Biggest Ebola Outbreak In History- ~10000 Cases

[Elheru Aran]

You may also be thinking of the classic case of the Tuskegee syphilis experiment. Though that was domestic, it is certainly one example of unprincipled research being conducted upon unwitting non-white volunteers.

Some of the current public frenzy about Ebola apparently rises from a 2012 study in which, in experimental conditions, Ebola Reston was transmitted from infected pigs to monkeys via airborne vectors; no physical contact. However, apparently that's because pigs are particularly good at coughing up viral matter which then becomes airborne in saliva. Not so much with humans.

[Elheru Aran]

Ghetto Edit to add:

Debating currently on Facebook, I'm starting to draw the conclusion that the majority of the frenzy is because nobody really knows enough about the various biohazard precautions being taken and they don't trust the professionals to take care of the situation. So, they hyperventilate. "They're gonna screw up, Ebola is gonna spread and kill everybody"

Never mind that dealing with highly infectious diseases is what those people do for a living...

[Broomstick]

You may also be thinking of the classic case of the Tuskegee syphilis experiment. Though that was domestic, it is certainly one example of unprincipled research being conducted upon unwitting non-white volunteers.

Actually, no - I mean, I know about the Tuskegee syphilis "study", of course, but I was specifically thinking of use of non-US people for medical experimentation. Impoverished people in both Africa and India have been used for that, and not just by the US. It's not too dissimilar from the exploitation of poor Indians for donor kidneys for wealthier other people.

Some of the current public frenzy about Ebola apparently rises from a 2012 study in which, in experimental conditions, Ebola Reston was transmitted from infected pigs to monkeys via airborne vectors; no physical contact. However, apparently that's because pigs are particularly good at coughing up viral matter which then becomes airborne in saliva. Not so much with humans.

During the Reston, Virginia outbreak for which E. reston is named it did unquestionably involve human animal caretakers being exposed and developing antibodies, presumably because the virus did enter their bodies. For whatever reason, though, E. reston does not cause illness in humans despite its close relation to other ebola varieties. Sure, it could mutate enough to cause illness - and the flu could mutate enough to generate a 90% lethal pandemic. In theory. Possibly. I don't stay up late at night worrying about it.

Debating currently on Facebook, I'm starting to draw the conclusion that the majority of the frenzy is because nobody really knows enough about the various biohazard precautions being taken and they don't trust the professionals to take care of the situation. So, they hyperventilate. "They're gonna screw up, Ebola is gonna spread and kill everybody"

Also, too many Hollywood disaster movies in which the experts are either wrong or screw up.

[Wing Commander MAD]

Ghetto Edit to add:

Debating currently on Facebook, I'm starting to draw the conclusion that the majority of the frenzy is because nobody really knows enough about the various biohazard precautions being taken and they don't trust the professionals to take care of the situation. So, they hyperventilate. "They're gonna screw up, Ebola is gonna spread and kill everybody"

Never mind that dealing with highly infectious diseases is what those people do for a living...

A lot of that probably has to do with the fact that as Broomstick mentioned, the people who handle this stuff for a living and are highly trained professionals apparently lost samples of small pox that were recently found. Likewise from I believe from roughly the same time frame when that came to light we had potentially dozens of workers in I don't know how many facilities potentially exposed to anthrax because highly trained professional(s) didn't follow procedure. People tend to remember when things go wrong, not the 99% of the time when all goes as planned and nothing happens. Really this simply came at the worst possible time for the CDC. About the only thing that could make this worse is if this was happening directly when these things came to light, and not a couple months later.

[Broomstick]

Most people do not work in endeavors that require exact and precise procedures 100% of the time and therefore do not understand that it is inevitable that at some point things will go wrong. You see this in aviation, in security, and in medicine. In the real world Stuff Happens - things break, people make mistakes, and only God controls the weather (or, for that last bit, insert phrase of choice meaning "random shit we can't control").

Ideally, you make the procedures and safety protocols redundant enough that when something does slip through the cracks the event does not become catastrophic. Such as not drinking random vials of frozen shit you find in the back of a freezer and handling it carefully, rather than having a smallpox-and-toast combo for breakfast.

Since smallpox was declared extinct in the wild there are two "oops" I'm aware of. One was in 1978 when there was an "escape" from a lab in the UK, resulting in two smallpox cases and two deaths - one from smallpox and one suicide due to the guilt experienced by the man responsible for laboratory safety. The other was the more recent "vials found in freezer" incident which, fortunately, resulted in no illnesses. Two incidents in more than 40 years is pretty low, give that the first was when smallpox samples were more common in laboratories and the second was a discovery from samples forgotten from before smallpox was eliminated in the wild. Strictly speaking, since 1978 no smallpox has ever escaped from containment.

Anthrax is more problematic because it's as common as dirt - literally. It is trivially easy to find anthrax in soil samples. Also, a lot more research is going on with it, therefore, more opportunities for screw-ups. Frankly, I find bringing this incidents to light reassuring on a certain level - mistakes aren't being hidden, they're being dealt with. Then again, I doubt I'm a typical member of the public.

