If you’ve ever found research on chronic pain confusing, maybe even seemingly contradictory, you’re not the only one. Doctors, scientists and researchers who study chronic pain receive advanced training and follow strict protocols, but sometimes even they wrestle with some of the challenges of studying pain. 

Why is that? We spoke with one of our Medical Advisory Board members, Dr. Steve Cohen, to ask for insight on leading chronic pain research. Dr. Cohen is also the Chief of Pain Medicine at Johns Hopkins and Director of Pain Medicine Research at Walter Reed National Military Medical Center. Our Chief Medical Officer, Dr. Jacob Hascalovici, also contributed viewpoints on understanding today’s chronic pain research landscape. 

The difficulty of measuring pain

When we asked Dr. Cohen about some of the challenges of studying chronic pain, he said, “It’s one of the most challenging conditions to study, because it’s a subjective outcome.” 

Although pain specialists can use technology like X-rays and magnetic resonance imaging (MRI) or tests that measure levels of chemicals in the body related to inflammatory responses, stress, nutritional deficiencies and other factors related to chronic pain, they don’t yet have a tool to measure pain itself. That’s why patient self-rating tools are so important for getting an overall picture of a person’s individual pain landscape. 

After all, you are the only one who knows exactly what it feels like to be you. Given that, it’s difficult for two patients with the same condition to compare their pain. If you and a friend are both dealing with lower back pain and report a pain level 7 out of 10, is that 7 exactly the same for each of you? It can be tough to say. 

That’s partly because the way the body interprets pain is very individualized. To return to our example, the two of you may both have fibromyalgia of a similar severity, and yet one of you may call it a ‘4,’ while the other rates the pain as an 8. (Or maybe one of you wishes you weren’t being asked to pick a single number to encapsulate what you’re feeling!) The way you internalize and process stress, the traumatic or supportive experiences you’ve had in the past, the quality of support you’re getting from friends and family, your own unique biochemistry and physiology, what you’re eating, how much you’re exercising and otherwise following healthy eating, exercising and sleeping principles, and certain elements of your outlook and personality, can all influence how you perceive and process painful stimuli. 

Pain, as researchers are increasingly confirming, is more like a story your body tells than a read-out on an instrument panel. That story can have highs and lows, can vary by hour, day, week or month, and can be interpreted by you in different ways, depending on what else is going on for you, how much stress you’re under, and how much support you’re getting from those around you. Pain, in other words, is very much situational. 

So, when researchers set out to investigate chronic pain, they must try to define pain and to include different ways of measuring and describing it, to make sure they’re talking about the same phenomenon at roughly comparable levels, or else the studies they conduct to determine whether or not certain medications and procedures “work” with chronic pain will not be as useful as they could be.

The reality of randomized controlled trials

For treatments, including medications, to be considered clinically effective, researchers often need to demonstrate that they perform better than no treatment at all. In clinical research, scientists accomplish this by running randomized controlled trials (RCTs). These typically randomly divide a patient group into two (or more) smaller groups: 1) the treatment group, which receives the new treatment being tested, and 2) the control group, which is given a placebo (a sugar pill, cream or other intervention that does not contain the treatment ingredients but resembles the new treatment being tested). 

If you were a scientist running an RCT for a new pain medication (let’s call it Adolor), you would need to find out if Adolor worked better than nothing. So you would recruit a group of people who are experiencing chronic pain, and you would randomly split some of them to your treatment group (those who get Adolor) and your control group (those who get a similar-looking pill that doesn’t contain any pain-treating ingredients.) 

Then, when it comes to analyzing the results, if the treatment group reports lower pain levels than the control group, then this suggests that Adolor likely does help treat chronic pain. You might continue testing Adolor, but your RCT could help demonstrate a connection between taking Adolor and experiencing less chronic pain. 

(To be clear, it wouldn’t be incontrovertible proof that Adolor “works” — proof is a very strong word in clinical research, since it’s technically impossible to prove something works all the time for everybody, so it’s more precise to say that Adolor, in our example, is correlated lower pain levels, and is thus likely to be involved in causing these lowered levels. Correlation is not always causation, however.)  Many treatments, including medications and injections, also sometimes work only for a certain type of pain (e.g. joint pain, nerve pain, etc). 

That’s the basic process. Even a “basic” RCT, however, is full of potential twists and turns. For example, placebos (the fake “treatments” you give the control group), do often result in patients reporting lower pain levels, even though placebos are “fake” medications. Why is this? The placebo effect, as it’s called, happens for a number of reasons, including: 

• Believing you’ve taken “something,” and that it must have some kind of effect (This is a kind of neurological priming. Since the brain expects something to happen after taking a pill, it may “decide” that it’s consequently experiencing lower levels of pain.)
• Taking “something” in context. Even if the “medication” control group patients take doesn’t contain active ingredients, the placebo is often still being given in a medical context, meaning in a clinic-like setting with the involvement of medical professionals, so the brain may conclude that a treatment of some sort has occurred.
• Paying attention to the fact that something has changed. Control group patients know they’ve added something new to their routine, and will thus likely be paying closer attention to how they’re feeling. This can lead to being very aware of what’s going on, including feeling lower pain levels, whether that’s due to actual improvement or to convincing yourself there is evidence you might be feeling better.  Sometimes placebo effects even occur for side effects, such that someone might feel nauseous or dizzy even though they did not get the “real” treatment; this is called a ‘nocebo effect’. 

