Welcome back to the Healing Pain Podcast with Eric Sun, MD
We’ve got an episode that is full of some good news that I can’t wait to share with you. When I first started this podcast over a few years ago, the CDC had released information about the opioid crisis as well as the overprescribing of opioids and its contribution to addiction, abuse and fatalities. Since then, the CDC guidelines have been pushing toward early non-drug or what is known as non-pharmacologic treatment options for patients with pain. We are definitely making some progress. Along with these guidelines, there’s a huge opportunity for physical therapists to help people who are struggling with chronic pain. One way we can do this is to intervene quickly and early in the care of people with pain. I spoke about this a bit in Episode 133 with Dr. Marcia Spoto, who discussed physical therapists as primary spine care providers.
In this episode, we’ll explore how the research is supporting early physical therapy for people with pain. A study conducted by our guest, Dr. Eric Sun, found that patients who receive early physical therapy within the first 90 days use less opioids for pain. Eric is an Assistant Professor of Anesthesiology, Pain Medicine, and Health and Research Policy at Stanford University and Medical Center in California. He spends part of his time working as an anesthesiologist and part of his time as an epidemiologist, crunching numbers and big data on important topics. It was published in 2018 in the Journal of the American Medical Association. I know you’re going to enjoy this show. Make sure to hop on over to the IntegrativePainScienceInstitute.com/podcasts. You’ll see a box where you can subscribe to the Healing Pain Podcast so I can send you a new podcast to your inbox. Let’s begin with Dr. Eric Sun and learn about how early physical therapy can lead to less opioid use.
Watch the episode here:
Is Early Physical Therapy Associated With Less Opioid Use By Patients With Musculoskeletal Pain? with Eric Sun, MD
Early Physical Therapy Plays An Important Role In Reducing The Risk Of Transitioning To Chronic Opioid Use For Patients With Shoulder, Neck, Knee, And Low Back Pain.
Eric, welcome to the podcast. It’s great to have you here.
Thanks for having me on.
You did a great paperback in 2018. It was published in the Journal of American Medical Association. It’s called the Association of Early Physical Therapy with Long-term Opioid Use Among Opioid-Naive Patients with Musculoskeletal Pain. I want to have you on because it’s an important paper talking about the topic of opioids and physical therapy. Tell us first a little bit about your practice and what you do. You wear two different hats.
I do wear two different hats. I trained as both an economist and as a physician. I spend a little bit of my time taking care of patients. One day a week I take care of patients in the operating rooms. I’m not a pain physician. I take it on an operating anesthesiologist. I do that about one day a week and the rest of the time I’m doing epidemiology and economics research mostly centered around pain and opioids.
I would think your training in economics is coming in handy with regards to the opioid epidemic and is informing a lot of the research that you’re doing.
I’m editing a lot of the stuff I’d bring from economics first of all just knowledge of these sorts of datasets. I know they exist. I know how to use them. A lot of those statistical tools when it comes, I try to bring in. That’s part of it. Those have been very helpful.
Tell us about what the aim of this particular study was.
The goal of this study was to see if early physical therapy could reduce opioid use longer-term for patients who developed musculoskeletal pain in the areas specified, knee, back, shoulder and neck. Basically, most people know there’s a concern about opioid use in the US and the opioid crisis. Our question was for people with newly developed musculoskeletal pain. Is early physical therapy a potential way of addressing this crisis? Can it reduce the risk for people who will go on to longer-term opioid use?
You had some great collaborators on this as well. We’ve had some of them on our podcasts. Are some physical therapists on the papers as well?
That’s correct. The physical therapist, the senior author is Steven George. Chad Cook and Adam Goode who worked with Stephen as well.
They’re great PTs that do research into pain. Tell us what type of study this was that use the term opioid-naive in the title, which is interesting. Explain what that is and how that helped inform this study.
The reason why we focus on opioid-naive patients, first of all, it’s a lot of the people who get musculoskeletal pain. It’s easy to make a statement in that case. Some of the patients in this study were defined as people who didn’t use any opioids in the year before. They had their musculoskeletal pain. They had not been exposed to opioids. Someone like that has after the opioid pain, it’s after they get the pain on opioids long-term you can say that’s definitely changed because before they weren’t on opioids. Now they are. That’s a real change. We are working on looking at people who weren’t opioid-naive.
