I'm a longtime reader of hacker news. I'm looking for a technical cofounder. I know this sounds bad but hear me out. I think a little bit about my background would be helpful. I graduated from UCSD in 2005 with my degree in computer science. I worked for a couple of years, mainly with PHP, before I applied to medical school. Currently I am a first-year resident at UCSF Fresno. After the year is over I will be heading to USC to start my radiology residency. I still try to code every once in a while and I know some Ruby on Rails, HTML/CSS, and JavaScript.
With all the time I've spent in the health care system, I seen a lot of opportunities to start a company that would solve some of the problems I've seen in the hospital.
Many of the problems in healthcare involve two systems that need data from each other. For example there is a machine that measures pulmonary function and will generate a PDF report. This PDF report has values that need to be entered into an EMR/EHR. This is time-consuming for some of the doctors and doctors should not be wasting their time with data entry. There is a stack of these reports and 10 minutes are spent entering in data and another one minute is spent interpreting the data.
I'm fairly certain I can write a program to do this and make an extra $500 per month from the hospital but starting a company around this idea seems to be a better option. With that being said, I'd rather have an engineer handle the coding and I would handle talking to hospitals and making sales. Many hospitals in the area need this sort of work done and I know we could generate income in this sector.
I'm looking specifically for somebody in California who can spend 2-3 hours per day coding on these projects. This person would probably be working on their own projects in their spare time. If this sounds like you please drop me an email at mtran115@gmail.com and we can talk some more.
First, be certain that you can make $500/mo by building it yourself and get the hospital paying $500/mo.
Then, if the opportunity looks good, start organizing (aka building an organization, a company) around that.
You'll short circuit a lot of pain that way.
It sounds like you're seeing a lot of problems in healthcare. That's good you're noticing them. It sounds like you also want to solve some of them, mostly because you're fed up with them being problems. That's also good. The best kinds of solutions often come from scratching your own itch.
Another similar recommendation: get more specific about what you want to solve. "Disrupting healthcare" is a platitude anyone can agree with because the news runs a segment every day about inefficiency in the health system. If you see 10 problems, does 1 stand out way above the rest? Work on that problem.
Just building a quick & dirty solution and shopping it around like smalter suggests is the best thing you can do to prevent working on a problem you'll burn out of in 6 months. You'll very quickly realize whether you really want to do that thing or not.
Don't just try selling something that interacts with medical data unless you have serious quality control (you may need various ISO certifications in place) and a good grasp of privacy law. That's what many of the frustrating hospital purchasing rules are trying to ensure.
The reason that the healthcare world has so many problems with obvious software fixes that haven't been fixed is because they're very rightly paranoid about using unsafe software, because sometimes it kills people even when it's designed by a corporation with lots of quality control (cough cough THERAC-25), and not just a side-business with a doctor and a hired coder.
I'm currently chewing on ideas from "Lean Startup". It suggest to try and validate your idea before even building a product, for example with a interactive prototype. If the OP has a good contact with his hospital, he can show the prototype and ask: "If I build this, will you buy it? Will you pay upfront for a discount?"
I looked briefly into building it myself (just in case this thread did not get a good response), and I can parse out the text using the pdf-reader gem and now I just need to use some regex to match what I want and then stick those values into a report.
Thanks again for the advice.
When I worked for a construction project management startup, this too sounded reasonable: many software systems for this audience were in the six-figure range... we were only asking $100/month or so. ("And a free trial!!!")
We learned that when your monthly price point is the size of an accounting error, no one can take you seriously.
This software was also holistic in nature: everyone, everyone, everyone has to use it for it to work well and as a core part of their organization. It's a superset of the marketplace business problem.
To address these we had to charge more. Much more. This raised the stakes on both parties and also filtered out less serious buyers who would probably fail with our software.
The construction companies who were serious buyers really considered thoroughly whether they could implement the change; we worked with them for, sometimes, months going through how the change would take place. And training, training, training. Usage metrics. Account hand-holders. Keep them using the system.
--
As others have said, medical software is some of the hardest to break into. Medical staff HATE CHANGE, and need to. Change in the short term leads to mistakes and mistakes cost lives or careers. The system is bureaucratic, politicized and slow moving. Anti-disruption.
