When you ask Stanford pain specialist Sean Mackey how he deals with chronic pain patients in the middle of the opioid epidemic, he has a consistent answer: “It’s complicated.”
Mackey, who describes himself as “a centrist on opioids,” says that’s how doctors should approach the decision to prescribe opioid painkillers to chronic pain patients. That many doctors didn’t approach their initial prescriptions in a more nuanced way in the 1990s, he argues, is one of the factors behind the current drug crisis. But he also says many doctors now take a too simplistic view against opioids, causing what he characterizes as “a public health crisis in pain.”
In the past, I’ve spoken with two of Mackey’s Stanford colleagues, Anna Lembke and Keith Humphreys. They approach the opioid epidemic as addiction and drug policy experts, respectively, and that colors how they approach the current drug overdose crisis. (You can read one of my full interviews with Lembke, where she speaks about the root causes of the opioid crisis.)
But Mackey, who is closer to the chronic pain side of things, came highly recommended by Lembke and Humphreys. Mackey went out of his way to say, “You will not find anybody who has more respect for Anna Lembke than me.” And he said that “Keith is great, incredibly smart.”
With that out of the way, he does have some big disagreements with his colleagues.
For one, he takes issue with the claim that research has yet to prove that opioids can effectively treat chronic pain. He agrees that we need more research, but he says he has personally seen some of his patients get better through careful opioid-assisted treatment — although, he clarifies, he uses the drugs not as a first-line response but as “a fourth- or fifth-line agent” for specific individuals.
This is what Mackey emphasizes again and again: Pain patients are different individuals, and different individuals will have different pain management needs. He notes that only a minority of his patients actually use opioids, and many rely instead on some of the 200-plus other non-opioid medications now available in the pain treatment field.
But when patients can be safely prescribed opioids to their benefit, Mackey says, he will prescribe the drugs to them — but, of course, monitor them to ensure they don’t progress to misuse and addiction.
What follows is my conversation with Mackey, edited for length and clarity.
How do you explain the start of the opioid epidemic?
What we had back in the ’90s is a confluence of multiple issues.
You had an increasing awareness of the importance of treating pain in our society. People want to go lay blame on “Pain as the Fifth Vital Sign.” I think that’s overly simplistic. It was an increasing awareness that pain is a real issue and it should be treated. And I think that was a great message, because what we had before was basically, you know, “Pain is good for you. Suck it up and deal with it.” Run the counterfactual to that. We do not want to go back to that time in our history.
That was compounded with the fact that in our medical training, we don’t get much education about how to treat pain. The average medical student gets seven hours of medical education around pain. The average veterinarian gets 40. Which is great if you’re treating your dog; it’s not so great if you’re treating your grandmother. So we don’t get much education in medical school about pain. [As Lembke points out, that’s still seven and 40 hours more than doctors get for addiction treatment.]
Compounded with the fact that we didn’t have many treatments back then. I’ve now got over 200 different medications that I can legitimately apply toward pain. The vast majority of them are non-opioids. Back in the late ’90s, you had a limited supply of drugs and meds that you could actually use for pain.
Compounded with the fact that you had some key select physicians that had published a small paper that got a lot of attention. There was a messaging that we should use opioids for chronic, non-malignant pain. And there was a perception that they were safe. That was — incorrectly — ascribed to these small-scale observations and studies. You can’t take a letter to the editor with 38 people in it, say, “No one in this had an addiction,” and apply that to a broad population. So we had a misunderstanding [and] a lack of awareness of the real risks of addiction back then.
Compounded with the fact that the drug companies saw this as an opportunity to sell. This is what companies do. This is, quite frankly, what makes America what it is. You’ve got companies that are selling a product. So they went out and took advantage of the messages, and they sold a product. And they were very successful in that. Some of those companies clearly misbehaved. Clearly. They were fined for it. They should have been fined for it.
Now you got these physicians — predominantly primary care doctors, dentists, emergency department docs — who’ve got little education around pain, who are aware that there’s an increased awareness to treating pain, who’ve got very few tools that they can use to treat pain, who now are hearing a message that, “You know what? It’s okay to use opioids to treat pain.” So they use the opioids, but they don’t have the skills for how to monitor it; they were never educated on how to actually figure out who these were going to work for and figure out for who these weren’t.
You’ve [also] got all of these primary care doctors who have increasing production pressures and have less and less time to spend with their patients. We were squeezing them for time, and as such, they only have a few minutes. And it was easier to prescribe an opioid than do anything else.
What you’ve got is a perfect storm of all of these things coming together. You see this incredible increase in the use of opioids over time. The baseline vulnerability of addiction to these agents probably hasn’t changed in our society; it’s the same. But the problem is that we exposed a population to a much larger amount of these drugs, and that baseline vulnerability that might not otherwise have been exposed is now exposed. And now we have more people with a substance use disorder.
The research has found no good data supporting the use of opioid painkillers for chronic pain, and there are obviously some bad risks to opioids. Anna Lembke has told me about how in her practice, she’s seen people’s pain actually get better when they get fully off opioids, because they experience less withdrawal pain and lose the hypersensitivity to pain that can come as a side effect with opioids.
So how do you reconcile all of that in your practice? What are some of the factors you’re considering?
A couple things you need to understand: When you hear the term “the data doesn’t support this,” what that really means is that “we don’t know.” But as physicians, we don’t like saying that we don’t know. Because then that puts the onus on us; that means we have to carry the issue of why we don’t know. So the way we reframe it is, “Well, the data doesn’t support it.”
The challenge we have with just about every single treatment for chronic pain is that we don’t know — or, if you prefer, the data doesn’t support it.
There are very few studies that show long-term improvements of chronic pain with opioids. There are some. But in the CDC [Centers for Disease Control and Prevention] guidelines, they chose to ignore any study that wasn’t at least one year or more. So they ignored all of the FDA randomized controlled trials that led to approval of opioids, because they didn’t meet that criteria. It all depends on how you want to look at the data.
But within that context, if the CDC were to apply that same set of rules to all of the treatments we have for pain, we would have come out with the same conclusion: “The data doesn’t support it,” or, if we’re being intellectually honest, “We don’t know.”
That’s why we need more research. We really do. We need more research on the long-term consequences of opioids but also just about every other treatment that we do for chronic pain.
But it’s important we understand the meaning of the data, because it’s not so clear-cut. When you hear that phrase “the data doesn’t support it,” you immediately mentally translate that to “you shouldn’t be using this.” That’s not the reality, but that’s how it often gets twisted.
So we’re faced with a situation in which we really don’t know the long-term outcomes of just about any treatment that we have for chronic pain, and yet 100 million Americans have chronic pain. We don’t need to treat them all, because [many] self-manage; they’re like my father, who has huge, terrible chronic pain but won’t talk to me about it and won’t see a doctor about it.
Okay, so how do opioids factor into that in your practice?
I don’t use opioids as a first-line agent. I don’t use them as a second-line agent in most chronic non-cancer pain. I don’t use them as a third-line agent. I may not use them as a fourth-line agent. But I may use them as a fourth- or fifth-line agent.
If I do, I do a trial of them. I say, “We’re going to give this a trial, and we’re going to see if they work for you, if they improve your quality of life [and] physical function. We’re going to monitor you. And if, in the end, they don’t [work], we’re going to take you off. And if they do, we’ll cross that bridge when it comes.”
I think you’re right that the right interpretation [of] the data is that we don’t know yet if opioids are effective for long-term chronic pain. And I think you’re right that people do tend to view the lack of evidence as definitive proof that opioids don’t work.
But when I asked Keith Humphreys about this, he gave me a funny example that I think can apply here: You could assume that jumping out of a plane without a parachute is safe because there is no randomized clinical trial for whether wearing a parachute makes you safer.
Basically, you can’t assume that opioids are good for chronic pain just because studies have yet to show they’re bad. We may not know if opioids are effective for long-term pain, but we do know that long-term use of opioids has a lot of risks. So we shouldn’t try to use these drugs until we know they’re effective.
How do you respond to that?
That’s an interesting straw man argument. I understand what Keith is saying, and that parachute analogy has been used repeatedly.
I would never jump out of an airplane without a parachute. That’s very clear. I don’t have to have a randomized controlled trial to make decisions on everything.
This is the challenge, though, that we face in medicine: Because we don’t have clear and convincing data on many of the treatments that we have, we have to then rely on the next step down in the quality of evidence. If it’s not randomized controlled data, it’s going to be observational data — ideally, some controlled observational data. If we got that, we can draw conclusions.
And if we don’t have that, we work our way down to [individual] expert opinion. That’s probably the weakest data of all. That’s highly influenced by the lease.
The fact is if you go looking, there’s clearly data out there that opioids improve pain. These drugs would have never been approved by the FDA if they didn’t.
The question isn’t whether they improve pain; the question is whether they improve it long term, because all of these FDA trials are only 13 weeks [in] duration. That’s where the data falls apart. That’s where we have to say, “You know, we just don’t know.”
We’ve got some studies out there of longer-term duration showing improvements in pain and function. We’ve got others to show that it doesn’t.
The key here is to figure out in which patients these medications will work, and in which ones they are going to be problematic. We’re getting better at that, but we still got a long way to go.
So I wanted to ask you a hypothetical: You’ve tried some of the first-line, second-line, and third-line treatments, and now you have a patient for whom you’re considering opioids. What are some of the things you would weigh for this kind of patient to see if the benefits outweigh the potential risks?
Do I think they have the type of pain condition that may respond to an opioid?
So if somebody has some bad arthritic condition. There’s this elderly guy I treated sometime ago. He was in his 70s, and he was hobbling around on two bad knees. But he wasn’t a candidate for knee replacement surgery. I put him on a tiny amount of methadone. He came back to see me the next time, and he gave me a huge hug, because for the first time in years he was able to go fly-fishing with his son.
That’s an example. He had no prior psychological history to suggest that he was at risk.
So what are those risk factors? The most obvious is prior history of addiction. Family history of addiction. History of depression or anxiety is also a risk factor; we find that these psychological risk factors induce a degree of vulnerability in a patient that puts them at higher risk if they’re put onto opioids.
Do they have medical conditions that may be impacted by the opioids? Do they have sleep apnea that opioids could make worse and cause them to have an untoward event?
We take into account the consequences of opioids from a medical standpoint. A lot of patients will get constipated on opioids, and we have to make sure we treat that.
It can result in a reduction in sex hormones, decreased libido, and decreased vigor and energy as a result of decreased testosterone. So we have to be mindful of that.
Recognizing, by the way, that darn near every drug I prescribe a patient has some sort of untoward side effect. There’s no free lunch here.
But it’s factoring in all of this. This is why it’s complicated. It’s not nice and simple.
So when you’re putting a patient on opioids after trying different treatments, given the concerns about tolerance, how does that regimen look? I’ve heard from some doctors that they limit patients to taking opioids once every few days. Are you doing anything like that?
Tolerance is a physiological consequence of prescribing a multitude of drugs. Opioids, beta blockers, calcium channel blockers — for the same amount of medication you get less effect, or you need more medication to get the same effect.
Tolerance is highly variable from person to person. I could put somebody on an opioid, and they can be on the same dose and never change. I could put somebody else on them, and they find that over time they’re escalating their doses. It does not imply addiction — no more so than you put somebody on a beta blocker for blood pressure and they need an increase over time because of tolerance.
We do monitor it. The challenge here is all of these issues take time.
What we’re really interested in is monitoring for bad behaviors. Are people behaving themselves on these medications? Are they trying to seek them out from other sources? Are they using more than they’re prescribed? This is what we’re trying to monitor, and it does take time.
So it’s a very individualized problem.
It’s an effort that takes time and energy. I’m very sympathetic to our primary care physicians who are under such huge production pressures and don’t have much time.
That’s where, ideally, good pain physicians come in. That’s where we can play a role.
Let me be clear: not just to be the people who write opioid [prescriptions]. [For] the vast majority of people, I don’t write opioids. Like I said, there’s over 200 medications I can use, [and] 90 percent of them are non-opioids.
Are there patients, then, who receive opioids from a primary doctor, and you actually phase these patients out of opioids into other treatments?
I got every scenario you want to lay out.
I’ve taken people off as many opioids as I’ve put them on. So Anna [Lembke] talks about taking people off opioids and them doing better. I have tons of those stories — people who were not really benefiting from the opioids.
I’ve got patients who were sent to me from primary care docs. They send them to us a lot, asking, “Is this person appropriate for the opioids they’re on?” So we do a comprehensive assessment. Sometime we say they should get off the opioids; it’s not really doing much for them, really not improving their function [or] quality of life, and we’ll help get the person off opioids.
We’ve got others in which we see that the doctor is really scared of writing [opioid prescriptions] because all of the national attention. This person is going to work every day, spending time with their family and friends, they’re behaving themselves, they’ve been on a stable dose for years, they’ve tried other treatments and they didn’t work. [Opioids] seem to be working for them; we would recommend that you just stay the course.
There’s no cookie-cutter patient. There’s huge individual variability.
You mentioned these other treatments. I get a lot of messages from pain patients. And they seem to believe that if you take away opioids, they’ve got nothing else. But as you’re telling me, there are all these other different treatments that might work on different individuals. What are those?
We can break down our treatments for pain into six broad categories.
One: the medications. I mentioned there’s a large number of different medications.
Two: interventional procedures. We’ve got nerve blocks. We’ve got probably 100 or so different injections, blocks, and minimally invasive surgeries we can do for pain.
Three: physical rehabilitative and occupational therapy approaches for pain. They help improve function, core strengthening, and endurance. We find movement is probably one of the best treatments for chronic pain.
Four: pain psychology. Tremendously underappreciated, underutilized in this country. The mind-body approaches are gaining more and more interest in society, and can be very effective.
Five: complementary alternative medicine approaches. Acupuncture, mindfulness-based stress reduction. Good data there.
Six: self-management. This is learning skills on how to better address your pain, how to better manage it and cope with it.
Let me be clear on the self-management, because that sometimes gets translated in the patient’s mind as, “They’re saying I just need to suck it up and deal with it.” That’s not my message. We need to give people skills on how to better manage the pain that they have. It will not eliminate the pain, but it will help them to lead more productive lives.
So there’s those six broad categories that we use. Opioids are just one tool within that broader toolbox.
What are some of the skills you teach patients for self-management?
We find goal setting is incredibly important. We want to help people identify clear, functional goals.
Most people, they want their pain to go away, but really they want control of their lives back. They want to be able to engage more with family and friends. They want to go back to work.
So we help them set goals and monitor those so that they can incrementally get there.
“Incrementally” is the key word here. What happens with chronic pain is you get into these bad cycles. You have a good day, and you go out like gangbusters, and you try to do everything. Then what happens is you find that you’re paying for it for the next several days, and you can’t get out of bed.
So we set pacing. We help people learn how to pace their abilities, and learn how to better understand their body.
Health education is a key one. We find through the data that people who are better educated about their bodies and their health, they actually do much better with their pain. They have much lower pain, and they have increased quality of life and function. So a lot of this is through health education as well.
We teach them about appropriate medication prescribing. This can relate to opioids, but it’s also in general. That just means it’s better to take medication on the clock, rather than just when you have pain. It’s better to have a continuous supply [and] a level of medication in your system rather than riding a roller coaster again.
All of these are part of self-management. There’s a lot more to it. But you didn’t hear anything in my message there that just says, “Suck it up.”
So medical marijuana. We do have studies showing that medical marijuana laws are correlated with a reduction in opioid overdose deaths. What do you make of this?
I don’t know. We don’t know.
Here’s what we do know: In small-scale studies, there have been findings that medical marijuana can help improve pain. The challenge is that we don’t have good, larger-scale studies on this. The reason is that marijuana is scheduled as a schedule 1 drug by the DEA [Drug Enforcement Administration], and as such, it is in the same category as LSD, PCP, and heroin. What I have called for when I was president of the American Academy of Pain Medicine a couple years ago was that we reschedule it as a schedule 2 drug so we can study it.
We need to effectively study this drug. Because we’re in the same situation — we’re drawing the same inferences about medical marijuana as we did back in the day with the opioids. We’ve got small studies. We’ve got an inkling of a signal there, that there may be something beneficial.
What we’re doing right now is running this huge social experiment on the United States through the legalization of marijuana. We’re going to find out what the impact is.
My hope in this is that the CDC and other agencies will appropriately collect the data so that we can really try to draw inferences as to whether or not there’s benefits from this. We clearly need more research.
I’m not anti-marijuana; I’m not pro-marijuana. I’m saying I don’t know and we need good research to figure this out.
If marijuana or cannabinoids are great pain relievers — they’re safe, effective, and have minimal side effects — you’re going to see me clearly stand in line and use those as a tool. But we don’t know.
Do you think the government response to opioids has been too harsh? What do you think is the ultimate solution to this crisis?
We should be treating this as a public health issue.
As such, we need public health solutions for it. I’m not convinced that we’re going to police or legislate our way through this. Public health approaches to these types of epidemics have been successful in the past. They can be successful in the future.
We also have to recognize that there’s a public health crisis in pain.
We have an opportunity to address both of these very successfully. So I co-chaired, on behalf of [the US Department of Health and Human Services], the National Pain Strategy. We released this strategic plan last year on how to address this country’s problem with chronic pain. If we can address this country’s problem with pain, we will address a good part of this opioid issue.
Not all of it, because we really have different issues with the opioids. We have people who were prescribed opioids for pain problems and went on to have other issues. But we also have another group of people who just started using heroin or opioids not because of pain.
But let’s look at the solutions. We have solutions through the National Pain Strategy. Because if we focus just on the opioids, if we just try to police away the opioids, it’s not going to work. And when we eventually do address this problem with opioids, we are still going to be left with a country and a planet that’s got large issues with chronic pain. Why not address both?