Osteopathic Manipulation for Low Back Pain May Work... A Bit.

But probably not enough to matter.

Welcome to Impact Factor, your weekly dose of commentary on a new medical study. I’m Dr. F. Perry Wilson of the Yale School of Medicine.

Low back pain.

If you have it, you know what a burden it can be. If you have patients with it, you know how frustrating it can be to try to treat. The laundry list of therapies is extensive: NSAIDs, physical therapy, muscle relaxants, and of course opioids are all frequently trotted out with limited success. It’s no surprise then that some patients turn to osteopathic manipulation to find relief.

Now, osteopathic manipulation is not chiropracty – we’re not talking back cracking here. According to the American Osteopathic Association, osteopathic manipulation is “moving a patient’s muscles and joints using techniques that include stretching, gentle pressure, and resistance.”

On my completely arbitrary scale of biologic plausibility, ranging from 0 (homeopathy) to 10 (Vitamin C to Prevent Scurvy), I’d put OMT at around a 4.

Could stretching and moving peoples joints and muscles help with low back pain?  Sure. But, and I think I’ll have this on my gravestone – biological plausibility is the start of medical research, not the end.

So I was delighted to see this paper, appearing in JAMA Internal Medicine which took 400 patients with low back pain and randomized them to osteopathic manipulation versus sham therapy.

There’s lots to pull apart here, but briefly this is a single-center study in France. Patients had been through the gauntlet of prior therapies as you can see here.

The primary outcome was the change in score on the Quebec Back Pain Disability Index at 3 months. This is a survey which scores the effect of back pain on your life from a scale of 0 – no limitations, to 100 – maximal limitations. Prior studies have established that the minimum clinically important change on this scale – the amount that means the patient is really feeling better – is 20. 

The QBPDI decreased by 4.7 points in the OMT group, 1.3 points in the Sham group – a significant difference.  But in terms of the percent of patients who hit that 20 point improvement bar, not much emerged – just 10.4% in the active group versus 7.6% in the sham group hit that mark – a non-signficant difference. In fact, the majority of patients in both groups had no change or a worsening in their QBPDI score.

In other words, this study has something for everyone. Practitioners of osteopathic manipulation can point to the significant difference in the groups as evidence their therapy works. Skeptics can point to the numbers who achieved real relief and say there’s no meaningful effect.

I tend to fall in the middle on these things.  If I did a trial of a new blood pressure drug, for example, and found that it significantly lowered blood pressure – but by just 5 mmhg, I might not rush to prescribe it – but I’d still be interested in why it worked – particularly if it’s a new mechanism of action or something.

So – why did OMT do better than sham?  One possibility? It really works – moving muscles in a special way relieves back pain.  It’s biologically plausible, right?

But there are other explanations for the observed effect. First, let’s talk about the sham.  How do you do “sham” manipulation?  The same providers who did the OMT did the sham, and provided “light touch” to the various areas that would have been manipulated in the active arm.

This acknowledges the fact that simple human touch, even without any particular manipulation involved, might make you feel better. Is it a good sham?  My metric for that is whether a reasonable person could figure out if they are getting the sham or the real procedure.  The best acupuncture shams, for example, involve sticking real needles around the body – just not in the particular points dictated by that practice.

The purpose of the sham is to balance out the placebo effect, but this is lost if people know they are getting placebo. You can measure this extremely easily.  You just ask the patients whether they think they got the real thing or the placebo. It’s unclear if the researchers took this step in this study – it was not reported in the data I saw.

Of course, there are other ways to break the blind – the practitioner (who was necessarily not blinded) could give it away. Kudos to the researchers for recording a subset of these appointments and having them professionally adjudicated. There were differences here.

Adjudicators found the “listening quality” and “reassurance” was higher in the OMT group compared to sham, for example.  These things can clue participants into their treatment assignment.

So the study, while nicely executed, still leaves us a bit unsure. If OMT works, it doesn’t work that well – at least compared to placebo. But there’s a difference between scientific and clinical practice. To the patient, there is no such this as a placebo effect – they just feel better. But to make the right choice, we need a different trial – one comparing OMT to an active comparator – like physical therapy – or even NSAIDs. Please not opioids. I’m more than ready to cross that treatment off the low back pain laundry list.

A version of this commentary first appeared on medscape.com.