Hacker News new | past | comments | ask | show | jobs | submit login
The Healthcare System: An Apple Tablet's Biggest Opportunity (readwriteweb.com)
16 points by ca98am79 on Jan 26, 2010 | hide | past | favorite | 28 comments

Having started PatientKeeper (http://www.patientkeeper.com), I feel like I have an informed opinion here.

1. Tablets have been around for a while. They didn't take off. I'm not sure why an iTablet would do any better.

2. The integration costs of healthcare are crazy. Tablets won't help.

3. Older physicians have zero interest in using tablets. The dominate the decision-making and the teaching hospitals.

4. If it doesn't fit in a coat pocket, it is hard to get in front of physicians.

5. Battery life and connectivity are crazy important -- if devices don't last a shift, you're screwed. If radiology equipment messes up wireless, you're lost.

6. Entrenched workflows, especially those with legalities around paperwork, are very hard to extract.

Finally, I have to say that the "wishlist" in this article is laughable.

1. Vital signs can be measured automatically. It turns out to be a large set of data that a professional needs to interpret to find useful. Realtime makes a great demo, but no one cares in practice.

2. Shift changes are huge, but tables are a tiny fraction of making them work. Shifting the responsibility between provider has never been a technology problem -- it's a communication problem.

3-5. Really? Videos for the patient are a critical element of workflow?

How about something real: checklists and best practice management. That makes a real difference.

Perhaps the author ignored it because paper works just fine.

MSIV here.

Sanj's post is rather insightful, and here's why: If it doesn't fit into my coat pocket, then it will end up being something that probably doesn't become 'mine'. Instead of having one per doc, there will be a few for each floor.

Since individual providers don't then have a personal relationship with any particular tablet, they all have less incentive to care for them (charge them, make sure they get put back in the right place, not treat them like coasters). No, the nurses will not do this for us.

This brings us to the point where we have a floor with several tablets that, if not stolen or covered in "someone else's" coffee, have no battery. If there is a backup battery system, all of those backup batteries are also drained. So far, so bad. But it gets worse: when the next set of docs rotate onto service, they see a totally dysfunctional heap of tablets and won't go near them. I say this from experience with the current iteration of this idea: 'cows' (desktops mounted on rolling platforms with a battery).

In my opinion, the best way to succeed is to make a product that each individual provider will develop a relationship with. You will succeed if you can make me treat my tablet like my iPhone because it (a) is 'mine' - not communal; (b) is portable - fits into my fairly giant white coat pockets, or is otherwise easily transportable; (c) gives me something that I want - not what some non-medical person thinks doctors might want.

Regarding point (c), sanj's statement about realtime vitals shows that he has discovered something that AirStrip hasn't: realtime vitals in most cases are for nurses, not doctors. I'm not going to be actively monitoring someone's heart rate unless I'm at the bedside. They are promoting the wrong feature, IMHO.

Sanj, it sounds like you have good insight into what doctors actually want. Please keep it up!

Nothing pocketable will be more useful than the paper it aims to replace. At a minimum, you'll need larger pockets. And even then, the reality of weight, battery life and connectivity are at odds with a per-physician solution.

I actually work as a director of medical informatics at a large, well known, medical institution that field tested and uses patientkeeper. Sanj's assertions are spot on. I won't adress all of them individually, but they all speak to real problems in the "digitization" of healthcare.

For those of you who think any singular device will be the electronic panacea people are waiting for in healthcare you are out of your minds and have no clue about the current state of medical informatics. Getting medical records systems into the healthcare environment with entrenched users is probably harder than getting the us legal system to use them.

As I have said in other forums, it is not any singular device or platform that will make this work. The current landscape is that of a patchwork of systems silos, none of whom talk to one another. There is practically zero integration at a platform level and any integration that is done is ad-hoc and custom.

