From Flu Wiki 2

Forum: Proposal for an Open Source Avian Flu Treatment Database

Melanie &ndash

original post removed at author’s request. a superior version exists here in the opinion of the author.

Michael Donnelly – at 12:41

Sorry for the formatting probs. I read the text formatting rules and I thought I understood them. Apparently not.

Comments welcome.

-Mike D.

Michael Donnelly – at 12:43

Note that even if you don’t agree with the cytokine storm hypothesis, the question of effective treatment is still open, and good data with a central clearinghouse are needed in any case.

DemFromCTat 13:18

Michael, would you like to/like me to post this in the Opinions section as well? it won’t scroll away. And the Opinion section is more easily editable.

Dubina – at 14:21

Michael, a few comments,

“…the problem right now is that the virus everybody fears does not exist yet. Without it, we can’t know if a vaccine is effective. So the vaccination strategy presents scientists with a situation that sounds a lot like a Catch-22.”

One article that I read recently said, “…the first wave will be disasterous” (because no closely matched virus will be available to make and administer an effective vaccine before the fact).

Another article recently linked here touted Biondvax and another, Acambis, both apparently “all-strain vaccines”. An effective all-strain influenza vaccine would solve a lot of seemingly intractable problems. If all-stain vaccines can be made, one imagines they might be made in parallel to other vaccine development efforts now hoped to happen in the next 4 to 5 years.

Regarding your proposal for an HPAI treatment database, have you considered what might be done to prevent cyber creeps? (Michael Kinsley’s term for a gang of malcontents that descended on one of his wiki experiments and effectively ruined it…in effect, a DOS attack.)

On a related point, the China Internet suffers tight governmental control at official sites, but isn’t so easily controlled at private blogs, or at the level of the home PC. Given the main surveillance problem at this point is Asia, and given cyber creeping has been raised to an art form there, particularly in the PRC, might it stll be possible to elicit public support for surveillance information by way of an online reporting / database such as the system you’ve outlined here?

Michael Donnelly – at 18:17

I’ve thought about the whole cybercreeps issue, and I am not too concerned about it.

In fact, the reason I like the Wiki model is that it seems to be in effect self-correcting. Any additions are in effect quality-checked by other users. One might put in a safeguard sort of like this: anyone who wants to file a report has to put some contact info in there, and then the people who maintain the database could check up on their identity somehow.

Also, The thing is that if you assume that these people will have no overall bias, their impact overall will just be to add statistical “noise”, which any statistician would tell you doesn’t make the eventual conclusion less valid, just takes more cases to reach a firm conclusion.

Plus, I have a feeling once the world is feeling the pressure from an outbreak, the overall number of creeps out there will dwindle pretty fast.

But the time to strike on this is now. We need this resource to be in place BEFORE the outbreak occurs, and get word out about its existence. There are a number of “influence nodes” on the web that would make good places to do this (Slashdot, Fark, others?), including web media outlets (like Wired, Google, and others). I have a feeling that once the wisdom of the endeavor is revealed, and the panic sets in, organizations like the Chinese govt will come around fast.

In the end, the usual authorities who take care of data gathering (WHO, CDC, etc) can’t do it fast enough. The only type of system that can IMHO is an open source system, with complete transparency.

Michael Donnelly – at 18:20

Dubina: Could you post my thing in the opinion section? That’d be welcome I think. THx -MD

Dubina – at 18:28

Hey, I’d be glad to, but what thing are you speaking of, exactly? Top post? Thread? And why me, not you? (I’ve not posted to the opinion section yet.)

Michael Donnelly – at 18:41

Dubina: Yes, the top post. I didn’t realize I could put it in the Opinion section myself. I had the impression you were an editor. I’ll try to put it up myself. But someone has since edited the top post, so it reads much more clearly now.

DemFromCTat 19:12

I’m the editor who suggested it. You can do it (it’s a wiki). I’ll do it if you’d like. I helped format the first post.

Dubina – at 20:05

Dem,

I’ve been wondering….

What’s the difference between a thread in which an idea evolves and an edited wiki post, I suppose in which all that has gone before has been revised such that it may not reflect its roots?

DemFromCTat 20:15

Here is the opinion post: link.

A thread has discrete ideas in a sequenced comment section whereas a wiki post is the final product only. Threads can certainly lead to ‘finished product’ wiki posts, but are much more interactive, yet clearly more raw material.

DemFromCTat 20:18

Put it another way: wiki posts are there to produce and edit the final product, whereas discussion threads are to talk about a topic, not necessarily to produce a final product. Final product posts are constructed to stand alone. How you got there is unimportant. Discussions are all about how you got there.

DemFromCTat 20:30

I recommend Opinion for those who have long things to say, btw, and the Forum for short things. And the forum piece can be linked to the opinion piece for those who wish to comment. We can link this page to Dr. Donnelly’s new opinion piece.

Dubina – at 21:10

But hpw do you get from a discussion thread to the final product? On the NewsHour, I think, an example of the wiki method showed three iterations with two most recent amendments displayed with original text. Who decides what’s “right” in the end? That seems an important question.

DemFromCTat 23:15

The community does. The wisdom of crowds. it’s constantly edited to be best product.

A final product from a discussion has to be posted elsewhere in the wiki by someone. If it’s a signed opinion piece, the etiquette is to just correct spelling and format. but if it’s a contribution piece like in the consequenses sections, have at it.

