Abstract

Objectives . Interventions that take psychosocial factors into account are recommended for patients with persistent back or neck pain. We compared the effectiveness of a brief physiotherapy pain management approach using cognitive–behavioural principles (Solution-Finding Approach—SFA) with a commonly used method of physical therapy (McKenzie Approach—McK).

Methods . Eligible patients referred by GPs to physiotherapy departments with neck or back pain lasting at least 2 weeks were randomized to McK ( n = 161) or to SFA ( n = 154). They were further randomized to receive an educational booklet or not. The primary outcome was the Tampa Scale of Kinesiophobia (TSK) (Activity-Avoidance scale used as a proxy for coping) at 6 weeks, and 6 and 12 months.

Results . Of 649 patients assessed for eligibility, 315 were recruited (219 with back pain, 96 with neck pain). There were no statistically significant differences in outcomes between the groups, except that at any time point SFA patients supported by a booklet reported less reliance on health professionals (Multidimensional Health Locus of Control Powerful Others Scale), while at 6 months McK patients showed slightly more improvement on activity-avoidance (TSK). At 6 weeks, patient satisfaction was greater for McK (median 90% compared with 70% for SFA). Both interventions resulted in modest but clinically important improvements over time on the Roland Disability Questionnaire Scores and Northwick Park Neck Pain Scores.

Conclusions . The McK approach resulted in higher patient satisfaction overall but the SFA could be more cost-effective, as fewer (three vs four) sessions were needed.

Introduction

Musculoskeletal spinal pain is very common and its related disability results in large costs to individuals, the health service and society at large [ 1 , 2 ].

Accumulated evidence has shown that patients with back pain or neck pain need to be encouraged to return to normal activities as soon as possible in order to reduce their pain-related disability [ 3 ]. However, it is unclear as to how this can best be achieved. Systematic reviews of physical therapy and exercise therapy have reported limited and conflicting evidence [ 4 , 5 ]. Several studies of back pain have suggested that psychological factors are better predictors of outcome than any physical ones [ 9 , 6–8 ]. Psychosocial factors may also influence the development of neck pain [ 10 ]. There is evidence that behavioural interventions are effective for managing chronic back pain [ 4 ].

Management strategies based on exercise and cognitive–behavioural (CB) principles appear to be effective in decreasing functional disability due to back pain [ 11–13 ]. A recent trial of patients with neck pain showed that usual physiotherapy was marginally better than a one-off brief physiotherapy intervention based on CB principles [ 14 ]. However, a more extensive and focused CB intervention to help patients identify their problems and find solutions might be more effective (the Solution-Finding Approach —SFA). The SFA was originally used for mental health problems [ 15 ], although more recently it has been used in the management of back pain [ 16 ]. Specifically, SFA is designed to facilitate desirable changes in patient's cognitions and behaviours, and encourage the adoption of long-term maintenance strategies.

The McKenzie (McK) method is one of the most commonly used approaches by physiotherapists for the management of spinal pain. This is a biomechanical approach based on classification of the condition and the individual's ‘directional preferences’, which guide the prescription of specific exercises. Improvement is assessed through the phenomenon of ‘centralization’ of the reported pain. Evidence for its validity has been demonstrated in a number of studies [ 17 ]. It purports to be an ‘active’ treatment whereby patients are enabled to take control of their situation [ 18 ], and it is consistent, therefore, with the guidelines that recommend a return to normal activities rather than rest [ 4 ]. There is limited evidence to support its effectiveness [ 19 ].

We therefore conducted a randomized trial to compare SFA and McK in terms of improving ability to cope, reducing disability and reducing pain.

Methods

Study participants

Patients were referred to National Health Service (NHS) physiotherapy departments by GPs in West and East Yorkshire. Those aged ≥18 yrs with back pain or neck pain of more than 2 weeks duration, considered to be of non-systemic origin, were all eligible. At the initial screening stage, we excluded patients with a score of less than 4 on the Roland Disability Questionnaire (RDQ) or less than 10 on the Northwick Park Neck Pain Questionnaire (NPQ) in order to exclude those with minor disorders.

Patients were also excluded if they had had any physical therapy within the previous 3 months or were currently planning to use private physical therapy alongside the NHS physiotherapy. Other exclusion criteria were referral from a hospital consultant, possibility of serious pathology or pregnancy.

Procedures

Patients who met the inclusion criteria and agreed to take part were screened by an independent researcher. If they were found to be eligible, informed written consent was obtained prior to being randomized. Randomization utilized a two-stage computer-generated permuted block design set up by the University of York telephone randomization service before recruitment started and was conducted through this remote service by staff who were independent of the research team.

In stage 1, patients were randomized to SFA or McK, stratified by referral centre and whether they had back or neck pain as the primary condition. In stage 2, they were randomized to receive either an educational booklet ( The Back Book [ 20 ] or The Neck Book [ 21 ], as appropriate) or no booklet. Questionnaires which included demographic information and self-completed validated outcome measures were then administered.

