Marburg Surveillance Project

Various ^ | May 4, 2005 | Vanity

[Judith Anne]

Welcome to the Marburg Surveillance Project.

This thread will be used for all of the latest Marburg Outbreak News and comments. This is the place to post all comments about the Marburg outbreak, all articles and links to articles about the Marburg outbreak.

We're going to use just one thread instead of having to go from article to article as we have in the past. We'll use this thread as long as we can.

[unseen]

Yes but there is a window between one doubling and the next where this goes from a problem to a major problem. Governments will always wait for the last second before making any tough decision. All Politicians are spineless wimps it is their nature. So if the numbers are not correct or the size of the problem is underplayed the Governments will miss the crucial time frame and many more thousands, millions could die. So in a big picture sense no the actual numbers do not matter but in the little battle of this virus the lives of many people hang on the truthfulness of the numbers.

[2ndreconmarine]

The reason that the total number (current) doesn't matter is that the main growth is in the future.

You are absolutely right. You also said it better and more intuitively than I could.

The mathematical answer is that the partial derivative WRT total present number is small compared to the partial derivative WRT growth rate. A small change in growth rate makes a huge change in outcome. A small change in present values makes a small change in outcome.

[Judith Anne]

...the lives of many people hang on the truthfulness of the numbers.

Yes, and knowing how inaccurate I think they are, you can understand my concern.

[Doctor Stochastic]

I misused the term "fittest" here. If we assume the "fittest for infection genotype" is the distributional mode on infection, perhaps some change in the body could dampen that particular genotype until the immune system could get going. I think the body does this itself by running a fever.

It's similar to what "particle filter" (not hepa) computations do to find control parameters in a dynamic system. If the high probability group isn't wanted, that's the one that's killed. I doubt it would work.

[Doctor Stochastic]

But your comment on government intertial is spot on. A crisis would develop almost instantly.

[Judith Anne]

Please ping the list to posts 1685, 1687 and 1688, and the discussion following. I'm not home.

Thanks...

[2ndreconmarine]

Thank you ........

You're welcome. My pleasure.

[Judith Anne]

Strike that, looks like you already did. *Duh!*

[Doctor Stochastic]

I actually posted to the wrong one of you. Unseen made the governmental comments. However, you are both correct.

[Dog Gone]

So in a big picture sense no the actual numbers do not matter but in the little battle of this virus the lives of many people hang on the truthfulness of the numbers.

Those are certainly important, but they are not as important as the location of those numbers. Since containment is the only way to effectively fight this outbreak, location means everything.

As long as this bug is killing people in an isolated province, it's a fascinating exercise in responding to a deadly virus, no offense to those who've died intended.

Speaking selfishly, I want to know when Marburg is an actual threat to me, and that's probably sometime after it spreads from Uige and its province into other areas.

Unfortunately, I don't see any steps being taken to prevent that.

[2ndreconmarine]

It appears there is a lot of research going on about how the VP40 matrix protein for Marburg (and Ebola) are causing (allowing) the virus to more effectively penetrate the host cell walls and replicate. From the research I have read, it seems that this is a continuing processes in the evolution of Marburg that is increasing its virility.

That would fit the mathematical model nicely. Notice that I did not specify the mechanism for increased growth; I only postulated that there was some genetic makeup that would enable it. Perhaps it is this gene. Moreover, if this gene is specific to the growth rate / virility, then other aspects of the genetic makeup could vary without affecting the growth. That would also fit well with the broad genetic optimum assumption in the model.

This is one of the things that appears to have the epidemiologists on edge.

Clearly.

Now I know that this outbreak is Marburg but, as Kelly_2000 pointed out to redoglum in post 1217, in part, "...it is Marburg and something entirely new." .....

Since the first signs of the outbreak occurred in October 2004 how many cases infected others before knowing it was Marburg? This is, essentially, a rhetorical question. However, if I understand your modeling correctly it appears to fit your modeling represented by Figure D.

Second point first. Excellent point and I think the data support this. Indeed, my earlier plots showed this explicitly. Namely, the epidemiological growth was faster in the earlier stages of the epidemic. When the virus was identified, better isolation was instituted. I have not modelled the data that way this last time because there are too many unknowns.

The issue of classification of the virus, i.e. : Marburg, a variant of Marburg, Ebola, Angola Hemorrhagic Fever, etc. is largely a moot point IMHO. I believe there is no significant difference. The classification serves no purpose in understanding or modelling; it is merely taxonomy.

