Two opioid papers to pay attention to

Earlier this week, a pair of papers on opioids came out.  One leaned in the direction of my priors (1) and one is making me re-evaluate my priors (2).  

The first study is a randomized trial comparing one year outcomes for people in the VA with significant back, knee or hip pain symptoms.  One arm received standard opioid therapy while the other arm of the trial received non-opioid pain medication.

Question  For patients with moderate to severe chronic back pain or hip or knee osteoarthritis pain despite analgesic use, does opioid medication compared with nonopioid medication result in better pain-related function?

Findings  In this randomized clinical trial that included 240 patients, the use of opioid vs nonopioid medication therapy did not result in significantly better pain-related function over 12 months (3.4 vs 3.3 points on an 11-point scale at 12 months, respectively).

A year out, there was no functional, clinical or statistical difference in the treatment arms.

There is a growing body of evidence that opioids are not particulary better pain killers in some common situations than less dangerous alternatives.  This is an important but not an amazingly surprising result in this paper.  Serious caveats need to be put into place about generalizabilibility but for the relevant population, this is more of a confirmation than a shocking result.

Now the second paper looks at the impact of Naxolone, an overdose counter-acting drug and asks if widespread availability of the drug changes mortality rates?

Naloxone access may unintentionally increase opioid abuse through two channels: (1) saving the lives of active drug users, who survive to continue abusing opioids, and (2) reducing the risk of death per use, thereby making riskier opioid use more appealing…. We exploit the staggered timing of Naloxone access laws to estimate the total effects of these laws. We find that broadening Naloxone access led to more opioid-related emergency room visits and more opioid-related theft, with no reduction in opioid-related mortality….We also find suggestive evidence that broadening Naloxone access increased the use of fentanyl, a particularly potent opioid.

This is a big deal as there has been a massive public policy push Naxolone/Narcan out into the community as far and as quickly as possible.  The theory of change is that fast and broad access to counter-OD drugs will save lives on net by reversing ODs.  And from there, a logic model can be constructed where the OD reversal can be an entry point to treatment or diversion services.

That is not the story that the evidence in this paper is showing.

It is showing that in areas with presumably high density of Narcan carriers, the cost of overdosing borne by an addict has gone down and thus riskier behavior leading to more overdoses increases.  This is very similar to the economic logic of increased risk taking with the proliferation of better anti-HIV medication or more macabre that the way to decrease auto accidents is to put a big knife in the steering column of every car.  The last fifty pages of the paper looks at a wide variety of robustness checks and sensitivity analysis.  The methodologists that I trust are telling me that the methods are solid and the evidence base is strong.

I think the policy implication is that if there is one last dollar available to spend on opioids, it would be better spent on preventing people from being addicted either by limiting initial exposure to opioids upstream or by funding more and better addiction treatments.

Finally, this is a major piece of new information.  It is making me move my priors but it is still conflicted with other recent studies (Rees et al, 2017)(3) and there may be a problem of abstracting the drug market and addiction details too far.  We need good evidence to inform effective policy.  This is strong but not completely conclusive evidence.  We should think hard about the value of these laws versus other policy options even as we seek more evidence as to what works.


(1) Krebs EE, Gravely A, Nugent S, Jensen AC, DeRonne B, Goldsmith ES, Kroenke K, Bair MJ, Noorbaloochi S. Effect of Opioid vs Nonopioid Medications on Pain-Related Function in Patients With Chronic Back Pain or Hip or Knee Osteoarthritis PainThe SPACE Randomized Clinical TrialJAMA. 2018;319(9):872–882. doi:10.1001/jama.2018.0899


(2) Doleac, Jennifer L. and Mukherjee, Anita, The Moral Hazard of Lifesaving Innovations: Naloxone Access, Opioid Abuse, and Crime (March 6, 2018). Available at SSRN:

(3) Rees, Daniel I., Joseph J. Sabia, Laura M Argys, Joshua Latshaw, and Dhaval Dave.
2017. “With a little help from my friends: The effects of Naloxone access and Good Samaritan
laws on opioid-related deaths.” NBER Working Paper No. 23171.