All consenting participants were asked their treatment preference independently of randomization. It was made clear that their preference would not change which arm of the study they would be allocated to. This approach allows for the influence of individual patient preferences on outcomes to be evaluated within a fully randomized trial using the preference as a covariate in the analysis [ 11 , 22 ].

Although it was not possible to blind the patients or physiotherapists to the treatment allocation, baseline data were collected by an independent researcher using self-administered questionnaires. The independent researcher was not a physiotherapist and had no involvement in the provision of treatment. The principles of equipoise were maintained on the basis that it was not known which treatment is most effective. The follow-up data were collected by postal questionnaires; the data entered onto the computer was anonymized, and in order to avoid potential bias, the treatment code was not broken until the main analysis was completed. Recruitment to the study started in March 2003 and ended in July 2004. The last patient was followed up in May 2005.

The primary outcome measures were the physical Activity-Avoidance subscale of the Tampa Scale for Kinesiophobia (TSK) [ 23 ], RDQ [ 24 ] and the NPQ [ 25 ] for back pain and neck pain patients, respectively. In addition, the Multidimensional Health Locus of Control (MHLC) [ 26 ], SF-12 (General Quality of Life Health Questionnaire) [ 27 , 28 ], Pain Self Efficacy Questionnaire (PSEQ), a subscale of the Chronic Pain and Coping Inventory (CPCI) [ 29 ] and the Hospital Anxiety and Depression Scale (HADS) [ 30 ] were used. The EQ-5D (Euroqol–– Health Questionnaire) [ 31 ] was used for economic analysis. All participants were followed up by postal questionnaire at 6 weeks, 6 and 12 months following randomization. Satisfaction with treatment was measured by a visual analogue scale (VAS) from 0 (least satisfied) to 100 (most satisfied) at these time points.

The SFA was based on CB principles. It allowed the physiotherapists to use their skills and experience to guide patients to deal with their problem. Patients allocated to this approach received an initial 1 hour assessment which included a physical examination, explanation about the condition, reassurance and goal setting. One or two follow-up sessions were offered to guide progress. The emphasis was on helping patients identify the main problems, work out solutions and then agree realistic goals for what they wanted to achieve.

Physiotherapists who took part in the study were provided with two study days by the research team to familiarize them with the SFA. The training included: enhancing communication skills, an introduction to CB principles, role play, the use of mock physiotherapist–patient consultation videos and a discussion led by a clinical psychologist. The teaching programme was backed up with a Trial Manual written for the purpose and based on previous research. The physiotherapists were encouraged to keep a logbook and practise newly learned techniques on at least five patients prior to participating in the study.

The McK system has a clear protocol, which is dependent on assessment of the problem through repeated movements of the spine [ 32 ] and a clinical phenomenon referred to as ‘centralization’ [ 17 ]. Specific spinal movements, which abolish the symptoms or shift them centrally away from the peripheral parts of the body, are noted by the therapist. Patients are individually assessed and prescribed direction-specific exercises dependent on the assessment findings. The success of the McK approach is considered to depend on the patient's active participation in the prescribed exercises. It may help patients overcome their fear of physical activity by demonstrating that movement can relieve pain; it may reduce anxiety and depression, and help them take control of their situation. These factors may enable the individual to cope better and return to their normal activities sooner, thus preventing long-term disability. All participating physiotherapists were given a day's training in procedures to be used in the McK arm of the study to update the physiotherapists in the principles and application of the McK method. However, therapists had all undertaken postgraduate training courses in the approach beforehand.

In total 28 physiotherapists from seven different physiotherapy departments participated in the study. Each physiotherapist provided both interventions, with one exception who provided only SFA.

Ethical approval for the study was obtained from the relevant Research and Ethics Committees (Hull & East Riding and Leeds). The trial has been assigned the International Standardised Randomised Controlled Trial Number ISRCTN48919562.

Statistical methods and analysis

We planned to recruit 312 patients. Allowing for an attrition rate of one in six, this required a sample size of 260 (130 in each group). This would provide 80% power to detect a differential effect size of 0.35 at a 5% two-sided significance level using an independent sample t -test. We based the calculation on a direct comparison of SFA and McK rather than any tests of interactions, for which additional information would be required. From our previous research [ 14 ], we anticipated a within-group standard deviation of ∼5.7 on the full TSK scale. Combined with a difference of two points on the TSK scale between McK and SFA, which we regarded as the minimum difference of clinical interest, this led to an effect size of 0.35. The more complex analyses reported next should increase power.

Our aspiration was to undertake a full Intention To Treat (ITT) analysis, that is analysing by initial treatment allocation. However, in common with most trials, due to attrition we were unable to follow-up every participant. Nevertheless, all those who returned usable data were analysed within their initially assigned groups, even if they did not receive their allocated treatment and any missing data were assumed to be missing at random.

