Slashdot Mirror
Advice for an Open Source Development Grant?
IgD asks: "My colleagues and I are developing an open source medical records system. A senior supervisor approached us and let us know a third party is offering a decent amount of money in the form of a grant for any legitimate medical research project. We were all but promised the money if we could come up with a proposal. Has anyone in the Slashdot community received a grant for open source software development? Are there any good examples of such a grant available? How could one measure the results of open source development for publication?"
Is your goal ge get the job done, or to open source your code, or to get money? Imho they are very different goals.
Less is more !
Refer to other medical research proposals. I'm assuming that you're not trying to get this grant on the sole basis of this being open source, right? The open v/s closed source nature of your project should be pretty irrelevant to your research proposal.
Mmmm.. Donuts
for someone with the time to do so, would be to make a portal with as much info on this subject, and links to sites that may be able to help you get started / funded for such a project. I'm sure the OSS community would appreciate a site with such information. Just a thought...
Have you checked around Source Forge for similar projects?
Have you check www.linuxmednews.com?
There are lots of projects in progress currently, perhaps you could work with one of those and help them out?
Alric
There already is an open-source EMR (gnumed.org), but no doctors are using them, because the switchover is just too hard. My dad's a doctor, and he was promiced a free EMR, just pay for the hardware. In fact, it was the FIRST EMR ever made, but he would have to hire extra people to load the rooms of files he had into the computer. It was just too hard.
My guess would be that writing this grant would be like writing any other grant. The idea is to propose an idea, outline your thesis (in this case, why you are making a case for this software and why open source), talk about the background of this project, why it is important what the implications are if this were to succeed and how you plan on going about completing this project. Outline the costs and give a timeline, clearly state your goals and wrap it up.
Visit Jonesblog and say hello.
I've received sponsorship money from some pretty big companies for the development (or augmentation) of an open-source project I wrote.
Now, this wasn't a "here is some money, go write something useful" type of sponsorship, but more along the lines of "we like your work but need a certain feature added. Here is the money, add the feature".
It probably doesn't help you though, since they (the money) came to me...
Ron Gage - Westland, MI
>How could one measure the results of open source >development for publication?"
.. Your reseach being opensource has no impact either positive or negative on the reviewers mind. Keep in mind the goals and make sure the job is well done .. Just because it is open source does not mean the results can be mediocre .. some of the best reseach is open source.
The Same way one measures other research results
Since you're doing medical research, the National Institute of Health's SBIR program seems most relevant. You can also find the application forms and guidelines if you look around.
May I ask you why you decided to open source it?
Unfortunetly despite of what Eric writes in "The Cathedral and the Baazar" Free Software developers are still most likely to recieve their reward in fame not $$. It may feel good but the Free Software community has so far failed to work out a way to consistently financially reward the actual developers (not hardware compnaies who bundle Linux with their servers!). Even if your project become popular all you'll get from users is bitching and moaning not meaningful bug reports. Just watch the burnout suffered by Fink, MPlayer and Router floppy authors.
US-UK-Israel: The real Axis of Evil
If you want to make an open sourced system just for the sake of it being open then you're going to have a hard time. I firmly believe that some things should be open sourced, but it's not always prudent to do so. Every situation is different and open source isn't always the answer. Companies need to make money, even the open source friendly companies release software closed source.
If you can get grant money based on the finished product then open source is a good idea, but if you need to provide a product that will be sold, i think companies will be a little more frugal handing you money to create something open.
They're in the open source medical software biz as well.
This guy is way out there
Then, talk about the benefits of an open source application, and talk about it's longevity and low cost of future manipulation. But only briefly, don't get too technical on them. Instead, attack the human aspect. Do some research and find out what the doctors and medical staffers themselves really hate, not their bosses/directors.
How could one measure the results of open source development for publication?
You haven't been in academia long -- the answer is you make up the results like everyone else!