I believe that a guerrilla approach to insurgent medical software is ultimately what will work - things like https://www.radiologyprotocols.com/, where a radiology tech wanted a common repository of knowledge for others in his field. Then, through word of mouth and using it with his colleagues, it gained first use in his hospital, then international use, and now it's starting to take off in the US. (Reminds me of "Big in Japan" first, or conversely Japanese artists having to become popular in the US before being taken seriously back in Japan.)
--
Okay, advice.
Sell it before you make it. If you have multiple hospital/medical contacts, play with pricing. Try outrageous things. If not, just try to work out any sort of deal and have them sign a piece of paper that says they'll pay.
Then build it.
Iterate with the first customer until it's "So Good They Can't Ignore You".
After this, you might get better traction outside the US. There are hospitals in other countries who are desperate for good software and also don't have the money for the large EMR software many US hospitals use.
GOOD LUCK!
If a Dr. sacrifices their productivity for a short time, with the result that he increases the productivity for all Dr's; then the Dr. has done the opposite of cheating.
>I can only imagine what he will be doing 7 years from now.
OP could follow your advice and be an average Dr. Or OP might actually succeed in improving wasteful processes in hundreds of hospitals.
I already work 60-80hrs/week, should I be working more? Would 100hrs/week make you happy?
> I can only imagine what he will be doing 7 years from now
6 years from now if I continue on this road keep my head down and work hard I'll be making 300-400k/year as an attending physician. I can work 4 days per week and make 200k. I don't care too much about money as long as I can pay my loans off and eat out once in awhile. That's all nice and dandy but I'd rather spend some of my free time building something that helps the healthcare system.
I'd love to talk to you more about this because these are problems I'd love to see solved. I think its a real up hill battle though.
Make friends with a lawyer. Read up on everything legal you can. Pay your lawyer friend when you need lawyer advice and lawyer action. But don't give them not only full-time, but founder, level status unless they are truly a full-time founder.
Plenty of others share their early-stage lawyer advice freely so startups don't need to get in this situation. Check out:
* http://rockhealth.com/resources/startup-elements/hipaa-compl...
This is part legal, and part QA, I believe.
I think I can avoid the whole thing by having my software just be a tool, not a final read, and doctors will still have to double check the numbers against the PDF hardcopy. It will just take 1 minute or less to glance at 10 values or so and make sure they match up compared to typing it in themselves
Manual entry errors happen at predictable rates, and with good processes you can get them down quite low.
Software error rates are not so predictable (unless you have full control over all variables... and that's surprisingly hard to get). You could deploy this software in the context of a slightly different version of the radiology software which used a different font in PDFs (or whatever) and find that all 8's are now read as 3's; no human would do that.
If a human started getting input that wasn't sufficiently readable, they'd talk to someone about it. Could the software do that?
One, never approach a community that lives and breathes startups using words such as "fairly certain," and "try." They have no power here.
Two, start the company, make a respectable business plan, and then whore yourself out. No one is going to be interested in a joint venture with a few paragraphs of information to go on.
You don't have to be perfect, nor do you even need a full roadmap of ideas or plans, but people need to see that you're willing to put effort into it. No one makes a successful startup, especially in the medical field without considerable research and effort. If people don't see that you're willing to put in the time, they'll never get behind you.
Good advice. Although sometimes, the best way to get a fire lit is just to set something up and wait for IT to complain...
EDIT: To anyone in the Bay Area (or really anywhere) also thinking about joining this industry, feel free to reach out to me too. My cofounders and I have learned way more than we want to know over the past few months about the healthcare industry + the startup side of things and are happy to pass on our knowledge.
As others have rightly pointed out, the technical issues are perhaps 10% of the problem. On top of that, because each EHR is configured so differently, your solution will be a consulting effort not a product (read: expensive)
My suggestion - thoroughly evaluate how difficult this would be to get approval for at your hospital. If you get past this stage, talk to other hospitals. Learn what the implementation process will be at each hospital. This will likely dissuade you but, if not, you've found a niche.
The reality is that this is not an IT problem but a process flow problem. Somebody with a lower pay grade should be entering this data, say an MA or a 1st year resident, until the IT team can prioritize this known issue.
What this means is that usually in the medical world you can't just build a product or service and sell it to EHR/EMR providers. Their (e.g., Epic, eClinicalWeb) customer is hospitals and clinics. So you need to get hospitals and clinics to demand/buy what you offer. Only then EMRs will pay attention and only because increasing amounts of their customers are knocking down their door asking for your product.