The TSK was divided into two subscales, Activity-Avoidance and Somatic Focus, in accordance with the factor structure reported by Roelofs et al . [ 23 ]. (This research was not available when the sample size for the trial was calculated.) Activity-Avoidance was chosen as the main outcome measure. To enable data on functional disability from back pain and neck pain patients to be pooled together for analysis, RDQ and NPQ scores were standardized. For all time points, the mean baseline condition-specific score was subtracted from the raw score and the result was divided by the standard deviation of the baseline condition-specific score. Other measures were analysed in the established manner.

Each of the outcome measures, at 6 weeks, 6 and 12 months were analysed jointly, assuming no structure for the matrix of correlations for outcomes at the three time points. A joint analysis of longitudinal data from all time points is now accepted as methodologically superior to separate analyses at each time point [ 33 , 34 ]. Baseline scores were adjusted by including them as covariates. The analyses for each outcome used all available data without imputation of missing data. Thus, the assumption that the missing data were missing at random (MAR) was made. As methods that do not make this assumption lead to more complex analyses and require further untestable assumptions [ 34 ], we believe the MAR assumption to be a pragmatic solution, as suggested elsewhere [ 35 ]. An initial model containing a three-way interaction between treatment allocation (McK or Solution Finding), booklet allocation (received or not) and time (a factor with three levels) was fitted using maximum likelihood. A likelihood ratio test was performed to determine whether the three-way interaction could be removed, using the 5% level of the relevant chi-square distribution as the test criterion. If removal of the three-way interaction did not significantly worsen the fit of the model, the process was repeated for the two-way interactions.

As a sensitivity analysis, the procedure described in the above paragraph was repeated assuming: Additionally, terms representing a main effect of treatment preference (a factor with three levels: McK, Solution Finding or None), and its interaction with main treatment allocation were added to the simplest models permitted and tested for statistical significance.

  • a heterogeneous first-order auto-regressive outcome correlation matrix structure and;

  • a compound symmetry outcome correlation matrix structure.

These analyses were performed using the Mixed Models procedure on SPSS Version 12 for Windows.

Results

A total of 649 patients with neck pain or back pain were screened, and 315 of these were eligible and consented to randomization (49%). These included 219 back pain patients and 96 neck pain patients ( Fig. 1 ).

F ig . 1.

Trial profile.

F ig . 1.

Trial profile.

The mean age of participants was 45.0 yrs (range 18–90) and 60% were women. Sixty-seven percent of patients had pain lasting for more than 3 months. Baseline characteristics are shown in Table 1 .

T able 1.

Baseline demographic characteristics and outcome measures by group allocation a

 Back Neck 
 
 

 
 McKenzie Solution finding McKenzie Solution finding 
 
 

 

 

 
 Booklet given Booklet not given Booklet given Booklet not given Booklet given Booklet not given Booklet given Booklet not given 
Number (% of whole sample) 54 (17.1%) 58 (18.4%) 57 (18.1%) 50 (15.9%) 28 (8.89%) 21 (6.67%) 19 (6.03%) 28 (8.89%) 
Age (years) 44.84 (16.17) 45.15 (14.93) 43.15 (14.40) 43.05 (14.33) 44.81 (15.17) 49.07 (17.69) 46.76 (15.53) 47.67 (13.76) 
Number (%) women 31 (57.4%) 34 (58.6%) 36 (63.2%) 29 (58.0%) 15 (53.6%) 13 (61.9%) 14 (73.7%) 18 (64.3%) 
Number (%) with chronic pain 40 (74.1%) 41 (71.9%) 33 (60.0%) 32 (66.7%) 14 (51.9%) 12 (57.1%) 14 (73.7%) 21 (77.8%) 
Number (%) Preferred Solution Finding 5 (9.3%) 3 (5.3%) 1 (1.8%) 0 (0%) 1 (3.6%) 1 (5.0%) 1 (5.3%) 2 (7.1%) 
Number (%) Preferred McK 3 (5.6%) 9 (15.8%) 8 (14.0%) 6 (12.0%) 2 (7.1%) 2 (10.0%) 3 (15.8%) 5 (17.9%) 
RDQ 10.849 (5.468) 11.155 (4.866) 11.296 (5.034) 12.458 (4.951) NA NA NA NA 
NPQ (0–36) NA NA NA NA 23.750 (4.265) 23.095 (5.195) 23.474 (3.777) 22.429 (3.967) 
TSK Activity-Avoidance 18.635 (4.598) 19.148 (3.887) 19.891 (3.966) 20.694 (4.388) 20.679 (4.922) 19.450 (5.226) 20.211 (4.590) 19.107 (4.701) 
SF12 Physical 39.764 (5.094) 40.588 (5.202) 38.978 (6.044) 38.768 (4.895) 41.044 (6.630) 40.137 (7.308) 42.800 (4.458) 42.289 (4.014) 
SF12 Mental 46.635 (10.285) 46.210 (11.625) 46.732 (10.243) 43.267 (12.485) 41.117 (13.243) 43.283 (12.209) 43.579 (12.070) 42.421 (13.614) 
TSK Somatic Focus 10.170 (3.257) 11.241 (3.536) 11.053 (3.573) 12.122 (3.308) 11.143 (3.628) 10.200 (3.806) 10.833 (3.276) 11.071 (3.185) 
MHLC Internal 30.547 (6.565) 31.103 (6.782) 30.526 (7.419) 31.149 (6.118) 29.407 (6.500) 28.524 (6.063) 30.722 (7.637) 31.929 (5.463) 
MHLC Chance 17.096 (5.577) 16.807 (5.313) 16.125 (5.599) 17.837 (5.336) 16.519 (5.011) 16.000 (5.196) 16.053 (5.286) 16.571 (5.088) 
MHLC Powerful Others 33.132 (4.772) 31.281 (5.226) 31.691 (5.517) 32.292 (4.626) 34.481 (6.154) 34.333 (5.808) 33.684 (6.237) 33.464 (4.895) 
PSEQ 40.075 (11.863) 40.842 (12.016) 40.193 (12.636) 39.020 (13.278) 37.000 (14.371) 41.238 (11.523) 38.368 (14.167) 38.500 (11.227) 
CPCI (Guarding subscale) 22.918 (16.842) 21.148 (14.740) 22.909 (15.313) 26.245 (18.134) 21.760 (16.937) 20.286 (19.262) 12.938 (12.783) 15.840 (14.392) 
HADS Depression 5.154 (3.556) 4.534 (3.256) 4.491 (2.989) 5.920 (3.238) 5.036 (3.834) 4.286 (3.823) 5.158 (2.911) 5.464 (2.874) 
HADS Anxiety 7.096 (4.165) 7.379 (4.316) 7.491 (3.747) 8.160 (4.302) 7.111 (4.032) 7.762 (3.520) 7.263 (3.709) 8.536 (4.299) 
 Back Neck 
 