But seriously, I'm not sure what is unusual about this situation. You apply for the grant, saying you want to research and develop XYZ system. At the end of the time period for the grant, you'll have to show that something happened, whether it is getting 1,000 developers working on it (this is good because you can clam the investment was matched 1,000 times in donations!) or having 12 private clinics and 2 hospital systems evaluating it and participating in system testing.
Whatever, you make up everything you can think of to measure (lines of code, contributors, patients tracked, data points, countries involved, languages ported to, web site hits, days of uptime, number of compatible legacy systems), keep track of it all, and at the end of the grant you write a paper saying how fantastic all the good stuff was, or why the whole thing failed and should never be attempted again.
If you really look into currently published stuff, you'll see that 98% of it is just proving and restating the obvious in a way that people can reference for future publications, so that they don't have to waste time on the obvious when the 2% of real research takes place.
I do applaud you and encourage you (and anyone else with the stomach for grant-writing) to pursue it, you'd be surprised how easy money is to get for useful projects if you can just keep up on the paperwork and wait months and months for every step to happen.
Get a half-dozen ongoing grants and you can basically have a small company that does pure non-profit open-source development year-round (and one full-time MBA to manage the grants!).
Recursive: Adj. See Recursive.
Why use open source as a medical record system? Why use the open source model? As one of the replying pundits put it "move out of mom's house and get a real job".
... actually the SCO model for those of you who remember SCO in its heyday before barritry became popular ... a place to find some security. More important is that from a physician standpoint what could be better than CQI, continuous quality improvement. The open source channel makes that not only possible, but nearly mandatory.
The reason for using open source software is that it is indeed a collaborative effort. For those of you that think that this is one setup shopping, its time to look at the serious projects that have made some inroads and continue to develop.
The business model for open source, indeed for those of us who remember a time before the internet (yes there WAS that time), when software started to be packaged with machines. Why, one might ask should one pay for software when it comes for free? Today $450 but tomrrow it's packaged. For those of us who watched carefully we knew that the real money to be made in the world of software would be in support and support applications. To a certain extent those who continue with proprietary and exculsionary sorts of software may well find themselves moved over because of freely available and very robust software.
The advance of linux and linux clones is such an example of the incursion of open source software. Free? Hardly. Freely available? Always.
From a standpoint of software design and development, the open source model gives those of us who wish not to be constantly hit up for nickles and dimes
So, those of you who have regarded the only path to enlightenment that of the Gates family or Big Blue, look again at some of the companies doing open source development.
As for the question about grant seeking; The FreeMED Software Foundation is seeking grants to employ coders and others to better the software. Since the Foundation is a non profit, seeking the development, promulgation and distribution of opensource software, people who are motivated to see better software development can contribute to the making of better software. In a way, donated dollars dictate direction.
There is much more about the open source movement and the intellectual freedoms that such development permits. Check them out. Check out the FreeMED Software Foundation (www.freemedsoftware.com) or the open source news list at LinuxMedNews (www.linnuxmednews.org).
Freely Submitted:
Irving J. Buchbinder aka DrGnu
FreeMED Software Foundation
Laws do not persuade just because they threaten. --Seneca
Now we have some funds to further the distribution. I will tell this guy it DOES pay to make Open Source Software.
This is a test. This is a test of the emergency sig system. This has been only a test.
My most successful string of open source solutions involved a small break with an audio-video company. I set up a small Linux network, with a small CRM that was based on, believe it or not, a web based PHP driven "application" that I designed to catalogue my DVDs, VHS tapes, CDs, games, books, and comic books. mySQL is the backend, Apache its "OS". I added a few useful modules from popular CRMs that I found on Sourceforge. All of this meant nothing to my client, they were only happy that it fit their every need (they were managing customer data, billing, scheduling, and reports across a series of applications like Excel and Quick Books Pro, and good old pen and paper). Needless to say, not only was my solution extremely scalable and cheap, but it removed the hassle of having to have a file cabinet handy and three or four programs. All they did was click a little link on their KDE panel, and up came Mozilla and their portal to my program.
