This pilot study utilized a pre-experimental single group, repeated measures design in order to gather data about effect sizes, clinically significant changes, and patient satisfaction. Since the trial was designed to test the MBFT protocol and the feasibility of undertaking a larger, more comprehensive trial, a control group was not used.

Inclusion criteria required that participants be aged between 18 and 65 years of age and have persistent LBP, defined in accordance with the International Association for the Study of Pain's (IASP) Task Force on Taxonomy as “pain that continues beyond the usual course of healing, usually taken to mean continuous or intermittent pain for at least three months” (Merskey and Bogduk, 1994). A cut-off score of 85 on the Orebro Musculoskeletal Pain Questionnaire (OMPQ) was used to exclude those with lower functional impairment and disability (Hill et al., 2010). Additional exclusion criteria included: (1) current use of medications containing an daily oral morphine equivalent dose of 100 mg; (2) predominant neuropathic pain conditions including dominant lower leg radicular pain with clinical signs of neuropathic pain as defined by the IASP (Treede et al., 2008); (3) low back spinal surgery in the past 12 months; (4) current substance abuse or dependence; (5) high risk of suicide; (6) active psychotic conditions, and (7) limited English language fluency.

Participants were recruited from the general population through advertisements in health clinics and hospitals in metropolitan Perth, Australia. A convenience sample of physiotherapists, psychologists, and pain specialists in the Perth metropolitan area were also contacted and invited to provide referrals for the study cohort.

This study was approved by the Curtin University Human Research Ethics Committee (approval HR72/2009) and adhered to the Declaration of Helsinki. Participants registered their interest in the study by contacting the Curtin University Psychology Clinic. Participants were then telephone-screened via Clinical Psychology (RS) prior to recruitment. Participants who met the inclusion criteria were then further individually screened by a senior specialist musculoskeletal physiotherapist (HS) experienced in pain management, who made a clinical judgment about their likely ability to safely complete the MBFT intervention (i.e., no specific neurological deficit or other serious pathology indicating a precaution or contraindication to participating). After being assessed and accepted for inclusion into the study, participants completed a pre-intervention questionnaire battery (outlined below) on day 1 of the intervention, and this battery of measures was completed again immediately at post-intervention (on the last day of the intervention) and at 6 months post-intervention.

The MBFT group intervention was developed specifically for this study, as outlined in the Mindfulness-Based Functional Therapy Facilitator Manual (unpublished document available on request). The intervention was comprised of an integrated weekly 2-h group session run over eight consecutive weeks and involved: group discussion; homework review; psycho-education; and mindfulness meditation aligned with cognitive behavioral physiotherapy re-education of normal relaxed movement and postures. All MBFT sessions were co-facilitated by the same clinical psychology and physiotherapy team (RS and HS). For pragmatic reasons, there were two intervention groups, with 10 participants in the first group and 6 in the second. While taking part in the intervention, participants were instructed to continue their usual activities and their usual medical care, including medications.

The mindfulness component of MBFT was adapted from the well-known Mindfulness-Based Stress Reduction (MBSR; Kabat-Zinn, 1990) and Mindfulness-Based Cognitive Therapy for Depression (MBCT; Segal et al., 2002) protocols. This involved in-session meditation exercises (for example, body scan meditation; awareness of breathing meditation; walking meditation; awareness of thoughts meditation), reflection and discussion on the meditation experience, psycho-education about the mindfulness meditation exercises, and daily home mindfulness practice using audio recordings of the guided meditations used in MBSR. The in-session mindfulness exercises were adapted specifically for a persistent LBP population, with tailored instructions on working with pain and the associated cognitions and affects. The physiotherapy component targeted basic functional movements that are commonly compromised in patients with persistent LBP, and which can contribute to the maintenance of pain (O'Sullivan, 2005; Dankaerts et al., 2006, 2009). The aim was to train participants to practice mindfulness in functional daily movements and postures, and using a graded approach, to progressively moderate persistent maladaptive protective or avoidant movement behaviors (O'Sullivan, 2005; Dankaerts et al., 2006, 2009). These functional postures and exercises included: body awareness of breathing; movement control involving progressing from non-weight bearing to weight bearing (e.g., rolling, sitting, bending, lifting, standing, and walking); transitioning between postures (e.g., sitting to standing); and targeting specific individual pain provocative tasks identified by the participant. An emphasis was placed on relaxed breathing and gentle, appropriately graduated movements avoiding excessive abdominal bracing, breath-holding, or inappropriate strategies such as hand support to get from sitting to standing. Graduated mindful virtual and real rehearsal of previously avoided, or pain provocative movements and postures, was encouraged. The aim of practizing these functional movements as virtual movements was to further reinforce non-provocative sensorimotor integration (Flor and Diers, 2009). A graduated cardiovascular program which emphasized time-contingent (McCracken and Samuel, 2007) rather than pain-contingent (Abbott et al., 2010) pacing was also implemented.