 

 
 McKenzie Solution finding McKenzie Solution finding 
 
 

 

 

 
 Booklet given Booklet not given Booklet given Booklet not given Booklet given Booklet not given Booklet given Booklet not given 
Number (% of whole sample) 54 (17.1%) 58 (18.4%) 57 (18.1%) 50 (15.9%) 28 (8.89%) 21 (6.67%) 19 (6.03%) 28 (8.89%) 
Age (years) 44.84 (16.17) 45.15 (14.93) 43.15 (14.40) 43.05 (14.33) 44.81 (15.17) 49.07 (17.69) 46.76 (15.53) 47.67 (13.76) 
Number (%) women 31 (57.4%) 34 (58.6%) 36 (63.2%) 29 (58.0%) 15 (53.6%) 13 (61.9%) 14 (73.7%) 18 (64.3%) 
Number (%) with chronic pain 40 (74.1%) 41 (71.9%) 33 (60.0%) 32 (66.7%) 14 (51.9%) 12 (57.1%) 14 (73.7%) 21 (77.8%) 
Number (%) Preferred Solution Finding 5 (9.3%) 3 (5.3%) 1 (1.8%) 0 (0%) 1 (3.6%) 1 (5.0%) 1 (5.3%) 2 (7.1%) 
Number (%) Preferred McK 3 (5.6%) 9 (15.8%) 8 (14.0%) 6 (12.0%) 2 (7.1%) 2 (10.0%) 3 (15.8%) 5 (17.9%) 
RDQ 10.849 (5.468) 11.155 (4.866) 11.296 (5.034) 12.458 (4.951) NA NA NA NA 
NPQ (0–36) NA NA NA NA 23.750 (4.265) 23.095 (5.195) 23.474 (3.777) 22.429 (3.967) 
TSK Activity-Avoidance 18.635 (4.598) 19.148 (3.887) 19.891 (3.966) 20.694 (4.388) 20.679 (4.922) 19.450 (5.226) 20.211 (4.590) 19.107 (4.701) 
SF12 Physical 39.764 (5.094) 40.588 (5.202) 38.978 (6.044) 38.768 (4.895) 41.044 (6.630) 40.137 (7.308) 42.800 (4.458) 42.289 (4.014) 
SF12 Mental 46.635 (10.285) 46.210 (11.625) 46.732 (10.243) 43.267 (12.485) 41.117 (13.243) 43.283 (12.209) 43.579 (12.070) 42.421 (13.614) 
TSK Somatic Focus 10.170 (3.257) 11.241 (3.536) 11.053 (3.573) 12.122 (3.308) 11.143 (3.628) 10.200 (3.806) 10.833 (3.276) 11.071 (3.185) 
MHLC Internal 30.547 (6.565) 31.103 (6.782) 30.526 (7.419) 31.149 (6.118) 29.407 (6.500) 28.524 (6.063) 30.722 (7.637) 31.929 (5.463) 
MHLC Chance 17.096 (5.577) 16.807 (5.313) 16.125 (5.599) 17.837 (5.336) 16.519 (5.011) 16.000 (5.196) 16.053 (5.286) 16.571 (5.088) 
MHLC Powerful Others 33.132 (4.772) 31.281 (5.226) 31.691 (5.517) 32.292 (4.626) 34.481 (6.154) 34.333 (5.808) 33.684 (6.237) 33.464 (4.895) 
PSEQ 40.075 (11.863) 40.842 (12.016) 40.193 (12.636) 39.020 (13.278) 37.000 (14.371) 41.238 (11.523) 38.368 (14.167) 38.500 (11.227) 
CPCI (Guarding subscale) 22.918 (16.842) 21.148 (14.740) 22.909 (15.313) 26.245 (18.134) 21.760 (16.937) 20.286 (19.262) 12.938 (12.783) 15.840 (14.392) 
HADS Depression 5.154 (3.556) 4.534 (3.256) 4.491 (2.989) 5.920 (3.238) 5.036 (3.834) 4.286 (3.823) 5.158 (2.911) 5.464 (2.874) 
HADS Anxiety 7.096 (4.165) 7.379 (4.316) 7.491 (3.747) 8.160 (4.302) 7.111 (4.032) 7.762 (3.520) 7.263 (3.709) 8.536 (4.299) 

a Data are means (SDs) except where indicated otherwise.

The 12-month follow-up rate for McK was 88% and for the SFA 77%. In each group, respectively, there were 82 and 72% who had valid data on the main outcome measures. The patients lost to analysis were similar to the completers in terms of gender but were slightly younger: 40 compared to 46.5 yrs old ( t = 3.39, df = 312, P = 0.001). Only 52 (16.5%) of the 315 patients in the trial expressed a treatment preference.

All patients who were included in the study attended for their first assessment. Patients with back pain attended on average 4.1 sessions of McK compared with 3.2 sessions of SFA. Patients with neck pain attended 4.6 sessions of McK compared with 3.2 sessions of SFA. Fewer than 5% in each group failed to receive the minimum protocol treatment of one face-to-face session with the physiotherapist.

Comparisons between McK and SFA from the simplest model permitted by the process described in the Methods section are presented for each outcome in Tables 2 and 3 . For most outcomes, no effects involving the main treatment or booklet factors were found to be statistically significant. There were two exceptions:

  • a significant interaction between treatment allocation and time was found for TSK Activity–Avoidance subscale. The direction of this interaction ( Table 2 ) implies that, compared to SFA, McK was slightly more effective in reducing fear and avoidance of activity that the individual believed might be damaging by the 6-month time point.

  • a significant interaction between treatment and booklet allocation was found for the MHLC Powerful Others subscale. As there was no significant three-way interaction between treatment, booklet allocation and time, this interaction was found to be consistent at all time points. The SFA supplemented by The Back Book or The Neck Book was slightly more effective than the McK approach in reducing a reliance on ‘Powerful Others’ including the physiotherapist ( Table 3 ).

T able 2.

Changes in primary outcome measures

graphic 
graphic 

a The interaction between treatment allocation and time was not significant for most outcomes. So, raw differences between McK and SFA over time points were therefore averaged to obtain an estimate that did not change over time for these outcomes.

b Refers to treatment by time interaction for TSK Activity-Avoidance and to treatment difference elsewhere.

c Calculated as absolute value of difference reported in sixth column divided by the average of the baseline standard deviations reported in Table 1 .

d Final three columns refer to standardized RDQ measure.

e Final three columns refer to standardized NPQ measure.

T able 3.

Changes in secondary outcome measures

graphic 
graphic 

a The interaction between treatment allocation and time was not significant for most outcomes. In these cases, raw differences between McK and SFA over time points were averaged to obtain an estimate that did not change over time for these outcomes.

b Refers to treatment by booklet allocation interaction for MHLC Powerful Others and to treatment difference elsewhere.

c Calculated as absolute value of difference reported in sixth column divided by the average of the baseline standard deviations reported in Table 1 .

d As a significant interaction between treatment and booklet allocation was found for this outcome, treatment comparisons are presented separately for those who received the booklet and those who did not.

In view of the statistically significant effects found for TSK Activity-Avoidance differences, Table 2 shows the confidence intervals and effect sizes calculated from the model including an interaction between treatment and time and a term representing the main effect of booklet allocation. For MHLC Powerful Others ( Table 3 ), differences, confidence intervals and effect sizes were calculated from the model including an interaction between treatment and booklet allocation and a term representing the main effect of time. For all other outcomes, differences, confidence intervals and effect sizes were calculated from the model containing only terms representing the main effects of treatment, booklet allocation and time.