If I were advising Obama, et al, I would Incentivise (with a capital I) data interchange over any specific data format or platform. Do not pick winners and losers, in my opinion it is kind of un-American. Let everyone have a fair shake in the market but encourage them to open their systems to integration from third parties through open read/write (bidirectional) api's. This would go a long way to getting new innovative systems into institutions with entrenched legacy systems.

Speaking to integration costs, point 2, I could not agree more. Working at any institution other than a small sized private practice requires an untold amount of training, education and ongoing maintenance to get the users to use any system of even minimal complexity.

"Wishlist" 1, again spot on. Doctors do not monitor vitals in real time. They just do not have the time, nor is it pertinent to quality of care. What doctors do want are intelligent alerts. But how do you get an intelligent, dynamic, real-time monitoring system that needs access to raw data? See my point above re open api's.

"The current landscape is that of a patchwork of systems silos, none of whom talk to one another. There is practically zero integration at a platform level and any integration that is done is ad-hoc and custom."

I don't understand this part of your comment. I'm not aware of any hospital information system which doesn't support HL7 to some extent.

HL7 is simply used as a data format, like json or XML. Newer versions of HL7 are actually XML-ish in format.

Saying "HL7" on it's own is nowhere near a solution. What I mean is that systems/platforms need to be encouraged (incentivized) to support bidirectional data exchange in any data format or transport mechanism the market will adopt. Like, say, Twitter. An API that would create and read internal data elements on the individual platform. This would allow systems providers to maintain their proprietary advantages but expose their systems to larger integration.

Trust me, the current state of affairs are anti-competition, anti-innovation and down right piss poor for patient medical care.

The HL7 specification defines specific message segments for specific kinds of clinical data. It defines the events which should trigger different types of messages. It defines a minimal protocol to be used to transmit messages (which nearly everyone uses). It is not at all analogous to JSON nor XML. HL7 3.x isn't just XML-ish; it allows the use of standard XML (in addition to the classic delimited format).

Again, I don't know of any hospital system which doesn't offer one or more specific HL7 interfaces (ADT, results, billing, etc.). It is very common for hospitals to have clinical document repositories which accept multiple formats.

While I agree with your general sentiments about the state of the industry, I don't believe the problem to be a lack of standards or opportunities for integration. Additionally, the incentives you speak of will happen at the end of the year. See my comment elsewhere about CCD/CDA (CDA is part of the HL7 standards). ARRA funds will be available at the end of the year for systems which implement HITSP standards (I'm oversimplifying a bit).

I wanted to add that I believe the industry's biggest problem to be a lack of software development talent. No top MIT graduate is torn between working for Google and working on automating the paper workflow of a hospital. Although there are several standards available for specific types of information exchange they are rarely implemented properly. They are also poorly designed. The CCD standard is quite atrocious.

That speaks to the intolerable nature of working in one of these bureaucracies. I could not agree with you more. There is a lack of talent in this field and a lot of it has to do with the leadership. Decision makers not only do not understand the potential of suggested solutions but also are unwilling to take the risk on homegrown or non-vendor supported solutions. In doing so they alienate the talent that works for them and forces them to look elsewhere.

I'll outline a hypothetical problem... Large institution passes on homegrown solution (no vendor support problem) buys large commercial system. Large commercial system takes years to integrate into med center culture and legacy systems, problems abound. Innovative students/professors/physicians/staff suggest enhancements to better... everything and do things the system does not and was not intended to do. None of this gets done due to short sighted, risk averse administration and lack of integration support from vendor. Nobody at med center knows how to decipher the gargantuan database schema or access their own data (programmatically) outside of the commercial gui. Frustrated talent walks out the door. Rinse, repeat.

Now, if these systems came with documented bidirectional open api's then people could just do whatever they wanted (with appropriate access controls, duh). If someone wants to write a new and better client for Twitter they do not have to ask twitter for permission, they just do it. Although vendors do use HL7 for ADT related activities, there is way more to a modern day EHR than that. And before people start beating me up about protected HIPAA data, simply control access tokens to pump fake data during development.