26 November 2005

Michael Donnelly – at 00:51

OK, so if someone is going to do an open source treatment evaluation DB, what should be on the list of treatments that are likely to be tried?

The “accepted” list so far (from fluwiki page on treatment):

Tamiflu, Relenza (neuraminidase inhibitors) Amantidine, Rimantidine (M2 blockers) Aspirin, ibuprofen, acetaminophen, other NSAIDs? corticosteroids? ORT for dehydration

I’ve been searching PubMed for possible cytokine blockers: DHA, curcumin (used prophylactically) thalidomide (and related compounds?) haloperidol (other antipsychotics?) pentoxifylline antidepressants (amitryptilline, desipramine) Enbrel (etanercept), Remicade (infliximab) beta blockers (propranolol etc)

Emphasis on these additional items should be on drugs that are already in widespread distribution, so experimental compounds lke ox40, and other cytokine blockers in the pipeline that might be rushed to trial aren’t going to be realistically tried by many people.

One additional complication is the fact that dosages would need to be tracked in some fashion.

The interesting thing about an open source drug trial DB is that it could be useful generally, and might in fact provide a useful adjunct to the standard double-blind placebo controlled study, at least in this case, where positive outcomes will presumably be easy to determine.

Dubina – at 05:04

Dem,

I understand the wisdom of crowds, as I’ve gone round and round contracts and other documents that had to be wordsmithed by lawyers before they could published. That wasn’t usaully easy to do.

I complained several weeks ago about the many facts and opinions being aired here on particular issues with no competent resolution in the end…seeming to me that readers were often left to pick and choose their own poison.

I think perhaps I also said something about the audience lifeline on “Who Wants to be a Millionaire”; in that circumstance, the wisdom of crowds is very good when the questions are very easy, but it tends to dissipate as the (multiple choice) questions get harder.

Then too, some threads here have a lot of stuff to say, but it’s not always easy to say exactly what they’re about.

What I’m trying to say eventually is that while I think it’s great that some of these very pertinent threads are being sorted into bankable information, I still don’t see how the wiki edition works in practice.

Maybe you could post an example or two showing how a thread (real or hypothetical) might be progressively boiled down to its essence.

Let me also say why I think this is important.

Several days ago, LexEcon said of something I’d written, “Darn it, stop getting in my head”…(actually, a nice compliment). But after studying the stuff that you collect on your sitemeter, what I thought I saw was just that capability: the data you get lets you really honest-to-God get in peoples’ heads here, especially into their collective head.

Just over a year ago, I watched a PBS Frontline program, “The Persuaders”.

http://www.pbs.org/wgbh/pages/frontline/shows/persuaders/

I was surprised to learn how public opinion could be so malleable given one might know just what buttons to push.

Likewise, before a few days ago, I had no knowledge of sitemeter-like data mining. Well, actually, I lie a little; some months ago, I configured my PC anti-spyware to quarantine and delete your tribal fusion dataminer whenever it came snooping around…trying to get in my head, as it were. I thought then that a dataminer on my PC might be up to no good, but now I think I see what it was up to.

This brings me closer to the point: About FluWiki tells us …

The purpose of the FluWiki is to help local communities prepare for and perhaps cope with a possible influenza pandemic. This is a task previously ceded to local, state and national governmental public health agencies. Our goal is to be:

This seems a very good statement of purpose, and so far as I know and can see, fluwiki is doing a great job in that regard. I see fluwiki getting better every day, feeding on itself, changing faster than point mutations on a flu virus in a henhouse over a fish pond. Pretty impressive, I’d say.

http://www.fluwikie.com/index.php?n=About.About

But I’m also mindful of a couple of lines from George Orwell’s 1984.

Who controls the past controls the future. Who controls the present controls the past.

I asked you some weeks ago how you knew what government was doing behind the scenes. You said, simply, and to the point, “because I see it”. I won’t pretend to know everything that your sitemeter tells you, but now I think I have a better sense of what you meant.

So the issue here is that you have your hands on a lot of raw political power.

Like Frontline’s Persuaders, you know what we know, and when we know it. You know where we live, and you know what turns us on and off. I think your executive summaries or whatever it is that you do to boil us down would make very interesting reading. While I trust that fluwiki’s motives are pure for now, I wonder if you have masters, and if you do, who they are, and who or what funds the site.

One other reason that I wonder is that I went looking for that information, but couldn’t find it.

DemFromCTat 10:17

Because it doesn’t exist. Here’s a Flu Wiki history, courtesy of the Reveres.

We don’t data mine. As noted, I pointed you to the sitemeter to show you what we can glean, but i spend no time doing so. i don’t collect emails, IPs or any other data. There are no corporate masters. The three of us (Melanie, the reveres collectivelky and I) pay out of pocket to run the wiki. it is what it says it is.

Some of us have day jobs, like the Reveres, that put them in the mix ;-). Some of us go to conferences and talk off- and on-line to academics, CDC people, WHO people, state people, etc. and it is from talking about that with folks that I have concluded there’s lots going on behind the scenes.

So to go back to your question about Forum to wiki final products, we had early on a nice chat about languages (see the Legacy Forum in the left sidebar) and lugon spearheaded a Spanish translation. We had a chat about scenarios people could participate and thus was born Brainstorming. Some of these are still shells because people have yet to contribute meat to the potatoes, but that will come in time. The framework is there.