When the sensitivity analysis with two alternative correlation structures was performed, similar results were obtained.

The two interventions did not differ significantly at 6 months or 1 yr in terms of days reported by participants as being unable to work or carry out usual activities (respectively, Mann–Whitney two-tailed tests, P = 0.383, P = 0.786). McK patients reported significantly higher levels of satisfaction compared to patients randomized to SFA (McK VAS median= 90, SFA VAS median= 70, Mann–Whitney two-tailed test, P = 0.008). Inclusion of the treatment preference as a factor in the modelling did not influence the findings reported earlier for any of the outcomes.

No adverse events were reported.

Discussion

There were no differences in outcome between the two treatment groups at any time points except at 6 months, when McK was found to be slightly more effective than the brief intervention in reducing avoidance of activity as measured by the TSK Activity-Avoidance scale. This was considered a priori to be a proxy for coping with the pain. However, this was only a difference of one point on a scale with a range of 24, representing a 4% difference. Clinically, this could not be regarded as important for the group as a whole. Patients allocated to SFA appeared to have benefited slightly more when this was supplemented by a booklet. They showed a greater reduction in reliance on health professionals according to their scores on the MHLC Powerful Others subscale. However, this difference in change on the MHLC was only 1.6 out of a range of 40, which represents a 4% difference in change. In view of this very small difference, and also the large number of outcomes that have been tested, we conclude that there were no clinically important differences between the two interventions in terms of the outcome measures used. However, patients were more satisfied with McK than SFA. A more ‘hands-on’ approach such as McK physical therapy or chiropractic has been found previously to be related to higher patient satisfaction [ 36 ]. There is some evidence to suggest that better treatment outcomes may be obtained from physical treatment when the treatment is tailored to a specific classification category [ 37 , 38 ]. It is a possibility that subgroups of patients may exist in the back pain population which respond more favourably to a specific intervention, e.g. a specific back exercise determined by ‘directional preference’ [ 39 ].

In another trial of a brief intervention for neck pain patients, preference appeared to have some influence on outcome [ 14 ]. In this present study, few patients expressed a preference so the power was reduced. There was no indication that patient preference influenced outcome in this present study.

Another recent trial of patients with back pain lasting no more than 12 weeks compared a brief physiotherapy intervention based on pain management principles with manual therapy [ 13 ]. Similar to our study, patients allocated to the ‘hands off’, pain management intervention had on average three sessions whilst the manual therapy patients had four sessions. As in our study, no differences in change between the two groups were found. The baseline scores on the RDQ for back pain were slightly lower in our study and the change scores were also smaller, which is not surprising as a majority of patients in our study had had persistent pain lasting longer than 12 weeks. However, our results are generally consistent with the findings of Hay and colleagues [ 13 ] and lend further support to the hypothesis that an intervention which is designed to help patients find out how to cope with their pain may be at least as effective as conventional physiotherapy which uses a biomechanical approach. It may also be cost-effective as slightly fewer sessions were given; in the present study, three sessions for the SFA compared with four or five sessions for McK.

That we included patients both with neck and back pain could be considered a weakness of our study. However, it can be argued that both are musculoskeletal in origin, are difficult to label in a specific way, and emanate from the spine, a view which others researchers have taken [ 3 ]. We included all patients with neck pain and back pain lasting more than 2 weeks as we wanted to increase the generalizability of our findings.

As part of this trial, we conducted an observational study to check adherence to the two intervention protocols. All participating physiotherapists were given training in advanced communication skills and the application of CB principles to enable them to use the SFA. Each physiotherapist acted as his or her own control, and so individual personality differences were not confounding factors. However, we were concerned that therapists might inadvertently use these new skills when delivering McK. Eighteen patient–physiotherapist consultations were recorded on videotape and were scrutinized and scored using a scale developed for the purpose. The majority of physiotherapists were able to switch between approaches without significant overlap in techniques, thus demonstrating differences between the procedures. However, behaviours such as psychosocial questioning and goal setting were inadequately utilized in both approaches. This finding is consistent with other studies which suggest that physiotherapists do not adequately utilize available opportunities to involve patients in goal setting [ 40 ]. Some physiotherapists showed a lack of confidence and hesitancy in using the SFA, for example they needed to rephrase a question several times before they could articulate the point or question they wanted to pose. Clinical supervision may be needed in addition to increase the use of open questions and psychological exploration [ 41 ]. The two days of training in CB skills delivered in the SFA were not consecutive, and might not have been sufficient to provide the necessary skills and confidence. Further research is needed to find out how CB principles can effectively be incorporated into physiotherapy treatment such as the McK approach, or a brief physiotherapy intervention such as the SFA for managing subacute and chronic back pain and neck pain.