Speaking to your other point about HL7, it's inherent standards, CCD/CDA and HITSP, I agree, there are certain proposed solutions but how are they actually employed? Whatever incentives ARRA is offering they just are not enough or geared in a fashion that may make a dent, imho.

I don't know if you can tell, but I am highly disillusioned as to the state of the state in this industry. Best as I can tell leadership in this area will come from out of the way, non academic, regional for profit hospital conglomerates that will force change to drive profit. Once the rest of the field sees their results they may join in in any standards those leaders employ. Hopefully they will include the bidirectional open api concept. I'm not holding my breath.

EDIT: When I say "if these systems came with documented bidirectional open api's" I am speaking to your point re "they are rarely implemented properly". These existing api's may get the job done, but they are rarely implemented and if they are, they are certainly not implemented properly. Perhaps that should be incentivized as well?

Thank you for a great post and your insight.

I don't know this, maybe you do, but isn't most health care directed software written for Windows / PC as well? Would having an iTablet limit the software options available to physicians as well?

Most systems have some sort of web interface these days.

1. the iPod touch is already a far better tablet than any iteration of, say, Windows tablets. If the iTablet shares the iPod's ease of use, reliability, durability, battery life and toolchain, there's no reason to expect those old lessons apply.

2. agreed. but they promise to be far more effective and far less costly than putting PCs all over a hospital.

3. they didn't care for computers, email or smart phones either.

4. Any effective device is either going to be nurse- or bed-/room-centric. Frankly, even great interfaces aren't so fast that it's worth the time for most doctors to navigate the UI to bring up a patient's file himself. Nurses move, but doctors move more, their waste is more expensive and their frustration/errors more dangerous.

5. battery and connectivity issues are less a problem in nurse-centric scenarios and a non-issue for place-centric.

6. think of it as a replacement for the ever-present clipboards and physical files. Instead of handing off a file, the nurse hands over her tablet. The legal aspects will have to be updated for a digital solution of one type or another. That ship is set to sail with or without tablets.

Also: checklists and best practice management would be far easier to capture, manage and report on, if it was being digitally captured. Paper works fine for the physician because she can fudge it after the fact. Were it captured at time of service and the lists themselves generated dynamically based on the particular case at hand, more errors would be caught, sooner.

(I've worked on several smaller projects in the healthcare industry and spent more than my share of time in hospitals, ERs and doctor's offices.)

Somehow I don't see how a new hardware platform with a shiny interface is going to fix the gnarly problems with electronic health records. A nice UI would certainly help with filling out forms but that's only the tip of the EHR iceberg.

The last few times I've been to doctor's offices/hospitals, I've noticed that a huge amount of doctors'/nurses' times are spent mouse clicking.

The docs' offices all have computers in them with EPIC or similar patient record systems, and in the ER they have mobile workstations on carts positioned throughout where docs and nurses input data after every patient interaction.

Healthcare IT depts can already put computers very close to the point of care to provide history, lab results, and input functions. A tablet would just be something for caregivers to lug around (how would they carry it anyhow?) and potentially be stolen.

The main problems in healthcare IT (as I understand it, anyhow) are not input devices-- it's software quality, usability, integration, and above all, cost.

I bought a used HP tc1100 tablet. I think Windows XP's Tablet software is amazing. When I hover over an entry field, a button appears to bring up the writing box, and the recognition is very close to 100%. (I always print, however. I hear the recognition for cursive writing isn't nearly as good.) The industrial design on the tc1100 is on a par with Apple's, IMHO.

If Windows tablets can be this good while embracing the old Windows UI paradigms, Apple had better be mindblowingly good. I would be very pleased if I can have good pressure sensitivity and accuracy with an eraser stylus, plus multitouch.

I think someone made a big mistake by not trying to sell this stuff to the public and concentrating on business instead. Surfing with the stylus is very nice, I find.

It does seem obvious that such a device may be good for EHR and such. Timing will be the problem. This year is the big push for implementing EHR. The systems that will get implemented are ones in existence from numerous commercial vendors and there is one solid open source solution, VA's vista.