Movement re-training exercises were blended into each session to complement the mindfulness exercises (e.g., targeting walking in the session where walking meditation was introduced). The intention of this approach was thereby to integrate the psychology and physiotherapy components of this intervention, rather than present each as separate components, by separate disciplines as appears common in many multidisciplinary pain rehabilitation programs (Eccleston et al., 2009). The educational component of MBFT used didactic and interactive delivery of current evidence-based information about pain management, covering: pain neurophysiology, central nervous system sensitization, the biopsychosocial model of pain; treatments for spinal pain; effective use of medications; importance of return to normal movement/exercise in modulating pain; the fear-avoidance model of chronic pain; and the role of stress and negative affect in maintaining pain.

The following measures were chosen to align with current IMMPACT (Dworkin et al., 2008) recommendations for the clinical importance of treatment outcomes in chronic pain clinical trials.

The Örebro Musculoskeletal Pain Questionnaire (OMPQ; Linton, 1999) is a 25-item screening tool used to assess the risk of future disability in pain patients. It includes items covering physical functioning, emotional functioning, and pain intensity. The OMPQ has been found to have test-retest reliability of 0.83 and to correctly identify 80% of acute pain patients who would not return to work (Linton and Boersma, 2003; Linton, 2005). The OMPQ was used in screening and questions 9-10, which are 10-point numerical rating scales (NRS) of pain, were averaged to form a measure of pain intensity.

The Oswestry Disability Questionnaire (ODQ; Fairbank et al., 1980) is a 10-item measure of low-back related functional disability. The ODQ has been widely validated, has a test-retest reliability of 0.83 to 0.99 and an internal consistency coefficient of 0.71 to 0.87 (Fairbank and Pynsent, 2000).

The short form of the Depression Anxiety Stress Scales (DASS-21; Lovibond and Lovibond, 1995) is a 21-item measure of emotional functioning, with 3 subscales. The DASS-21 has been shown to have high internal consistency, with Cronbach's alphas of 0.94 for Depression, 0.87 for Anxiety, and 0.91 for Stress (Antony et al., 1998; Henry and Crawford, 2005).

The Mindful Attention Awareness Scale (MAAS; Brown and Ryan, 2003) is a 15-item measure of present-moment awareness of actions, interpersonal communication, thoughts, emotions, and physical states. The MAAS has been found to have good convergent and discriminant validity, excellent test-retest reliability (r = 0.81, p < 0.0001), and good internal consistency, with a coefficient alpha of 0.87 (Brown and Ryan, 2003).

The Pain Catastrophizing Scale (PCS: Sullivan et al., 1995) is a 13-item self-report measure of the degree to which people experiencing pain adopt exaggerated negative interpretations of their pain. The PCS has good criterion-related validity and excellent internal consistency, with a reliability coefficient of 0.92 (Osman et al., 2000).

The Chronic Pain Acceptance Questionnaire (CPAQ) is a 20-item instrument that measures acceptance of pain. The CPAQ has been found to predict patient functioning and is internally consistent, with a coefficient alpha of 0.78–0.84 (McCracken et al., 2004).

The Rand 36-Item Health Survey (SF-36; Ware and Sherbourne, 1992) is a 36-item measure of health status and health-related quality of life. The SF-36 is comprised of eight subscales: (1) physical functioning, (2) role limitations due to physical health, (3) role limitations due to emotional problems, (4) energy/fatigue, (5) emotional well-being, (6) social functioning, (7) pain, and (8) general health. Subscales have internal consistency coefficients of 0.83–0.91 (Davidson and Keating, 2002). The pain subscale was not used as a measure of pain intensity in this study because it takes into account both intensity and interference. Since emotional functioning was measured by the DASS, the emotional well-being subscale was not included in analysis.

The Client Satisfaction Questionnaire (CSQ; Larsen et al., 1979) is an 18-item service evaluation measure that has a high degree of internal consistency (0.91), and is substantially correlated with service utilization and degree of client-reported change (Larsen et al., 1979).