Some of those people who were taking opioids for whatever reason coming in the year before, their pain, before the musculoskeletal pain occurred. It’s a little more challenging in that case because of what’s changed. If you were using the average ten MMEs before this musculoskeletal pain and now using twelve, is there a change? If not using, A, is there a change? It makes that a little bit harder. You have to think about the changes. I think about this study and the reason why we chose it is because here it’s pretty clear. You weren’t using them before and now you are. That’s changed.
It’s a large study. How many people did you look at in this particular study?
It’s around 80,000 people, if I recall correctly or 90,000 people. Basically, it’s a retrospective study. It’s based around insurance claims data. That’s how they identify all these things. When a healthcare provider provides care, they have to fill out a claim, submit it to the insurance company in order for everyone to get paid. That will typically include things like diagnosis codes. That’s how we can identify that you came to a physician with a complaint of musculoskeletal pain. Crucially, in the case of drugs, when you fill it out, obviously in time you go to the pharmacy if the pharmacy is submitting a claim. The nice things about these data census, a lot of people have private insurance so they can be quite large samples. It’s large, around 90,000 patients.
In this study, you looked at single-site pain. First, why single-site pain versus potentially someone with fibromyalgia that has more global multisite pain? What were the sites you also looked at in this study?YES! Early physical therapy prevents long-term opioid use patients with musculoskeletal pain! Click To Tweet
For this study, we looked at the neck, back, knee and shoulder. We chose those because that’s where most pain occurs. When it comes to multisite pain, our thought was it’s a little bit easier to talk about each one individually. These are things we want to work on down the line, but some issues you might have with multisite pain. If a person comes in and says they have knee pain and tomorrow they have back pain or let’s say they come in next year, they say they have back pain, how do you characterize all of that in terms of what’s going on? It’s a little bit harder. Where it’s clean in some sense, a guy came in and says he’s knee pain. That’s the only pain he says he’s had from the next few months. A lot of us don’t make things analytically simple. It’s definitely multisite pain because you mentioned people who used to opiate this beforehand, we’ll treat them, look in the studies or different things we’re working on.
Are these patients that either showed up at the emergency room or potentially their primary care provider that had pain?
Yes, someone who came in and said, “I’ve got pain.” That’s what we were looking at.
How was early physical therapy defined because as most PTs know, sometimes it can take a while before a patient shows up to physical therapy from the ER, from seeing their primary care physician? How would you define that early physical therapy?
The baseline definition I believe was early physical therapy within 90 days. In other words, from the day you came in and saw the doctor, did you get physical therapy within 90 days of that? We also did look at within 30 days. The results are generally fair with some of them.
I’d like to see everyone in 30 days or before. Tell us about some of the results that you came across. I know there are a couple of different interesting findings in the paper from the research.
We asked ourselves two questions. What we do is you have this episode of musculoskeletal pain. What happens if we ignore the first 90 days after that because that’s short-term? What happens in the subsequent to that? If you will, post-operative event days 90 to 365. We said, “Did you fill an opioid prescription during that time? If you did, how much did you fill?” What we wanted to see is people with who got early PT. PT within those first 90 days. Did it reduce the opioid use later on? In particular, did it increase the probability that you would feel no opioid long-term? If he did feel opioid, did you use less? I would say more or less the finding of the studies that we found effects, not for all the pain categories but most of them. What we found is basically for all the pain sites, early physical therapy was associated with say a 10% increase in the probability of remaining opioid free long-term. That’s what we found. For most of the pain sites, for people who did use opioids long-term, we found that it was associated with a 10% reduction. We found that physical therapy was associated with an overall reduction in opioid use long-term.
The one side of those four groups I looked at the neck, back, shoulder and knee. The one group that didn’t show such great results was the neck category. You talked about things like whiplash doesn’t necessarily respond well to physical therapy and other interventions. That was one group that didn’t do so well, but the rest of them, there’s good evidence that early physical therapy can help with opioid use long-term.
That’s what we found. The bottom line is that it does seem to be associated with using less opioid in the longer-term, which is what most of us would want.
How does this study help inform the application of the American College of Physicians or the Centers for Disease Control with regard to prevention guidelines in the management of pain?
It provides more evidence for the use of non-opioid. Both guidelines, here we’re looking at it’s not an opioid, it’s not a drug either. What we’re saying is maybe we’re providing a little bit of evidence that these things are helpful.
What’s the feedback you’ve gotten? From PTs, I’m certain it’s positive. What about from maybe the medical community and the prescribers who were following your work?
In general, it’s been fairly positive. In general, this is an issue that concerns a lot of people. For the most part, physical therapy is certainly fairly benign as things go. It’s not so much downside. From a lot of people, there’s evidence of some upside, little downside. This should be something that we do more of, the general feedback.
You mentioned you have a follow-up study to this or a similar study?
We’re trying to extend the study along the two dimensions you mentioned. One is looking at multisite pain. That was probably a more challenging one that we have to think about how would we set that up analytically. We’re also wanting to look at who we’re using opioid beforehand, looking at those people. Those are the two follow on studies to that.
I know multisite pain is different. There are different challenges to both the care of the patient as well as research with regard to that. From a research perspective, what advice do you have to other researchers with regards to looking at the epidemiology of things like chronic pain and opioid use because it can be quite complicated?
The key can be quite complicated. There are so many ways to study it. Data sets like the ones I use are the advantages, lots of people and good follow up because most of the people who are insured, I can see everything they do. The downside of these data sets is always that there are a lot of things I don’t see. They’re probably important. I don’t see pain scores. I know diagnosis codes, but sometimes those don’t capture the full story. I know that you got an X-ray, but I don’t know what it said, for example. The challenge is to find ways to integrate in my mind, the more granular data that you might find in registries or within your own EMR. How can you integrate that with this data? They both in my mind offers different facets of the issue. Registry data, EMR data had this nice granular data, but sometimes the follow-up is not so good. Those patients might not always come back to you. They may stop answering the calls. It’s a registry. Finding some way to integrate the two would be helpful down the line because that will evaluate look at things more granularly, followed patients over long periods of time. That’s where one challenges and hopefully some that we can address over time.
In your study, you had mentioned that there have not been a lot of studies like this performed and the ones that haven’t been formed, it’s equivocal. We don’t have a clear idea yet.
For a lot of reason, this hasn’t been looked at as much. We’re glad that it made some contribution. For whatever reason prior to us doing it wasn’t been looked at as much.
Where would you like to see the research go on? Does the topic of early physical therapy interest you particularly?
There are so many things that we are trying to do to address the issue of transitions to chronic overuse. I’m interested in more broadly the entire material that exists that we think may or may not help. I’m interested in looking at whether it does help, what’s the data out there? In many cases, particularly comes to longer-term outcomes, those are hard to do. A clinical trial with one-year follow-up, that’s hard to do. There’s far less evidence there. Basically evaluate them. What works, what doesn’t, that’s the one I’m interested in.
These data sets can be harder for us in the US because we don’t have a national healthcare system. The 80,000 people in your study is a large amount. We didn’t have that in our country to look at things on a large, large scale.
CMS, for example, is very good. You can get their data as well, so that’s helpful. That’s data from Medicaid patients. It despaired in and segregated like that. That’s out there. The real tough one especially when I was looking at substance use disorder and things like that, it’s going to be people who don’t have insurance. Those are the hardest people to track in these data sets because they don’t have insurance. All these data sets are based around that. That’s one of the challenges and that’s probably one group that unfortunately, at least right now the data we have is hard to do research.
Should we start to look at this data as looking at back pain, knee pain, shoulder pain? Would that be helpful? I know you had four different groups in this study.
I talked to my colleagues who know this better clinically because they know what’s interesting in a way. The thing about these sites, especially in back pain is most pain is going to be these four things, particularly back pain. That’s at least why we chose that for this paper. Maybe this is for the count of those four sites, that’s probably most pain. It’s interest in doing a study.