They quickly refferred me to their lawyer's office, my own dentist, and another small business that specialized in boat repair. I quickly made about $12,000 in my spare time, and not one bit of the software I used cost anything. I have yet to return to any of their sites to fix anything since, and this was over a year ago. The only thing I did was give them each a call when Redhat made RHN available so that they could sign up and have their systems updated for them remotely, for very little $$$. The circle of Linux business life eventually brought money back to Red Hat, whom's OS I used for free as an ISO download, at all of these sites.
You gotta love it.
The important part to open source is the record format. After that, I can live with open or closed source software to access it. An open file format is imperative though, for any useful portability of records for patients.
BTW: Communism is where the government evenly distributes wealth, not when alot of people work together on a project. Might be helpful if you have to go through a high-school history course again ;)
A very good person to ask about this kind of thing would be Jon Hall (Maddog) at Linux International.
I believe he has been involved with Open Source Development grants and LI might even have such a program.
Bill
I don't know if anyone could even use it. Under HIPAA law you have to have a business associate's agreement with all vendors, and ALL vendors must supply support for the products or they aren't compliant. That's not even about the actual security or code flow of the program, but the whole project itself.
I do HIPAA audits, and I couldn't give them a a good rating on the risk analysis if they used it, because of that. No support==non-compliant. I could suggest they buy it from your company if your selling it, I don't see it being used by an IT staff somewhere without a vendor.
Maybe we DID take the blue pill. You wouldn't remember anyway.
LinuxFund.org has been kind enough to supply two of my projects (GNUpdate and Gaim for Qtopia) with funding and hardware. It may not directly relate to your question, since it sounds like you already have someone that may be giving you the money, but you can look at their policies and requirements, and request more information.
I don't see any reason why you shouldn't be able to get NIH funding and/or grants from other sources to build a low-cost or freeware EMR system with a well-written proposal and a knowledgeable Principle Investigator directing the project (you probably need someone with a PhD in Medical Informatics or an M.D./D.O. to be your PI). If you don't have a qualified PI to head the project yet, find one. You don't want to be perceived as a novice. The people you contact will likely be far more helpful the more you sound like you know what you're doing.
My advice to you (I have received several grants and fellowships) is to really be tight on the proposal. If the grant application asks you to write it in blood, write it in blood. I have also been on committees that review grant proposals. We would kick proposals, unread, for not having the correct format on the TITLE PAGE! When I complained that this was a minor thing, the committee chair looked over her glasses and said, "If it isn't perfect, then it doesn't deserve our consideration or our money." Be tight. That is my advice.
Comparing it to Windows will be a moot point, since El Dorado is going to have a 40% larger code base than XP.
Not that I want to know, but the goals and interests of the funding agency should be reflected in your proposal. For example, if the funding agency is interested in third-world health, then play up the low cost of the software for impoverished/disadvantaged clinics. If the donor is a big pharma company, then play up how they could "give this away" to potential customers or use it in clinical research. If the third party is more interested in academic research, then show how the system can support data collection, double-blind studies, etc. I'm not sure that the "open source" angle will have much traction unless you can show that the open source process multiplies the impact of the donor's money.
The more you know who the donor is, the better your proposal will sound to them.
Two wrongs don't make a right, but three lefts do.
In creating medical software you have to use code sets. They are somewhat modular, but I don't know how modular they are to be able to change them between countries. That would be up to the writers, and is probably something they really need to look into. If they can make the codesets modular, then that would help in more world-wide acceptance.
I think the same about the funding. I really was commenting on the whole "open source" of it.
It won't have the "rouge install" factor behind it, because it can't. You can't use it without some type of vendor support, or at least where PHI is concerned.
If it violated the service contract, someone could end up in jail.
Maybe we DID take the blue pill. You wouldn't remember anyway.
