Tuesday, December 11, 2007

Blogging on Peer-Reviewed Research

I recently came across a paper entitled "A Case of a Potential Manipulation Responder Whose Back Pain Resolved with Flexion Exercises" by Stephen May, MSc and Richard Rosedal, PT, Dip MDT. They present a single patient case study of a LBP patient that met 4 of 5 of the manipulation CPR that drastically improved with flexion exercises but did not receive spinal manipulation.

The authors conclusion(s):

We have presented a case study that was positive for 4 of 5 items of the clinical prediction rule for manipulation responders, but this patient was successfully treated with flexion exercises. The clinical prediction rule may not represent a discrete subgroup but may include patients who can be defectively managed in other ways.
I think the authors are presenting an important point. CPR's must be, for the most part, discreet from other classification, or subgroups of patients. However, this single patient case study is very weak data to lead us to abandon the Manipulation CPR.

Now there is no escaping that this single patient got better. However, one could do a case study in which a person met 4 out 5 of the manipulation CPR and the DID NOT get better at all. That's why larger randomized studies are done and then replicated.

Speaking of replication. The authors suggest that the results of the CPR studies may not work in general as all studies, including the replication study, were done on US military personnel. Specifically -

..this is why once a CPR has been derived in one population it must be validated in a different patient population with different clinicians. The manipulation CPR ahs been so validated, but both derivation and validation studies were performed by specially trained clinicians on US military personnel in military facilities, and the applicability of the results to other patient groups and clinicians is unknown.

In truth, the validation study was done 8 clinics in different regions of the US where "Most participating sites were health care facilities within the U.S. Air Force". Therefore, like in my setting near an Air Force Base, not only military servicemen, but also their families and dependents are treated. Secondly, in the US at least, spinal manipulation is a core skill and intervention taught in physical therapy schools.

In my opinion, the authors true colors about the manipulation CPR are shown with this quote:

If this patient had been treated with manipulation by itself, as she appeared to fit the CPR criteria, this management may have failed to provide the patient with her own ability to control and abolish her symptoms when they returned.


Although it has been directly stated that CPRs are not meant to replace clinical judgment and should be used to complement clinical reasoning, it may also be argued that the use of CPRs minimizes the clinical reasoning process, reducing decision making to a "tick-box" activity to bypass more complex and high-level reasoning that is often required in clinical practice.

The first is a direct dogma held by many "McKenzie" therapist who usually proudly boast not needing to touch a patient to get them better. It's also a Red Herring. To imply that the use of manipulation precludes a competent PT from educating and providing the same means to "self control" their pain is almost pretentious.

The second passage is a classic "Guru" based argument. You can't be providing good results if it's easy or straight forward. One must complete expensive Con Ed after Con Ed to get additional letters after their name and finally portray to the patient the complexity of their pain.
I'm almost saddened as, I feel, McKenzie approach works very well specifically because of it's attempt to subgroup patients, which is what this CPR is striving for. Not because it has a complex model that only those with the money and time (not just access to a well done RCT) can achieve.

Finally, CPRs are used all over in medicine with great results. The Ottawa ankle rule, Canadian C-spine rules, Acute Heart failure index, Short term risk after after a PE, Deep Vein Thrombosis, etc. They are there to help us choose the most appropriate treatment for a patient, which leads to improved outcomes, lower costs, and increased patient satisfaction.

I encourage Mr.'s may and Rosedale to continue with their hypothesis and hopefully tease out what type a subgrouping is more appropriate. I know we've already got authors looking at a potential extension oriented subgroup of LBP patients.


Post a Comment