Hacker News new | comments | show | ask | jobs | submit login
Healthcare tech ideas we'd like to fund (rockhealth.com)
75 points by malay 1520 days ago | hide | past | web | 29 comments | favorite

As someone both with a chronic disease, and who works in Health IT -- personal health tracking is something I'd very much like to see (and am working on in my spare time).

Part of the problem with chronic illness is that our (well, my!) dumb monkey-brains don't cope with it too well. I degraded, slowly, over the course of two years -- ending up in hospital very recently. On a day-to-day basis, over that course of time, you just don't notice you're getting worse -- what's an extra minute of cramping, introduced over the course of a month? Having a tool which tracked me, and would let me see (in a chart plotted over time) that I've crossed a line and need to seek help would probably have helped me avoid a week-long stay in hospital, and a commitment to take some (relatively) nasty drugs for the next couple of months.

When I got released, I went and bought a Fitbit Aria for weight-tracking (and am building something that'll query their API and expose my weight data, alongside other data I'll track via a mobile app), and it's excellent. The ease-of-use is the main thing; not needing to manually enter my weight after weighing myself means my data's always there. I'm not sure how I'll accomplish that for the other stuff I want to track, but it should be interesting figuring it out!

So, in short: this is an area I'm stoked to see investment in.

How would something track (to use your example) cramping? Short of having a device where you press an "I have cramps" button then press it again when they go away, I'm not sure how any type of software could support your specific example.

That being said, personal health tracking is definitely a area of interest, and Fitbit and others have proven that there's money to be had and good to be done.

Cramping's one of my symptoms; I'm being a bit coy. To be more blunt (the disease is Ulcerative Colitis), the main thing is bowel movements per day. At the moment, I've got an NFC tag sitting in the bathroom that I swipe to log a motion (as the path of least resistance)... but (as mucky as it sounds) I'd like to hack something together with an Arduino/Raspberry Pi and a weight sensor attached to the toilet, which could detect that I'm sitting down and log a movement for me automatically (in much the same way the Aria automatically logs my weight!).

This'd need to tie into a mobile app for manual logging anyway (I'm not exactly going to rig this up at work!) with a backing web service, but yeah. There's definitely room for something clever in this space :).

That's a really good idea. I could see a small Arduino board with wifi and a FlexiForce[0] (just the first thing that came up in Google for "Arduino weight sensor"), and like you said a mobile component that would either be a stopwatch to actually log or a time entry for past movements (or when you didn't have your phone, etc).

My sister has Crohn's and one of college roommates had UC so I'm somewhat familiar with it.

My contact info is in my profile if you want to talk about this more.

[0] http://bildr.org/2012/11/flexiforce-arduino/

That device is represented in every phone and tablet. It just needs the software to capture it.

It doesn't even need to be real-time. An app/text message that asks you every three hours "been cramping up lately?: No, once or twice, several times, lots"* would collect a lot of relatively accurate information, and easily show you 'whoa, I cramped up almost every hour this week, two months ago it was once a day!'

*in my answers I'm assuming that we are tracking the incidence count, rather than duration/severity - this is easily adapted if necessary.

1) A mega platform on top of existing EHR's won't be an easy thing. It would require them to play ball and large players, like Epic, probably won't. 2) Chronic disease management is important, but really challenging. People don't really spend their days thinking about their illness. They don't look forward to having their blood sugar assessed. Any kind of disease management program is going to require that the patient be minimally engaged, yet data still effectively collected and assessed. The only evidence of real consumer engagement in health has been with respect to personal fitness. 3) HIPAA in a box is going to require some involvement of the government. They've thrown these regulations out there with rather vague specifications as to how they would be applied to ehealth initiatives. If the government wants to enforce this, which they have begun doing, they have to provide better guidance.

Just a few thoughts. In general though, Rock Health has grown into a great program and anyone who is pursuing a health related startup should give it strong consideration.