There are many examples of this available. My group (at a medical school) has collaborated on a number of grant projects (government and private granting agencies) where money was awarded for the development of open source software. In fact, the grants we have worked on require that the software be made publicly available (although most of what we do is of little interest outside of clinical/educational medicine).
If you want another great example, contact the IT group at the University of Delaware. They developed (under a grant and collaboration with a few other schools) uPortal. This is an open source portal system which is packaged and serviced by several different vendors (RedHat style) like SCT and Campus Pipeline. Since the development they have found they get a better response from other granting organizations as well as vendors (like Blackboard and WebCT to develop modules for it). So not only do you get the immediate benefit of money to assist in the development (to pay salaries, buy computers, buy software, etc.) but you get many continuing benefits from it as well.
One example I'm familiar with is the NIH Office on Neuroinformatics (no link provided, I don't want to Slashdot my funding agency!) supports the development of software for things like brain imaging and databasing. Their funded projects include lots of open source and GPLed projects, some directed by Slashdot readers (well, at least one). I don't know of any place where you can find successful applications, but you can at least browse some project descriptions.
This is just one example, I'm sure there are many others even just at the NIH (incl. at the new NIBIB).
Open Mash was funded for a while by the NSF, a common source of academic funding. I'm not sure if you and your colleagues are in an academic setting, but the Open Mash web site does have the proposals and reviews of those proposals available on the site. Check out the Papers and Publications area.
I'm not exactly sure what you're asking when you talk about measuring the results of an open source project for publication. But any proposal would have to talk about why the project you are proposing has value in-line with the goals of the committee or group you are submitting the proposal to. Knowing exactly what those goals are can be difficult. This is one of those situations where you really have to tailor your writing to the specific reader(s).
Many university hospitals and other large hospitals all have some sort of electronic record system, and many are converting to an all electronic system.
I am not aware of any that are open source, Most I imagine, as everything is in medicine, has to be tested relentlessly and thus is costly and a real pain in the butt to get certified for new security regulations.
From a doctors viewpoint (mine), they are a pain to use - they slow you down, patients wait LONGER!. It's much easier to dictate and have the secretary slip the note into the chart. A positive aspect of the EMRs is that charts don't get lost (yah, but computers crash).
Many of the younger doctors are used to EMRs and will probably goto that in their private offices, 'cause they're used to it. I could go on for hours, but I'll stop here.
..........FULL STOP.
No need to get angry. Read what you wrote:
don't know if anyone could even use it. Under HIPAA law you have to have a business associate's agreement with all vendors, and ALL vendors must supply support for the products or they aren't compliant. That's not even about the actual security or code flow of the program, but the whole project itself. I do HIPAA audits, and I couldn't give them a a good rating on the risk analysis if they used it, because of that. No support==non-compliant. I could suggest they buy it from your company if your selling it, I don't see it being used by an IT staff somewhere without a vendor.
The parent of that posting is the original article. Shall we quibble about what the word it means? Open Source? This project based on Open Source? The ONLY information you have on the project based on the original article is that (a) it is open source and (b) that it has the inside track on funding. If you read the sourceforge entry, the only additional information you can gain is that they plan to use Windows/Delphi/Kylix on the client side and MySQL on the server side. (eeeyuck! but that's a technical preference of mine, nothing to do with HIPAA audits).
So, whether you meant that you "couldn't give them a good rating on the risk analysis if they used [open source]" or that you "couldn't give them a good rating on the risk analysis if they used [project you know nothing about except that it uses open source and has an inside track on some funding]" -- the bottom line of what you wrote implies that (since the only substantive thing you know about the software they propose is that it will be an open source package for keeping medical records) this medical records system will be inadequate.
Obviously, there is a big difference between what you wrote and what you meant to say .