This is not to say that placebos are “scammy” or that placebos prove pain “isn’t real.” Instead, researchers are coming to terms with how complicated and nuanced experiences of chronic pain actually are, and how powerful the mind, moods, expectations, stress and other factors can be when people analyze their exact pain levels. 

In fact, researchers have studied whether or not people report improvements in pain even when they know they’re receiving a placebo. (The researchers went so far as to label the pill bottle patients used ‘Placebo.’) In 2014, researchers reported that these patients, who had been fully informed the fake “medication” they were taking was not expected to have an effect, still said they felt less pain after taking it. Ultimately, the clearly marked placebo reduced pain about half as effectively as an actual pain medication. 

Again, this doesn’t mean that medication is a complete mess or a fraud. In fact, people and doctors can actually use the placebo effect to their advantage by taking nothing more than a vitamin, for example, if they take it in the context of expecting a decrease in pain levels. This is an example of how expectations, awareness and following a care ritual can reduce perceived pain levels, even without curing the original cause of the pain. 

It does, however, complicate RTCs. For example, it’s possible for a medication, say a pain cream, to have some effect on pain levels, but not as great an effect as the placebo effect. Does that make it an ineffective medication or one that should be disregarded? It’s difficult to say. The details of a study’s design and enrollment protocols can affect its outcomes deeply. Even the relationships between the patients and the researchers can affect the outcome, Dr. Cohen reminded us. 

Another twist is that you are not the average patient. You are you, with a unique genetic background and unique current circumstances. RCTs try to recruit a large number of patients to show the likelihood that a particular medication will work or not overall, meaning that it is likely to reduce pain at least somewhat for many people most of the time. Your specific situation may not fit this general scenario though, and so a drug that works for most people may not work for you (and the other way around: some treatments that may not overcome the placebo effect or otherwise demonstrate strong efficacy overall in a general patient population may, in fact, work for you.) 

Your genetics, stressors, support from friends and family, allergens, resiliency levels and even certain mental behaviors can influence whether or not a particular pain treatment, such as a compounded pain cream, will work for you. 

To sum it up, factors that influence an RCT and may affect the reliability of its reported outcome include:

• Its overall design and the design of its various sub-stages
• Whether or not a true control group can be used (the NIH does not always approve the use of control groups, depending on what type of medication is being tested, and some trials are halted partway through if a particular medication appears either very effective, ineffective, or too potentially harmful)
• Whether or not patients enrolled in RCTs truly represent “the average patient”
• Whether or not patients enrolled in RCTs represent a patient like you closely enough
• Funding levels and mechanisms
• The relationship (including likeability, trustworthiness and mutual respect) between doctors, researchers and patients involved in research
• Patients’ desires to “do well” in a study by reporting positive effects
• The placebo and nocebo effects
• Doctors’ desire to care for patients and to observe an effect
• Whether or not the RCT is truly “blinded” (meaning whether or not patients can truly be given a convincing enough “fake” medication in the control group. For example, it would be difficult to persuade people they are using a cream containing capsaicin, a pain reliever, if they did not detect its characteristic burning sensation)
• Researchers’ desires to publish results (it can be very discouraging for researchers to find “negative” or inconclusive results)
• Issues with an RCT’s statistical analyses
• Issues with academic journal editors’ ability to accurately analyze a study and determine whether or not it has been designed and run well 
• The possibility of confounding factors (for example, maybe patients in the Adolor treatment group were also using CBD, but weren’t reporting it)

To sum it up

In other words, even large, well-run RCTs can run into difficulties trying to conclusively demonstrate a medication’s efficacy. They’re the best method we have, clinically speaking, and they’re better than nothing by far, but it’s best to keep their limitations in mind when reading chronic pain research and looking into potential treatments that may work for you. 

When discussing pain creams, for example, Dr. Cohen pointed to how, overall, it can be tough to nail down whether or not particular ingredients work reliably and perform consistently above the placebo effect in the general population. Certain ingredient combinations administered to specific patients, however, may indeed achieve positive results. It’s very difficult to study every ingredient combination in every type of patient and condition, however (not to mention prohibitively expensive), so patients may want to remain well-educated and open to different potential treatments to help pinpoint what will be ideal for them. 

Again, does this mean medicine and clinical science aren’t trustworthy or are just too complicated to do well? Not exactly. By the time a treatment reaches the stage of development where it’s ready to be tested in an RCT, it’s already likely to have passed multiple other trials to demonstrate its basic safety and to help researchers better understand its potential benefits. 

We have to keep in mind that yes, research is very complicated, and that research is iterative. New ideas come along frequently, and researchers must constantly analyze ideas and hypotheses against what we are learning overall while remaining open to new approaches. As we conduct more research, we learn more, and sometimes end up changing our views and recommendations about certain treatments. The bottom line is that doctors try to prescribe the treatments they think have the best chance of working for the specific patients they’re seeing. The placebo effect or patient-provider effect, in these circumstances, isn’t always bad, and it can be helpful to adopt a bit of a research mindset, conducting experiments, so to speak, with yourself as a subject to see what fits best for you. Pain is a very nuanced process, and so treating pain can be very nuanced, too.