Most doctor's offices already have scheduling and billing systems. The gov's incentive funds are oriented to purchase existing approved EHR systems. Its a difficult process to get a new app approved for the gov incentive funds.

So the shiny new tablet, as cool as it may be will most likely not be the compelling peice that makes the sale. The compelling piece is integration with the doctor's already existing scheduling and billing system and that which is approved for gov funds.

If Apple or someone else can do for doctors what iPhoto has done for snapshots, then they've got it! Host EHR software in the cloud. Just make something which doctors and assistants can just buy, which will generate forms or integrate with existing systems in some fashion. Granted, many of these systems are archaic. Perhaps too many of them will be too crusty to be capable. What percentage of the scheduling and billing systems can use Microsoft or Adobe forms for input?

Here, a completely closed ecosystem in the vein of the iPhone with its App Store can be a tremendous boon. This will help assure Docs that their hardware doesn't have malware installed. If Apple builds in some form of Hypervisor executable DRM, all the better.

The doctors could deduct the cost of such a tablet as a business expense. If it does media, magazines, is a passable e-book reader, can be a home automation and FrontRow remote, then I think Apple could sell tons of them!

So basically if Apple can come up with a system that can interface with all existing records systems used by doctors and hospitals and can act as a complete personal media device and home automation control then they would sell a ton of them?

Seems a bit of a tall order even for Apple. A lot of hospitals have homegrown records systems or a whole collection of interconnected commercial products. The physicians have also gotten used to the existing systems, and in many cases have been involved in customizing the system to their specific needs. If the Apple hardware is nice (which it probably will be) then that is a good first step, but it will require the involvement of a lot of third party developers to create better medical record systems.

It's the integration with existing systems that's the real problem. Industry standards in that area are still incomplete and not fully implemented by all vendors, so usually integration requires a lot of custom work. Hosting EHR software in the cloud doesn't solve this.

I am thinking when you say Host EHR software in the cloud that you mean hosting the data.

What are the implications of HIPAA in the cloud, given that Amazon says that S3 is not valid for credit cards?

Let me shed some light on the subject by explaining what is actually happening currently. CCD/CDA documents (XML documents containing clinical information) are being pushed as the new standard for exchanging medical information (mostly by IHE which is in turn being backed by HITSP). These documents are stored in an XDS repository. Currently these repositories tend to exist at the IDN level (large groups of hospitals and physician offices). The real question is how we are going to get these repositories exchanging documents with each other.






Thanks for the good links.

But I am still hazy on how this set of documents and the storage of them is impacted by HIPAA.

My regular doctor (I'm in Canada) makes frequent use of tablets, but I've never taken the time to find the make/model. This article reminds me to ask, next time.

The make and model of the tablet itself barely matters. They are likely Windows tablets. The software is what matters. The only question in my mind is whether or not a tablet with the capabilities of an iPod Touch/iPhone will enable better software.

See this old post: http://tinycomb.com/2010/01/09/breaking-apples-tablet-is-for... It appears that a company named Motion Computing Inc (and i guess they are not the only one) already sells a tablet that targets the healthcare industry.

My employer provides solutions for electronic patient reported outcomes for clinical drug trials, and Motion is our provider for tablets for use in clinical settings. We are anxiously awaiting information about the Apple device, and the HP device shown at CES, as the Motion tablets can be a bit overkill for some applications.

Our industry, at least, is very much ready for more/better tablets.

I sign tablets when a courier comes to the door, and it's such a random squiggle that I doubt such a tablet is accurate enough for healthcare.

I'm expecting Apple to announce a tablet tomorrow.

Apple has been completely tight-lipped on the matter, and the only thing they've said about tablets is that they don't know how to make a good one.

So why am I, why is everyone, expecting a tablet? What evidence is there that should support that expectation? Apart from the question "what else could it be"?

And given how much effort Apple puts into keeping things under wraps, how did so many people and media outlets get so convinced a tablet is 'due' from Apple?

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact