A major study published in The Lancet journal concluded that cannabis does not provide pain relief or decrease opioid use. The study made headlines around the world because lots and lots and lots of other research found the exact opposite, including the landmark National Academies of Sciences, Engineering and Medicine report in 2017 that cited "substantial evidence" for pain relief. What's the deal? PRØHBTD dug deeper to find out.
The Argument Against Analgesia
Recruiting participants from Australian pharmacies, the researchers tracked 1,514 patients who take opioids (including fentanyl, morphine, oxycodone, buprenorphine, methadone and hydromorphone) to manage chronic non-cancer pain. The patients were 58 years old on average, and the regular cannabis consumers had experienced pain for an average of 13 years, compared to 10 years for non-users.
The participants took part in initial interviews to set baseline levels for opioid use, cannabis use, pain severity and pain interference (i.e., its effect on sleep, work, social interactions and daily life) on a scale of one to ten, and the researchers tracked changes over the next four years via annual interviews and self-completed questionnaires. Upon the completion of the study, nearly one in four participants utilized cannabis for pain relief, but those who used cannabis reportedly experienced greater pain severity. Per the researchers, "People who used cannabis had greater pain and lower self-efficacy in managing pain, and there was no evidence that cannabis use reduced pain severity or interference or exerted an opioid-sparing effect."
On the surface the study looks tight: 15 high-profile researchers tracked more than 1,500 pain patients over a four-year period. Of course, looks can be deceiving.
(Moving forward, "cannabis consumers" refers to those who consume 20 or more days per month, and not to the occasional consumers who also participated in the study.)
Cannabis Set Up to Fail
The participants self-reported the data—e.g., pain severity, pain interference, anxiety levels, etc.—which the researchers defended as reliable: "Although our data were self-reported, this method of collection is reasonably reliable, particularly when there are no disincentives for being honest." The study made a point to defend self-reported data as reliable, but it then exclusively singled out the self-reported data on cannabis analgesia (or pain relief) as unreliable.
"Participants who used cannabis reported that the mean effectiveness of cannabis on pain was 7 out of a possible score of 10," said the study. Likewise, a strong majority of cannabis consumers in the fourth and final follow-up credited cannabis with relieving pain severity (82 percent) and pain-related distress (73 percent) and improving sleep (63 percent) and general relaxation (64 percent). Among those who stopped consuming cannabis by the four-year mark, only 12 percent did so because they felt it was ineffective.
So how does one look at this data and say cannabis failed to provide pain relief? Per the study, the self-reported efficacy of cannabis is unreliable because the "people who used cannabis in the past month reported greater pain severity and interference."
Outright rejecting a category of "reasonably reliable" self-reported data based on another category of self-reported data begs all kinds of questions.
Seven Pertinent Questions about the Study
Question one: So opioids don't relieve pain either?
The researchers dismissed cannabis-derived analgesia because "people who used cannabis in the past month reported greater pain severity and interference." If cannabis had no analgesic effect and all the participants took similar opioid dosages, the severity scores should be nearly identical. Even more so considering the cannabis consumers took more opioids on average, which they did by a significant margin. So either the cannabis consumers had confounding factors and/or higher rates of pain to begin with, or the researchers need to explain why those who took more opioids experienced more pain severity than those who took less.
Question two: What would have been a valid score for self-reported cannabis analgesia?
A score of one still suggests a small measure of efficacy, but the consumers gave cannabis a seven, and 2015 findings based on the same study data said the cannabis consumers did experience "greater pain relief." The seven score was the highest of any self-reported data using the one-to-ten scale, so if the highest self-reported number isn't high enough to suggest pain relief, what score would? Did the researchers need to see an eight, nine or even a perfect ten? The researchers' basis for dismissing the self-reported seven—a basis only applied to this particular score—suggests they would have likely overruled any score that suggested efficacy.
Question three: Would the efficacy score be even higher if the consumers had quality-controlled cannabis?
Dr. Wayne Hall, the Director of the Centre for Youth Substance Abuse Research, is one of the contributors to this study, and in a separate policy paper he said, "Backyard illicit cannabis medicines may be cheap to produce, but their quality is uncertain and variable." So the high self-reported efficacy score came from "uncertain and variable" cannabis that had to be illegally produced, procured and consumed. Since the participants only had access to backyard illicit cannabis, it might have suppressed the self-reported cannabis analgesia score. Imagine if the pain patients had access to high-quality, lab-tested cannabis that involved less uncertainty and variance: The efficacy score could've been even higher.
Question four: How do the researchers justify overruling a high efficacy score based on a minor difference in pain severity scores?
From the start of the study to the final follow-up, the pain severity for cannabis consumers increased from 5.1 to 5.3, while non-consumers dropped from 5.1 to 4.7. That's a difference of about half a point, which some would argue is not significant enough to invalidate an efficacy score of seven. Complicating the issue further, the pain interference score actually improved for both: Cannabis consumers dropped from 6.2 to 6.0 and non-consumers from 5.6 to 5.3. In other words, the researchers suggest a fifth-of-a-single-point increase in pain severity for cannabis consumers invalidated a high efficacy score and a decrease in pain interference. That's a tough sell.
Question five: Why did the severity scores invalidate cannabis efficacy, and not the other way around?
"Pain is a very subjective thing, so actually getting very good measures of pain and understanding where the effects are can be quite complicated," said contributing researcher Michael Farrell, Director of the National Drug & Research Centre. Dr. Farrell noted the "complicated" and "subjective" nature of pain scores in response to an interviewer's reference to "heaps of evidence that cannabis works for things like pain." Unless he wants to advertise bias, Dr. Farrell should then agree that the pain severity scores he used to argue against efficacy are also complicated and subjective and thus not entirely reliable. His warning about the accuracy of pain measurements is especially pertinent for this study in which the scores were all within a fraction of a point—both in terms of their individual movement and in direct comparison to each other. If the study allows itself to disregard key self-reported data, a solid cannabis efficacy score seems more credible than a comparison of subjective severity scores that differ by a fifth of a point.
Question six: Do reductions in opioid use explain the higher pain severity score?
For non-consumers across the four-year study term, annual opioid use dropped from 7,000 units (morphine equivalent) to 5,500. Cannabis consumers, on the other hand, experienced a much more significant reduction: Opioid use went from 9,000 units to 4,900, and 21.5 percent discontinued all opioid use. The researchers said cannabis didn't help reduce opioid use because only 29 percent thought it had at the four-year mark, yet consumers did experience the highest reduction in actual opioid use, and a significant decrease in opioid use might explain the minor increase in pain severity. Moreover, it seems suspect that the researchers would reject a high efficacy score based on a "subjective" and "complicated" fifth-of-a-point increase in self-reported pain severity and then accept the less-favorable self-reporting on opioid reduction even though the non-subjective data shows a major 46-percent reduction.
Question seven: Why did consumers give such a high efficacy score if cannabis doesn't actually work?
Dr. Farrell actually answered this question. "One of the things we think happens when people report [cannabis] benefits is the sleep and sedation effects it has," said Dr. Farrell. "Often, when you get a good night's sleep, your pain is a lot more tolerable." But does he actually believe cannabis improves sleep? Dr. Farrell and three other researchers from the present study published research in 2018 that described the sleep-benefit evidence as "low quality." Think about this for a moment: A contributing researcher explained away the high analgesia score by saying the consumers merely experienced a different cannabis benefit, for which he previously claimed there is only "low-quality evidence."
To a Hammer Everything Looks Like a Nail
Another question involves the possibility of bias. Most of the researchers seemed to start this study with the pre-existing view that cannabis is a threat.
The biggest cannabis skeptic of the bunch seems to be Dr. Hall. In a 2014 Addiction study, he argued, "Regular cannabis use in adolescence approximately doubles the risks of early school‐leaving and of cognitive impairment and psychoses in adulthood." Dr. Hall also cited the gateway drug theory and included cancer, heart attacks and bronchitis among the potential cannabis risks, while the Daily Mail quoted him saying, "If cannabis is not addictive then neither is heroin or alcohol" and "it is often harder to get people who are dependent on cannabis through withdrawal than for heroin."
Dr. Louisa Degenhardt, also with the Centre for Youth Substance Abuse Research, joined Dr. Hall in a study that claimed "frequent cannabis use in teenage girls predicts later higher rates of depression and anxiety" and another that said cannabis may contribute to schizophrenia. The duo also teamed up on a study that said "probable adverse effects include a dependence syndrome, increased risk of motor vehicle crashes, impaired respiratory function, cardiovascular disease, and adverse effects of regular use on adolescent psychosocial development and mental health."
Another researcher with the Centre for Youth Substance Abuse Research, Dr. Gary Chan, joined Dr. Hall in studies that associated cannabis with 1) poorer physical and mental health, 2) mental health impairment and polydrug use and 3) lower income and educational levels. Likewise, Dr. Milton Cohen claimed medical cannabis offers false hope and that its benefits have been oversold to the public.
Shifting to researchers from the National Drug & Research Centre, Dr. Farrell once co-wrote an editorial that said "legalization... would probably reap a new generation of problems," while Dr. Marian Shanahan led a 2014 study that said cannabis legislation does not provide a boost to the economy or a net social benefit, but she did list off a litany of potential harms. Finally, lead researcher Dr. Gabrielle Campbell co-authored a study on new therapeutic community admissions that found "the alcohol and opioid groups were significantly less likely than the cannabis group to have a history of attempted suicide."
Some of the contributors did note the possibility of "modest" benefits from medical cannabis, though most of their previous work focused on potential harms. Among all the contributors, those from the Centre for Youth Substance Abuse Research seemed to have the largest body of published work that highlights the potential risks and discounts the possible benefits. Predictably, the Centre's website features a marijuana page that does exactly the same thing.
Overall, these researchers are dedicated, respected and credible, and many of their studies contain valuable findings. However, when it comes to cannabis, they seem to suffer from confirmatory bias, in that they were "inclined to uncritically accept evidence that accorded with their pre-existing beliefs." And we must credit the preceding description of bias to Dr. Hall who used those very words to describe researchers who say medical cannabis helps reduce opioids use.
Did Someone Say Big Pharma?
In June, Virginia's Attorney General filed a lawsuit against the maker of OxyContin, saying the company knowingly lied about its harmful risks. The opioid epidemic led Republican Congressman Morgan Griffith to say the FDA needs "to be looking hard for less addictive pain relievers, which may include marijuana. If you told me I was going to be in a lot of pain and I had a choice between… an opiate and smoking marijuana… I would smoke the marijuana before I would take the opiate."
That's not what Big Pharma wants to hear.
PRØHBTD does not question the integrity of the researchers, but dismissing the analgesic benefits of cannabis is a major benefit to opioid producers, and most (if not all) of the researchers for this study have received funds from major opioid producers, albeit not for the study in question. The relationship between the researchers and opioid producers needs to be stated, as it was in the study's Declarations of Interest.
Mundipharma pedals OxyContin in countries outside the U.S., and it's given research funds to Milton Cohen, Briony Larance, Nicholas Lintzeris, Peacock and Degenhardt. Per the New Yorker in 2017, "As OxyContin spread outside the U.S., the pattern of dysfunction repeated itself: to map the geographic distribution of the drug was also to map a rash of addiction, abuse, and death. But the Sackler family has only increased its efforts abroad, and is now pushing the drug, through a Purdue-related company called Mundipharma, into Asia, Latin America, and the Middle East… The company has organized junkets, and paid doctors to give presentations extolling OxyContin's virtues. In fact, certain doctors who are currently flogging OxyContin abroad—'pain ambassadors,' they are called—used to be on Purdue’s payroll as advocates for the drug in the U.S."
Suboxone and Subutex are brands of buprenorphine used to treat chronic pain and as an opioid replacement for those in addiction recovery. British multinational Reckitt Benckiser (and its buprenorphine division Indivior) produces this opioid, and it previously gave grant money to Raimondo Bruno, Suzanne Nielsen, Campbell, Farrell, Degenhardt, Larance and Lintzeris. Per a 2016 headline in the Guardian, 35 U.S. states sued Reckitt Benckiser for "profiteering" from opioid treatment and "shamelessly preying on patients in need of help" by keeping Suboxone prices artificially high.
