The current issue of JMMT includes a very nice review of evidence supporting Spinal Manipulative Therapy (SMT) for low back pain. This interesting review is a solid piece of work, and useful in bringing to light the current picture of where we stand: namely that SMT should be used for low back pain, but questions still exist about the best way to use it.
The authors of the review have recently published some evidence that calls into question some of the recent guidance we have been getting in terms of clinical prediction rules for spinal manipulation. In one study published this year, they conclude:
"The clinical prediction rule proposed by Childs et al. did not
generalize to patients presenting to primary care with acute low back
pain who received a course of spinal manipulative therapy."
I was interested in speaking with Dr.Childs and getting his opinion about the recent review and what he considers the current big picture in SMT. He was kind enough to oblige my questions!
NPATT: The authors of this review have suggested a return to a paradigm focused on making a specific patho-anatomical diagnosis for low back pain patients. This seems in contrast to the treatment-based classification approach where treatment decisions are made according to patient presentation, not a patho-anatomical cause. What does history, and current evidence tell us about finding a specific diagnosis for low back pain?
John Childs: Pathoanatomy is only relevant for guiding treatment decisions in roughly 10-15% of cases of LBP. Even in these cases, the pathoanatomy may be serious (ie, cancer, fracture, etc) so the percentage of cases in which pathoanatomy is relevant for guiding physical therapy treatment decision-making is even less. This is the reason why concepts of “treatment-based classification” and “subgrouping” patients with LBP based on clinical examination findings have become such an important research priority over the last 10 years. The elusive search for the pathoanatomic diagnosis and “magic bullet” treatment lies at the root cause for the disaster of LBP management in the U.S. and a re-focus around identifying pathoanatomy would be a big step backward rather than forward.
NPATT: The authors speak at length about the generalizability your CPR validation. Can you respond breifly on the issue?
Childs: The manipulation CPR as developed by Flynn and validated by Childs can only be generalized to similar patients that were included in these 2 initial studies and only using similar treatments. In both the Flynn and Childs studies, the SMT intervention was standardized and limited to a single high velocity thrust technique. Childs et al also included a comprehensive exercise strengthening program to the intervention. In contrast, Hancock et al allowed therapists wide latitude in which manual therapy techniques to use. The large majority of therapists (97%) elected to use lower velocity mobilization techniques and the 4-week intervention did not include an active exercise component (strengthening or otherwise), thus it’s difficult to compare the 2 studies. It is not surprising at all that Hancock did not find any differences in outcome based on whether patients fit the manipulation prediction rule because they tested an altogether different treatment approach. Their study does provide strong evidence that lower velocity mobilization procedures, in the absence of an active exercise component, is likely ineffective for patients with LBP.
NPATT: Along the same lines, the authors suggest focusing SMT treatment on specific painful segments. I've also seen evidence arguing that SMT is not, by nature, able to address specific segments and is a more global approach. Couple this with a proposed neurophysiologic mechanism, and I'm not sure targeting a specific segment is realistic or necessary. What is your opinion on this?
Childs: There is no evidence to suggest that SMT directed to a specific segment can be done in a reliable and valid way. Even presuming it could be done, there is no data to suggest that a “specific” approach is more clinically effective than a general approach. Many studies have demonstrated that the effects of SMT are likely occurring above and below the targeted segments. Much research is now being directed at understanding the mechanism through which SMT acts to improve pain and function. It seems that selection of the right patient is a more important priority than which technique is used. There is data now in the lumbar spine to suggest that SMT incorporating high velocity thrust manipulation is more effective than mobilization in the subgroup of patients who fit the manipulation rule.
NPATT: What's the big picture here for SMT and low back pain?
Childs: Pick the right patient and use SMT frequently in conjunction with an active exercise strengthening program. Don’t lose sleep if the models of “diagnosing” presumed biomechanical dysfunctions confuse you or don’t make sense. In all likelihood, they are mostly invalid and not useful for decision-making anyways. The key is being able to match patients to the right treatment based on key clinical examination variables associated with a successful outcome from a particular standardized treatment approach.
Thanks for the great conversation, John! This is an important debate and I'm eager to see where it ultimately leads us.