Unsafe at any level!
-
@Speakerphone Dude said:
who speaks like an awkward teenager with a Tourette syndrome
Swears make Jehovah cry.
-
@boomzilla said:
It's more like saying, "We used to think 120 systolic pressure was fine, but now that's bad. You should have 115 or lower. And then when we get enough people below 115, we'll switch to 110, because there's no safe level of blood pressure."
Why not? It worked for cholesterol.
-
@morbiuswilters said:
Wow, that sounds really stupid.
Hence the use of the word 'bollocks' in my post. There's another post from that site(dated Jan 26 '12 - will edit post to link to it later if I remember)which elaborates on the attempt to address the UK Government's 'not even wrong' aspect of this:New guidelines issued by the Department of Health say that the indicator used to assemble these statistics should be based on just alcohol-related primary diagnoses.
This may seem like a small change, but it has big implications: it will cut numbers dramatically. The claimed figure of over a million admissions a year is based on scrutiny of all the conditions recorded for each patient, not just the one for which the patient is actually admitted
[mod - added link - PJH]
-
@morbiuswilters said:
Wow, that sounds really stupid.
It's statistics. They cannot possibly measure alcohol consumption in the entire inpatient population, and they certainly cannot claim a patient is drunk unless they have lab work to prove it - you could get sued for writing "patient was drunk" in the patient chart unless you can back it up with facts.
So what they do are study a smaller set of the population, those that DO have known alcohol measurements, and they use those to determine a degree of correlation between their visit diagnosis (which is known for all patients) and intoxication. So long as the sample group they use to estimate the degree of correlation between alcohol consumption and clinical diagnosis is representative of the population at large, the method is very accurate.
It's like any kind of polling or statistics - you cannot measure every individual, so you measure a subset and extrapolate to the entire population. As long as your subset is of appropriate size and composition it's fairly accurate.
-
@Cat said:
So what they do are study a smaller set of the population, those that DO have known alcohol measurements, and they use those to determine a degree of correlation between their visit diagnosis (which is known for all patients) and intoxication.
This is patently not what they're doing. Another abuse they use: If, e.g., you're admitted to hospital twice for 'self harm' you'd be counted twice x 0.2, even though you're only one person so should only be recorded as once x 0.2, exaggerating the numbers even more.
-
@Cat said:
@morbiuswilters said:
Wow, that sounds really stupid.
It's like any kind of polling or statistics - you cannot measure every individual, so you measure a subset and extrapolate to the entire population. As long as your subset is of appropriate size and composition it's fairly accurate.
I would definitely believe that this is how the numbers were developed. And I would agree with Morbs that it's really stupid. But they get some very precise numbers at the end of the year, which makes for good news stories and budget proposals, so it all works out in the end!
-
@PJH said:
This is patently not what they're doing. Another abuse they use: If, e.g., you're admitted to hospital twice for 'self harm' you'd be counted twice x 0.2, even though you're only one person so should only be recorded as once x 0.2, exaggerating the numbers even more.
For some of the acute cases, that seems to be their methodology. For the effects of chronic alcohol consumption, they use relative risk factors. For example, males aged 16 to 24 who drink are about 2.25 times as likely to develop oral cancer than those who do not drink. 80% of males in this age group are drinkers. From this, you can estimate the population of will develop 2 oral cancers for every 1 that would be expected if the entire population were non-drinkers; in other words, half of the oral cancers in this age group could be eliminated by eliminating drinking.
The actual methodology is a more complex than this. They don't just consider drinking as a binary - it's actually abstainance, low consumption, moderate consumption, high consumption, very high consumption, and they look at the distribution of those levels of consumption in the population, and the relative risks of each level. They then look at the (known) prevalance of a disease condition in the general population and extrapolate the prevalence of the same disease in a population (of identical age and gender distributions) where nobody drank.
And they are trying to calculate the number of ADMISSIONS not the number of PEOPLE, so it's absolutely correct to count both. Think of it like this: if you are admitted six times for liver damage from alcohol, *all* of those admissions were caused by drinking, not just the initial one. Other statistics in the report are based on people, but this one was specifically trying to count the number of hospital admissions related to alcohol.
Overall, the methodology is sound - there's no other way to measure the effects. You can't tell at the level of the individual whether their disease was caused or not caused by alcohol, but over large populations you can.
For example, males who smoke are 23 times as likely to develop lung cancer compared to non-smokers; that means if you had 2300 male smokers with lung cancer, 2200 of those 2300 had cancer because they smoked, and 100 would have had lung cancer anyway. You couldn't tell which of the people was in the 2200 versus who was in the 100, but over a large enough group you could come up with an accurate estimate of how many cancers could have been prevented among the group had they made a different lifestyle change. That's basically the methodology.
-
@Cat said:
Overall, the methodology is sound - there's no other way to measure the effects. You can't tell at the level of the individual whether their disease was caused or not caused by alcohol, but over large populations you can.
Even if their methodology is really sound (and I'm definitely not prepared to concede that), the numbers are reported like they're about acute things, like alcohol poisoning or injuries that happened while impaired. To lump stuff like chronic liver disease with this sort of thing is simply dishonest and misleading. It's like using cancer rates as a cautionary tale about house fires as a result of smoking in bed.
Which is just another long winded way of saying that it's pretty stupid.
-
@Cat said:
And they are trying to calculate the number of ADMISSIONS not the number of PEOPLE,
Not when the government are using those numbers and interpreting them as the amount of the population that are binge drinking. Which is what the UK government, and the associated fake-charities (and the Daily Mail) that are against having fun are doing.
-
@Cat said:
For example, males who smoke are 23 times as likely to develop lung cancer compared to non-smokers;
Citation please. Otherwise you're spouting bollocks.
-
@PJH said:
@Cat said:
For example, males who smoke are 23 times as likely to develop lung cancer compared to non-smokers;
Citation please. Otherwise you're spouting bollocks.That's roughly the number I've seen. I suspect it varies for different types of cancer, etc, but see here:
@Mayo Clinic said:Relative risk
Relative risk gives you a comparison or ratio rather than an absolute value. It shows the strength of the relationship between a risk factor and a particular type of cancer by comparing the number of cancers in a group of people who have a particular exposure trait with the number of cancers in a group of people who don't have that trait.
For instance, relative risk might compare the lung cancer risk for people who smoke with the lung cancer risk in a similar group of people who don't smoke. You might hear relative risk being expressed like this: The risk of lung cancer for men who smoke is 23 times higher than the risk for men who don't smoke. So the relative risk of lung cancer for men who smoke is 23.
Note that the relative risk I've seen in studies for second hand smoking is usually less than 1.2 (i.e. a 20% increased risk), which, from an epidemiological standpoint means that it's not an interesting correlation.
-
@boomzilla said:
@aliquot said:
I read the rest of the article, the idea is to lower the lead-blood-level threshold for action to below the "Shit, this kid has lead poisoning" level, so that public health officials can take action before the kids actually get lead poisoning.
I'll give you the benefit of the doubt, and assume you just didn't read closely enough instead of innumeracy, and will inform you that the current standard, 10 is well below the threshold for poisoning, which is 45. This is in decimal numbers, BTW, so the current regulatory threshold has a 350% buffer before the clinical problem of poisoning is reached.
You are mixing up two different things here. Acute clinical lead poisoning != the only bad effect lead has on children. Here's the quote you're referencing in the article: @NYT said:
Actual lead poisoning, he said, is defined as blood levels above 45
What he's calling "actual lead poisoning" means your blood level is so high that the lead has systemic effects - and if high enough, can actually kill you. At 70-100 µg/dL kids can have comas and seizures. At 45 µg/dL you can get anemia, porphyria-like symptoms, lethargy, gross neurological effects and so on. 6 weeks of 45 µg/dL and you will get "lead lines" in your bones which show up on X-ray. IANAD but it appears 45 µg/dL is the clinical threshold for "severe" lead exposure meaning you need chelation therapy to get the lead out of your blood.
micrograms per deciliter of blood (µg/dL). At that level in young children, he
said, “their life is at risk, they need to be seen clinically, and
interventions absolutely need to be taken immediately."So that's what the 45 µg/dL number refers to. But, children exposed to much lower levels of lead have lower IQs and higher prevalence of ADD. These neurological effects are irreversible, so when you take the lead away the kids don't get smarter. The 10 µg/dL threshold in children is about this, not about acute ie "actual lead poisoning." The reason the CDC has changed the threshold from 10 µg/dL to 5 µg/dL is that a lot of new science has come out showing neurological effects in children from blood lead levels < 10 µg/dL.
If you're actually interested, here's the report from the advisory committee that recommended the lower threshold: http://www.cdc.gov/nceh/lead/ACCLPP/Final_Document_010412.pdf It summarises the more recent science showing permanent, irreversible damage to children from blood lead levels < 10 µg/dL.
@boomzilla said:
@aliquot said:
Preempting lead poisoning in kids doesn't seem like a waste of money to me.
So, at what level of lead does it become "poisoning?" It's nearly as much of a cliche in toxicology as correlation/causation is in statistics, but "the dose makes the poison." How much will it cost to reduce the lead from minuscule amounts to microscopic amounts? What are the benefits?
I skimmed the report from the advisory committee and learned that even with blood level 1 ug/dL - 4 ug/dL there is a measurable drop in IQ of about 3-4 points. Definitely something "unnoticeable" in one child, only noticeable in a large study with a control group. At 10 µg/dL the drop is about 5 IQ points. And a population drop of 5-10 IQ points can make a big difference in the number of people who are moderately retarded.
Most parents whose kids have BLL (blood lead level) of 5-10 µg/dL will have no idea because the symptoms are subclinical. Kids with 10 µg/dL BLL won't be anemic, have gastrointestinal problems or seizures or gray skin, but will be on average 5 IQ points stupider and 1cm shorter. If their parents knew they had elevated blood lead levels, they could do something about it before their kids grew up a little stupider and a little shorter. The article links to a blog post that explains how lowering the threshold to 5 µg/dL will result in more screening, more lead abatement from property owners, more family lead education and follow-up testing.
What's the cost to society of many people being a little bit stupider, and a big difference in the number who meet the cutoff for moderate retardation? I don't even begin to know how to calculate that.
tl;dr : The CDC lowered the threshold from 10 µg/dL to 5 µg/dL because there's new science showing irreversible cognitive effects from childhood blood lead levels <10 µg/dL. 5-10 µg/dL of lead in the blood won't make a kid sick the way 45 µg/dL will, but it will make the kid stupider.
-
@aliquot said:
I skimmed the report from the advisory committee and learned that even with blood level 1 ug/dL - 4 ug/dL there is a measurable drop in IQ of about 3-4 points. Definitely something "unnoticeable" in one child, only noticeable in a large study with a control group. At 10 µg/dL the drop is about 5 IQ points. And a population drop of 5-10 IQ points can make a big difference in the number of people who are moderately retarded.
OK, assuming all of that is correct, then TRWTF is either the reporter or the guy he quoted for putting out a message that is the opposite of what they meant.
-
@aliquot said:
it will make the kid stupider.
So, why aren't we upping the ammount of lead then? We need stupid people
-
@serguey123 said:
So, why aren't we upping the ammount of lead then? We need stupid people
Don't know about where you live, but there seems to be a more than adequate supply of them around here.
-
@Scarlet Manuka said:
@serguey123 said:
So, why aren't we upping the ammount of lead then? We need stupid people
Don't know about where you live, but there seems to be a more than adequate supply of them around here.
It is our main exporting good, we always need to breed more
-
@morbiuswilters said:
I should do that, but I've been too busy orchestrating my move to Paradise...
It really isn't. It *is* pretty, but most of that is ruined by the tourist trade. You have to get off Oahu...