>* A mega platform on top of existing EHR's won't be an easy thing*

I built something like this in 2009 and got pretty much noplace with it. Our app, Contineo, was an HL7 compliant, back-end agnostic, mobile front end that would work with any HL7 compliant backend.

We were stonewalled by Epic, lead along by Siemens (who had begun talks with us on licensing our technology), we embarrassed Eclipsys as they didnt have a fully HL7 compliant system....

It was really difficult. So we opensourced the tech and gave it to Medsphere to integrate into OpenVista.

I still think there is a significant opportunity in disrupting healthcare in the large hospital, but I don't believe its via the EHR at this time, but through the patient room and patient experience.

If there are any folks out there that want to collaborate on something - I'd love to meet up.

(I applied to Rockhealth's second class and got too interview - but we didn't get selected.)

hey, would love to talk with you a bit more as a fellow med/tech enthusiast.

Me too, 2nd year med student and tech enthusiast.

Include me as well. !

These are some good points. On the EHR question, I think what we would like to see is a company that starts using the work coming out of Standards & Interoperability (S&I)[1]. One can reasonably expect that much of this work will surface in Meaningful Use requirements (since one of the primary objectives of MU is interoperability) and thus would force vendors, including Epic, to comply. The platform here would do the "hard work" of integrating under the S&I frameworks to multiple endpoints and exposing the integration through developer-friendly APIs, similar to what Eligible has done on the administrative side. We see this type of "easy" integration leading to more modularity for EHRs (which is supported by Meaningful Use requirements) and ultimately better end products for hospitals and patients.


Why do you think past efforts haven't "panned out" yet? Is it:

a) they're working on it, but just not there yet

b) the major EHR do everything in their power to delay progress by providing token access

c) it's actually very hard to sync medical records across so many different standards / systems

d) not enough patient / doctor demand / use-cases, i.e. people don't transfer as much as we think they do, diluting the value of inter-operability

? or something else? all of the above?

This area of HIT fascinates me because I hypothesize that it's a necessary precursor to reform. But, knowing little about the space, I'm not sure where to start my investigation.

None of the above, IMHO.

In San Francisco, Kaiser could electronically share patient data with every other system running Epic with very little technical work and price. We're talking hours of work here. That would let them share data with UCSF, PCMH and every other Epic site.

They don't. Why not? It's because sharing patient data makes it easy to poach patients, and that's bad for business. There are similar situations in other cities too.

I don't blame them because the government hasn't made data exchange required yet. It's coming, with some of the requirements for Meaningful Use Stage 2 & 3 and the work done by the S&I framework. APIs will help too, because HL7 is confusing to newcomers and costs money. Vendors could make this easier too, but it's not like Epic or Cerner could compel anyone to share patient data.

I have a more in depth answer about this topic here: http://www.quora.com/Electronic-Health-and-Medical-Records/W...

I would think HIPPA is one of the issues here. Federal law says it is a big no no to freely share health info without a raft of signed paperwork, basically.

Typically customers handle this through some type of waiver system. Generally at the initial visit in an organization a patient signs a ton of releases that state that the parent organization allowed to share patient information as needed through the course of treatment. Most people want to share their information, but there are systems in place to restrict access as needed by law/best practices. Patients can opt out of data, and some data sharing systems support this through various technical choke points. If you were a nurse at one organization, you may not want people snooping through your medical records at another medical organization. The Direct Protocol supports this through pseudo-anonymity of accounts.

I worked at an insurance company for five years. Part of my job was to get authorizations so I could request medical records. I had annual hippa training. I was not allowed to read any medical records or request any medical records that were not directly involved in doing my job. One standard of hippa is "minimum necessary." That means I am only entitled to as much information as is absolutely necessary to do the job and not more. So I suspect that some system to share records has serious challenges.

1) A mega platform on top of existing EHR's won't be an easy thing. It would require them to play ball and large players, like Epic, probably won't.

They definitely won't play ball with a platform attempting to integrate all of them. That's just a risk; there's no upside. It's possible that none of them would welcome a "platform" at all.

But survey the landscape of EHR vendors. There are a few unique positions of strength out there. Each aspect of uniqueness defines a hypothetical market. The mere existence of some of those markets would develop an asymmetry that simultaneously penalizes competitors and solves a big problem for the partner. Build the spark of the market and you have a significant ally with a strong incentive to nurture your business.

Idea quality indicator: you want the existing player aligned with you, not buying your company (then killing it). They should need you externally.

I'm building a business on one of these hypothetical markets. I made my choice based on my personal capability and fit with the details of the problem. But there are a huge number of different ways to improve both the quality and efficiency of healthcare. (You don't get to solve big problems if there are no big problems.)

"...make everyone of us a doctor..."

That's a very, very bad idea. Should we educate people about their medical problems in a simple and smart way? Yes. Should we delude people into thinking they are doctors? Absolutely not.

"IANAL" is one thing but there aren't enough asterisks in your keyboard to save you from that bottomless pit of liability.

I tend to agree in general. And expecting advice over an internet forum is ridiculous in the best of all worlds.

But frankly, you don't need a high school diploma + a Bachelor's + med school + residency in order capture a huge fraction of a Real Doctor's value as a classifier. Machine-learning-derived classifiers can do even better. This is where there are huge gains to be made.

Every time where cutting out the BS actually matters (pun intended), TPTB "discover" that you can churn out functionally-equivalent doctors with significantly less training. See: military training of doctors, WWII. If you suggested turning away people from doctor training back then because they didn't have a 4-year degree already, they'd laugh or have you committed.

One of their ideas is quite close to something I'm working on.

However, I wasn't exactly thinking about using an accelerator, and I haven't head about them. A quick search on HN doesn't give a lot of results.

Does anyone here has had some personal experience with them? Feel free to contact me by email if you do not want to post a reply.

[I'm not interested on the PR I can find with google, but by personal experiences, opinions]

This is helpful:


We moved from Wisconsin to be a part of the V4 class, and it was amazing. To be surrounded by great business and technical minds who all give a damn about trying to make healthcare a bit better sustainably is really cool. I'd recommend it.

I was in the first batch at Rock Health, and now continue to help out as a mentor. It was a great experience for me, and it's come a long way in the last couple years.

As with all accelerators, your mileage may vary. If your key needs are technical mentorship or great design leadership, Rock Health won't instantly solve those problems for you, though there are really smart folks working alongside you in the office who may well have faced some of the same challenges before.

What Rock Health does bring is a pretty amazing network within health care, and surprisingly direct access to the people and companies that can help make things happen for you. Need a trial in a hospital? Rock Health makes it (comparably) easy to get one started with top-tier places like UCSF and the Mayo Clinic. Want a meeting with a major payer? United Healthcare will be in the office on Thursday, Kaiser Permanente and Aetna are coming next week. Not sure what particular regulations mean for you? A prominent consultant holds office hours twice a month, and one of the heads of Medicare is coming by soon.

As someone whose career was built designing great products on the consumer web, I got keenly interested in some of the healthcare challenges once I met my wife (a pediatrician). But so much of the space is inaccessible to outsiders. Rock Health was the perfect venue to learn, build, and connect with the digital health space, and I'm a far more capable technologist as a result.

It's still a new program, and there are certainly improvements to be made (and, as I mentioned, it's come a long way), but on the whole, I'd give serious consideration to what Rock Health can bring to your work. For me, it was priceless.

If you don't hear from one of our alumni because they happen to miss this thread, feel free to e-mail me (e-mail is in profile) and I would be happy to connect you with someone.

This is well worth applying to -- amazing program

wait. they want to invest in one of the most heavily regulated and friction filled business domains out there and the seed is 100k!?!?!?

this is a joke right, as in this is really the onion under the covers...

Funny, I actually have an idea for their top listing. Funny to see that listed, I mean.

"Your Black Swan idea"? I hope they realise that...


...she dies at the end.

I think they're probably referring to Taleb's concept of black swan (http://en.wikipedia.org/wiki/Black_swan_theory)

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact