Saturday, January 23, 2016

Do Recent Events Support Drug Prohibition?

Do those recent events refute drug policy libertarianism?

The number of drug-related poisonings is up dramatically over the past 15 years, and much of the increase appears to be attributable to prescription opioid painkillers and benzodiazepines. Very recently, heroin related poisonings have shot up. Very, *very* recently, fentanyl poisonings have shot up dramatically. (Fentanyl is reflected in multiple cause of death code T40.4, “Other Synthetic Narcotics”, a category that actually includes a few other drugs. There’s a steady increase in T40.4 deaths over the past 15 years, then a sudden surge in the 2014.)

Supposedly these recent trends prove that we need to be tough on drug use. If this viewpoint is correct, those recent trends refute drug policy libertarians who believe partial or full legalization can occur with little or no negative consequences. There are stronger and weaker versions of this viewpoint, so I’ll deal with the stronger and better articulated version. It goes something like the following: “The recent increase in drug poisonings is a result of relaxing constraints on prescription opioids, along with other potentially dangerous medicines such as benzodiazepines. This proves that there is a very large demand response to relaxing prohibition, contrary to what drug-libertarians believe. Prescription painkiller users can’t handle their shit, so they pop so many pills that they overdose. We would have saved ourselves these overdoses if we’d stuck to our guns on strict prohibition. The taste for opioids whetted the appetite for something stronger, and in very recent years we have seen a heroin epidemic. This refutes the drug-libertarians’ belief that people will always stick to the low-concentration versions of a drug when given the choice. Weak drugs are indeed a ‘gateway’ to stronger drugs.” In this narrative, a supposedly lax policy toward painkillers led to a huge increase in deaths, and an unforeseen side-effect of people moving to even more dangerous drugs. There are many assumptions and assertions in here to unpack. Overall I don’t find it a convincing story at all. It’s not a good description of what’s happened in the recent past.

Have constraints on prescribing painkillers slackened recently?

The above narrative would be credible if there had been a recent relaxation in controls on painkillers, but that’s not the case. If anything the opposite is true. There were some very high-profile prosecutions of pain doctors in the early 2000s, which is just around the time the increase in overdose deaths started happening. The DEA really started going after doctors who (supposedly) prescribed too many painkillers. These events even got the attention of the National Association of Attorneys General, who sent a letter to the DEA in 2005 telling them to cool their jets. Still, the DEA’s actions had a cooling effect. The early 2000s, when prescription opioid overdoses started to rise, was an era where doctors were increasingly paranoid about prescribing painkillers. Whatever caused the rise in painkiller use, it certainly wasn’t federal drug policy. This is consistent with a major tenet of drug libertarianism: usage rates increase and decrease over time of their own accord; government policy has little effect on these trends. 

One can speculate about why prescriptions rose despite an apparent crackdown. Did pain specialists retire or stop practicing, and naive primary care practitioners pick up the patient load? Did primary care practitioners prescribe more recklessly? Were they more easily manipulated by drug “seekers”? If so, than the recent rise in painkiller prescriptions is, if anything, a negative consequence of strict drug prohibition. In this case recent events refute strict prohibition, not drug-libertarianism. Or maybe the DEA got slapped back so hard by the Attorneys General and doctors’ lobbies that they backed off entirely, in which case there really *was* a slackening of government policy. But then why were painkiller overdoses rising up until 2005, when the Attorneys General sent their letter to the DEA? I’m really not sure what happened, but it definitely looks like there was an official crackdown at exactly the time that painkiller use started to explode.

Does the availability of low-dose drugs satisfy or merely whet the appetite?

The other piece of the above narrative, that low-dosage painkillers whetted the appetite for harder stuff like heroin, is highly speculative. A major tenet of drug policy libertarianism is that users will opt for the softer, low-dose version of a drug whenever possible. We have decades’ worth of observations on this point, and it appears to be *mostly* true. Bolivians chew the coca leaf, while drug smugglers prefer the more compact powdered cocaine. People preferred beer and wine before and after alcohol prohibition in the 1920s, but they preferred hard liquor during prohibition because its compactness made it easier to distribute. When given the chance they stick with the leaf and the beer. They don’t inexorably escalate to crack and bourbon. Prohibition drives people do the hard stuff; liberalization leads people to the softer stuff. Recent events may actually bolster this point. If the recent leveling-off of prescription painkillers is due to a law enforcement crackdown (I confess I don’t know the real cause), that would cause some of the former abusers to substitute toward heroin. Some of the recent overdoses are due to fentanyl poisoning. Fentanyl is so powerful in such low doses that people have OD’d from handling the drug in the production process, and illegal suppliers have been spiking their heroin with it. These overdoses are almost entirely avoidable. Quite a lot of harm reduction, precluded by a prohibition regime, would be possible in a legalization regime. In a legal market, doses can be specified, advertised, regulated, standardized, titrated to the user’s tolerance, tested for impurities (like freaking fentanyl), etc. That’s not to say there would never be any overdoses, but there would almost surely be fewer.

Another point to consider is that our long-standing policy of drug prohibition has pushed people toward drugs with higher toxicity, away from drugs that aren’t very dangerous. No marijuana overdose has ever killed anybody, a fact that is widely known by potheads and admitted even by many staunch drug warriors. LSD, while a potent psychedelic at low doses, has no known lethal dose. Other psychedelics are similarly non-toxic. MDMA, Ecstasy, is also very non-toxic compared to other drugs. (No, it doesn’t cause you to burn up from a fever or turn your brain into Swiss cheese.) Many of these are potentially therapeutic. So while a very limited view of the past 15 years might superficially indicate a failure of drug libertarianism, under a longer-term view drug prohibition has pushed people toward more dangerous drugs. I’ll admit that this is *a bit* speculative; we don’t know if recreational opioid users and recreational pot smokers (or LSD or MDMA or psylocybin users)   are the same people. But there does seem to be a great deal of substitution; when one drug becomes unavailable people switch to something else. Insofar as we’ve missed the opportunity to direct people to safer, non-toxic recreational drugs, we’ve missed a chance to save lives.

Is the recent surge in overdose deaths attributable to drug addicts and recreational users?

An implicit assumption of the “recent events refute drug libertarianism” narrative is that most of the recent increase in painkiller use is actually misuse by addicts or recreational users. That assumption appears to be contradicted by the SAMSHA and Monitoring the Future surveys, which do not show any increase in the illicit use of pharmaceuticals . As I’ve written in previous posts, many of these deaths are accidental drug interactions by unwary patients (perhaps encouraged by their unwary doctors) who didn’t realize what they were doing. Surely recreational use accounts for some of the drug poisoning deaths, but it can’t really account for the *increase* in deaths. (Unless the surveys are completely missing an increase in illicit use, or the increase in opioid use represents the same people taking greater volumes of the drugs.) If the additional deaths are accidental interactions by unwary people with legal prescriptions, then the past 15 years don’t offer an example of relaxing drug prohibition. It’s not as though we’ve instituted a government policy of giving addicts and recreational users free access to their favorite drugs.


Superficially, the past 15 years look bad for drug policy libertarians. I must admit a 3+ fold increase in overdose deaths surprises me. Most people, drug libertarians and drug warriors alike, probably didn’t realize the problem could get so bad. But only a shallow reading of recent history impugns drug libertarianism. A deeper look shows that this recent trend occurred under a regime of strict (in fact, tightening) drug prohibition. Most of the increase is probably not attributable to recreational or compulsive use of these drugs. The heroin overdoses were to a very large degree avoidable, as countries that have taken a “harm reduction” approach to drug policy have seen both overdoses and addict populations plummet. I’d like to see our country try a combination of legalization and harm reduction. That would be a wonderful experiment, and we’d learn a lot from it. To determine whether libertarian drug policy works or not, we’d have to actually *try* it first. 

Thursday, January 21, 2016

Is it Bad if Prescription Painkiller Deaths Increase?

It’s been widely reported that the number of deaths from prescription opioids has increased in recent years, and I’m willing to believe that the increase is real. My question of the day is: Is this really a bad thing?

While the total number of deaths has gone up, the rate of deaths as expressed “per legal prescription” or “per gram prescribed” has remained steady.  There were ~9,500 kg sold in 1999 and 4,707 “Other Opioid” deaths; there were ~24,300 kg sold in 2012 and 11,676 “other opioid” deaths. In other words, there were 0.50 deaths per kilogram in 1999 and 0.48 in 2012. Admittedly this quantity is slightly volatile and fluctuates between 0.4 and 0.5 over the period, but it doesn’t appear as though deaths per legally prescribed gram have increased significantly. (My figures for deaths are from the CDC’s Wonder database, and my figures for kg of opioids sold are awkwardly read off a graph that a former CDC employee kindly sent to me by e-mail. The original source for that information is the National Vital Statistics System, the DEA’s system for tracking potentially harmful medications. See my note at the bottom; be cautious about anyone’s tabulation of “prescription painkiller deaths.” They are always subject to many caveats. I am completely unsatisfied with how they are tabulated in this post or by the CDC.)

Any medicine entails some degree of risk. The risk of dying from any kind of medicine may be very small, and we may mitigate it as much as we feasibly can, but the risk never gets to zero. In this case, more people are making use painkillers, but the risk hasn’t changed. If the risk was acceptable in 1999, why isn’t it acceptable in 2014? Why the scare-mongering? A patient being treated for pain is at no greater risk today than 15 years ago. I’m picturing someone who has just made the sober decision to take a potentially dangerous medicine. He wakes up one day and realizes that there are four times as many people like him, taking similar medicine. Does he say, “Aaaaah! There’s more of me!”? Or does he say, “Great! The trade-off is worth it, and more people are getting the treatment they needed.” Scare-mongering about an increase in overdose deaths without looking at the risk *per user* is much like scare-mongering about an increase in murders without dividing by population.

Of course some people claim that most (or *all*) of the increase in painkiller use is attributable to irresponsible doctors handing the stuff out like candy. Or perhaps it’s due to an increase in “seeking” behavior by painkiller addicts (though this is a problematic story because the increase in illicit use doesn’t show up in SAMSHA or the Monitoring the Future surveys). On the other hand, some people insist that pain was under-treated in the past and we’ve just now caught up with the problem. These points are debatable, and indeed they are much debated. But the debate is silly without this piece of information: “For whatever reason you use prescription painkillers, the risk to a random user is no higher than it was 15 years ago.” Whatever threshold we have for “unacceptable risk” should apply in all time periods. Maybe a death rate of, say, 1 per every 2,000 prescriptions written is acceptable for intense pain, but not for mild pain or recreation use. Or maybe that level of risk is acceptable for all pain levels *and* recreational use. Maybe grown-ass adults should be able to decide their own thresholds for themselves. I’m not trying to settle this debate or answer the question I posed in the title of this post. I certainly have an opinion, but the point of my post is not to convince you to share it. I’m merely trying to clarify what the trade-off is in this debate.

All that said, I believe there is low-hanging fruit here to pluck. A constant death rate really *is* unacceptable if there are simple, unexplored solutions that would reduce it. Most of these deaths, as I have said before, are drug interactions, *not* single-drug overdoses. We’d go a long way toward bringing down the death rates with better education, perhaps starker warnings, for pain patients taking multiple drugs. Some genuine harm reduction is possible here. But I suspect it’s counterproductive to wag our collective fingers at painkiller users while saying, “Don’t do this!” or at doctors while saying, “Don’t let him use this!” or at policymakers while saying, “Don’t let him prescribe it *and* don’t let him do it!” Prohibition is a blunt and ineffective instrument for addressing this problem. 


Note: I’m just using the “other opioids” category to represent prescription painkiller deaths, which isn’t quite right, although it is the largest single drug poisoning category that the CDC has a code for. Some tabulations include methadone and “other synthetic narcotics,” which may or may not be valid. At any rate, any tabulation will show that “prescription painkiller deaths” are up by roughly the same factor as “Opioid sales” or “number of prescriptions.” So whether it’s really 0.5 deaths per kilogram or actually 0.75 or 2.0 or 10 deaths per kilogram, I don’t really care. My point is that it hasn’t changed much. (Try doing your own by reading the figures off of the chart at the bottom  of this page.)

Monday, January 18, 2016

Fentanyl and Heroin Overdose Deaths

There is a recent disturbing trend of rapidly increasing heroin overdose deaths. Many of these are actually cases of heroin spiked with fentanyl. Jacob Sullum discusses it here. The trend can also be seen in the CDC’s own numbers on drug overdose deaths. (plot below). You can easily see the rapid rise in heroin-related deaths. The CDC’s multiple cause of death database has a category called “other synthetic narcotics”, which (according to the CDC) is mostly fentanyl (multiple cause of death codeT40.4). See the thick light-blue line on the plot below. It looks like there’s been a steady increase over the past 15 years, then a sudden leap in the past year. We’ll see what 2015 brings when those data are released. Fentanyl is an extremely potent synthetic opioid that is 2 orders of magnitude more potent than morphine or heroin. Given the potency, it’s incredibly easy to overdose on it. Illegal suppliers generally aren’t *trying* to kill their paying customers. But the very low cost and the extremely high potency are a deadly combination. Mistakes are inevitable in a black market, where the natural regulatory forces of the market operate poorly and the explicit regulation by government does not operate at all.



A slight caution: Don’t make the mistake of *adding* the heroin and “other synthetic opioid” deaths together. A death with more than one cause listed on the death certificate will show up multiple times in the CDC’s multiple cause of death data. There is some overlap between these deaths, which is kind of my point. I suspect that many deaths from recent years labeled as “heroin” overdoses were probably fentanyl overdoses. It’s just that some localities haven’t caught on to the trend and so the toxicology reports aren’t showing it.

I wrote previously about the recent increase in prescription drug poisoning deaths (here and here). I think that some proportion of those deaths are unavoidable. For any useful medication, *some* amount of risk has to be deemed acceptable. We can’t insist on a zero percent death rate, not if we’re being reasonable. Whatever risk level (expressed as, say, a mortality per 100k users) we deem acceptable, we shouldn’t freak out when more people take that risk and the total number of deaths rises. While the number of deaths and the total kilograms of opioids sold have gone up in recent years, the number of deaths per kg of opioids sold is roughly constant. The risk as expressed per legal dose (or per legal prescription) has remained steady. In other words, if the risk was acceptable in 1999, it should be acceptable in 2014. (About 0.0005 deaths per gram sold, if the “Total Kg Sales” figures I have are accurate, in every year from 2002 and 2014.)

Heroin, on the other hand, has become more deadly. There were 1.3 deaths per 100 users in 2002 and 2.5 deaths per user in 2014, though admittedly both the numerator and the denominator used to calculate this figure are volatile. (Estimates of number of heroin users taken from SAMSHA surveys, estimates of heroin deaths from CDC multiple cause of death data, all publicly available.) *Most* of these deaths are entirely avoidable. Most overdose deaths happen because people don’t know what they’re getting. Some heroin overdoses would surely happen even in a regime of legalization, but most overdoses are a consequence of prohibition for a host of reasons. One reason is that the drugs can be spiked with toxic chemicals that the user doesn’t know about. Another reason is that people don’t know what dosage they’re getting. If a network of heroin dealers is busted, someone else quickly (*very* quickly) moves in to take over. The new suppliers may be dealing something of higher purity than their predecessors, and unwary users may accidentally overdose. A third reason, much overlooked, is that people who are arrested for heroin use may temporarily get clean while in jail. Upon release, they have lost their tolerance to the drug, but they may return to using at levels they were used to. (This post by James Schneider suggests that the third reason is the most important, the second second, and the first least important, although that might have changed in very recent years.)

The solution is simple. Heroin users need only to ask the pharmacist about the concentration and the recommended dosage. They should ask about potential interactions with their current medications or other recreational drugs they are currently taking. They should check the box to ensure it came from a reputable distributor, that there are no toxic additives (such as fentanyl), and should obtain a clean needle for…oh, shit. I forgot. None of that is possible under drug prohibition.

This is one we can lay squarely at the feet of the drug warriors. In principle, it’s possible to reduce the harm from even the most reviled monster drugs, but that would require the ability to sell them at known doses, without adulterants, through knowledgeable, responsible dealers who warn their users of the *real* hazards. Drug prohibition has made this scenario impossible. Let’s stop pretending there’s a law enforcement solution to this problem. A legal “harm reduction” regime, like what’s in place in Switzerland, will get us most of the way there.

Saturday, January 16, 2016

Prescription Opioids and Benzodiazepines, and a Failed Opportunity to Meaningfully Inform the Public

As I said in a previous post, I am extremely skeptical of the CDC’s reports on rising prescription painkiller overdose deaths. More pointedly, I think their scare-mongering is irresponsible and their proposed policy responses are mistaken. My best guess is that the recent increase in “drug poisoning” deaths is partially real, but most of the increase is attributable to drug *interactions*, not overdoses per se. I think they have missed an opportunity to meaningfully inform the public about a real hazard. They could be saying “Don’t take opioids with benzodiazepines or alcohol. It will kill you!” Instead they are saying things like, “Let’s crack down on doctors who prescribe too much and patients who use too much. Let’s restrict the supply.” A patient taking prescription painkillers could read one of these CDC reports without actually being informed of the risks and how to avoid them. (See here and here . In the both links, the CDC actually manages to say that drug interactions between opioid painkillers and benzodiazepines are common, but it fails to emphasize the importance of this phenomenon.)

Here is something interesting I’ve found in the CDC's mortality data:



Data for the above chart comes from the CDC's Multiple Cause of Death mortality database. The "Other Opioids" category includes most of the prescription painkiller deaths. Look at how closely the Benzodiazepine trend line follows the Other Opiods trend line (the thick lines, in blue and green). Both show a rise over the period from 1999 to 2011. They even show the same dip in 2012 and 2013, and a fairly significant bump in 2014. It’s unlikely to be a coincidence. The linear correlation is 0.995. If you look at the raw death counts, they match up reasonably well. Between 1999 and 2000 there were 184 additional other opioid deaths and 183 additional Benzodiazepines deaths. In 2002 to 2003 the numbers are 455 and 249. In 2004 to 2005 the numbers are 544 to 455. In 2011 to 2012 the numbers are both negative; they drop by 575 for other opioids and 321 for benzodiazepines. It’s not exactly a 1-to-1 matching, but the correlation is very strong. The linear correlation between these numbers (change in other opioid deaths and change in benzodiazepines deaths) is 0.83. Not perfect, but strong.

One can make a similar observation state-by-state. I pulled MCD data from the CDC’s Wonder database. For every state with any volume of deaths, the correlation between benzodiazepine deaths and Other Opioid deaths is in the high 80s or 90s. So every state with a significant amount of data is seeing the countrywide pattern. See how the scatter plots show a linear pattern for each of several states. 




Combine this with the fact that just about *every single* category of drug is on the rise. (See chart above and in previous post.) Most of these aren’t overdoses. Most of these categories of drugs are hard to actually “overdose” on. Painkillers aren’t enslaving their users and forcing them to compulsively swallow pill after pill until they drop dead, nor is this happening with other pharmaceuticals or street drugs. Rather, uninformed users are taking bad combinations of drugs, the interaction of which is fatal. So the increase in prescription opioid deaths is showing up in other categories of drugs.



The impression you might get from reading some of the CDC’s public service warnings (like this one) is that irresponsible and recreational use of prescription drugs is on the rise. This is a problematic story, in that it appears to be contradicted by government surveys on drug use. According to those surveys (SAMSHA  and Monitoring the Future), rates of *illicit* use of pharmaceuticals are flat, or even declining. One could surmise that there’s a bias in the surveys, whereby someone who misuses a legal drug doesn’t report the use as illicit. Perhaps the number of abusers is growing, but they aren’t reporting their use as “illicit use” because they have a legal prescription. But the survey wording is pretty clear; it asks the respondent to count as “illicit” any use that’s recreational or other than recommended by the doctor. See page 30 and page 32 of the Monitoring the Future report, where a change in wording of the question caused a sudden jump in self-reporting; clearly the respondents are *somewhat* following the survey instructions or this wouldn’t have happened. Supposing there *is* a reporting bias, it’s very hard to claim that such a bias explains the survey results. If respondents are less likely to report illicit use because they are getting more and more legitimate prescriptions, there should be a *downward* trend in the use rates from ~2000 to 2011, when legal prescriptions for opioids were steadily increasing. And responses admitting to illicit use should have increased after 2011, when legal prescriptions leveled off (and recreational users presumably resorted more often to illegal supplies of these drugs). In fact, illicit use of pharmaceuticals decreased after 2011, in both surveys and for all age groups. One would also have to argue that the same reporting bias exists for two different surveys, and for every age category of respondents. If you *do* think that a trend-inducing reporting bias is possible, consider that it might be a factor in the “drug poisoning” figures put out by the CDC, and that it likely overstates an existing trend, or creates one out of nothing. Several CDC reports assume that “diversion” is a huge and growing problem, but apparently that’s not the case. We should see that in the “illicit use” numbers.  We simply do not. From years 2002 to 2014, the correlation between death rates and illicit use rates is actually slightly negative, although the correlation between deaths and opioid sales (in kg per 100k) is in the high 90s. More people have been using legal prescription opioids, probably mostly as directed. If illicit use of prescription painkillers is a problem, it’s no more so than it was 11 years ago, and it’s not the driver of the recent surge in poisoning deaths.

There are more legitimate prescriptions than there were 15 years ago, but properly administered narcotics almost never turnspatients into addicts 
“Although there were 11,882 patients who received at least one narcotic preparation, there were only four cases of reasonably well documented addiction in patients who had no history of addiction. The addiction was considered major in only one instance. The drugs implicated were meperidine in two patients, Percodan in one, and hydromorphone in one. We conclude that despite widespread use of narcotic drugs in hospitals, the development of addiction is rare in medical patients with no history of addiction.” This is consistent with the survey results. We are prescribing more opioid painkillers, but we aren’t getting hordes of zombified addicts as a result."

If the recent surge in drug-related poisonings is real, the most likely explanation is that these are drug interactions. Even the CDC appears to know that this is an important drug interaction. From the link: 
“Benzodiazepines were involved in 31% of the opioid-analgesic poisoning deaths in 2011, up from 13% of the opioid-analgesic poisoning deaths in 1999.” 