Very rare is the case where an EMR will see the market opportunity you're carving out and buy your product. They don't care. They'll either buy your company once you get big enough to represent a sizable addition to their revenue (single millions is not even close) or just build an imitation feature once they're forced to.
We have a group of friends who are familiar with healthcare and have programming backgrounds.
Have lived and worked in some of the finest hospitals in US (viz: Ny Presbyterian and KCH brooklyn) and also in rural India this has given me a very good understanding of health problems that can be fixed with technology. Have already done some good amount of work and things are already in place. Need someone who can complement me with the right skills.
HealthEnclave.com
Am looking for some one good with-
- Django / Python / Flask - Tastypie / Django Rest framework - Fronend Desing (HTML5, CSS3, Jquery, Bootstrap, Less) - MongoDb / PosgreSQL
additional skills: (not necessary but it would be great if you have 'em)
- Objective C (iOs dev) And JAVA for (Android) - Some understanding of ML , Data analytics - Electronics prototyping with Arduino
I am familiar with and have some amount of practical working knowledge of all the above listed technologies.
Thanks Asingh [at] HealthEnclave [dot] com
I've dropped you an email. We're based in Singapore, but we're happy to work together.
Btw, is HealthEnclave meant for wellness or do you plan to develop it along a more serious angle?
Looks like a content farm?
With that said I am not trying to make a quick buck here.There are real life pain points that I am trying to solve. The longer it takes for me to create my vision the longer people will have to live with pain, suffering, anxiety (because of lack of proper information to make choices upon).
If a good surgeon who took Fees for his services die tomorrow then it would not matter because his patients who just go to another surgeon. But if this doc creates something that solves a real life issue that no one is able to solve or maybe he creates the best one.. Only then would IMO would it even matter that this person even existed
This. From my limited experience, even with all the systems able to talk HL7, apparently what message segments the system sends, the message segments the system reads etc. are different from system to system. The result is entire subsystems or heavy customizations for redirection, massaging, and transformation of messages. Something to think about systems integration.
Another approach is to buy subsystems from as few vendors as possible (naturally subsystems from one vendor would all talk among themselves very nicely) - but that is obviously not without any repercussions.
I'm a pharmacist based in Singapore working in IT side of things for the pharmacy department in a public hospital here (means: pharmacy system, EMR, inpatient automated system, decision support). Still learning but would be happy to keep in touch.
This problem is probably a goldmine if a decent middleware could be written that grabs data from the device, transforms it to a standard format, and then loads it to EHR/EMR/HCM whatever software the hospital is using.
I can create a very few extremely simple solutions in my mind that will be quite precise. You don't even need much coding. In the worst case OCR can do with simple extraction script if the PDF is too unruly. And are a weekend job. But selling it to someone will be a pain.
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRS...
What you describe is an incremental improvement.
HL-7 is the standard for other medical apps. Even though these standards exist, the problems are still there and solutions are sometimes too pricey for hospitals.
Thanks for the encouraging words ... it's been a long road.
We're based in Singapore and previously built www.MedF1.com, a B2B platform for medical supplies ordering for private clinics. We did a health report generator from some Quest Diagnostic lab results previously for clinic chain in Singapore. Dropped you an email - ryan.teo@gmail.com.
Interested in work with healthcare disruptors, we can design the software (mobile, web) and hardware (Ardiuno, Linux).
I hate responding to trolls, but your comments through the thread have been a bit abrasive to me. Accusing someone of 'cheating the system', saying someone has to work 70 hrs a week, only having one shot, saying they should quit, telling them they must work 13hours a day to learn.
I'm assuming you're in (or rather I hope you are) medical school and not just someone applying. Here is the rub. Some people don't have to try to do well. There are people who never studied for the MCAT and scored 36+. They didn't have to really try to do well.
A radiology at the place he mentioned is very competitive, so it is safe to assume this guy was at the top of his class. The intelligence/knowledge difference between the top of the class and the bottom is enormous. Not everyone is the same. Don't fault him for what he wants to do in his free time.
Side note when looking at your profile: looks like both hnofficehours.com and hackernewsers.com are down/no more? Dead links.
Otherwise, you might have to parse the data from the serial port.