[Wing Commander MAD]

Most people do not work in endeavors that require exact and precise procedures 100% of the time and therefore do not understand that it is inevitable that at some point things will go wrong. You see this in aviation, in security, and in medicine. In the real world Stuff Happens - things break, people make mistakes, and only God controls the weather (or, for that last bit, insert phrase of choice meaning "random shit we can't control").

It doesn't help that the vapid parasitic 24-hour news cycle profits from exploiting that either. I'm sure this will eventually blow over until the next time Shit HappensTM, then the cycle shall begin anew. I do feel bad for the poor SOBs the CDC has fielding PR during all this. I don't envy that job.

I do wander, though, how many cases of "ebola" various hospital ERs have had to treat since the news decided to pick this up?

[Chardok]

The head of the CDC said that an Ebola outbreak in the U.S. is inevitable - Man, that's some scary stuff, even if it is limited in scope. 29 more people died between tuesday and wednesday alone.

[Simon_Jester]

I am revising my prediction at the beginning in light of more data coming in; the rate at which new cases occur does not appear to be slowing down... yet. This is something I banged together in OpenOffice while trying to get a sense for what's going on:



Anyway, the interpretation is that this is a graph of how many new cases were reported. Data wasn't available (to me, anyway) on every day of the outbreak, but you'd expect the rate of reporting to be deducible- if 80 cases were reported over a 4 day period that averages to 20 cases per day. We notice a couple of spikes during which hugely more cases than normal were reported, but the background rate is still in the 30s... and it was in the 20s when I issued my prediction.

I was previously optimistic that the graph was concave downward. That would indicate that while more and more people are catching the disease, at an increasing rate, the rate is leveling off, suggesting that ultimately public health measures were going to catch up with the spread.

But I'm no longer sure that the graph is concave downward, even ignoring the spikes.

EDIT: Then again, the trend COULD be concave downward if it weren't for the big spike around August 1 (the big one), but there's no way to know unless we have a few weeks with no spikes.

[Further notes, "Day 1" on this graph is March 1, to do the data analysis in a hurry I abstracted out everything except "how many people have caught Ebola" and "how many days since March 1"]

[Broomstick]

The head of the CDC said that an Ebola outbreak in the U.S. is inevitable - Man, that's some scary stuff, even if it is limited in scope. 29 more people died between tuesday and wednesday alone.

With international travel being accessible to pretty much everyone these days, yes, at some point you'll have an infected traveler show up in the US. Or anywhere else you care to name.

On the upside, the disease should be better contained in a nation with modern infrastructure than where it typically breaks out in Africa.

[madd0ct0r]

that's a nice exponential with reporting delay you have there simon

[Simon_Jester]

I'm not entirely sure of that, but it hinges on your interpretation of the last few datapoints, which are horribly skewed by the spike around August 1.
___________________________

Note for those in the audience who don't have calculus, if the number of cases is increasing exponentially then the graph of the rate of new cases will look exponential too, because the rate of change of an exponential function is itself an exponential function, and one with the same doubling time

On the other hand, other things, like a logistic curve look exponential at first before they level off.

[madd0ct0r]

The graph is reported cases, so there's going to be some natural variation around the period between infection and report. Couple that with a rich get richer mechanism - on a day with New cases, they check relatives and neighbors so potentially find new cases that were previously unreported, or are in early stages. At best this spike is robbing cases from the future, but it looks to like it's the missing exponential part

[slebetman]

Haj season is approaching. 2-3 million people living literally shoulder to shoulder then going back to their countries of origin. If it spreads there it'll spread everywhere.

[Elheru Aran]

Haj season is approaching. 2-3 million people living literally shoulder to shoulder then going back to their countries of origin. If it spreads there it'll spread everywhere.

Given that (rumour) Saudi Arabia already had someone diagnosed with Ebola pop up, I would not be surprised if they start screening people at the airport as they get off the plane or something. Given the amount of people that go on Hajj, that would be a really nasty way to ensure that Ebola gets spread around... especially as a lot of these aren't just 'regular' people, they're people who can afford to go on Hajj, which means you're talking a lot of upper-class folks... who probably can afford to travel... so, yeah, bad situation.

Perhaps they'll quietly prohibit Hajj travellers that have recently been to certain countries? Not sure they can do that though...

[Broomstick]

Haj season is approaching. 2-3 million people living literally shoulder to shoulder then going back to their countries of origin. If it spreads there it'll spread everywhere.

Given that (rumour) Saudi Arabia already had someone diagnosed with Ebola pop up, I would not be surprised if they start screening people at the airport as they get off the plane or something.

Last I heard, the gentleman in Saudia Arabia suspected of having ebola has now died. Test results to confirm or rule out ebola are still pending. He had returned from a business trip to Sierra Leone. I'm not overly concerned about that, as he had been treated in isolation and the Saudis will be handling the body properly to prevent infection from spreading.

The problem won't be people in hospitals, it's people sick outside of medical facilities that are most likely to spread the infection.