Nicholas Karayannis, who works at Stanford has done a similar paper on back pain. He’s zoned in on back pain specifically, but yours is more global. I was like, “That’s cool because we have more joints that we can talk about.”
Imagine this data is that because this is big and that’s what we chose for common ones. The nice thing is if you were to take more rare pain diagnoses, these days it’s big that you can still have again. It wouldn’t be practical for some nerve pain diagnoses to do a clinical trial. You’re going to find those people, let alone follow them for X number of years. These data sets are all that you need to do that. This is big. You look across millions of Americans, you’ll find even rare things. You still get the numbers right. That’s one of the nice things about these data sets.
I’d love to share some of this work with insurance companies to see if it changes.
We haven’t written on whether it saves money. That’d be the interesting thing because what’s the ROI. There’s the upfront cost of physical therapy. That would be an interesting area to study. It’s hard to say how it turns out because physical therapy costs money. We find opioid reductions, but opioid drugs are cheap. It’s not a direct effect, but I probably thought I looked at was this cost of a physical therapist for the savings and opioids. It’s not saving money in that sense. You’d have to think more globally about are you reducing ER visits. Those things are going to be expensive. That’s definitely a good point in sending some would be down the line.
Addiction, mental health issues that come up with that you can look at your spots to put that together and say, “Ten to fifteen visits of physical therapy could be cheaper than opioids, plus all the other care that goes along with it.”
That’s important because as you know physical therapy, the coverage is not always the best. Wide variation across insurers as to what they will cover for alimony and for how long trial. I have policy relevance.
In your study, did you mention how many average visits someone received for physical therapy?
I did. We did look at the number of visits. I know we did look at it. For shoulder, it looks like it was six sessions in the first 90 days. Mostly looking around five to six sessions for the people who did get physical therapy.
They’re coming twice a week. They’re getting about three weeks of physical therapy, which is interesting because when you think of a chronic patient that may be the tip of the iceberg for them. They may need to obviously come for a longer period of time and maybe with longer physical therapy, those MME maybe even less.
Limitations of the data we have are, first of all, all of these patients said it’s not a new diagnosis. In that sense, it was acute. You could say we didn’t follow them long enough to make it chronic, follow them for a year. It doesn’t ultimately become chronic, but it’s going to be the case that somebody on most people, it’s an acute pain of some sort.
Eric, I love this study. It’s in the Journal of the American Medical Association, Association of Early Physical Therapy for Long-term Opioid Use Among Opioid-Naive Patients With Musculoskeletal Pain 2018. We’ll stay connected and follow your other work. If you’re doing this work, it interests many people with pain, physical therapists, psychologists who are interested in figuring out how we can use cognitive-behavioral strategies to help people with pain. How can we learn more about you and stay connected to your research?
You can go probably go to my Stanford website.
Are you part of the Stanford Pain Management Team?
I am. I don’t know if my website’s with them. My website is Profiles.Stanford.edu/Eric-Sun. That’s where I list all my papers and that’s probably where they catch up with the stuff I’m doing. I don’t do blogs or Twitter and maybe I should.
It’s great work. We appreciate you doing some work looking at physical therapy and opioids. If you’re a physical therapist or you’re someone interested in curbing the use of opioids, make sure you share this podcast out with your friends and family. We’ll see you next time.
- Episode 133 – previous episode
- Dr. Eric Sun
- Association of Early Physical Therapy with Long-Term Opioid Use Among Opioid-Naive Patients with Musculoskeletal Pain
About Dr. Erin Sun, MD
Eric Sun is an assistant professor of anesthesiology, perioperative, and pain medicine (adult-MSD) and, by courtesy, of health research and policy (health services research) at the Stanford University Medical Center.
The Healing Pain Podcast brings together top minds from the world of pain science and related fields to discuss the latest findings and share effective solutions for persistent pain.
If you would like to appear as an expert speaker in an episode of The Healing Pain Podcast contact firstname.lastname@example.org.
Love the show? Subscribe, rate, review, and share!
Join the Healing Pain Podcast Community today: