There is a new web standard for blogging that is gaining some legs: "Blogging for Peer-Reviewed Research Reporting" or BPR3 for short. The standard was developed by several well-known science bloggers and "strives to identify serious academic blog posts about peer-reviewed research by offering an icon and an aggregation site where others can look to find the best academic blogging on the Net." Indeed! Increased exposure to our posts sounds great. So, without further ado, here is our first official BPR3-based report:
A recent study published in the highly renowned Lancet journal reported on the effects of either "spinal manipulative therapy" or NSAIDS added to "first line care" for low back pain. Their conclusions:
"Patients with acute low back pain receiving recommended first-line care do not recover more quickly with the addition of diclofenac or spinal manipulative therapy."
The study examined this relationship using 4 treatment groups: spinal manipulative, placebo manipulative (sham US), diclofenac, and placebo diclofenac. The design was inherently solid, but let's look into such a strongly worded conclusion and see if it was warranted.
The outcome measure:
The primary outcome measure was days to recovery, defined as the first pain-free day, or the first of seven consecutive days in which pain was 0-1/10 on the Numeric Pain Rating Scale. I found this interesting, as most low-back pain literature does not stipulate that a 0/10 pain level be achieved prior to deeming one recovered. In fact, a recent article in the European Spine Journal suggested that success after spinal surgery is a variable thing indeed, and most often does not dictate absence of pain to deem the procedure a success. Perhaps the authors of the Lancet study are suggesting that spinal manipulation and NSAIDS be held to a higher standard than spinal surgery? Perhaps that is an unexpectedly noble suggestion, although I do not think that was the intent!
The subjects:
The population studied included all patients with low back pain of less than 6 weeks duration, who were then randomly allocated to the respective treatment groups. I'm sure right away your brains are screaming: "They didn't classify the patients???" No, they didn't. The allocation to treatment groups was completely random and stands in the face of the large body of evidence that exists supporting Treatment-based Classification for Low Back Pain. I was thinking hard for a witty comment on this, but in picking up the November PT Journal I found this had already been done for me. In Dr. K. Shepard's 38th Mary McMillan Lecture, she cited a letter to the editor to the NEJM written by physical therapist, Lynn Harding who was responding to a research report which similarly lacked the use of classification system when assigning patients. He wrote:
"A study that randomly assigns patients with low back pain to various conservative treatment protocols will produce the same results as a study that randomly assigns patients with abdominal pain to undergo appendectomy, cholecystectomy, or exploratory laparotomy. Neither study makes any sense."
Well said, but not apparently well-read by the reviewers of this present report.
The Treatment:
The definition of spinal manipulative therapy was poorly controlled. In the details of the results, we learn that:
"Most participants had several low-velocity mobilisation techniques (232/239, 97%) with a small proportion also having high-velocity thrust techniques (12/239, 5%)."
I can now question several more things: Why was spinal mobilization chosen over spinal manipulation by so many therapists, since literature supporting its use is clearly lacking? Certainly, this could not be the treatment that was described as "consistent with contemporary best clinical practice"? Secondly, there is a growing body of evidence that suggests spinal manipulation and spinal mobilization are two very different interventions with very different effects. Why were they grouped together? Finally, if both interventions were used, but one used dramatically more than the other, and the one that was used is poorly supported by research, then have we learned anything about either of those treatments?
The study also left out the concept of active treatments all together. There are some pretty solid guidelines supporting active treatments for this patient population, as well as good evidence which supports combining manual therapies with the active treatments. Yet the authors chose to study the manual techniques in isolation without providing us a good reason as to why.
The Summary:
Alas, all was not lost. If one looks carefully there is a nice line to use here. The diclofenac group reported 11 adverse reactions, while the spinal manipulative group reported zero adverse reactions. I guess I like that part of the study.
I continue to be perplexed how so many of these "magic bullet" studies make it to press. They operate under the guise of generalizability, but are poorly reflective of optimal clinical practice. There are various reasons for this, from reviewers unfamiliar with the research to the inherent difficulty of designing such RCT's that reflect the pragmatic and flexible nature of good clinical practice. Regardless of cause, when studies like this come out, there is a duty to suppress our "peace genes" and turn up the volume of the critique. Eventually, someone may hear us.
Stay tuned on this one...perhaps we can host a debate with one of the authors of the study.
ERIC
