This study reveals that the CB-CFT group demonstrated superior outcomes compared with the MT-EX group across every domain measured at post-intervention and at 12-month follow-up (). Both groups showed significant improvement in short- and long-term follow-ups; however, the CB-CFT group was superior based on clinically meaningful changes as defined by MIC, defined as >10-point change in ODI and >1.5 on PINRS (Ostelo et al., 2008
The effect sizes of conservative interventions from previous Cochrane reviews reveal findings similar to the MT-EX group (Assendelft et al., 2004
). In response, calls have been made for a paradigm shift, away from a biomedical ‘injury’ model, to viewing LBP as a multifactorial biopsychosocial disorder, and directing treatment at beliefs and behaviours that promote pain and disability rather than simply at the signs and symptoms associated with the disorder. Calls have also been made for the need for a multidimensional classification-based approach to direct management of NSCLBP in order to make treatment more person-centred (Borkan et al., 2002
; O’Sullivan, 2011
). This is supported by reports that disability levels in chronic pain are better predicted by cognitive and behavioural aspects of pain rather than sensory and biomedical ones (Campell and Edwards, 2009
). CB-CFT addresses all of these objectives.
Although satisfaction rates were high in both groups, odds of being completely satisfied were over three times higher in the CB-CFT group at 3 months and five times higher at 12 months. The degree of patient satisfaction is seen as a reflection of the quality of care and as an important outcome in its own right (May, 2001
). The importance of communication, patient-centred approaches and goal matching has been well documented in the literature as important for the therapeutic relationship, enhancing compliance and patient outcomes (Linton, 2005
; O’Sullivan, 2011
CB-CFT had a strong cognitive focus with an emphasis on reframing the persons’ understanding of their back pain in a person-centred manner, with an emphasis on changing maladaptive movement, cognitive and lifestyle behaviours contributing to their vicious cycle of pain. This was performed by means of reflective communication, providing a contemporary understanding of pain mechanisms, correcting faulty pathoanatomical beliefs, goal setting, verbal, written and visual feedback (viewing their own back) and a strong emphasis on normalizing movement behaviours within a graded functional approach.
Although this is the first RCT to address maladaptive movement behaviours specific to the patient’s presentation within a cognitive framework, the exact benefits from targeting specific movement training cannot be isolated from the other aspects of the intervention. The behaviours that were targeted were prioritized based on the movements or postures that patients reported that they most feared, avoided and/or that provoked them. These identified movements were the targets for the movement retraining aspect of the intervention based on the patient’s classification and were integrated to the goals of the patient. The use of visual feedback such as mirrors was central to this process. The aim of this approach was for each subject to acquire self-management strategies for their disorder by developing positive back pain beliefs, pain control, developing adaptive strategies of movement that enhanced functional capacity and the ability to engage in regular physical activity.
These findings are supported by previous reports of benefits with different targeted behavioural approaches to managing NSCLBP. Moseley et al. (2004)
reported reduced pain and enhanced function associated with pain education. Asenlof et al. (2009)
reported superior long-term outcomes for treating NSCLBP with an individually tailored behavioural intervention targeting cognitions, motor behaviour and activity, compared with physical therapy. The use of visual feedback when training movement in patients with LBP has also been shown to reduce pain and influence functional capacity (Sheeran et al., 2012
; Wand et al., 2012
). Hill et al. (2011)
reported superior outcomes when management was targeted on the basis of psychosocial risk factors. Vlaeyen et al. (2002)
reported benefits with a graded exposure approach to management in a series of NSCLBP patients with high levels of fear. CB-CFT incorporates all of these aspects within its intervention.
Given the multidimensional nature of the CB-CFT intervention, it is not clear as to the exact basis for the superior outcomes. We hypothesize that the mechanisms for change are likely to be multifactorial given the patient-centred body–mind behavioural approach, in contrast to a more treatment-orientated signs and symptoms approach in manual therapy. On one hand, this behavioural approach may have impacted on cognitive factors known to affect pain sensitivity and disability such as developing positive beliefs, reduced fear, increased awareness, enhanced understanding and control of pain, adaptive coping, enhanced self-efficacy, confidence and improved mood. Evidence for this is supported by the reduction in fear of movement and improved mood observed following the intervention. On the other hand, the functional behavioural aspects of the intervention were targeted at enhancing body awareness, relaxation of guarded muscles, normalizing maladaptive movement patterns, body schema retraining with the use of mirror feedback, extinguishing pain behaviours, conditioning and increased functional capacity. These factors have been associated with levels of pain, disability, fear and catastrophizing (Wideman et al., 2009
; Lewis et al., 2012
; Wand et al., 2012
). We also acknowledge that the active engagement required of subjects for this behavioural approach may present a barrier for those unwilling to self-manage their disorder (Carr et al., 2006
). This may be dependent on the levels of acceptance and readiness to engage in behavioural change, although these factors were not formally assessed.
Although it was not a primary aim of the CB-CFT intervention, the results demonstrate a 2.95-times less likelihood of taking sick leave for their LBP at 12 months compared with the MT-EX group. Previously, only studies using cognitive behavioural therapy in multidisciplinary treatment models have shown an effect on sick-listing for this patient group (Airaksinen et al., 2006
). However, as these numbers were extracted from the OMPQ and were self-reported, these findings must be interpreted with some caution. The patients in the CB-CFT group also sought less additional treatment for their pain, implying they may have been more empowered to self-manage their disorder, suggesting significant cost–benefits.
A limitation of this study was the number of patients that either did not start or complete treatment. While there was a comparable proportion of non-completers in each group (), 8 of 59 (13.5%) of the MT-EX group failed to commence their allocated treatment, compared with only 1 of 62 (1.6%) of the CB-CFT group. This may be due to the fact that seven of these subjects had reported previous manual therapy treatment with poor effect, which would have potentially biased for a poorer outcome in the MT-EX group. It should be noted that there was no statistically significant difference between completers and non-completer based on baseline characteristics, and as we performed our analysis conditioning on baseline scores and confirmed the absence of a confounding effect of age, gender, BMI, LBP duration and work status on our results, we are confident that our estimates of treatment effect are not substantially biased by these missing cases. Also of note, no dropouts reported adverse effects from either intervention arm. Furthermore, due to a lack of power, we were unable to determine the influence of the subject classification on the outcome.
We acknowledge that there are also a number of additional methodological considerations that may have influenced the results of this study. Firstly, the patients were recruited from a variety of sources both primary and secondary care levels, as well as newspaper advertising, which could have influenced the kind of patients who entered the study. However, the wide inclusion criteria in the study suggest that a common and representative group of patients with chronic localized LBP without objective sign of pathology to the spine were included. Webb et al. (2003)
reported that an Oswestry score of >25% is considered the cut-off score for classifying ‘disabling back pain’. The patients recruited were, on average, just below this [24.0 (MT-EX) and 21.3 (CB-CFT)] at baseline, and hence it can be said that the patients sampled in this study had moderate back pain and functional impairment sufficient to result in activity limitation and sick leave for many (see ). It is also acknowledged that therapists in both arms of the study were not blinded to the intervention, and although all therapists had considerable experience, the influence of therapist enthusiasm and expectation for change was not controlled for.
Furthermore, the multidimensional nature of the study limits any conclusion as to the specific effects of the different components of the intervention. Future research that investigates matching versus non-matching of interventions for patients with chronic mechanical LBP may help identify the effects of specific aspects of the intervention (Kent et al., 2010
). Also given this was a pragmatic trial, the intervention dose was not controlled in either group, although both groups received remarkably similar attention. While this intervention appears to be successful for the population we tested, further studies are needed to investigate those with higher levels of pain and disability, patients that are long-term sick-listed as well as in other cultural and occupational groups, in order to determine the generalizability of the findings. This approach also needs to be compared with other cognitive approaches and tested within a multi-centre trial framework in the future.
The results of this study support that a behaviourally orientated targeted approach to manage NSCLBP (CB-CFT) was more effective at reducing pain, disability, fear beliefs, mood and sick leave at long-term follow-up than MT-EX.