You have a better chance of getting struck by lightening than you do of overdosing on acetominephen. That, to me, is hardly a case for changing the administration parameters of these drugs.

Way to go on being useless, FDA.

Good point Barbie! The odds of getting sick from salmonella-tainted peanut products is pretty low, too. Can’t see why anyone got worked up about the FDA missing that one.

Actually, after bragging about being OCD about such things, it turns out Barbie didn’t do her homework. Lightning deaths in the US: 82 per year, on average. Deaths from acetaminophen overdose in the US: 458 per year. You’re more than 5.5 times more likely to die from an acetaminophen overdose than from a lightning strike. But hey, 450+ preventable deaths per year is statistically insignificant, so why bother, right?

Notwithstanding the real discussion here, where the hell did the phrase “shit purple nickels” come from?

Would “shit a brick” have been better? Anyway, “shit purple nickels” was an expression a friend of mine used to use when we were growing up, and a quick Google search confirms that it’s not unheard of:

http://www.google.com/search?q=“shit+purple+nickels”

Actually, I did do my homework, and my claim was consistent with my database, according to the national lightning safety institute (yeah, I said wtf at this as well, but it actually does exist):

http://www.lightningsafety.com/nlsi_lls.html

1000 strikes on people per year, which is what I said (…blah blah probability of getting *struck* by lightening…)

Compared to a populace of 300,000,000+ people, 450 is not statistically significant. Considering the number of reported acetominephen related medical emergencies, which I’m taking as 56,000 from the CDC report you sent me, it amounts to 0.8% of that population that is allegedly breaching the healthful limit of acetominephen consumption. Statistically speaking, a percentage of this magnitude can be considered as lying outside the normal distribution. That’s where I’m coming from when I say that “it is statiscally insignificant,” not a disdain for public health.

Also, I’m fairly certain the exact context I gave for my OCD claim was that my OCD does not extend outside my lab.

Oh well, I guess that’s what I get for bragging. Continue to jest at my expense; I live to entertain.

SB -

Without, I freely admit, having done more than quickly scan the exchange over at your place, a few quick notes:

1. It’s not a proper comparison to look at people struck by lightning versus people who died from an acetominephen effect/reaction. You either have to compare people who died in both cases (82 vs. 458, T. says) or who were affected in both cases (1000 vs. 56,000, you say).

2. Comparing the number of acetominephen-related deaths or medical emergencies to the population of the US is also an invalid comparison. You’d have to compare the number of deaths/emergencies to the number of people taking acetominephen, which is obviously a good deal smaller.

3. Identifying 56,000 as the number of people “allegedly breaching the healthful limit of acetominephen consumption” is just wrong. The number has a specific meaning: the number of ER visits related to acetominephen. Calling it something else, something which presumes both that everyone who exceeded the “healthful” limit was affected seriously enough to justify going to the ER and that, more importantly, every one of those people actually did, is obviously improper. (And no, embracing the first of those presumptions does not advance your case: that some people can take a toxic dose of something without apparent ill effects – or with ill effects wrongly ascribed to a different cause – is hardly a radical assertion.)

4. That raises another point, which is that you should not be so quick to dismiss the number of deaths. How many of those 56,000 ER visits would have resulted in death had that emergency treatment been unavailable? Obviously there is no way to know but realizing that it was the availability of emergency treatment that kept the death toll from being higher (even as we don’t know how much higher) is certainly relevant in considering the dangers of the drug.

5. Finally, “statistical insignificance” is often in the eye of the beholder. Those killed on 9/11 amounted to a trace over 0.001% of the population. How would you have reacted if someone called that death toll “statistically insignificant?”

To join the fray while focusing on the most meaningless of tangents, I think the “struck by lightning” comparison is okay (meaning, we need not further dissect those who died from lightning strikes). I recognize Larry’s point, and agree in terms of logical, but the idea of “more likely to be struck lightning” is so engrained in our language that it is almost a universal standard for putting the unlikelihood of anything into perspective. Consider this an exception to one-to-one logical requirements on the basis of colloquialism.

Just my thoughts. Carry on.

I have to read through the other discussions before making any more meaningful comments on the issue.

Actually, Dig, the reason I brought up the lightening strike anagology was specifically because it illustrates a highly unlikely probability. All apples-to-monkeys comparisons aside, lightening strikes and 9/11 victims etc, the statistical significance of the number of acetominephen-related health emergencies remains unchanged.

You want me to admit that I think 450 people dying, or 56,000 medical emergencies is “nothing to be concerned about.” I’m not going to say that. I don’t feel that way, and I will agree with you that it is not a nice number of medical emergencies. However, all that is implied when I say it is a statistically insignificant number is that it is on the order of outliers, and thus it does not demonstrate that there is a significant problem in the usage of acetominephen. It does not prove that there is a significant lack of understanding in the warning and usage guidelines by people who are taking the drug.

Right, but you’re caught up in semantics. You’re playing with semantic fire when you jump back and forth between the term “statistically significant” and then to a discussion that deals with whether those deaths are “significant” in the broader sense. Referring to them as statistically insignificant just primes the reader to feel that you are implying those peoples lives are insignificant as well. Then your implications that those who suffer do so on account of their own ignorance, coupled with your known libertarian tendencies don’t do you any favors either.

And, I hate to knock you off your pedestal, but “statistical significance” isn’t some uber-advanced concept that scientists like yourself have to explain to the intellectual hoi polloi like us. Statistical significance isn’t a concept reserved for scientists; you should have a pretty firm grasp of it if you’re, say, a second year sociology student at a city-run college who is pulling solid Cs…

The issue isn’t that TG or I don’t get that the ODs are statistically insignificant, the issue is that we see that, but then say, “Who gives a fuck; that’s not the point!” As, TG said, are the deaths preventable, what new obstacles/inconveniences would be caused by a regulation to fix them. It’s a question of a value proposition, not percentage of a whole, not of outliers. And, a good deal of regulations and laws are aimed to prevent “outlying” events in the first place, so I’m not sure why calling something an outlier is relevant in this context anyway.

So ,forget statistical significance, let’s leave that shit for Good Will Hunting. Talk in terms of a value proposition – why does the regulation of acetometaphine (w/e) + Percocet lose out on a cost benefit analysis? Once we figure that out, I’ll tell all y’all about some crazy shit called “square roots…”

*Percocet was involved in the weekend when the name “digglahhh” was born.

“And, I hate to knock you off your pedestal, but “statistical significance” isn’t some uber-advanced concept that scientists like yourself have to explain to the intellectual hoi polloi like us. Statistical significance isn’t a concept reserved for scientists; you should have a pretty firm grasp of it if you’re, say, a second year sociology student at a city-run college who is pulling solid Cs…”

Well, last I heard, I was being told that “you can’t use “statistically insignificant” interchangeably with “insignificant”, which I haven’t been doing; my placement of “statistically” out in front is quite intentional. So, it would seem to me that you are, in fact, in need of further clarification.

The fact that you think regulatory measures are aimed at preventing outliers means that you haven’t a clue what being an “outlier” actually means.

…and I don’t think you’re a bunch of dumb hoi polloi, I just don’t think you have a good understanding of statistics, and the importance of statistics where epidemiology is concerned.

No. It’s you who is jumping back and forth between science speak and everyday language.

That statement was in regards to laws and regulations in general, not in terms of determinations made by regulatory bodies in the drug development world. Many laws and regulations by both public and private organizations are implemented to address instances that happen at statistically insignificant (or very close thereto) rates.

Something being a statistical outlier is not a reason not to address it, if the outcome it is leading to is death.

As far as I’ve read, nobody has accussed your or using “significant,” as a subjective, abstract concept, interchangablly with “statistically significant,” an empirical concept which says nothing about the inherent value of the acts/incidences, etc. that the stats are measuring. What has been said is that even granting statistical insignificance (which is not so open and shut if you open up the definition of AEs beyond simple deaths], that should not be the deteriminent of whether this regulation is worth implementing.

What people are accussing you of implying is that because the rate of such incidences (as you conveniently, and some would say myopically, define them) is significantly insignificant (a point which has not been disputed on any grounds other than your limiting the scope of AEs to deaths), that means, ipso facto, the regulation is not needed, or shouldn’t be bothered with.

And, everybody knows what an outlier is, SB, it’s any Canadian kid with a hockey stick born in the first quartile of the calendar year…

There’s a lot of jumping from math-speak, to regulatory-body speak, to everyday speak in this discussion. But, the core of the issue seems much simpler; it’s a value proposition.

…well that would be a big blow Barb, since I work two jobs, one in drug development, and the other in statistics.

“And, everybody knows what an outlier is, SB, it’s any Canadian kid with a hockey stick born in the first quartile of the calendar year…”

Yes, I’ve read that Gladwell book, too, Digg.

“…I work two jobs, one in drug development, and the other in statistics.”

Oh, well, excuuuuuuuse me, had I known that you worked in statistics, I never would’ve brought up your misconception of outliers and significant figures in stats.

Well, last I heard, I was being told that “you can’t use “statistically insignificant” interchangeably with “insignificant”, which I haven’t been doing

Perhaps not, but as digg points out, you have been trotting out “statistically insignificant” as if that aspect is relevant to the overarching policy discussion. It’s not, at least not particularly. (Although digg summed it up much more elegantly than I did…) In other words, if you don’t think the alleged statistical insignificance of the numbers should inform the policy decisions, then why bring it up at all?

And just to put things into perspective, your opening salvo was to link to the recommendations of a panel of highly-qualified experts who have a lot more subject matter expertise on the issue in question and to call them “stupid.” I initially attributed this to a general aversion to regulation, which seemed to deeply offend you; at this point, as near as I can tell, your preferred course of action for the FDA in such cases is akin to something between analysis paralysis and “don’t just do something, stand there!”

Oh, you’re funny, Barb.

That statement was actually a bit disingenuous. I don’t work on the R&D side of drug development, so I’m not a scientist. And, my “stats” job is really based in sports. Your comment was just funny to me, so I had to comment.

I’d like to know where I made an error about statistical significance though. I didn’t say that I disagreed with your assessment, just that statistical significance (or lack thereof) of the deaths from ODs from acetaminophen is irrelevant to whether there should be a regulation barring what are ostensibly FDCs of acetaminophen and Perkys or Vikes.

My only other stated beef is that you continue to cling to this strawman that the only AEs that are worthy of being counted are deaths.

There’s also disconnect in your sample sizes for your comparison, but Larry already covered that. But, again, I don’t care whether those are legit comparisons or not because that’s not the issue.

One more time, why does instituting this regulation not make sense from a cost-benefit analysis standpoint? That’s been my question for like three posts now. You wanna try to answer that?

So far, from your responses, you’re answer seems to be, “because what it is seeking to prevent happens extremely infrequently.”

Is that your basic counterargument? If not, please elaborate.

I’m not sure I like this “Barb” thing…

“One more time, why does instituting this regulation not make sense from a cost-benefit analysis standpoint? That’s been my question for like three posts now. You wanna try to answer that?

So far, from your responses, you’re answer seems to be, “because what it is seeking to prevent happens extremely infrequently.” ”

…um, nooooo. I haven’t given a response to your question, so that doesn’t make my viewpoint cut-and-pasteable into a…what are you fond of calling out…oh, right, that might be *your* strawman right there. I’ve said that the events we’re discussing do happen extremely infrequently, but that’s not my answer to your question.

I haven’t answered your question because, honestly, I missed the cost-benefit analysis question being posed because I was snarking off to you and because of…ooooh look shiney!

That’s an interesting angle. I’m gonna think on that.

It really isn’t an aversion to regulation that is driving me here, but my question has been “prove to me that the system is not working.” So far, no one has come up with a reasonable answer *to that question*. Instead, it seems, everyone wants to talk about what is deemed a cause for calling for remediation of the current system. So, lets talk about that then. Digg suggests cost-benefit analysis, so I’m gonna think on that while I change the spark-plugs in my car.

In the mean time, since I’m trying to get my car in decent shape by the end of the day and it’s 104 degrees outside, why doncha help a sista out and start talking cost-benefit analysis to me, baby.

It really isn’t an aversion to regulation that is driving me here, but my question has been “prove to me that the system is not working.” So far, no one has come up with a reasonable answer *to that question*.

That’s because you’re asking the wrong question. It’s not a matter of whether the system is working, but of how well it’s working, whether or not we can easily do better, and if there’s any downside to trying to do better. That’s what the FDA’s shooting for, and they’re “stupid” for doing so.

Instead, it seems, everyone wants to talk about what is deemed a cause for calling for remediation of the current system.

Well, since you’re the one who has from the beginning been acting as if the FDA is on the wrong side of that analysis, why don’t you make a case to support your allegation that what they’re proposing is “stupid?”

***ponders for a second or two, decides she’d rather go back to being covered in engine grease than playing “semantics twister” with tgirsch***

I, too, vote for the engine grease. Enjoy the holiday weekend.