A lesser-known opioid painkiller is tapentadol, which Johnson & Johnson sells as Nucynta in the U.S. and Seqirus sells as Palexia in Australia. Amy Peacock and Degenhardt both received educational grants from Seqirus. Likewise, Hall reported grants from the Therapeutic Goods Administration, an Australian regulatory body that's been criticized for being too cozy with the Big Pharma drug hawkers.
Just as taking lobbyist money doesn't mean a politician's compromised, receiving money from opioid producers for previous work doesn't mean a researcher is compromised, even if many opioid producers probably are. Some might argue that mentioning the opioid connection is unfair, but this study promotes opioids indirectly by devaluing cannabis as a pain reliever, and receiving grant money from opioid producers could be considered a financial interest in its use. To that extent, bringing up the opioid connection is at least as fair as Dr. Hall criticizing cannabis studies because some "promoters of medical cannabis... have a financial interest in its use."
The researchers tend to have ties to opioid producers and a predisposition against cannabis. However, medical cannabis supporters in Australia include a majority of general practitioners.
Cannabis and Pain
Per the study, no difference in analgesia exists between cannabis consumption and abstinence, but if cannabis is totally ineffective as a pain reliever, other factors must be considered to explain why the pain scores were not identical. Think about it: If cannabis didn't make a difference, why are the scores different?
One possibility is that non-consumers were more likely to utilize alternative pain therapies like local anesthetics, specialized yoga, hydrotherapies, chiropractic care, electrical stimulation, acupressure and acupuncture, but that's quite a leap without hard data to back it up. The more plausible possibility is actually much simpler: Those who consumed cannabis suffered from more severe pain.
Compared to the average for all participants at the start of the study, cannabis consumers had suffered chronic pain for three years longer, were taking more opioids for a longer period of time, and were willing to risk arrest to self-medicate with an illicit drug. (Australia did not have a medical cannabis program at the time and barely has one now.) They were 12 years younger on average (48 vs. 60), which means their pain started at a median age of 35 compared to 50 for non-consumers. Long-term chronic pain that started at age 35 could suggest a higher likelihood of injury and/or physical disorder as opposed to progressive age-related issues. The study itself acknowledged that "people who later initiated cannabis use were more likely than those who never used cannabis to... have greater pain severity." All these differences could mean those who consumed cannabis simply suffered from more severe pain.
Dr. Farrell himself said that pain scores are subjective, so maybe the initial 5.1 scores for both consumers and non-consumers were equals numerically but not in terms of the actual severity. Pain severity is not a constant, and those with longer histories of pain and higher-dose opioid use (i.e., cannabis consumers) may have simply experienced more severity (i.e., a fifth of a point) over the course of four years. And then there's this admission by the researchers: "It could be that in the absence of cannabis use, pain severity and interference might have been worse."
Nevertheless, the nature and extent of cannabis-based analgesia remains unclear as a clinical matter in the absence of high-quality studies. Based on the majority of recent research, cannabis appears to provide some relief, but researchers aren't certain about how much relief or how to maximize the benefit through specific dosage, delivery and cannabinoid content. For example, CBD- and terpene-rich cannabis might be more effective at providing analgesia, but the illicit consumers in Australia likely had little information on the backyard cannabis they were smoking or ingesting.
What researchers really need is for their governments to allow quality research involving the consumption of specific cannabis products, and what the public needs is legal cannabis medicine that's regulated to ensure quality, accuracy and safety (e.g., childproof storage, no contaminants, etc.). High-quality research paired with a regulated product will lead to clinical guidelines that can maximize the therapeutic value of medical cannabis.
No study is entirely accurate or inaccurate, so don't outright dismiss this study or the other researchers' work. Cannabis-use risks do exist, and both PRØHBTD and the Lancet-study researchers will surely agree on the danger inherent in self-medicating pain outside a comprehensive treatment plan. If cannabis helps relieve your pain, use it responsibly, but it should be a complement to, not a substitute for, professional medical care.
David Jenison (firstname.lastname@example.org) is Editor-in-Chief at PRØHBTD.