Data analysis was carried out in the following order: (1) data screening; (2) computing descriptive statistics for all variables using PASW Statistics 18.0 for Macintosh (Spss Inc, 2010); (3) testing for group changes between pre- and post-test, and between pre-test and 6-month follow-up, using a series of paired t-tests in PASW Statistics 18.0; (4) calculating effect sizes for pre- to post- intervention and pre-intervention to follow-up changes using Cohen's d statistic; (5) conducting power analyses for each test using G*Power (Erdfelder et al., 1996); and (6) testing the clinical significance of the study outcomes using the methodology outlined by Jacobson and Truax (1991). This latter formula defines clinically significant change as movement to within two standard deviations of the mean of the normal population on a particular variable. Normative data from the functional population was used to calculate cut-off scores to determine whether participants were in the normal population at post-treatment and follow-up (Lovibond and Lovibond, 1995; Bowling et al., 1999; Fairbank and Pynsent, 2000; Van Damme et al., 2002). This approach to clinical significance testing is more stringent than several other methods of measuring minimal clinically important differences (Dworkin et al., 2008).

A total of 23 people were referred and screened for the study, 7 of which did not meet the inclusion criteria. Sixteen people (12 females and 4 males) participated in the study. Two participants dropped out and, due to missing values, their data were excluded. A further two participants' data were excluded due to a large number of missing values. Pre- and post-intervention data were available for 12 participants. Two participants did not return follow-up questionnaires, leaving 10 participants' data in the follow-up analyses.

Descriptive statistics for outcome measures at pre-test, post-test, and follow-up are presented in Table 1.

Prior to testing for group changes in outcome variables over time, data were screened for suitability of performing paired samples t-tests. The majority of variables met assumptions of normality, despite the Shapiro-Wilk test of normality being violated in five of the variables. A visual inspection of histograms indicated four of these were either normally distributed or were not identified as having a significant amount (greater than 0.01) of skewness or kurtosis. Given that the t-statistic is relatively robust against these minor violations, the data were deemed appropriate for analysis (Tabachnick and Fidell, 2007).

Table 2 summarizes the changes in outcome measures between pre- and post-intervention, and between pre-intervention and 6 month follow-up. Statistically significant improvements in 9 of the 13 variables (functional disability, catastrophizing, stress, pain acceptance, physical functioning, role limitations due to physical condition, energy, social functioning, and general health) were evident comparing pre- and post-intervention with alpha 0.05. At the Bonferroni-corrected alpha level of 0.004, energy and catastrophizing were the only variables showing significant pre-post changes, with pain acceptance and general health approaching significance (p = 0.006 and p = 0.005, respectively). Between pre-intervention and follow-up there were significant changes in 10 of the 13 variables (catastrophizing, mindfulness, pain acceptance, depression, stress, physical functioning, role limitations due to physical condition, energy, social functioning and general health) with alpha 0.05. At the more conservative Bonferroni-corrected alpha level (0.004), physical functioning and social functioning were significant, while pain acceptance, depression and mindfulness were approaching significance (p = 0.004, 0.006, and 0.008, respectively).

Between pre- and post-intervention, large effect sizes were obtained for pain catastrophizing, role limitations due to physical condition and energy, as shown in Table 3. Between pre-intervention and 6-month follow-up, large effect sizes were evident for pain catastrophizing, mindfulness, pain acceptance, depression, stress, role limitations due to physical condition, and social functioning.

Table 4 summarizes the results of reliable change and clinically significant change analyses for pre-post intervention and pre-intervention to follow-up changes. Immediately after the intervention, a majority of participants failed to show reliable change on measures of functional disability, mindfulness, acceptance, anxiety, physical functioning, energy, social functioning and general health. However, half the sample had clinically significant change and were therefore classified as recovered on catastrophizing, depression, and role limitations due to physical condition. At follow-up, the majority of participants failed to demonstrate reliable change on functional disability, catastrophizing, mindfulness, depression, anxiety, social functioning and general health. However, the vast majority of participants were recovered on physical functioning and role limitations due to physical condition, while half were recovered on stress and energy. A further 70% of participants had reliable change on pain intensity and were therefore classified as improved.

Of the 16 participants who entered the study 2 dropped out, one for logistic reasons and one due to a lack of self-perceived benefit. The overall attendance rate of sessions (including dropouts) was 87%. Insufficient data was returned to report on homework adherence. Scores on the Client Satisfaction Questionnaire (CSQ) ranged from 24 to 32, where 32 was also the maximum possible score. The mean score was 30.15 (SD = 2.85). All of the participants rated the intervention as satisfactory or higher, with the vast majority (85%) rating the intervention as highly satisfactory. Qualitative data recorded on the CSQ were also largely positive, with comments such as: “I was very happy with the program and found it helpful, many thanks”; and “I am very grateful firstly for this program being recommended to me and for meeting the facilitators who are so dedicated to their professions and to sincerely caring for our wellbeing.”

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.