Therefore, PCI-32765 acknowledging the differences in the definition of spinal manipulative therapy, our findings are consistent with the results of this review. The finding that those provided with Strain-Counterstrain treatment registered a significantly greater improvement in global rating of change at the end of the intervention period is unlikely to be clinically relevant because the difference between groups was only 0.5. Approximately 40% of individuals with acute low back pain are likely to recover rapidly without

intervention or with first-line intervention of simple analgesia and advice (Pengel et al 2003). This may be one reason for the small effects of additional treatments such as Strain-Counterstrain and other spinal manipulative therapies (Hancock et al 2008). This may also have clinical implications for provision of spinal manipulative therapy to patients with acute low back pain. For trials to demonstrate substantial effect sizes for acute low back pain treatments, it may be necessary to exclude individuals with a highly favourable prognosis regardless

of treatment (Stanton et al., 2008). Clinically, it would be reasonable to withhold relatively expensive treatments such as Strain-Counterstrain from these individuals while providing adequate analgesia and advice knowing that they are likely to recover quickly (Hancock et al 2008). Another consideration for sampling in studies of treatments for non-specific acute low back pain is that the condition is unlikely to be homogenous within a sample (Brennan et al 2006, Kent and Keating 2004). While all DNA Damage inhibitor Mephenoxalone participants in this

study had a minimum of 4 digitally tender points identified using Strain-Counterstrain procedures, this does not confirm that they were a homogenous sample and it is likely that the source of acute low back pain varied among the participants. A possible strategy to manage sample heterogeneity in future studies assessing Strain- Counterstrain treatment for acute low back pain would be to develop an algorithm, specifically for Strain-Counterstrain treatment, to identify individuals more likely to respond to this form of treatment. Such algorithms have previously been shown to improve outcomes for non-specific acute/subacute low back pain (Brennan et al 2006, Childs et al 2004). Personal clinical experience suggests that for such an algorithm, factors favouring Strain-Counterstrain treatment might include: recent and sudden onset of symptoms; no more than one previous episode of acute low back pain; more than 4 but less than 10 digitally tender points identified at anterior and posterior sites claimed to be associated with low back pain; pain localised to the lumbosacral region; and less than 45 years of age. Our findings should be considered within the context of the limitations of the study design.