As to the rest, the Forum is only half the wiki. Many govt. types skip the forum altogether and go to the rest of the wiki, especially the Links section. But it would take someone with initiative to post or move a Forum piece/idea elsewhere. If it’s Dr. Donnelly’s idea, he can post elsewhere. As an editor, i might move it to the section I think fits best, but it’s his work. OTOH, although he offered to try and post to Opinions (my suggestion) I did it for him as a helping hand, just as I helped clean up the formatting.

I don’t do a periodic ‘executive summary’ of Forum pieces for anyone like a mysterious corporate master somewhere. ;-)

With the wiki, what you see is what you get. If I see a comment in a Forum (like Boxun), I might drop a link about a relevent wiki page to encourage people to actually use the wiki (read or post), not just comment in the Forum. But that’s about it.

Other editor’s jobs are to correct factually wrong info, watch for copyright violation, plaigarism (uninteresting but necessary and not up for debate), pie fights, formatting errors like sidescroll, etc.

Melanie – at 14:29

dubina,

Let me second Dem’s comment. We are very deliberate about keeping this wiki completely independent. Any funding that we might receive in the future will only be accepted with the understanding that we remain completely independent. The editors receive no direction from anyone and the wiki community sets its own direction.

viralprotein – at 15:21

Michael,

Back to the original intent of this thread; I am curious about another potential Avian Flu Treatment.

While most people view Mosquitoes as little Flying Hypodermics because of their ability to inject viruses like West Nile and Malaria it turns out that they can’t deliver influenza.

According to a paper published in Cell, the mosquito’s immune system contains a TEP1 protein that destroys all influenza viruses by latching onto the pathogen’s surface so that phagocytes can recognize and engulf the foreign invader.

Science Direct/Cell March 2004

Since there is a homolog for TEP1 (telomerase associated protein 1)it would appear that raising the plasma levels of TEP1 in mammals could be a promising anti-viral treatment.

Eccles – at 15:36

If I may inject a question based on my own ignorance -

It is my understanding that the telomeres are the structures which determine the longevity of a cell line. When the telomere is gone, the entire line dies. Telomerase, from the name, is an enzyme associated with telomeres. Might not playing with an enzyme associated with telomeres have potential unintended and substantially negative consequences for the life span of the individual?

(And thanks for not flaming me if this is a really stupid question. I am not well educated in medical matters.)

Michael Donnelly – at 16:34

viralprotein:

The original intent of this thread was to propose that we do something to prepare for an outbreak, and the inevitable trial of many drugs that will occur in that situation. Given that people are going to be desparately trying lots of different drugs, someone should make an attempt to gather systematic data about successes and failures. Hence my proposal.

But as regards your point: my emphasis is on drugs that exist now, and are in widespread distribution. Not sure if TEP1 counts on with either of those counts or if there is a drug out there now that increases TEP1.

I’ve been focusing, as some here have, on the possiblity of using cytokine blockers, hence the second part of my list above. As of this date, the cytokine hypothesis is still controversial, plus there’s the fact that there are dozens of cytokines, and the cytokine cascade is a complex process. It’s not clear which ones play a role, if in fact ANY do, in mortality from infection by this flu

Interesting addition to the of potential cytokine modulator and an example of the complex nature of the cytokine system:

statins (e.g. atorvastation) which may increase some cytokines while decreasing others

viralprotein – at 18:19

Eccles,

Telomerase is responsible for repairing telomeres and increasing the lifespan of a cell. Discovering the TEP1 protein was a Life Extension drug would be a serious fantasy. Unfortunately we’re going to need more than one protein to pull off that trick. Still I’ve seen some stuff in the lab that gets pretty close.

But we aren’t going to give it to humans without first beta testing it on a few mice.

Eccles – at 18:30

Thanks VP That explains it for me. And, as I’m rapidly becoming one of those who could use an extensive retelemerization (the sooner the better).

I’ll be rooting for you.

viralprotein – at 18:56

Eccles,

LOL, get in line…

DemFromCTat 19:38

Michael, all it would take is for you to create such a repository on Flu wiki is creating a page. You could, for example, do so on Brainstorming (one of the categories to the left, created for the interactive wisdom of crowds).

Now, how the page would get filled is another story altogether. You’d need to publicize and attract those who’d have data to fill in the blanks.

05 January 2006

seeker – at 14:33

Michael, et al,

I have absolutely no med training or schooling, but I came across info about cytokine suppression with probiotic strain which may be helpful in treating irritable bowel. I don’t have the knowledge to figure out if this particular strain would be beneficial in any way, and it appears that probiotics in general are making headway into the self-care health market.

Please excuse me if this is not germane to the avian flu discussion. I just wanted to add it into the mix if it were another way to reduce the ravaging effects of a cytokine storm in a patient.

http://cdli.asm.org/cgi/content/full/11/4/686

http://cdli.asm.org/cgi/content/full/11/4/686

06 January 2006

Swann – at 03:24

“Now, how the page would get filled is another story altogether. You’d need to publicize and attract those who’d have data to fill in the blanks.”