41 replies
  1. 1
    Nicole says:

    Very interesting. I wonder how aggressively Big Pharma will push back against the results of the first one, as I imagine prescriptions of opioids for arthritis is a big source of revenue.

    Hip pain is, I believe, what Prince and Tom Petty had been prescribed opioids for.

  2. 2
    MomSense says:

    So the author(s) thinks that opioid addicts are making rational decisions to engage in riskier behavior because of the increased access to Naxalone/Narcan?

    How much time have they spent with addicts? Let’s just say I am very skeptical.

  3. 3
    dr. bloor says:

    @MomSense: Well, regardless of the precise mechanism, the data pretty plainly show that you hit the point of diminishing returns on Naloxone pretty quickly once it moves beyond the hands of trained professionals.

    David’s good point about prevention and access to good treatment being the best cure has been axiomatic in the mental health community for a long while.

  4. 4
    Nicole says:

    @MomSense: Well, and as is said, it contradicts earlier studies. I imagine additional studies will be done to see if the outcomes are replicated. But it’s worth doing those additional studies. No one is saying use of Naxolone/Narcan should be eliminated, but they are saying it might be, ultimately, more valuable to put more money into treatment and rehab, and programs to help people not get addicted in the first place (especially since evidence is growing that opioids don’t necessarily work better than less addictive methods, anyway).

    ETA: Or what Dr. Bloor said.

  5. 5
    Ruviana says:

    I read about this study over at Drum’s place yesterday and found his observations persuasive. Any comments about his comments?

  6. 6
    Robert Sneddon says:

    @Nicole: Opoids are reasonably cheap, comparatively speaking. Different formulations and tweaking of the basic molecule to provide patentable products are the only financial win for pharmas (“ask your doctor today…”) but a lot of the opoid catalogue is out of patent and available as low-cost generics.

  7. 7
    Victor Matheson says:

    Off topic, but Daniel Rees, the coauthor of the paper that suggests widespread availability of narcan does reduce opioid deaths also has one of the seminal papers on crime and sports. He finds that college football games are associated with large increases in property crime and crimes of violence, especially when upsets occur. Apparently it’s not just the players that get “upset”.

  8. 8
    Victor Matheson says:

    @Ruviana: I read Drum, but I haven’t read the paper. The problem with papers like this is getting the causation and timing right. Places with major opioid problems are more likely to adopt widespread carry of narcan. So, narcan use and opioid problems will be correlated.even if narcan adoption does cause reductions in deaths, there may be some hopeless cases already baked into the cake. Thus, you could have a situation where a place has high opioid use adopts narcan use and opioid deaths rise for a significant amount of time. Of course the authors know this, but it is a tricky econometric problem to solve.

  9. 9
    Glidwrith says:

    A relevant phrase “no reduction in opoid-related mortality” – does this imply that over a longer term, despite the availability of Narcon, that eventually the OD catches up to the addicts and they die?

  10. 10
    Schlemazel says:

    Quite some time ago I read studies done by insurance companies that found pain meds less effective for long-term pain & were sending people to “pain management clinics”. They taught people how to deal with pain and how to understand how much pain they were actually having. Turns out people tend to over estimate their actual level of pain and make it worse by focusing on it. They were, at the time at least, having very good success at weening people off drugs.

    Sorry I have no links or really much memory of the programs. As a chronic pain sufferer myself I learned more nuanced pain levels & how to deal with them. Outside the cancer treatment period I have never taken opioids at home & even OTC drugs only rarely.

  11. 11
    oldster says:

    “…the way to decrease auto accidents is to put a big knife in the steering column of every car.”

    As a car owner, I vehemently reject this proposal.

    As a frequent bike-commuter, I have to confess that there is much to be said in its favor.

    Sure, it makes it dangerous for drivers to slam into things. But no more dangerous than it has always been for bikers, pedestrians, kids in strollers, etc. etc.

  12. 12
    RSA says:


    I read about this study over at Drum’s place yesterday and found his observations persuasive. Any comments about his comments?

    I came across Kevin’s post yesterday too. He seems to be misinterpreting the scatter plots he shows. He writes,

    But I’m skeptical of these findings. To see why, take a look at this scatterplot of opioid deaths in the Midwest:

    But the scatter plot isn’t of opioid deaths; it’s of the residuals of a regression on opioid deaths. (The vertical axis is clearly labeled “Residuals”.) So when Kevin says that he sees no obvious pattern, that’s okay—it’s what you hope for in residuals. But he’s wrong a second time, because he’s ignoring the pattern of residuals mostly below zero before the Naloxone law and mostly above zero after.

    I could be wrong, too, but I think that’s what’s going on.

  13. 13
    Nicole says:

    @Robert Sneddon: That’s a good point- I’ve only been prescribed opioids a few times in my life, and only the first time did I think they actually did more for the pain than an NSAID alone, ergo I’m not really familiar with them. So, because I figure most of these things come down to money, the pushback (as I assume there will be one) is then Big Insurance preferring to pay for low-cost generics than for other, likely more expensive, forms of pain therapy.

  14. 14
    The Moar You Know says:

    So the author(s) thinks that opioid addicts are making rational decisions to engage in riskier behavior because of the increased access to Naxalone/Narcan?

    How much time have they spent with addicts? Let’s just say I am very skeptical.

    @MomSense: I am as well, but…that’s what the data showed. I’d like to see a much larger study.

    I didn’t really understand the scope of the problem until I lost a good friend of mine to this shit back in ’08. She didn’t die. It might have been better if she had. She’ll never function outside of her parent’s home again. When they go I expect she’ll relapse and die. The urge for this shit transcends everything in the addict. So I think a great deal of skepticism is warranted regarding an addict making dosing decisions based on anything…but again, that’s what the data shows. So let’s look harder at that.

  15. 15
    dr. bloor says:

    @Victor Matheson: There’s nothing so nervous as a fraternity sofa or a car parked on the street following a Michigan loss in the Big House.

  16. 16
    peej01 says:

    Confession: I do take opioids for chronic back pain. They do work for me, but not for my knee pain, weirdly enough. I can live with the knee pain.. I can’t take nsaids because I also have chronic kidney disease. I was taking the nsaids for years for the back pain, probably contributing to the kidney problems. So what am I supposed to do in this case? I do know that my insurance company has started to limit people to 1 week’s supply. That seems to be their solution.

  17. 17
    Mnemosyne says:

    I guess my question on the Naxolone thing is, are people who overdosed being given the drug and then sent on their way, or are they being steered into rehab while they’re (presumably) still shaken up by their brush with death? If they’re not at least being offered services, it’s not going to dp much to help people get clean.

  18. 18
    Barbara says:

    @Mnemosyne: Sometimes the best way to understand what it’s like is to read a narrative account that tracks these interventions. The New Yorker had such a piece recently and most addicts were offered further treatment even if initially through local emergency rooms and most declined. The lion’s share of the effort has to be in preventing new addicts.

  19. 19
    Mnemosyne says:


    I’ve seen alcoholics online spend a lot of time trying to figure out the right dose of vitamins that will allow them to continue drinking to excess, so don’t underestimate how much trouble addicts are willing to go to in order to continue using.

    And IMO (not a doctor!) a serious addiction often has a suicidal component to it, so it’s not that surprising to me that an addict who overdoses would continue trying to kill themselves.

  20. 20
    Mnemosyne says:


    I would also be interested to see statistics about drug courts and how much success they’ve had in steering people away from prison and into rehab.

    Part of the problem, of course, is our idiotic healthcare system. You can offer rehab services to an addict, but if a bed isn’t available, or if there’s a fee that the addict can’t afford, it’s a useless offer.

  21. 21
    brettvk says:

    Numbers versus humans. Defunding a Naxalone program makes sense when you’re allocating scarce public funds and you can point to a study like this. But if you’re a family trying to keep your son/daughter alive in the hope that they can someday pull themselves out of addiction, restricting access to a drug that can save them at one particular moment will not get your support.

  22. 22
    Barbara says:

    @Mnemosyne: It’s also important to recognize that overdoses can signal suicidal behavior. Narcan doesn’t reverse that.

  23. 23
    buckguy says:

    I’m reading the Doleac paper and it’s shot through with questionable assumptions and a poor grasp of behavior, policy and regression statistics. There are problems with their geographic clustering and if if you just look at the regression lines and the data points, it’s clear that they may have been selective with their statistical tests. The rational man assumptions of behavior that underlay the work have been successfully challenged from a variety of different perspectives. They have presented their data to public health audiences that have raised many questions and suggested that the data were noit ready for wide circulation.

  24. 24
    HILFY says:

    “Generalizabilibility”, isn’t that one “bili” too many?

  25. 25
    Old Scold says:

    Don’t want to dispute your methodologists, but they seem to be taking a dull econometrics knife to a public health gunfight, as per buckguy . And McMegan is on the case: so you know what that leads to.


  26. 26
    MomSense says:


    Well said.

    With the usual caveat about the limitations of anecdata, let’s say you are working at a homeless shelter and as the opioid problem in your area expands you see a lot more addicts at your shelter. Around the same time, Fentanyl hits and all of a sudden they are dying in the bathrooms at an alarming rate. As the opioid problem is worsening, you start requiring all of the staff to carry Narcan, you cut off the bottom 18 inches of the bathroom doors, you routinely administer Narcan under the bathroom doors when you see someone going in and don’t see their feet go across the small room from the toilet to the sink, and you still lose a lot of people because they don’t all use inside where it’s warm.

    So are the addicts coming to the shelters to use in the bathrooms because they know you have Narcan? It really doesn’t seem that way when you are in that environment. It feels like the opioid problem is spiking, fentanyl presence increases, overdoses increase, the measures people on the front lines take increase, and I think it would be very difficult to figure out causation without a hell of a lot more actual taking to human beings involved.

    And of course we need more prevention and treatment. We also need better education and protocols for prescribing providers. My oldest was handed a prescription for 14 Vicodin after his wisdom teeth were extracted. I ripped that paper up and called the dentist. One pill at most and he can switch to an extra strength Tylenol. I was furious.

    The people who genuinely need opioids for pain treatment are now jumping through unnecessary and shaming testing and questions because these drugs have been overprescribed and unnecessarily prescribed.

  27. 27
    Ruviana says:

    @RSA: Thread’s dead but thanks for the response.

  28. 28
    CrsngthDrknss says:

    As someone in recovery myself, I can attest to amount of time and effort addicts spend trying to figure out the best way to get high for the longest period of time for the least cost. So, when we are using, we’re definitely capable of rationalizing thought, if not actual rational thought. That doesn’t make the conclusions of this study anymore valid, especially given the other problems noted, but surviving active addiction does require a certain level of cunning that can resemble rational thinking in a limited sense. But am I the only one who was struck by the fact that they used a scale that goes to 11 in the first study? Obligatory:

  29. 29
    PhoenixRising says:

    the cost of overdosing borne by an addict has gone down and thus riskier behavior leading to more overdoses increases

    It’s the ‘and thus’ that is the sticky wicket. Narcan leads to addicts living to die another day, thus raising ER visits for ODs with no effect on death rates.

    Making someone who is using opiates (pills and/or heroin, because we’re doing such a great job of tightening access to pills made in factories inspected by the FDA that addicts are crossing from the junk in a bottle with someone else’s name to the junk imported by drug cartels) into a living addict instead of a dead one doesn’t seem to have an impact on whether they OD and die, statistically. And it’s easy to prove it. People in this condition are unlikely to be able to try geography therapy for their addiction; drinkers sometimes have success with moving somewhere else & starting new social networks after detox, but since opiates are illegal, addicts can’t leave their dealers/supply network as easily.

    So the ER that saw my cousin Rob 3 times and shot him with Narcan is served by the ambulance crew that picked up his body at his apartment. From one perspective, the trips to ER and referrals to detox (which he used in the final 6 months of his life) represent wasted health dollars. Narcan didn’t save his life or cure his dependence. It merely extended the time-frame for him to wrestle with addiction.

    Highly inefficient, compared to preventing opiate access in the first place for a guy who fell off a lot of roofs once the first fall got him a bottle of oxy? Yes. We should definitely put more resources into stopping addiction before it starts. But that’s gonna be a tough sale to make in communities beset by addiction in white adults 25-55, because every parent of that population is hoping her child will be the one who wakes up ready to change.

    This is not like HIV risk behaviors, though. HIV+ status is binary and permanent, vs something you expect to be more able to cut back on tomorrow. It’s true that better treatments for HIV, treatments that alleviate the quick ugly death from AIDS and transformed positive status into decades of manageable conditions you mom will probably not see the end of, raised transmission rates among gay men. However, the analogy breaks down when you unpack it: Having sex with other men is not inherently harmful, nor is it illegal. Nor does it cause immediate violent death in a certain % of those who practice it. It’s more like diabetics who eat a donut & shoot some glucose: this is a health risk, and I know what to do about it. Or my wife, who has a shellfish allergy, letting me order the crab cakes once she checks her purse for the epi-pen. Once you’ve controlled for the consequence ‘I’ll probably die from this pleasure’ some risk becomes tolerable to otherwise healthy folks.

    I guess I dispute the premise that someone who is already dependent on opiates is healthy in that sense, or able to make a cost-benefit analysis. But we don’t need the ‘thus’ to be proven (I don’t think it can be) in order to make smarter policy choices. We just need to be willing to tell old Republican white people that their addicted kids are going to be collateral damage as we orient our health policy toward preventing addiction.

    Good luck with that!

  30. 30

    @Mnemosyne: another problem is that our rehab system is largely built around a cult.

  31. 31
  32. 32
    Barbara says:

    It’s not analyzing rational behavior on the part of addicts so much as giving guidance for where public health dollars have the greatest benefit. Narcan is increasing as an expense for local governments. I suspect these researchers would have preferred a different outcome, one that supports even extreme efforts to save lives.

  33. 33
    Mnemosyne says:

    @Major Major Major Major:

    Bella Q corrected me on that one time: actual in-patient hospital treatment is much more science-based and you have real psychiatrists and psychologists working with people both one-and-one and in group therapy.

    The problem is that most people end up in the amateur group therapy that is NA or AA, and while it can be very helpful for a subset of those people, it definitely is not the answer for everyone. Most people need serious in-patient treatment, not amateur group therapy.

  34. 34

    @Mnemosyne: The thing is, those two treatments go hand-in-hand most of the time, and AA/NA is often legally mandated after an incident like a DUI or an OD (or part of a sentence-reduction agreement). And after your inpatient (or intensive outpatient) time, during which you’re usually also dragged to AA/NA sessions, it’s the only option left.

    It can indeed be very helpful for a subset of the subset of people who actually want to quit, but AA at least doesn’t have a success rate any higher than baseline spontaneous remission.

  35. 35
    Ab_Normal (at work) says:

    I’m in the same boat, except it was my stomach lining I killed with NSAIDs. Now I use a combination of celecoxib, PPIs, and, yes, the dreaded opioids to control chronic pain and let me actually work for a living. While I’m all in favor of studies, I’m afraid #1 is going to be used as an excuse to screw over spoonies who don’t respond well to OTC NSAIDs…

  36. 36
    Jim says:

    Those of us who get kidney stones frequently (acute pain, not chronic) found early on that opioid alternatives just didn’t hack it. In my own case, taking one hydrocodone at the onset of symptoms usually handles it; if the stone is slow to pass, two pills may be needed. You and others do use the term “chronic pain,” which I think is the key. Use of opioids for acute pain is far less likely to lead to reduced effectiveness and addiction. I’d hate to see these things outlawed completely (my experience is that the current normal prescription is for 10 pills max).

  37. 37
    WereBear says:

    As I understand it, a big factor in the opiod crisis is that these people don’t have much in the way of health care.

    They don’t get physical therapy or the right surgery or that expensive diagnostic test or enough time in the hospital/off work to actually recover.

    They get a prescription. Next!

  38. 38
    BruceJ says:


    Precisely. My wife was taking WAY too many OTC NSAIDS when the doc finally prescribed methadone for her ongoing pain from a shattered/rebuilt ankle. That worked, but now he’s taking her off it and putting her on all sorts of other things that aren’t working.

    They’re going to go back to the bad old days “In pain? take two aspirin and call me in the morning quit bothering me”

    But all of this is hypocritical anyway..the only reason they declared an ‘opiod crisis’ was because [gasp]!!! WHITE PEOPLE are dying of it. Somehow that magically changed addiction from a lawnorder problem to a medical problem…

  39. 39
    MoxieM says:

    @peej01: I live with chronic pain. (Back shit blah blah blah. The worst part is the stenosis: progressive, in my case irreversible, and most I can’t really walk well, or stand for reasonable periods of time.)

    I do not get any kind of narcotics from MDs, because overweight over educated past middle aged white women are grouped into that “who cares” sector, well documented in the lit.(fuck them and the pain they rode in on, not our concern.) Like the Rheumatologist who told me to my face that my Fibro was “psychological” and I needed to exercise (I already was), but then put Fibro down as a Dx in my record. Uh-huh.

    A better and more topical illustration: See: N=240, patients at the VA for the effectiveness study. I wonder what the gender breakdown looked like? And yes, there is evidence that male and female bodies process pain differently, just like heart disease, and a host of other things. Hmmm.

    I have had a few opioids prescribed (like 14 of them) once or twice, and Holy God, I felt like I had gone on vacation, gone to heaven, just become … normal … for a few shining hours.

    They may not work long term, but what they do for the people they help is to make a life worth living. (When I compare my experiences to my fairly close and intimate knowledge of healthcare qua care in Germany, I don’t wonder that about 25% of chronic pain patients are suicidal, many to completion.) And they don’t typically use opioids in DE — they support people in need via the health care system in creative and meaningful ways.

    I wake up every morning cursing that I didn’t die in my sleep, because I have to face another day like this. (I’ll hang around since I promised my daughter, who I see once a year.)

    Another example: I had a cancer scare last year, and I honestly hoped it would take me. I was sorry the biopsy was negative.

    No one should live like this.

    Oh, and I have a Ph.D. in Sociology of Medicine, hahaha for what it’s worth, which is not much since I can’t work, and couldn’t get any if I tried. (ask me how I know). So, the studies, I can read them.

    In my case the cheapest outcome for all concerned would be to — I dunno.

  40. 40
    MoxieM says:

    @Ab_Normal (at work): exactly. And if I understand the lit, most of the actual deaths are from street drugs (i.e., fentanyl). I don’t know how tenuous thelink is between prescribed opioids and the transition to street drugs.

    ETA: I did original research on injection drug users (also on HIV/AIDS harm reduction), ID users and dual-diagnosis, and pathways of treatment modalities. So it’s not as if I don’t have some sense of this landscape.

  41. 41
    DissidentFish says:

    To me denying Naloxone on the grounds that a study says its not efficient overall is inherently inhumane. It keeps people who don’t have to die from dying. I like that.


Leave a Comment

Leave a Reply

Your email address will not be published. Required fields are marked *