The one statistic you never hear about drugs | Hacker News Reader

The one statistic you never hear about drugs

alexkrupp.typepad.com

23 pointsAlex391729 comments

You don't hear that statistic because it's cobbled together from a bunch of completely different causes to make a shocking-sounding headline.

Basically, the take-home from the article is that smoking kills a lot of people, drinking too much is bad for you, and people die from adverse reactions to medication.

And that the deaths from illicit drugs were relatively low. It's the things you don't read about in the newspaper (not outrageous or tragic enough, just normal) that will get you.

That is actually a pretty interesting nugget that I missed on my first pass. I'm surprised that hasn't become a libertarian/anti-drug warrior rallying cry to convince those less moved by the "victimless crime" approach.

I work as a teacher. Drug deaths are higher frequency in younger cohorts. Emotional impact much higher. Libertarians probably staying away from that one.

That's true, and again, something that more data like this can be really helpful with. I disagree that it will steer people away from this argument though, since "victimless crime" is alive and well, despite the very moving and personal response of "parents of the children who die are very much victims." Especially given that those who libertarians are trying to convince are generally technocrats, assuming they don't bother attempting to convince those arguing from morals.

Doesn't distinguish dying from smoking at 65, from dying from an overdose at 18. These are massively different problems.

So the article boils down to, middle-aged people die from smoking; all else is noise in the statistics.

This is fairly important. Drugs and alchohol kill people fairly young, which gives them more impact on total people-years removed.

Tobacco, alcohol, and prescription drugs, many of them probably prescription painkillers, kill hundreds of thousands of people every year.

When you say "drugs and alcohol" you should indicate the illicit drugs part of that statement is under 1%. I doubt you are correct about alcohol killing the young primarily. This article doesn't provide a lot of detail except for a broad indication it's not underage drinking that kills the most: http://www.medicinenet.com/script/main/art.asp?articlekey=17...

>> prescription drugs, many of them probably prescription painkillers, kill hundreds of thousands of people every year.

But how many lives are saved by these drugs? To be clear, I am totally open to this number not being nearly high enough to justify the deaths caused, and very much believe our prescription drug system should be reformed. But the fact that prescription drugs actually save lives, and are not simply causes of death, gets ignored far to much for me. I want to know the true net cost of prescription drugs (an intentionally absurd analogy would be something like airbags kill 7 people a year with no mention of how many people those airbags saved).

On the other hand, Acetaminophen is used to supposedly make prescription painkillers less subject to abuse. I wonder how many people are killed by liver failure this way? That's pretty much intentionally killing people who could be put into addiction treatment.

Also a good point. My main point is when looking at drug deaths, the total should reflect (deaths caused-deaths prevented), rather than a raw count of total deaths. At least when looking at it from a macro, pure optimization perspective.

It does distinguish, at least between smoking and alcohol:

While many fewer Americans are killed from alcohol than from tobacco, the absolute percentages are somewhat deceiving; whereas the lifespan of current smokers is about 12 years shorter compared to those who have never smoked, the average death from alcohol comes a whopping 29 years prematurely.

This seems like comparing apples to oranges. Looking at the average lifespan of smokers v. the average age of an alcohol related death isn't right. It should either be the average lifespan of those who drink (or perhaps even more accurately, alcoholics) v. the average lifespan of smokers. Or, the average age of a smoking related death v. the average age of an alcohol related death (and within this, probably should add something that splits out causes between alcohol related decision making, ie drunk driving and alcohol related illnesses, although the point of the comparison is to highlight the dangers, so the fact that smoking doesn't result in drunk driving accidents is important).

> This seems like comparing apples to oranges. Looking at the average lifespan of smokers v. the average age of an alcohol related death isn't right.

It is an apples and oranges comparison, but that's just the data we have. I tried to word this in a way that highlighted the fact that the comparison was a little dodgy.

Thanks for the clarification. I'm just pretty wary of any conclusions drawn from this comparison, given the issues I highlighted.

That's probably a good idea. But hopefully surfacing the lack of good data is at least some productive.

This is probably the most important stat about drugs, however, it falls victim to a pet peeve of mine. For cigarettes and alcohol and drugs of abuse (the ones without accepted medical applications - street heroin, cocaine, etc.), the analysis is pretty useful. However, with prescription drugs, simply counting the number of deaths is misleading. ADR's are tragic, and we should be doing all we can to learn more to prevent them. However, a correct analysis, at least when attempting to optimize the way it seems he is trying to, is to total the number of deaths caused by drugs, and then subtract the number of deaths prevented by these drugs. All optimizations should be addressed through that lens. I think the author would agree with this, I just didn't see any mention of the fact that while drug deaths are tragic, there are also many deaths prevented by current application of drugs, and ignoring the opportunity cost of changing drug application (potentially reducing the number of lives saved) is potentially dangerous.

Yeah I didn't mention this in the essay, other than briefly alluding to the fact that not all were preventable, but if you look at the footnotes the sources I linked to go into this in more detail. It's difficult to come up with an exact number though: the primary source I'm citing says that (IIRC) only about a third of ADRs are preventable, but the FDA page citing that source says that most ADRs are preventable. And in a way they could both be correct; it's possible that someone could die from taking an emphysema drug that they needed in that moment, but maybe the emphysema itself was preventable.

Thanks for the clarifications. Don't get me wrong, this was a great effort at summarizing the data available, I was just adding the additional data that would be useful.

The information about medication non-compliance is interesting.

Non compliance doesn't appear to be related to cost (similar rate in England with low cost or free prescriptions and US with expensive prescriptions), or severity of illness (main reason for failure of transplanted organs is non-compliance with med regimen), etc etc.

About half the English medication budget (£4.5 bn of £9 bn) is wasted because people do not take their medications properly.

There's possibly big money if you can work out a way to get people to take their meds correctly. Drug companies want this; health providers want this.

Is there any data on why people don't take their meds? Do they just forget to take a pill every day, or is it willful noncompliance due either to fear/misinformation or adverse effects? And is the problem typically that the drugs don't work if not taken every day (i.e. missing a day compromises treatment fully) or do the noncompliant tend to just stop taking them outright?

This paper, linked in the footnotes, is a good starting place:

http://www.tandfonline.com/doi/abs/10.1080/17437199.2010.537...

Sanity check: This post attributes 19.1% of all deaths to smoking, but this is greater than the total number of smokers (18.1% according to http://www.cdc.gov/tobacco/data_statistics/fact_sheets/adult... which was the first search result I got). Since the author used that statistic and didn't immediately launch into a justification of how it's not completely crazy, I can't trust this article enough to read further.

"Since the author used that statistic and didn't immediately launch into a justification of how it's not completely crazy, I can't trust this article enough to read further."

The methodology is in the footnote. But at least cancer deaths from smoking deaths lag by several decades, so I don't see why it would be at all unusual for there to be more smoking deaths than current smokers.

That is the expected result for a sub-population that has a shorter lifespan.

My sibling comment addresses deaths lagging over time. But even in the steady state of (say) 18% smokers, you expect more than 18% of deaths to be smokers because their life expectancy is shorter.

If every time a smoker dies you blame it on smoking, then sure. But while smoking is really terrible, it's not quite that terrible; smokers still manage to die of unrelated things sometimes.

Appreciate all the data packed into this post. There are some amazing stats here. Agree that we definitely need this type of clarity to increase awareness and generate dialogue.

The drug abuse that pisses me off the most is the doping of livestock with antibiotics, to (a) make them grow faster, and (b) enable disgusting living conditions for the animals being housed.

There are a lot of things we do wrong with antibiotics (e.g. doctors who prescribe them for a cold, patients who finish a course prematurely because they "feel fine", and the aforementioned abuse of them in factory farming) and drug-resistant bacteria kill people who had nothing to do with the original abuse.

That should be included in those statistics as well.

Unfortunately that is almost impossible to measure with our current tools (at least I have never seen it done convincingly, and I generally agree with you that this is ignored too much. Although I have to qualify that GMO-food is better than no food, while recognizing that in almost all cases, that's not anywhere close to what the choice boils down to).

I think it was sama who said that factory farming is one of the things that in 50 years we will look back at in confusion over why anyone accepted it, and I tend to agree.

Reply on news.ycombinator.com