It’s quite plausible this is an understatement, as thorough toxicology screenings aren’t always done in all localities. This report, also from the CDC, indicates that a high percentage of both opioid and benzodiazepine poisoning deaths involve alcohol as well. 

“The analyses showed alcohol was involved in 18.5% of OPR and 27.2% of benzodiazepine drug abuse-related ED visits and 22.1% of OPR and 21.4% of benzodiazepine drug-related deaths.” 

In previous hysterical scare-stories related to OxyContin, it was revealed that 1) the number of poisoning deaths was vastly overstated and 2) actual poisoning deaths were most often from combinations of drugs, not from a single drug. See here
“…overdose victims tend to have multiple drugs in their bodies. Approximately 40 percent of the autopsy reports of OxyContin related deaths showed the presence of Valiumlike drugs. Another 40 percent contained a second opiate such as Vicodan, Lortab, or Lorcet, in addition to oxycodone. Thirty percent showed an antidepressant such as Prozac, 15 percent showed cocaine, and 14 percent indicated the presence of over-the-counter antihistamines or cold medications. Deaths like those could be the result of any of the drugs present, drugs working in combination, or one or more drugs plus the effects of other conditions, such as illness or disease. Indeed, the March 2003 issue of the Journal of Analytical Toxicology found that of the 919 deaths related to oxycodone in 23 states over a three-year period, only 12 showed confirmed evidence of the presence of oxycodone alone in the system of the deceased. About 70 percent of the deaths were due to “multiple drug poisoning” of other oxycodone-containing drugs in combination with Valium-type tranquilizers, alcohol, cocaine, marijuana, and/or other narcotics and anti-depressants.” 

Benzodiazepines by themselves are also very unlikely to cause an overdose

“Taken alone, they rarely cause severe complications in overdose; statistics in England showed that benzodiazepines were responsible for 3.8% of all deaths by poisoning from a single drug.”

And here
“Taking benzodiazepines with alcohol, opioids and other central nervous system depressants potentiates their action. This often results in increased sedation, impaired motor coordination, suppressed breathing, and other adverse effects that have potential to be lethal.”

So it’s probably hard to literally “overdose,” as in “take such a large dose as to be fatal”, on any single one of these drugs, but it’s easy to accidentally take a fatal combination of drugs.


So what’s the point?, you might ask. Am I just hair-splitting here? Am I pedantically correcting the imprecise language of a government agency, who insists on saying “overdose” when it should be saying “multi-drug interaction”? Isn’t 15,000 deaths per year a big problem, by whatever name you call it?  No, I think this is an important point. If you don’t properly specify what the problem is, people will miss the point. Governments will enact the wrong policy responses. A patient who is taking painkillers should be meaningfully informed by the CDC's news briefs and public service announcements. Someone taking a therapeutic dose of prescription opioids should not be worried that s/he will turn into an addict, or drop dead of an overdose. Both outcomes are exceedingly rare for a properly administered round of painkillers. But such patients *should* be concerned about potential drug interactions. They should be alerted to talk with their doctors about their current medications and drinking habits. A properly educated patient should be able to read these CDC news briefs and say, “Ah, I can do something *myself* to make sure *I* don’t fall victim to this growing problem.” Take another look at this somewhat hysterical brief; the section titled “Solutions” is entirely about what *other* people can do to solve the problem. If I’m a random guy reading this brief, there’s very little I can actually do to effect government policy. I might be convinced (wrongly) that opioids are inherently dangerous and that the recent rise in prescription rates is inherently problematic. I might also be convinced (again, wrongly) that the solution is to stop *other people* from prescribing and using painkillers so often. The link vaguely hints at talking to your doctor, but that’s not at all the thrust of the message. I don’t think I’m just picking on one bad PSA, either. The link is typical of what comes up in my Google searches on this topic. I consider it irresponsible fear-mongering. It misses the point, badly. It fails to meaningfully inform the reader of the problem. Indeed it *misinforms* the reader. 

Monday, January 11, 2016

Mortality and Age

I spent some time recently picking through CDC’s mortality data. My curiosity was piqued by recent reports of exploding mortality rates due drug poisonings. Luckily you can download the data from CDC’s Wonder database and cut it any way you like. Here’s what I found. Rates are given as deaths per 100k population. The CDC doesn’t have accurate population figures above age 85, so I can’t get those mortality figures from the CDC. They are excluded from my table. The CDC’s data query automatically calculates the mortality rate for you. I’ve added three columns: the change in mortality at each age ( (final – initial )/ initial), and the effect of one year of aging in both 2000 and 2014 (calculated as “this age’s mortality – prior age’s mortality”). So what does it tell us?
(Table at bottom of post.)

Despite recent alarmism, I see mostly good news. The vast majority of age groups (which you can get by each single year!) see a decline in mortality over the past decade and a half. For eleven age groups, people age 25 – 35, mortality has actually increased. That’s not to say things look worse if you’re a 30-year-old. Your life expectancy at age 30 isn’t just based on your morality at age 30; it’s also based on mortality at all years *after* 30. A randomly picked 30-year-old in 2000 could expect another 48.3 years of life (see here, page 25); a 30-year-old in 2013 could expect another 50.1 years of life ( see here, page 12). So this age category is not worse off than it was just over a decade ago. The mortality at any given age is a small piece of the pie, since someone is pretty unlikely to die in the 11 years that span 25 to 35. The improvement in mortality at older ages makes up for the loss in that decade.

What really struck me was how mortality changes as you age. When people talk about drug policy or gun policy, they’re talking about a few points per 100k. If you believe the CDC’s figures, there are about 15 drug poisonings per 100k. Gun homicides are significantly smaller.  These magnitudes, while disturbingly high, pale in comparison to the effect of aging even a few years. Aging from 30 to 35 increases your mortality by 31.9 points (deaths per 100k per year). Aging from 40 to 41 increases your mortality by 21.1 points.  It accelerates quickly from there. Aging from 50 to 51 increases your mortality by 48.1 points (possibly that’s a mild outlier). Aging from 60 to 61 drives up mortality by 75.5 points.

Look at the improvements for ages 0 – 20. These are massive improvements in our earliest years. The improved mortality at age 0, an age at which mortality is famously quite high, is especially good news. People in their mid-60s to mid-80s (where the CDC data stops) are also seeing some impressive gains. All-cause mortality is dropping 20% or more for these ages. A recent paper by Angus Deaton, 2015’s winner of the Economics Nobel, shows how a few age and race demographics are seeing some losses (or at least failing to realize the gains seen in other demographics). But the overall story is a hopeful one. Relevant to my point, a critique  I found of the Deaton paper points out that the 44 – 54 age cohort has gotten older. The mortality effect of sheer aging could explain some (not all) of the backsliding in this demographic (specifically the non-Hispanic white 44-54 demographic).

If we fixate on middle years, age 30 – 60, the average person was 43.6 in 2000 and 45.1 in 2014. Aging that year-and-a-half adds 34 points to your mortality (using 2014 mortality figures; obviously I have to extrapolate a little for fractional years). For comparison, the largest single cause of death in this age group in 2014 is ICD code C34.9, “Bronchus or lung, unspecified – malignant neoplasms” (lung cancer, I presume), which comes in at just under 20 deaths per 100k. A year and a half of aging has a larger magnitude than any single cause of death. I don’t want to overstate my point; admittedly the codes are pretty narrow, and the mortality for “all cancer” or “all heart conditions” would probably be larger. And you may be saying, “No crap. Older people die more.” Still, the magnitudes surprised me.

I don’t want to diminish the importance of social problems like gun violence and drug overdoses. The rates are small, but the total numbers of victims are in the tens of thousands, and each one is a tragedy. I’m not advocating that we compare every problem to “one year of aging” and dismiss it if it’s a lot smaller. Any particular policy approach to any particular problem needs its own cost-benefit analysis. If there’s a problem that kills only 10 people a year and it can be solved for, say, $100k, we shouldn’t flippantly ignore the problem because of its smallness. We should spend the money! And if a problem kills 100,000 people a year but doesn’t respond *at all* to a multi-billion dollar “Just Do Something” government policy, we should stop wasting our money. The overall magnitude of the problem is irrelevant to the question of how to address it (or who should address it, or whether or not to address it). If we simply ignored all problems smaller than “the effect of a year of aging on mortality” we’d leave a lot of important, and more to the point solvable, problems unsolved. Individual problems need to be understood and addressed specifically and one-at-a-time, but we should certainly quantify them and put them into perspective.  




Sunday, January 10, 2016

Why I’m Skeptical of the CDC’s Drug Overdose Figures

There has been much reporting of very large and very recent increases in the number of drug overdoses. If you believe the CDC’s raw figures, the total drug overdose deaths have increased from about 6 per 100,000 in 1999 to about 15 per 100,000 in 2014. Overdoses involving opioids (painkillers, heroin, methadone, etc.) have increased from about 3 to 9 per 100,000 in the same time period. A ~3 fold increase in overdose deaths is indeed alarming. But is it real? Does it make sense? Does the world really change that much in 15 years? What are the implications for drug policy? How is this possible when illicit drug use is basically flat over that time period? These are the questions that drove me do some pretty extensive research on this topic. I found that, unfortunately, nobody has written the definitive book, or even the longish article, on the question I wanted to answer. So I’ve done some digging and here’s what I’ve found.

Inaccurate Determination of Cause of Death

Steven B Karch. Learn his name. Read his textbook, Pathology of Drug Abuse. If you want to have an informed opinion on drug policy, you should understand how drug overdoses happen, which drugs are the biggest “offenders”, how a determination of cause of death is made, how often a mixture of drugs (rather than an “overdose” of a single drug) is the culprit, and a host of other issues discussed in his textbook. I first saw his name in this article by Radley Balko, which is also skeptical of the CDC’s alarmism on recent overdose deaths.
“Dr. Steven Karch, who has written a widely used textbook on drug abuse and pathology, says because tolerance for opioids can vary so much from person to person, there's no scientific way to definitively say that a death was caused by an opioid overdose. "There are plenty of people walking around with levels of opioids in their bodies that would be declared toxic if they were dead on a slab in a medical examiner's office," Karch says. "Toxicology is the least important part of making a diagnosis."

"I don't know where they got their numbers," Karch says of the CDC estimates. "There's no peer review of those figures. You follow the footnotes, and it looks like they're getting the information from medical examiners. But it doesn't say how the medical examiners are concluding that these were overdoses--if, say, they're just relying on toxicology results." Asked if that's usually how overdoses are diagnosed Karch says, "That fits my experience."

So determining the cause of death isn’t exactly a science. A toxicology screening showing very high concentrations of metabolites might not actually tell you anything about the cause of death.

He’s also quoted in this article in Time:
In other words, Karch says, the "lethal dose" of these drugs is impossible to standardize, even for patients who are taking a single drug rather than a combination of many. A 2000 study by Karch compared methadone patients who had died from clear causes, such as car accidents, with those who had died of a suspected overdose. Karch found that there was no predictable difference in the postmortem blood levels of methadone between the two groups. In fact some of the patients who died of other causes had higher levels of methadone in their bodies than those thought to have overdosed. "You can die from a drug and you can die with a drug," says Karch. "When you have four orders of magnitude separating either end of the curve, many of these deaths may not have to do with drugs at all."


Here is a paper written by two other doctors saying essentially the same thing. (The link is to the abstract; the full text is available online but I’m having trouble finding a stable ungated link to it).
Current data collected by medical examiners and coroners are incomplete and inadequate to evaluate the factors that lead to fatalities involving prescription opioids. Determining cause of death is critically important. Two methods are proposed to improve consistency and accuracy in the collection and analysis of decedent data in opioid-related poisoning deaths. First, an improved death certificate is needed to collect evaluative data, including: extent to which opioids were judged to 1) cause, 2) contribute to, or 3) be present in investigated deaths; extent to which opioids as a cause of death were found 1) alone, 2) combined with other prescription drugs, 3) combined with alcohol, or 4) combined with illicit drugs; the time of death; the presence or absence of a valid prescription; and the estimated quantity of opioids taken proximal to death.
Apparently anyone who has specialized knowledge regarding drug overdoses is skeptical of the CDC’s figures. If the overdose deaths are counted inaccurately, we shouldn’t be crafting public policy based on those data.

Most of the reporting I see on drug overdose death rates comes from a CDC database that simply counts up deaths that are given a particular set of ICD-10 cause-of-death codes (mostly codes X42 and X44, although several other codes are considered “drug poisonings”, depending on the particular tabulation). If overdose deaths are being routinely misclassified, then these numbers are potentially misleading. Consider the following disturbing possibilities:

1)   Death rates attributed to drugs are overstated (or understated) in every year due to an overall bias.
2) Trends in drug-related death rates are skewed due to changes over time in assigning cause of death, as more (or fewer) deaths are attributed to drug overdoses.

If only 1) is true, then the disturbing recent trend in drug overdose deaths could still be real, but the actual death rate in every year could be overstated (or understated for that matter). If 2) is true, then the overdose death rate in one of the years between 1999 and 2014 may well be accurate, but the *trend* could be spurious. It may be an artifact of, say, more frequent toxicology screenings, more thorough autopsies, or a greater propensity to attribute a death to an overdose. If someone drops dead of an unknown heart condition, perhaps we’re more likely to investigate that person’s drug history than we were 10 or 20 years ago. So perhaps deaths are simply being shifted from a heart- (or stroke- or allergy- or what-have-you) cause of death to a drug-poisoning related cause of death. Death rates related to heart conditions have declined in recent years, and this is a *big* category responsible for a lot of deaths. So it's plausible that a small redistribution from this category could cause a large change in another category. All of this is of course consistent with the death rates or the trend rates of those death rates being accurate, but we should definitely view them with more skepticism.

CDC’s Public Database on Causes of Death

With all those caveats, let’s actually take a peek at the CDC’s own cause of death data. The CDC maintains the appropriately named Wonder database which contains every death that generated a death certificate, from 1999 – 2014. It will surely be updated as new years of data roll in. One caution: if you’re using the Multiple Cause of Death database, it will double-count some deaths, since multiple causes can contribute to any single death. I grabbed the data for 2000 and 2014 for every single cause of death with its own ICD-10 code, and the sum was 6.5 Million in 2000 even though only about 2.5 million people died that year. Even so, it’s useful because it assigns particular drugs as causes of death. If you want to know how often heroin, cocaine, or benzodiazepines (3 of the top four killers) are implicated in deaths, this database will tell you that. So here’s what I found.


This table is the death counts and rates in 2000 and 2014 by drug type, sorted from highest to lowest 2014 mortality. (Apologies for the shitty formatting.) It looks suspicious as hell to me, because almost every single category sees a huge increase in mortality. Even the categories that are extremely small. So are we really seeing increases in every goddamn category of drug, or are toxicology screenings and patient histories just more thorough than they were a few years ago? Cocaine use has been in decline for over a decade, and yet death rates are up 26% since 2000? (See use rates for major drug categories here.) Methadone death rates are up almost 200%? I thought people were shifting toward pills. Note the first category in the table to show a *decrease* is “other and Unspecified Narcotics.” The decline is small, as are the total number of deaths in this category. I'm not suggesting that the increase in other drug mortalities is completely explained by a redistribution from the "Unspecified" category. But the decline is consistent with the notion that we know more about what drugs a decedent was taking. Look at some of the smaller categories. The dreaded Antitussives are up almost  a factor of 8. “Cannabis (derivatives)” up by a factor of 6, admittedly from a very small beginning. It is widely known that it’s impossible to overdose on smoked marijuana, so I have no idea what’s going on here. Barbiturates are up slightly; that’s weird because my understanding is that these are used less frequently than in the past (citation needed). This bolsters my suspicion that at least *some* of the recent trend is driven by simply having more information on the deceased, such as toxicology screenings and drug use history. In one narrative, “People are using more drugs, and at a constant death-rate per user, we should expect more deaths.” But would anyone toting this narrative predict that death rates would be up for almost every single drug category? Shouldn’t drug users be shifting *away* from other drugs as they have gained easy access to prescription painkillers? Of course, drug interaction is a common cause of “overdose.” People who die with prescription painkillers in their system are often killed by an interaction between painkillers and, say, alcohol, benzodiazepines, psychiatric meds, etc. So it’s possible that more people are dying overall as a result of prescription painkillers (the CDC’s narrative) and we’re just marking down whatever else is in their system. But I would have expected some dramatic declines in some of these categories if the CDC’s narrative is correct.
By the way, notice how flat the illicit use of psychotherapeutics is in this document, page 17 and 18. There are more legal prescriptions and more people taking meds, but illicit use (as defined in the document) is pretty flat (or possibly *down* even) over the past decade. That’s another strike against the “out-of-control pain meds” narrative.

The above table gives two snapshots in time. Let’s take a more thorough look at the years in between:





(Once again, shitty formatting. Sorry.) The “Other Opiods” category started to level off around 2010 at just under 4 deaths per 100k. Heroin has been skyrocketing. It doubled from 2000 to 2010 (about 0.5 deaths per 100k to about 1 per 100k), and then more than tripled (it’s 3.4 per 100k in 2014, supposedly). I can’t find good data on rates of prescription opioid use since 2011, but the sources I’m seeing with some quick Googling suggest that rates have continued their historical increased. This is odd, and inconsistent with the “more painkillers lead inexorably to more overdoses” narrative. Here’s a better narrative. Drug cops have been increasingly harassing and arresting pain patients and pain doctors. (See this excellent CATO report on the topic ). As soon as a patient starts to “look suspicious”, they get cut off by their doctor (who doesn’t want trouble from the cops). They then turn to street heroin to get their fix. Cops have also been arresting and prosecuting pain doctors, which has had a major chilling effect on that specialty. Many of them have been retiring, and young doctors are mostly unwilling to enter the profession. So pain patients instead get treated by primary care practitioners who don’t seem to know what they are doing.  (See this interesting thread in a forum where many of the posters are pain doctors. They are criticizing primary care practitioners for not knowing what they're doing when prescribing painkillers. Incidentally, note "algosdoc"’s comment #64, where he suggests that the cause-of-death coders don’t really know what they’re doing. This is consistent with my suspicion that many “overdose” deaths are misclassified.) In other words, in their zeal to “do something” about the large number of overdoses, drug cops have driven people away from prescription painkillers to street heroin, which is much easier to overdose on. They’ve put out of business the doctors who best understand pain medicine (and its associated dangers), and the slack  is being picked up by doctors who lack that specialized knowledge. Another narrative is that the street price of heroin declined relative to the price of prescription painkillers (citation needed) so people naturally made the substitution.

The recent increase in rates of heroin use appear to be real but smaller than the increase in the overdose rates. Apparently, heroin has become a more deadly habit than it was 10 years ago, in addition to being more common. If you know anything about how heroin overdoses happen, you should realize that most of these deaths are preventable and wouldn’t happen in a regime of legalization. Countries that have taken a “harm minimization” approach have seen overdose rates plummet. 

I’m not certain which narrative or explanation best fits the above observations. But clearly overdoses can move independently of the overall use rates. We should all be skeptical of the “roughly constant mortality per user, so more users means more overdoses” story, because it doesn’t fit.

Miscoding the Cause of Death

I still don’t fully understand the ICD-10 coding process, but it’s fairly obvious that coding errors are frequent. See the discussion on page 75 and the accompanying chart in the 2007 Summary of Vital Statistics for New York City:

The New York City Bureau of Vital Statistics reports drug-related deaths as a combination of the following ICD codes: F11-F16, F18- F19, X40-X42, X44. “F” codes represent drug-related deaths due to chronic drug use and “X” codes represent accidental drug-related deaths. The total number of 2006 drug-related deaths in New York City was virtually the same when comparing manual to automated coding methods (979 vs. 973, respectively.) However, the number of drug-related deaths in each of the subgroups differed drastically: manual coding resulted in a significantly higher number of drugrelated deaths due to chronic drug use or “F” codes compared to automated coding (903 vs. 149 deaths) and a significantly smaller number of accidental drug-related deaths or “X” codes (76 vs. 824 deaths).

In other words, a sudden drastic jump in the number of drug poisoning deaths occurs due to changes in coding methodology. 

It’s plausible that this happens elsewhere. This CDC article suggests that there are only a few cause-of-death coders in the entire world. It seems they frequently assemble for conferences. With a small, tightly-knit group, one can imagine a kind of group-think slipping in. One can imagine a scenario where a widely reported trend feeds on itself. If the cause-of-death coders (“nosologists”) think that overdoses are more common, perhaps they’d be more likely to classify a death certificate they see as an X42 or an X44 (the most common drug poisoning categories). When studies have been done to measure the accuracy of cause-of-death coding, they tend to find sizable errors, though the exact magnitudes vary wildly (citation needed).

What It All Means

I wish I could arrive at a solid conclusion, like “So the trend is a statistical artifact…” or “Therefore the trend is exaggerated by x%...” or “While the trend is real, actual rates in any given year are over/understated by a factor of z…”  I can’t. The information required to “correct” the CDC’s morality data does not exist. One would have to do a thorough examination of every single overdose death over the last decade and a half (and we’d also have to review every death that *might have been* an overdose death that we miscoded as something else). Presumably some kind of sampling could be done to avoid reexamining the entire set, but the task would still be enormous. The people making the cause-of-death determination are often doctors. The people casting doubt on their judgments are also doctors. It looks to me like the highest levels of training and education are thrown at this problem, and it’s still riddled with errors. It’s going to be hard to gather a team of super-duper experts to “correct” their work.

I’m not going to endorse radical skepticism, where “we can’t know anything about anything.” I don’t approve of that kind of nihilism. But given the low credibility of our existing data set, we should fall back on our Bayesian priors. There are some stylized facts that we know about drug use and behavior, which can inform any policy response. Rates of illicit drug use tend to move of their own accord, with little regard for official policy. In a given year, overall rates may be up or down a little, while specific drug categories may be up or down more dramatically. But these movements are almost totally uncorrelated with the government’s official drug policy. Populations go through cycles, where a drug is discovered, becomes popular, becomes stigmatized, becomes banned, becomes forgotten, then gets rediscovered (see Ryan Grimm’s excellent book This is Your Country on Drugs for a thorough description of this process). The news media catches on just after the peak of the cycle, when interest in the drug has already started to wane. Given all this, we should be very hesitant to enact any kind of policy response to the latest hysterical drug scare.