Now, taxonomy generally has a purpose, largely to break up complicated problems into simpler ones. However, it can be overdone and become restrictive, and I believe that is the case here. We are arguing about the name of the thing. Call it an umtyfratz if you want. It doesn't change the underlying dynamics.

There appears to be no structural, essential difference between any of these viral variants. They are all filovirii (i.e. their structure is filamentary). They all consist of an amino acid sheath surrounding an RNA core. They all work the same way. They all are hemorrhagic fevers. The genetic makeup is about the same length and complexity.

Clearly, the exact genetic makeup differs. However, it can equally accurately be viewed as a single virus type, such that there are specific genetic combinations that provide successful growth. My original model postulated a growth function that was a single peak: it was a Gaussian. However, by contrast, imagine a growth function with several peaks (and valleys). The peaks correspond to the different names of the disease (e.g. Ebola, Marburg, etc.). The valleys correspond to nothing; they die out. Some of the peaks are higher and broader than others. However, essentially, this is all the same virus with modest genetic changes. What is significant is that this particular genetic combination of this virus may have a faster growth rate and a broader distribution than others. Name it anything you please.

[infominer]

Thank you for your excellent analysis. Even with your highly detailed modeling and explanations, you always bring the discussion back to "brass tacks". Both are essential to following this thread and understanding if / when this outbreak can / will become personal.

Thanks again!

[little jeremiah]

slackjawed in awe placemark.

[Dog Gone]

I don't know what time zone you're in, but I don't think as clearly as you just did after about 3 pm my time.

[unseen]

Question......

Your analysis is for a select group starting with the first case of infection. However Human society is not one group but is made of several groups within groups. Thus first instance the Uige is a group but within Uige are neighborhoods and within neighborhoods are families. Taken to the other conclusion Uige is a subset of a province which is a subset of a country which contain many subsets groups i.e city. thus your growth rate would be dynamic would it not? In Uige the growth rate would match your graphs but in a subset just recently infected say a township in the capital city the growth rate would reset to the beginning would it not?. So that while all our focus is the Uige brushfire. There could be a smoldering fire in several other cities and countries in Angola and Africa respectively could there not? It was months before Angola was able to identify the virus. Would it not stand to reason that it would also be a period of time before a "hotspot" was found in another subset especially with a reclassification of the numbers.

[Dad yer funny]

ping to 1731

[2ndreconmarine]

In Uige the growth rate would match your graphs but in a subset just recently infected say a township in the capital city the growth rate would reset to the beginning would it not?.

Actually, no. That is not how the exponential works. An exponential shifted in time is still the same exponential with the same time constant but a different intial value.

So that while all our focus is the Uige brushfire. There could be a smoldering fire in several other cities and countries in Angola and Africa respectively could there not?

Yes.

We have already discussed this basic issue on these threads. The epidemiological projections that I have done are very simple and large-grained. In order to do the fine grain analysis you suggest, requires a full-blown numeric epidemiological code. Those codes have man-years of developmental efforts and include all of the variables you suggest (and a good many more). They are well beyond the scope of my efforts. So, the best I can do is an approximation that assumes general averages for all those other effects. It means that these projections cannot be as accurate. They are probably reasonable given the quality of the data though. That is, given the quality of the data, you won't do much better with the large codes because the improvement is small compared to the data uncertainty (I guess).

[unseen]

Agreed.

[xVIer]

Quarantine, has worked in the past but may not work now... Officially, there is little information known (admitted) regarding the habits of this virus, the vector remains unkown, and also any plants,animals, bugs etc. could be potential hosts/carriers... these facts create a number of unwelcome scenarios...we know that this is closely related to the flu and has a likely potential to recombine with the flu (this is unfathomable!)...so migratory birds, mosquitos, or even countless inanimate trade objects carry the threat of spread, hopefully trade would stop under a quarantine. For now we are trading freely, as far as I know. Angola is big in oil, however Uige, itself, is a coffee producer... that alone has tremendous implications...
We know from previous stories discussed here that, if not outright airborne, it is definately aerosolized, can survive on surfaces and in water for days, and has killed dogs and pigs,along with monkeys and humans. Will the MSM ever start asking questions about this?
I really hope we are all just paranoid...wouldn't that be great?! ;-)

[null and void]

I wouldn't worry too much about coffee. Any virus would have a hard time surviving all the processing.