If what you meant to do was to alert the author to the imporance of having an SDLC or an SLA such that the institution using it wil pass HIPAA muster, then tell them, that they use open source it or not secondary to the importance of the quality of their SDLC and support SLA, in order that the institutions who may adopt their software pass an HIPAA audit. This is perfectly reasonable. You might also suggest that he use a slightly better and more platform-independent system than Delphi, as many Delphi components do not port nicely to Kylix. You might also suggest that he consider using a more robust database than MySQL -- PostGreSQL perhaps, one that can handle triggers and stored procedures, as the sheer size of medical records storage problems are legendary . There are a host of technical suggestions you could make that would contribute to the final systems' ability to actually work in the real world.
In your reply, your objection is that This particular law requires vendor support. It wants an externally designatted source that will say "we support this software for technical issues". This could be just about anyone who downloads it and takes the time to learn it. It could be them. It will probably be the people funding them.
So? Why not. Could be you. Are you objecting to the fact that one group wrote it, and any number of others might sell support for it? In that case, what do we do with all those MSCEs that would support the Windows part of it, but did not have any part in developing MS Windows?
There is only one developer registered for the sourceforge project, and while one beta prototype has been released, one would expect that the author intends, by stating he is open sourcing it, to check the code into the sourceforge CVS repository. Granted, CVS is only an SDLC tool, not the SDLC itself, but it bodes well.
What else? Oh, that toy RDBMS and thick Delphi-based Windows client. Well, until we see the proposal, we really have no idea what he's going to do to redevelop the prototype into something ready for Cedar-Sinai, do we
Medical records NEED to be open source precisely because of the nature of the business of keeping medical records. No one wants to share. I dont care if you can't read mine. Without an industry push possibly backed by big players the medical records will remain on islands, lost, inefficient and ineffective.
my other sig sucks less
The Insight Segmentation and Registration Toolkit (ITK at itk.org) was funded by NIH/National Library of Medicine. From the beginning of the $10 million project, the contract required all code and data to be delivered in open-source form. It defined a consortium of 6 prime contractors (3 academic, 3 commercial) each of whom had a particular role (e.g., architecture, algorithms, software process). It has been a great experience (version 1.4 will be released in two weeks).
Isn't that what the "Paypal Donate" button is for (the one that never gets clicked...once) ?
You are a bit late. The Department of Veterans Affairs (VA) has been developing a fully integrated suite of applications to run its entire system of hospitals for years, and since it was done by public employees, it is available for free to anyone. It is written in M (or MUMPS) so you always get the source code. This package, called VISTA (used to be called DHCP) runs every aspect of their hospital system, from counting calories in your breakfast to full reporting to Washington, DC, and includes a true paperless medical record. The package has had HIPPA in mind for several years, long before the private sector started becoming concerned with it. And yes, it can bill Medicare as well as insurance companies using DRG's, CPT, etc. In fact, SAIC got a full copy of everything from the VA a few years ago and sold it back to the Department of Defense for $2B (two Billion dollars) and called it CHCS (Composite Health Care System.) If you want to find out more about it go to http://www.hardhats.org/ and all the info is there, including links to get the software from the VA. The package is so good, there are other countries which use it for their national health care system. This suite was driven from the bottom up by the users in each area of the hospital, so the Nurses decided what the Nursing package needed, the Psychiatrists decided what the Psychiatry package needed, etc. Read the site - it is impressive!
We're actually funded by 'norsk forskningsrad' (norwegian councel of science), and we're situated at my uni. If you're going to take money, there are a couple of questions you're going to have to ask yourself:
1. What does the grant-giver want in return? While some grants are nearly no-strings attached, in the end there's strings on everything; and no exceptions made..
2. How much more time will you spend writing documentation/progress reports/other stuff purely for whoever gave you money?
3. And more importantly, how will taking a grant change the pace of the project? Will taking money for making something mean you're taking on responsibility you'll not be able to meet with your other responsibilites? When you're taking money for doing something, its going to be *alot* harder putting it away for a month because your wife/gf/dog/actionman turns sick, and your female boss at work gets her period.
"" How about taking the safety labels off everything, and let the stupidity-problem solve itself? """