[Simon_Jester]

The graph is reported cases, so there's going to be some natural variation around the period between infection and report. Couple that with a rich get richer mechanism - on a day with New cases, they check relatives and neighbors so potentially find new cases that were previously unreported, or are in early stages. At best this spike is robbing cases from the future, but it looks to like it's the missing exponential part

Well, the thing about Ebola is that it has a pretty short 'burn;' there are no symptoms until suddenly there are lots of them. So cases of Ebola are nigh-undetectable until suddenly they're massively detectable, so it's hard to front-load detection unless there are LOTS of cases going undetected on a regular basis. In which case, given how infectious this thing is, it would have blown completely sky-high quite a while ago, I'd think.

[madd0ct0r]

could you send me the data file? I might try smearing the results a little (by about 24 hrs either way) to see wot new pops out.

[Simon_Jester]

Stuck it in my Dropbox, link should be here.

[Alyrium Denryle]

A lot of that probably has to do with the fact that as Broomstick mentioned, the people who handle this stuff for a living and are highly trained professionals apparently lost samples of small pox that were recently found.

A sample got misplaced by the FDA back in the 1970s... things were a lot more cavalier in the 1970s.

Likewise from I believe from roughly the same time frame when that came to light we had potentially dozens of workers in I don't know how many facilities potentially exposed to anthrax because highly trained professional(s) didn't follow procedure.

Bit of a lab fuckup there. What they actually sent WAS sterile, and they caught the error very quickly.


As for private enterprise, no... no... That would would get done anyway. It IS done anyway in university and government labs, which, with higher funding allotments would become far more extensive.

[Wing Commander MAD]

A lot of that probably has to do with the fact that as Broomstick mentioned, the people who handle this stuff for a living and are highly trained professionals apparently lost samples of small pox that were recently found.

A sample got misplaced by the FDA back in the 1970s... things were a lot more cavalier in the 1970s.

Likewise from I believe from roughly the same time frame when that came to light we had potentially dozens of workers in I don't know how many facilities potentially exposed to anthrax because highly trained professional(s) didn't follow procedure.

Bit of a lab fuckup there. What they actually sent WAS sterile, and they caught the error very quickly.


As for private enterprise, no... no... That would would get done anyway. It IS done anyway in university and government labs, which, with higher funding allotments would become far more extensive.

Uh, where did I say anything about private enterprise?

Anyways, the point was that highly trained professionals made mistakes, as is bound to happen since they're only human, and that that usually has a negative impact on public perception of the organization as a whole. Undeserved, but understandable. As I said, bad timing combined with a news media that profits from fear mongering.

[madd0ct0r]

ugh.

data is up to date to the 8th august report. I've had to leave most of that final report out of the graphs, since there are people walking around infected now who haven't been reported yet but were infected in the same time period. It gives a false sense of optimism so I cut the last few days off.



Ok, so what I did was i took the reported number of cases and counted back to their likely infection day. Since I don't know shit about ebola's incubation factors, I've taken a uniform probability distribution to spread the incubation period between 2 and 3 weeks.
Since we get infections per day as the base line, if we are winning, that should be going down. It's not really.

Failing that, I tried looking at the rate of change of the rate of infections (and then the rate of change of that and all). It swings about too much to get useful info out. The swings are getting bigger, but that's to be expected since the outbreak is also larger. If anyone's got a good correction for that I'd appreciate it, spent far to much time on this already.





https://dl.dropboxusercontent.com/u/641 ... akData.ods

[Simon_Jester]

Hm, yeah. When I first started the graph, I may have misread it or been misled by its very smooth shape... and that was back on the 26th, two weeks or so ago.

[Ziggy Stardust]

Ok, so what I did was i took the reported number of cases and counted back to their likely infection day. Since I don't know shit about ebola's incubation factors, I've taken a uniform probability distribution to spread the incubation period between 2 and 3 weeks.
Since we get infections per day as the base line, if we are winning, that should be going down. It's not really.

As I posted earlier in this thread, the incubation period for Ebola is highly variable, from as few as 2 days to as many as 3 weeks. The average, though, is typically less than 10 days, right around 1 week (that is, the distribution of incubation times is skewed; it may be more helpful to use an exponential probability model; Weibull distributions are common in public health data simulations). Ebola is not contagious during this incubation period; it only becomes contagious when the patient is symptomatic.

Failing that, I tried looking at the rate of change of the rate of infections (and then the rate of change of that and all). It swings about too much to get useful info out. The swings are getting bigger, but that's to be expected since the outbreak is also larger. If anyone's got a good correction for that I'd appreciate it, spent far to much time on this already.

How did you analyze it? Typically, when studying infection rates you use a logistic model. With a few more assumptions you could also fit an accelerated failure time model. If there is enough interest, I can do all of this tomorrow easily enough; one of the benefits of working in epidemiology is that I have code for this just sitting around.

[madd0ct0r]

oooh, that'd be cool.

I didn't bother running a logistical regression on it since the data was so swingy, and the overall number is the sum of cases over several geographical areas. Just tried now after implementing a roughly weibull distribution and spreading the incubation period. The link above should point to an updated spreasheet. Data's still pretty blah. The grey line marks the artificial drop off caused becuase the last report I've seen is the 8th august.

[loomer]

You may find this helpful. http://hisz.rsoe.hu/alertmap/index2.php