Most patients did not have a preference for treatment allocation prior to randomization. However, patients’ expectations and experience of having a hands-on treatment rather than a hands-off intervention may have been an important factor resulting in increased patient satisfaction for the McK approach. A brief physiotherapy intervention using CB principles may be more cost-effective.

graphic

Acknowledgements

The authors gratefully acknowledge the participation in this trial of all patients and staff in all the collaborating centres. Kathryn Gray is thanked for helping to prepare the manuscript. This study was funded by the Arthritis Research Campaign.

The authors have declared no conflicts of interest.

References

1
Maniadakis
N
Gray
A
The economic burden of back pain in the UK
Pain
 , 
2000
, vol. 
84
 (pg. 
95
-
103
)
2
Klaber Moffett
J
Richardson
G
Aspden
R
Porter
R
Costs and effectiveness: approaches to the management of back pain
Lumbar spine disorders: current concepts
 , 
1995
London
World Scientific
3
Nachemson
A
Vingard
E
Nachemson
A
Jonsson
E
Assessment of patients with neck and back pain: a best evidence synthesis
Neck and back pain: the scientific evidence of causes, diagnosis and treatment.
 , 
2000
Lippincott Williams & Wilkins
4
COST B13 Action
European Guidelines for the Management of Low Back Pain
 , 
2004
 
5
Klaber Moffett
J
McLean
S
The role of physiotherapy for non-specific back pain and neck pain
Rheumatology
 , 
2006
, vol. 
45
 (pg. 
371
-
8
)
6
Burton
A
Tillotson
K
Main
C
Hollis
S
Psychosocial predictors of outcome in acute and sub-chronic low-back trouble
Spine
 , 
1995
, vol. 
20
 (pg. 
722
-
8
)
7
Mannion
A
Junge
A
Taimela
S
Muntener
M
Lorenzo
K
Dvorak
J
Active therapy for chronic low back pain Part 3. Factors influencing self-rated disability and its change following therapy
Spine
 , 
2001
, vol. 
26
 (pg. 
920
-
9
)
8
Linton
S
A review of psychological risk factors in back and neck pain
Spine
 , 
2000
, vol. 
25
 (pg. 
1148
-
56
)
9
Bongers
PM
Dewinter
CR
Kompier
MAJ
Hildebrandt
VH
Psychosocial factors at work and musculoskeletal disease
Scand J Work Env Health
 , 
1993
, vol. 
19
 (pg. 
297
-
312
)
10
Ariens
GA
van Mechelen
W
Bongers
PM
Bouter
LM
van der Wal
G
Psychosocial risk factors for neck pain: a systematic review
Am J Ind Med
 , 
2001
, vol. 
39
 (pg. 
180
-
93
)
11
Klaber Moffett
J
Torgerson
D
Bell-Syer
S
, et al.  . 
Randomised controlled trial of exercise for low back pain: clinical outcomes, costs, and preferences
Br Med J
 , 
1999
, vol. 
319
 (pg. 
279
-
83
)
12
Frost
H
Lamb
S
Klaber Moffett
J
Fairbank
J
Moser
J
A fitness programme for patients with chronic low back pain: 2 year follow-up of a randomised controlled trial
Pain
 , 
1998
, vol. 
75
 (pg. 
273
-
9
)
13
Hay
EM
Mullis
R
Lewis
M
, et al.  . 
Comparison of physical treatments versus a brief pain-management programme for back pain in primary care: a randomised clinical trial in physiotherapy practice
Lancet
 , 
2005
, vol. 
365
 (pg. 
2024
-
30
)
14
Klaber Moffett
JA
Jackson
DA
Richmond
S
, et al.  . 
Randomised trial of a brief physiotherapy intervention compared with usual physiotherapy for neck pain patients: outcomes and patients’ preference
Br Med J
 , 
2005
, vol. 
330
 (pg. 
75
-
8
)
15
Mynors-Wallis
L
Gath
D
Lloyd-Thomas
A
Tomlinson
D
Randomised controlled trial comparing problem solving treatment with amitriptyline and placebo for major depression in primary care
Br Med J
 , 
1995
, vol. 
310
 (pg. 
441
-
5
)
16
Van Den Hout
A
To solve or not to solve
PhD thesis
 , 
2002
Netherlands
Universit of Maastricht
17
Aina
A
May
S
Clare
H
The centralization phenomenon of spinal symptoms–a systematic review
Man Ther
 , 
2004
, vol. 
9
 (pg. 
134
-
43
)
18
Holdom
A
The use of the McKenzie approach to treat back pain
Br J Ther Rehabilit
 , 
1996
, vol. 
3
 (pg. 
7
-
10
)
19
Clare
HA
Adams
R
Maher
CG
A systematic review of efficacy of McKenzie therapy for spinal pain
Aust J Physiother
 , 
2004
, vol. 
50
 (pg. 
209
-
16
)
20
Roland
M
Waddell
G
Klaber Moffet
J
Burton
K
Main
C
The Back Book
 , 
2002
London
The Stationery Office
(pg. 
1
-
21
)
21
Waddell
G
Klaber Moffet
J
Burton
K
The Neck Book
 , 
2004
London
The Stationery Office
(pg. 
1
-
27
)
22
Torgerson
D
Klaber-Moffett
J
Russell
I
Patient preferences in randomised trials: threat or opportunity?
J Health Serv Res Policy
 , 
1996
, vol. 
1
 (pg. 
194
-
7
)
23
Roelofs
J
Goubert
L
Peters
ML
Vlaeyen
JW
Crombez
G
The Tampa Scale for Kinesiophobia: further examination of psychometric properties in patients with chronic low back pain and fibromyalgia
Eur J Pain
 , 
2004
, vol. 
8
 (pg. 
495
-
502
)
24
Roland
M
Morris
R
A study of the natural history of low back pain: I Development of a reliable and sensitive measure of disability in low back pain
Spine
 , 
1983
, vol. 
8
 (pg. 
141
-
5
)
25
Leake
A
Cooper
J
Dyer
S
Williams
K
Turner-Stokes
L
Frank
A
The Northwick Park Neck Pain Questionnire, devised to measure neck pain and disability
Br J Rheumatol
 , 
1994
, vol. 
33
 (pg. 
469
-
74
)
26
Wallston
KA
Wallston
BS
DeVellis
R
Development of the Multidimensional Health Locus of Control (MHCL) scales
Health Education Monographs
 , 
978
, vol. 
6
 (pg. 
160
-
70
)
27
Jenkinson
C
Layte
R
Development and testing of the UK SF-12 (short form health survey)
J Health Serv Res Policy
 , 
1997
, vol. 
2
 (pg. 
14
-
8
)
28
Jenkinson
C
Layte
R
Jenkinson
D
, et al.  . 
A shorter form health survey: can the SF-12 replicate results from the SF-36 in longitudinal studies?
J Public Health Med
 , 
1997
, vol. 
19
 (pg. 
179
-
86
)
29
Jensen
MP
Turner
JA
Romano
JM
Strom
SE
The Chronic Pain Coping Inventory: development and preliminary validation
Pain
 , 
1995
, vol. 
60
 (pg. 
203
-
16
)
30
Zigmond
AS
Snaith
RP
The hospital anxiety and depression scale
Acta Psychiatr Scand
 , 
1983
, vol. 
67
 (pg. 
361
-
70
)
31
EuroQuol Group
EuroQuol* — a new facility for the measurement of health-related quality of life
Health Policy
 , 
1990
, vol. 
16
 (pg. 
199
-
208
)
32
McKenzie
R
May
S
The Lumbar Spine—Mechanical Diagnosis and Therapy
 , 
2003
2nd ed
Waikanae, New Zealand
Spinal Publications Ltd
33
Diggle
P
Heagerty
P
Liang
K-Y
Zeger
S
Analysis of variance methods
Analysis of longitudinal data
 , 
2002
Oxford
Oxford University Press
(pg. 
114
-
25
)
34
Fairclough
D
Analytic methods for ignorable missing data
In Design and analyses of quality of life studies in clinical trials
 , 
2002
Boca Raton, USA
Chapman and Hall/CRC
(pg. 
93
-
114
)
35
Brown
G
Prescott
R
Applied Mixed Models in Medicine
 , 
2006
2nd ed
Chichester
John Wiley & Sons
pg. 
82
 
