Julie Nowak wrote:
> I work currently as a french physiotherapist in a pain center to
> conduct a study on chronic neck pain. We had already made a draft
> study on this subject but the lack of patient was not possible to
> draw reliable conclusions. So we begin the study and hoped to have
> the advice of your organization to conduct a study reliable and
> relevant.
> This study focuses on self-management of patients with chronic neck
> pain. Patients were divided into two groups, a control group of
> liberal conventional physiotherapy for 3 months and the group with
> physiotherapy liberal more self-learning exercises on the neck. An
> initial assessment is conducted by a physiotherapist after inclusion
> in the protocol by the medical center pain. We test the amplitude,
> strength, kinesthetic , pain, muscle tension, etc.. Then we teach to
> the patient self-exercises like active mobilization, stretching,
> muscle strengthening, and friction massage to relieve pain.
>
> After 3 months the groups are reversed, so as not to delease a group
> and see if it is better to do self-exercise after liberal physical
> therapy or at the same time. The study will be made during 1 year.
>
> Two problems are presents with this Protocol:
>
> The conventional physical therapy is vast and varied, many methods
> exist and it seems difficult to compare different physiotherapists
> liberals among them, each method is unique. I don't know how to do to
> have homogen liberal traitment.
>
> The 2nd question is in the choice of stretching, in effect giving
> stretch identical to each patient can not seem to be the most
> effective method, but adjust stretch learning to each patient makes
> it difficult to compare results.
Many medical interventions rely strongly on the skill and experience of
the practitioner. Surgery trials are an excellent example. This
introduces some heterogeneity into the sampling process. Heterogeneity
adds "noise" to the data but if you are lucky it should not produce any
biases. So I would encourage you to select as large a sample size as you
can, report the heterogeneity as a limitation of the study, and hope for
the best. Any attempt to excessively standardize either the control
group or the treatment group will produce an artificial result that does
not represent how medicine is practiced in the real world.
You can view this as a trade-off between internal validity and external
validity. A rigorously defined intervention and control group makes the
statistical comparison good (internally validity) but makes it harder to
extrapolate to a real world setting (external validity). As a general
rule, studies of new and emerging interventions should stress internal
validity (proof of concept trials) and studies of interventions that are
already well established should stress external validity (pragmatic
trials). I suspect that you are comparing interventions that are already
well entrenched, but that is pure speculation on my part.
Bonne chance avec votre recherche.
--
Steve Simon, Standard Disclaimer
"The first three steps in a descriptive
data analysis, with examples in PASW/SPSS"
Thursday, January 21, 2010, 11am-noon, CST.
Details at www.pmean.com/webinars
|