Would hardware/software like the following links work? I know that designers will sometimes modify software for specific needs. http://tinyurl.com/8h352 and http://tinyurl.com/bzdyo

22 January 2006

revere – at 23:25

to all: I don’t have a problem with Open Access data bases (this isn’t really Open Source) but it doesn’t seem to me that it will be feasible to hae a valid analysis along the lines that Michael suggests. I say this as an epidemiologist. There are too many potential biases (systematic error sources, not the colloquial sense of “non-objective”) to be able to untangle effects from all the noise. What it will turn out to be is a gigantic anecdote pile, which, at some future time and with hindsight we might discover someone had a good idea. But a lot of what people to do and for themselves isn’t particularly safe, and having a big pile of unvetted suggestions that desperate people will go to worries me.

As a wiki co-founder I am very convinced that there is a vast amount of raw brain power “out there” that we never get the benefit of. The Wiki is an attempt to harvest some of that. But that doesn’t mean that you can just dump stuff into a bag and hope something will come out of it.

That’s my initial take on it. I am willing to be convinced otherwise. One suggestion, Michael, is that you spend a bit of time over in the Basic Science section and get a bit more up to speed on flu science if you want to pursue this.

23 January 2006

Rudi Cilibrasi – at 03:35

In terms of Open Source databases, for me the architecture choices are surprisingly simple these days. I choose the same sort of thing TextDrive does: Ruby language using the “Rails” framework. I believe that the most important first step for any open-source database proposal is definitely to build the SQL data model. To be Rails-compatible it should have id fields (int) as the first fields.

DemFromCTat 09:00

see if something already exists that can be adapted. try this.

Michael Donnelly – at 09:34

Actually, I’m a little surprised to see this forum bumped today. The proposal has changed considerably since I first put it out there, you can see the newer version in the Opinion section.

DemFromCTat 09:37

the link is here.

Michael Donnelly – at 09:42

Thanks for the suggestion DemFromCT, but its not just a question of adapting something that exists. Its also a matter of attracting a large number of conscientious participants, in order to gather the best data possible inthe shortest amount of time (to avoid the situation that Revere anticipates). That requires INFLUENCE, and PUBLICITY, neither of which I have. Right now my efforts are focused on enlisting help from some entity with that sort of capactiy. The technical aspects are secondary.

Michael Donnelly – at 09:47

Revere: might a more explicit proposal with a demonstration do the trick?

One of the points about the Open Access flu treatment database is that it is in a sense atheoretical. It just assumes that people out there will be forming their own theories, or buying into those of others (such as the cytokine storm idea), and then acting on those beliefs by trying a certain subset of drugs/herbs in the event of a pandemic. All I want to do is gather the data about their successes and failures in the most simple minded fashion possible. In a way, by restricting the amount of information they can give us (just getting info on their exposure, their wellness level, and the drugs/herbs they’ve been using) we leave them very little room to add anecdotal stuff in there.

MaMaat 10:22

I would like to suggest that the MIA interval (3 days) be longer, maybe as much as a week. Alot of people may be unable to communicate for more than several days because they are a.)too busy caring for others to check in b.)ill themselves c.)subject to power failure and unable to communicate with others if they are connected to a group. I’m only suggesting this because updating that quickly might be alot of uneccessary work for the people who are monitoring the site and correlating the information if participating individuals are frequently experiencing the above difficulties (likely I think). The persons handling the data would then have to go back and change alot of status info for individuals. This could quickly become overwhelming depending on the number of people involved, both those submitting and handling the data. Premature (my presumption, I am not a medical expert) MIA designation might also temporarily skew the reports if they are accessed frequently and the worst is assumed by those using the information to decide on courses of treatment. Secondly, even if the information is of limited value during the crisis, shouldn’t it still be collected? Isn’t it information from past occurences of flu outbreak what experts are using as models for treatment and statistics? The information collected could prove to be invaluable in the future. I think it should at least be attempted.

Michael Donnelly – at 10:44

I actually have suggested other ways around this problem. One way would be to have individuals register as groups, that way, one of them could update the status of others. In the heat of a pandemic, people will be totally focused on the issues of the disease, and I think they’ll be motivated to check in with the database regularly, to update their own status, as well as to check the ongoing results of the entire community.

08 February 2006

jimofPA – at 09:46

Since H5N1 induces a host cytokine storm much like the H1N1 from 1918, are any of the infected victims being treated with immunosupressive drugs such as methotrexate, tacrolimus, or anti-TNF-alpha drugs (Remicade, Enbrel, Humira)? Can chickens so treated survive H5N1? I’d love to know so that when this sucker goes pandemic, I can survive.

09 February 2006

Michael Donnelly – at 22:14

Wouldn’t we all, Jim? There are so many compounds that have been suggested as possible treatments, most b/c of their cytokine modulating properties: those you mention, plus haloperidol, the statins, thalidomide, too many others to mention. Problem is, no one seems to be in a position to collect the data under conditions that would be a good test of their efficacy. I mean, how could you, until the disease is already on the ground? Hence my original proposal, which is designed to be a way to get real time efficacy data, as a pandemic is in progress.

So far I haven’t heard of any interest in the idea, but notice no copyright claim. I put it out there for anyone who might wish to have a crack at it. I sure can’t do it.

BTW the proposal you see above has been significanty revised since I first proposed it. You can find the revised version in the Opinion section.

Michael Donnelly – at 22:20

I still think Revere’s comment re: the quality of the data is perfectly valid, but that by constraining the information sought to very simple info that is not subject to much bias, we might be able to get usable info out of it. It’s success, as he notes will depend on having a large number of exposed participants. Perhaps as many as 100K. For each compound on the list, you’ll need several thousand people using, plus you’d need a voluntary “control group”: participants who aren’t in fact taking anything.

