals are behind in technology anyway? This would be an unfounded claim.

BIDMC is not exactly some struggling rural hospital with cold war era x-ray machines.

I wasn't talking about the diagnostic equipment, I was talking about the patient history infrastructures.

I'm actually more knowledgeable about Canada's health infrastructure but I'm pretty sure they share some characteristics.

> I'm actually more knowledgeable about Canada's health infrastructure but I'm pretty sure they share some characteristics.

Not really. The heterogenity of healthcare in the U.S. means things like access to patient information ranges from Extensive to the Point of Being Spooky (Kaiser) to "Who are you again?"

>We replaced all the Google components on the devices so that no data travels over Google servers.

This sounds intriguing. I think aside from the privacy issues, google glasses are stupid because many people don't want to wear glasses, but being able to have a CyanogenMod-type firmware for google glasses would at least remove the privacy risk to the user, if not for other people in the area.

I've shared this video before because I love the concept of augmented vision in this type of capacity. It might seem a little farfetched now, but I could see how something like Google Glass could reshape the medical and research fields:

https://www.youtube.com/watch?v=orXws08ODiQ

A sanitized version with much less information was posted: http://geekdoctor.blogspot.com/2014/03/wearable-computing-at...

This sounds incredibly useful in a hospital/emergency room. A definite step up from the use case of walking around town blinking to take pictures. Anyway to see more screenshots?

Link seems to be broken now. Too bad.

BIDMC in general is at the top of the health IT development and practice game. Good for them for expanding into glassware!

Try the cached version:

http://webcache.googleusercontent.com/search?q=cache:xB8946F...

This link works correctly http://geekdoctor.blogspot.com/2014/03/wearable-computing-at...

Reply on news.ycombinator.com