36
Cherkin
D
Deyo
R
Battie
M
Street
J
Barlow
W
A comparison of physical therapy, chiropractic manipulation, and provision of an educational booklet for the treatment of patients with low back pain
New Engl J Med
 , 
1998
, vol. 
339
 (pg. 
1021
-
9
)
37
Hicks
GE
Fritz
JM
Delitto
A
McGill
SM
Preliminary development of a clinical prediction rule for determining which patients with low back pain will respond to a stabilization exercise program
Arch Phys Med Rehabil
 , 
2005
, vol. 
86
 (pg. 
1753
-
62
)
38
Fritz
JM
Whitman
JM
Childs
JD
Lumbar spine segmental mobility assessment: an examination of validity for determining intervention strategies in patients with low back pain
Arch Phys Med Rehabil
 , 
2005
, vol. 
86
 (pg. 
1745
-
52
)
39
Long
A
Donelson
R
Fung
T
Does it matter which exercise? A randomised control trial of exercise for low back pain
Spine
 , 
2004
, vol. 
29
 (pg. 
2593
-
602
)
40
Baker
S
Marshall
H
Rice
G
Zimmerman
G
Patient participation in physical therapy goal setting
Phys Ther
 , 
2001
, vol. 
81
 (pg. 
1118
-
26
)
41
Heaven
C
Clegg
J
Maguire
P
Transfer of communication skills training from workshop to workplace: the impact of clinical supervision
Patient Educ Couns
 , 
2006
, vol. 
60
 (pg. 
313
-
25
)

Comments

0 Comments