10 February 2006

anonymous – at 01:16

Dumb question, but is there a way to get scientists with access to H5N1 the collection of suggestions that have been made here in wiki so far? I know there are scientists and docs here, I just don’t know if they are interested in looking at suggestions posited by amateur researchers.

I guess one big problem is that cytokine-reducing agents may not do much to H5N1, in or out of a test tube; tests have to be done in a human, or at least some sort of host? My concern is, if and when communities become overwhelmed by the pandemic and are out of pharmaceuticals that may be able to treat cytokine storm, they won’t be aware of the large number of meds, herbs, or other that have shown some success in non-flu studies. How much would it cost for a local flu clinic to have a substantial amount of curcumin on hand, for instance? I know that may not help at all, but if all other known methods are unavailable, would it be prudent to have it on hand?

Obviously, I don’t mean curcumin specifically, but there are plenty of out-of-the-box methods that may help. I’m just wondering how to approach TPTB with these ideas.

Michael Donnelly – at 09:47

anonymous:

Your post asks several questions.

  1. 1: the scientists with access to H5N1 know about the suggestions. You can bet on it. Will they test all of them? No. Will they test any of them? Probably not, at least not in any kind of realistic situation (i.e. in an infected human). Yes the tests have to be done in a human for the results to lead us to better expectations about the efficacy of any putative treatment. Note that some reports are emerging already about Tamiflu being used on people, but we expected them to give tamiflu to people infected with brid flu. What about those other things? While they have tamiflu to use, and it remains the accepted standard of care, all these other things won’t get tested.
  2. 2: How much would it cost to get a large amount of curcumin (for example)? Not much, but then the unit cost of the substance isn’t really the question. The issue is how much of that substance is already in the supply chain. Even if we found out today that curcumin (or sambucol, or anything) cured bird flu, the speed at which we could get enough doses of the stuff would be very slow indeed. Money is no object when there’s none to be found.

My own preference would be to lean on those substances that are already in abundant supply, hence my interest in prescription meds that are already very widely available: e.g. the statins, prednisone, haloperidol, etc. If one of these could be shown to be an effective treatment, every community with a pharmacy would be able to get their hands on the stuff, because there is already so much of the stuff in circulation.

That said, some of the herbal preparations could be supplied pretty quickly from alternate sources. Take for example resveratrol (sp?) which as I recall is derived from grape skins and is found in varying amounts in the foods (and drinks) on the shelves of grocery stores everywhere. If someone determined that resveratrol did in fact help people survive, you could see people breaking out cases of red wine and so forth.

Curcumin isn’t widely available, but grocery stores have piles of turmeric available, both in pure form and mixed into curries. If curcumin was the stuff to get, we might figure out a way to use these other ingredients to get at the active ingredient.

Sambucol is created from black elderberry, and I’m willing to bet that if someone discovered sambucol is the stuff to use, people could go into the woods to where it is growing wild, harvest the berries and cook up mountains of the stuff to get out the sambucol.

But each of these scenarios requires actual empirical knowledge, not just supposition. Right now, everybody has their pet theories, and many of us plan to test them out on ourselves and our families if and when the time comes. My proposal was merely to create a way to gather data about their successes and failures, and in so doing, help us all figure out what can be done once the tamiflu has run out

28 February 2006

Michael Donnelly – at 22:49

BTW, My proposal has evolved quite a bit since I first proposed it, so anyone who thinks they might be interested, should check it out again. It’s currently about 5000 words, and I have yet to add the references in. Anyone who might care to help me put in references, I would greatly appreciate it.

I’d hate to think that all anyone knows about my proposal is from the initial post (above). Quite frankly, that thing stinks. Go here to find the new version.

02 March 2006

d00ley – at 22:33

A post of mine resulted in my being directed here (thanks to MaMa), and I am quite pleased, but not surprised to find that others are already on top of this and well under way to hammering this idea out. I own a small software company and though I’m not in the financial position to offer free development, I can offer needed development at cost — ~<$15/hr. However, I don’t think this will be necessary since I believe that there will be enough volunteer talent coming forward to get this built, if a pre-existing system isn’t already available. Also, do we have any idea how many FluWiki users, readers and active participants there are? What about other similar online flu communities? And, are each of these communities in communication with each other?

I’m pasting my original post below simply for the sake of context:

---begin previous post--- I’ve been researching Vitamin C for quite a while, and there is promising evidence that Vitamin C (ascorbate acid) can calm a cytokine storm. When needed, the body can readily use doses as high as 200–300 grams/day. At these levels, it is used in many bodily tasks and displays antiviral and antibacterial qualities. Below are two links that specifically discuss vitamin c as related to the Avian Flu.

http://www.orthomolecular.org/resources/omns/v01n12.shtml

http://www.cqs.com/influenza.htm

On a tangential, but similar note, I think it would be extremely valuable to begin to discuss now ideas regarding organized feedback as to the effectiveness of different medicinal, herbal and pharmaceutical tried DURING the pandemic (if it happens). If the Internet remains available, even if only for a week or two after the pandemic begins, the FluWiki community alone, through organized and diligent feedback, could nail down an effective treatment within the first week of it hitting large numbers. Just in the forums alone, I can see that there will be hundreds of different treatments put into action. It would be bittersweet, of course, some of us reporting ineffective treatments, but among all of us we should basically be able to have the largest and shortest trial of treatments in the history of mankind. If there is an effective treatment, I’m sure that we could find it, confirm it, and communicate it to ourselves first, and then to our neighbors, families, and so on. ---end previous post---

NS1 – at 23:26

have we examined and settled on a database platform yet? Does the fluwiki software have an underlying malleable database that is driven off customizable forms? Does this database support full relational inquiry (SQL)?

Do those posting have access to some sort of topical online database that is configurable automatically through the wiki?

05 March 2006

d00ley – at 13:01

As for wiki software, here’s a comparison chart. Not sure how useful, but here it is: http://en.wikipedia.org/wiki/Comparison_of_wiki_software

Michael Donnelly – at 16:22

Yeah, I really can’t comment on how the software should be designed or what it should be capable of. My part is just to imagine what the data that get input might be like, the data structures that would result, and the analyses that might conceivably be carried out on them.

The critical issue from this standpoint, is that there need to be clearly defined and agreed upon criteria to be used by all participants for a)exposure status, b) symptoms, and c) each treatment under consideration. Without these things, no one person’s experiences will be directly comparable to any other person’s, thus removing the possibility that we might achieve a large enough sample size to give us the statistical power necessary for this project.

That said, I think that if we restrict our focus to just the panflu (PF51), then those first two will be easy. See my proposal for some criteria that I think might be workable.

NS1 – at 20:03

Donnelly-

Check boxes with ranges for reporting and an “other” free format comments style box in each category for further explanation?

Michael Donnelly – at 22:03

Yep. I think what we’d do is have a committee decide on a large list of “standard treatments” where they are standard only in the sense that they have some agreed upon definition. Then on top of that, we’d create comments fields that allow participants to record qualitative data to go along with it.

06 March 2006

NS1 – at 01:18

Exactly!

NS1 – at 03:50

Categories of treatments / Prophylaxis with input of dosage amount and schedule. We’ll need to scrub the data for units of measure standardization, et al.

We also will need a method to accept ‘cluster’ treatments to save the supplier of information input time. Mother inputs treatment schedule for 4 children and then enters age/sex of each child. Submits separate report for husband and self due to dosage variations. Allow a saved profile for each submitter to copy treatment from last submission while requiring a new status.

Michael Donnelly – at 09:28

Right.

Prophylaxis w/b registered by allowing people to input the things they are taking/doing well BEFORE they develop symptoms.

We could make things easier for people by using cookies to remember who they are and what their most recent combination of options was. Then all they’d need to do is select/deselect according to the things that they’ve changed since the last time they provided info.

Indeed, this effort would probably REQUIRE that people are input as “clusters” in order to create a sort of buddy system. Because after all, if you input your data alone, and you get so bad that you can’t do the data entry, someone needs to do it in your place.

We could solve the problem of isolated individuals who want to take part, by creating a sort of “online buddy” another individual who lives close by, and put them in phone/personal contact with each other. By creating contact redundancy, we shouldn’t lose too many individuals due to communication problems.

NS1 – at 09:47

works for SEALs.

12 March 2006

Another guy with a PhD in Immunology – at 08:23

Hi MD,

I’ve just been reading your proposal and I think it’s just what the doctor ordered (sorry!). It’s clean, simple and probably the easiest of all ideas to implement. News of it would spread like wildfire if it was set up properly. I suppose it needs a REALLY stable server with great bandwidth!

Can I suggest Sambucol as an entry on the list of treatments? It is widely used in many countries and doesn’t require a prescription. There are loads of people who use it already for the common ‘flu.

I believe many treatments such as Sambucol, made by companies that are not pharmaceutical giants like Roche, may benefit greatly from this ‘experiment’. Clinical trials cost a fortune. Perhaps if the idea begins to materialise, you can convince some of these companies to help fund the project (it’s cheaper than a standalone study, right?!). Obviously objectivity and transparency must be kept to a maximum - with money comes corruption, right? But those companies that believe in their product and are happy for the general public to test it must surely be keen, no?

Sue – at 08:31

I had heard about Sambucol years ago and have experimented with it maybe since 1995. I’ve given it to several members of my family, including myself and none of us could see that it did anything to help. I was so disappointed each time.

Another guy with a PhD in Immunology – at 08:57

One thing that my family and I noticed was that it did absolutely nothing to help a cold but really did help with ‘flu. Ok, we didn’t do serological tests but the £50 note test (would you get out of bed to catch a £50 note fluttering by?) would have distinguished the two!

MIchael Donnelly – at 18:09

Another guy etc:

I have been revising the proposal quite a bit, and I am thinking about turning into a more general effort to gather data of all kinds (including, but not limited to, preventative “remedies” like sambucol). The idea would be to use the events of a pnademic to gather any and all data about anything that might affect infection rates, and once infected, mortality rates. So I’ll be changing my proposal significantly to have people report about the amount of contact they have with other individuals, about how often they wear masks, about what their daily intake of alcohol is, of caffeine, of…Oh, to heck with it. Just hang around a bit and wait for the next version of my proposal which will be arriving by the end of the week.

I am trying to pitch it to my employer (UW-Stout) to see if they’ll host the database and web pages during a pandemic, should one decide to break out. As far as I know, there is no other research intitution or group who has a plan in place to gather such data prospectively. Having a vast amount of prospective data about any of a variety of strategies people might use to prevent infection or minimize symptoms once infection has occurred, would be of inestimable value when planning for the future.

Dr C – at 18:17

Have you seen the Treatment and Prophylaxis thread here today? It appears that the online medical text on Influenza Report 2006 has posted thier new chapters on that and drugs.

An evolving database would certainly be useful though.

And here’s what I’ve picked up in the respiratory therapy area.

“From the Tropical Disease Hospital in Hanoi:

Pressure Controlled Ventilation maintaining PIP < 30 CM H2O (any higher creates PIE regardless of patient age) PEEP < 10 CM H2O I:E 1:2 if problems with oxygenation (which there will be) I:E 1:1 Rate to maintain PaCO2 ~ 60–80 but watch pH FIO2 as low as the patient will tolerate witout developing metabolic acidosis pH ~ 7.30″

http://www.influenzareport.com/index.htm

Timber – at 18:19

Michael —

Again, I commend you for your honesty and effort to do the right. The built-in resilence of the internet makes it a hopeful tool for this kind of database.

Go for it, and let me know how I can help : tdick@du.edu

MIchael Donnelly – at 18:29

Yeah, I have seen that thread lately. Haven’t looked at the new “influenza report” but if its anything like what I’ve read up to now, mostly it’ll be a lot of “here’s what we think will help, but really, we have no way of knowing what will actually make any difference”

I want to gather these data in the most theory-free manner possible, including any preventative or treatment for which a claim of efficacy can reasonably be made (where the criterion for “reasonableness” is left highly subjective). So dancing a jig in the full moon light would probably not be on the list, but regular prayer and use of homeopathics would. Still, part of the reporting would be open-ended, so if a bunch of people hit on the tactic of dancing jigs in the full moon light, a fact like that might pop out as the data come in.*

MaMaat 18:30

When people are registering initially it would be good if they stated any known pre-existing medical conditions and prescription meds that they might be taking for maintenance. These factors could have a significant effect on the course of their illness. It would also help in figuring out if there are any negative interactions for combinations of maintenance and treatment medications/therapies. Just a thought.

MIchael Donnelly – at 18:37

MaMa:

You got it! My upcoming revision will actually have people register by rpeorting relatively unchagning data to begin, such as preexisting conditions and prescription meds. So thanks.

MaMaat 18:43

Thank-you for working so hard to get this set up. Collecting this information and turning it into a usable resource will be a monumental task and you’ve done a great deal already. This is very important work and your effort is really appreciated!

NS1 – at 21:58

Donnelly-

I believe that your inclusion of activities as you’ve stated will be an important factor in the gambit as we analyze the results. Be sure that this section invites as much comprehensiveness as the supplier wants to provide.

Your work here is certainly one of the most important products to come from this collaborative environment.

Ask for help when you need it.

15 March 2006

Marco Italy – at 06:45

Donnelly, thanks for this project! I’m a programmer (SQL and Web Developer) so please don’t hesitate to ask for tech help if you need it! I’ll be glad to help you.

Marco

16 March 2006

Michael Donnelly – at 19:26

Thanks Marco. I’ll be posting an updated version of the proposal soon, so look for that, because I’ll spend some time discussing technological solutions for data collection. That’s where your expertise might be very useful.

Michael Donnelly – at 19:35

Still thinking about Revere’s “systematic biases” comment. If I can’t figure it out beforehand this project might be useless before it is even born. Hopefully Revere will take a second look once the new proposal is out, and suggest to me exactly what sorts of biases might come into it.

17 March 2006

NS1 – at 01:30

Donnelly-

I think the reveres, who are expert in these matters, will review what and how you gather the info and will provide valuable input on bounding questions, especially.

Any large body of anecdotal evidence is useful, even raw. However, my hope is that a group will form, review all data regularly, consider the treatments and supply opinions / interpretations including a top list of ‘vetted’ solutions.

The process is the heart of analysis:

Disclaimer here→

eyeswideopen – at 02:48

Fantastic idea. Hope there’s electricity to be able to log it in. This is real “save the world” thinking at its best. Simple, direct, intuitive implementation could make this very, very powerful.

Michael Donnelly – at 09:39

NS1: I was struck by the the “pile of anecdotes” comment as well, becuause in fact a lot of very well executed studies done now use indirect methods, i.e. patients are basically asked to report what they experienced (while taking a drug, etc). The only difference is that there is usually some qualified clinician doing the interviewing. Here, there’d be no clinician, just a form. But a well designed form should get us most of the way there.

In my mind, the critical elements for a valid data set are:

1. the study must be prospective, i.e. reporting is not done after the fact, so there is no question of memory distortions.

2. there must be clear cut quantification of symptoms, using a standardized scale, so that people can’t shade their ratings of symptoms according to their own expectations. We’ve already got a pretty good scale, from anon_22.

3. various items on the watchlists must be well defined according to clear criteria, so the values of these things for different individuals are comparable. So, for example, if we have a “green tea” entry in the treatments section, there are plenty of ways to use green tea. One subject might take a green tea extract pill, while others might brew it as actual liquid and drink a certain amount. If we allow both to say “yes, I used green tea” we may be lumping dissimilar things into a single group. That would make it tough to know what to infer based on their experiences.

4. The size of the subject group must be massive. It needs to number in the 100K range I think. Why? Because there will be such a wide variety of experiences, that we’ll need numbers like that to deal with statistical noise, and the relatively small number of people who will be takig any one apporach. Take for example, the green tea believers (you know who you are). How many people will be using this stuff to try to ward off the flu? Out of 100K, will it be 1000? 2000? Perhaps. And we’ll need that many because perhaps the tea will work better for men than for women, or for older people better than younger, so we’d need enough green tea users to further subdivide the group into age groups and such. It doesn’t ake many divisions before we’re looking at group sizes in the tens, which is not enough power to draw valid inferences.

5.the data collection system should have redundancy built into it so that it can keep going under sub-optimal circumstances. So perhaps there should be multiple data collection sites. There should also be multiple channels for collecting data (internet, phone, in person). Finally, each subject should be linked up with several other subjects so they can keep track of each other.

Thanks for everybody’s support, it’s really helping motivate me to revise the proposal again. Update soon.

Michael Donnelly – at 09:46

eyeswideopen:

As for your “save the world” comment, perhaps it will, maybe it won’t. But we have to at least try. To let an event like this without trying to use it as an opportunity to get good data, to me (a died-in-the-wool empiricist) that would be a moral failure.

We may discover that the panflu is so awful that nothing anybody does makes much of a difference. If that’s true, perhaps we might consider rechristening it “superflu” and the survivors should start making their way to Colorado.

19 March 2006

Michael Donnelly – at 01:29

OK, I have finally finished the third revision of the proposal. Those of you who have been interested in the past, I want to reach out to you and to everyone else, to have a careful read of the document. There’s still a lot of work that needs to be done, including:

1. adding links to source documents 2. creation of watchlists for exposure risks, prevention strategies and treatments 3. design of the reporting form

I get the impression that since my name has been at the top, that some people might hesitate to edit it. I have now removed my name from the top, and put it at the bottom, with a clear disclaimer about the nature of authorship of this proposal at the top. I hope that opens the door wide for others to help with the effort, because I am tired of it. I don’t suspect I will take another look at it for a few weeks at least.

I’m tired, so TTFN

TIA, Mike D.

eyeswideopen – at 01:40

Michael: Whether it actually succeeds in saving the world or not, well, that remains to be seen. But the idea is the sort that conceivably could. Kudos.

NS1 – at 01:41

We’ll be going over it in the next few days. Thanks for giving us this opportunity, Michael.

22 March 2006

Michael Donnelly – at 15:05

I am now shopping this idea around to some big-name groups. I have been in contact with some people with some expertise in this sort of project (not sure they would like to be named so I won’t), and they tell me that they’ll be looking over the proposal early next week. Hopefully someone will like the idea. Not sure how quickly anyone would be able to move on an idea like this, but right now I am thinking of a timeline that has this project in the can by July 31. Hopefully any pandemic holds off until then. Of course, if it doesn’t, and enough people know about the project, an outbreak might have the effect of putting it on the front burner for a lot of people.

Any help you can give critiquing the text or adding source links would be greatly appreciated. I hadn’t intended to do this so soon, but I think I’ll be adding to my analysis section, trying to describe the analysis approach in more detail, particularly with respect to the time-series analyses that would be possible. Very important for establishing cause-effect relations, particularly when the incubation period is well known.

23 March 2006

Michael Donnelly – at 11:19

Just to make it clear here: I am trying to get an academic center interested in taking on this project, because they can provide needed funding, expertise, and technology support. I am not confident that the wiki model can accommodate a project like this. That said, anyone who wants to try to make it work as a wiki project is welcome to. If I do find an academic research group who will take on this project, it will absolutely be done with the understanding that their will be open access to raw data, with results and analyses released on a continuous basis.

Still need help with the links and logic.

Owl – at 11:38

Micheal, where is IT right now? Am I missing a link?

“Still need help with the links and logic.”

Michael Donnelly – at 11:56

Link is at the top of this thread. Or, you can navigate over to the opinion section.

24 March 2006

Michael Donnelly – at 11:23

Added a conceptual model diagram with description to the proposal. See if it makes any sense to you.

Owl – at 18:12

Michael, that is one detailed document. I cannot imagine the time and effort that took to compile. My thesis was a cartoon compared to it…

I was wondering if there is anyway to look at medication reporting sites that might have daily data logs that would be similar to what you are trying to create. You are more apt to have access to these than I am. I remember seeing one. Or what about drug trials - they must have some on line reporting programs… just a thought.

25 March 2006

Michael Donnelly – at 09:12

Actually, it was surprisingly easy. It started as about a 500 word essay in late October. It’s been sort of like a seed. All I did was plant it and water it. It has taken on a life all of its own. I have no idea how it got to 15K words.

Working on a detailed reporting form now, so that’ll be added very soon.

Still looking for help with linked sources.

I haven’t seen any reporting sites such as you describe. I suppose they must exist. Problem is that I would expect them to be too focused, on specific drugs and specific effects and side effects. This reporting form if it happens will cover about 100 variables, many of them behavioral. So it won’t be cut from the same mold really.

26 May 2006

BroncoBillat 01:20

Old thread closed to speed Forum access

Check dates

Retrieved from http://www.fluwikie2.com/index.php?n=Forum.ProposalForAnOpenSourceAvianFluTreatmentDatabase
Page last modified on August 15, 2007, at 10:01 AM