Research Report. Graded Activity and Graded Exposure for Persistent Nonspecific Low Back Pain: A Systematic Review

Transcription

1 Research Report L.G. Macedo, MSc, is a PhD candidate at The George Institute for International Health, University of Sydney, PO Box M201, Missenden Rd, Camperdown, New South Wales 2050, Australia. Address all correspondence to Ms Macedo at: lmacedo@george.org.au. R.J.E.M. Smeets, MD, PhD, is Professor, Adelante Zorggroep, Hoensbroek, the Netherlands, and Department of Rehabilitation Medicine, School of Public Health and Primary Care, Maastricht University, Maastricht, the Netherlands. C.G. Maher, PhD, is Director, Musculoskeletal Division, The George Institute for International Health, University of Sydney. J. Latimer, PhD, is Associate Professor, The George Institute for International Health, University of Sydney. J.H. McAuley, PhD, is Research Manager, The George Institute for International Health, University of Sydney. [Macedo LG, Smeets RJEM, Maher CG, et al. Graded activity and graded exposure for persistent nonspecific low back pain: a systematic review. Phys Ther. 2010; 90: ] 2010 American Physical Therapy Association Post a Rapid Response to this article at: ptjournal.apta.org Graded Activity and Graded Exposure for Persistent Nonspecific Low Back Pain: A Systematic Review Luciana G. Macedo, Rob J.E.M. Smeets, Christopher G. Maher, Jane Latimer, James H. McAuley Background. Graded activity and graded exposure are increasingly being used in the management of persistent low back pain; however, their effectiveness remains poorly understood. Purpose. The aim of this study was to systematically review randomized controlled trials that evaluated the effectiveness of graded activity or graded exposure for persistent ( 6 weeks in duration or recurrent) low back pain. Data Sources. Trials were electronically searched and rated for quality by use of the PEDro scale (values of 0 10). Study Selection. Randomized controlled trials of graded activity or graded exposure that included pain, disability, global perceived effect, or work status outcomes were included in the study. Data Extraction. Outcomes were converted to a scale from 0 to 100. Trials were pooled with software used for preparing and maintaining Cochrane reviews. Results are presented as weighted mean differences with 95% confidence intervals. Data Synthesis. Fifteen trials with 1,654 patients were included. The trials had a median quality score of 6 (range 3 9). Pooled effects from 6 trials comparing graded activity with a minimal intervention or no treatment favored graded activity, with 4 contrasts being statistically significant: mean values (95% confidence intervals) for pain in the short term, pain in the intermediate term, disability in the short term, and disability in the intermediate term were 6.2 ( 9.4 to 3.0), 5.5 ( 9.9 to 1.0), 6.5 ( 10.1 to 3.0), and 3.9 ( 7.4 to 0.4), respectively. None of the pooled effects from 6 trials comparing graded activity with another form of exercise, from 4 trials comparing graded activity with graded exposure, and from 2 trials comparing graded exposure with a waiting list were statistically significant. Limitations. Limitations of this review include the low quality of the studies, primarily those that evaluated graded exposure; the use of various types of outome measures; and differences in the implementation of the interventions, adding to the heterogeneity of the studies. Conclusions. The available evidence suggests that graded activity in the short term and intermediate term is slightly more effective than a minimal intervention but not more effective than other forms of exercise for persistent low back pain. The limited evidence suggests that graded exposure is as effective as minimal treatment or graded activity for persistent low back pain. 860 f Physical Therapy Volume 90 Number 6 June 2010

2 Many patients with low back pain (LBP) have concerns that are inadequately explained with a traditional biomedical model. Contemporary approaches have recognized the multifactorial etiology of LBP and the necessity to adopt a biopsychosocial model when dealing with the condition. Cognitive behavioral models propose that pain-related fear, kinesiophobia, and unhelpful beliefs about back pain may be primary factors leading to increased pain and a decreased level of activity or functioning. 1,2 It has been suggested that these factors not only may represent psychosocial barriers to recovery but also may contribute to important biological changes, such as disuse and deconditioning after an injury. 3 Graded activity and graded exposure are interventions commonly used in the management of persistent LBP. These treatments incorporate behavioral and cognitive approaches to improve activity tolerance. The primary difference between these interventions is that, with graded exposure, patients are asked to create a hierarchy of feared activities. The exposure starts with the least feared activity, and the therapist helps the patient appraise the exposure and its consequences and then address irrational and counterproductive beliefs, leading to reductions in the anxiety associated with the activity. Once the negative associations are extinguished, activities associated with higher levels of anxiety are addressed in the same way. With graded activity, operant conditioning principles are used to reinforce healthy behaviors. The program focuses on functional activities and progresses in a timecontingent manner regardless of pain to achieve functional goals and increased activity. Principles of quotas, pacing, and self-reinforcement are key features of the program. Although both treatments have been endorsed in clinical guidelines for the management of persistent LBP, 4,5 the effectiveness of the 2 treatments has not been well established. Although no systematic review of graded activity or graded exposure has been published, a Cochrane systematic review of cognitive and behavioral interventions for chronic LBP has been completed. The review concluded that combined cognitive therapy/respondent therapy and progressive relaxation alone are effective treatment modalities for shortterm pain reduction in patients with chronic LBP but that there are no significant differences between the different forms of behavioral interventions. 6 Because most of the studies included in that review involved only psychological interventions, with minimal or no exercise component, the effectiveness of graded activity or graded exposure cannot be established from that review. Therefore, the purpose of the present study was to systematically review randomized controlled trials that evaluated the effectiveness of graded activity and graded exposure interventions for the treatment of persistent ( 6 weeks in duration or recurrent) nonspecific LBP at short-, intermediate-, and long-term follow-up evaluations. The outcomes of interest were pain, disability, global perceived effect, and return to work. Method Data Sources and Searches A computerized electronic search was performed to identify relevant articles. The search was conducted on MEDLINE (1950 to February 2009), CINAHL (1982 to February 2009), PsychINFO (1806 to February 2009), PEDro (to February 2009), and EMBASE (1988 to February 2009). Key words relating to the domains of randomized controlled trials and back pain were used, as recommended by the Cochrane Back Review Group. 7 Terms for graded activity and graded exposure were included in the search by use of MeSH (Medical Subject Headings of the National Library of Medicine) terms and specific guidelines for each database (Appendix). Subject subheadings and word truncations specific for each database were used. There was no language restriction. One reviewer screened the search results for potentially eligible studies, and 2 reviewers independently reviewed the screened articles for eligibility. A third independent reviewer resolved any disagreement about the inclusion of trials. Authors were contacted when more information about a trial was needed to allow the inclusion of that trial. Many authors were contacted for additional information related to inclusion criteria From the latter studies, we included 5 trials 11,12,14 16 and excluded 4 trials (2 authors confirmed that their trials were not eligible, 8,10 and we did not receive a response from 2 authors). Staal et al 16 provided data for the subset of participants who met the inclusion criteria for this review (ie, excluding participants with acute pain). Researchers who published relevant articles were contacted to help identify gray literature and articles in press. Citation tracking was performed by use of the database Web of Science on ISI Web of Knowledge,* and a manual search of the reference lists of previous reviews * Thomson Reuters, Level 3, 100 Harris St, Pyrmont, New South Wales 2009 Australia. Available With This Article at ptjournal.apta.org The Bottom Line Podcast Audio Abstracts Podcast This article was published ahead of print on April 15, 2010, at ptjournal.apta.org. June 2010 Volume 90 Number 6 Physical Therapy f 861

3 and eligible trials also was performed. The International Clinical Trials Registry platform from the World Health Organization and the Cochrane Centre Register of Controlled Trials also were searched with terms for LBP, graded activity, and graded exposure. Study Selection Studies were eligible for inclusion when they were randomized controlled trials or quasi-randomized controlled trials comparing graded activity or graded exposure to placebo, no treatment, or another active treatment or when graded activity or graded exposure was added as a supplement to other interventions. The reviewers followed a research protocol that was developed before the beginning of the review process and that included a checklist for inclusion criteria. Trials were considered to have evaluated graded activity when the treatment included the following 3 features: The Bottom Line The treatment involved principles of operant conditioning, such as reinforcement of healthy behaviors. Treatment goals were functional activities. The program included a baseline and then incremented activities in a time-contingent manner regardless of pain. Additionally, trials in which treatments were described as graded activity and in which various handbooks 2,17 were cited were deemed eligible. Trials were considered to have evaluated graded exposure when the treatment included the following 4 features: What do we already know about this topic? Graded activity and graded exposure differ from traditional exercise because these interventions incorporate psychological principles into activity prescription. It remains unclear whether either of these interventions is effective. What new information does this study offer? This review established that graded activity, but not graded exposure, is an effective treatment for persistent low back pain, although the effect size is small. There is no evidence that either is superior to a traditional exercise program for persistent low back pain. If you re a patient, what might these findings mean for you? Exercise is an effective treatment for persistent low back pain but the type of exercise does not seem particularly important. Feared activities were identified. A hierarchy of feared activities was created. Exposure started with the least feared activity. The therapist assisted the patient in appraising the exposure to feared activities and its consequences. The therapist addressed irrational beliefs, counterproductive beliefs, or both. Additionally, trials in which treatments were described as graded exposure and in which various handbooks 18,19 were cited were deemed eligible. Randomized or quasi-randomized controlled trials were included when they explicitly reported that a criterion for entry was nonspecific LBP (with or without leg pain) with a duration of at least 6 weeks (nonacute LBP) or recurrent LBP. There was no age or sex restriction. Trials were included when one of the following outcome measures was reported: pain, disability, quality of life, perceived effect, return to work, or recurrence. Data Extraction and Quality Assessment The methodological quality of the trials was assessed by use of the PEDro scale (values of 0 10), with scores extracted from the PEDro database. 20 Assessment of the quality of trials in the PEDro database was performed by 2 trained independent raters, and disagreements were resolved by a third rater. Methodological quality was not an inclusion criterion. Two independent reviewers extracted data from the included studies by using a standardized data extraction form. Mean scores, standard deviations, and sample sizes were extracted from the studies. When this information was not provided in the trial, the values were calculated or estimated by use of methods recommended in the Cochrane Handbook for Systematic Reviews of Interventions. 21 When there was insufficient information about outcomes to allow data analysis, the authors of the included studies were contacted. All 11,16,22 25 but one 26 of the authors 862 f Physical Therapy Volume 90 Number 6 June 2010

4 Database searches : February 2009 PsychINFO: 183 CINAHL: 89 EMBASE: 2,161 MEDLINE: 1,545 PEDro: 868 Hand search: 4 Total after removing duplicates: 3,971 Web of Science search 3 potentially eligible Indicated by experts 6 potentially eligible Reasons for exclusion: 33 did not use graded activity or graded exposure, 8 10, included patients without nonspecific low back pain, included patients with acute low back pain, were not RCTs, were secondary analyses of another trial, and 1 used graded activity for both treatment groups potentially eligible after assessing titles and abstracts 83 potentially eligible 15 original studies included in review (18 articles) Figure 1. Flow chart of systematic review inclusion and exclusion. RCTs randomized controlled trials. who were contacted to provide trial data responded to our queries. Means, standard deviations, and sample sizes were extracted for shortterm (less than 3 months after randomization), intermediate-term (at least 3 months but less than 12 months after randomization), and long-term (12 months or more after randomization) follow-up evaluations. When multiple time points fell within the same category, the one that was closer to the end of treatment for the short term, closer to 6 months for the intermediate term, and closer to 12 months for the long term was used. These references for time points were based on guidelines from the Cochrane Back Review Group. 7 Scores for pain, disability, and global perceived effect were converted to a scale from 0 to 100. When more than 1 outcome measure was used to assess pain, disability, and work status, the outcome measure described as the primary outcome measure for the trial was included in this review. June 2010 Volume 90 Number 6 Physical Therapy f 863

5 Data Synthesis and Analysis Results were pooled when trials were considered sufficiently homogeneous with respect to participant characteristics, interventions, and outcomes. I 2 was calculated by use of RevMan 5 27 to assess statistical heterogeneity. I 2 describes the percentage of the variability in effect estimates that is due to heterogeneity rather than sampling error (chance). A value greater than 50% may be considered substantial heterogeneity. 21 When trials were statistically homogeneous (I 2 50%), pooled effects (weighted mean differences) were calculated by use of a fixed-effects model. When trials were statistically heterogeneous (I 2 50%), estimates of pooled effects (weighted mean differences) were obtained by use of a random-effects model. 21 Results Study Selection The initial electronic database search resulted in a total of 3,971 articles after the removal of duplicates. Of these, 74 articles were selected as potentially eligible on the basis of their title and abstract. Through a Web of Science search of these articles, 3 other potentially eligible articles were identified. A total of 77 potentially eligible articles were considered for inclusion; only 14 were found to be eligible for inclusion in this review (Fig. 1). Twenty-four experts were contacted to provide information about gray literature and articles that we may have missed. We received 12 responses suggesting 6 potentially eligible studies; only 1 of these was included in this review. Therefore, we included a total of 15 original studies with results presented in 18 different articles. The International Clinical Trials Registry platform contained 8 randomized controlled trials that were potentially eligible for inclusion. Of those, 3 were already included in our study, 22,26,90 3 were still recruiting patients, and 2 were not eligible because patients with acute back pain were included. Methodological Quality The methodological quality assessment with the PEDro scale revealed a median score of 6 (range 3 9). Masking of the therapist was not included in any of the trials, as expected in studies of activity prescription, and masking of patients was included in only 1 trial. 23 Point estimates and between-group comparisons were present in 14 of 15 trials, 12 of 15 trials had comparability at baseline, and 10 of 15 trials had concealed allocation, intention to treat, and adequate follow-up evaluations. Study Characteristics Fourteen of the randomized controlled trials included in this review compared graded activity or graded exposure with another treatment or with no treatment (Tabs. 1 and 2). One trial with a factorial design (2 2) was also included; the first factor was graded activity, the second factor was advice, and each factor had 2 levels (active versus sham). 23 In 7 trials, graded activity was compared with a minimal intervention (usual care, waiting list, sham exercise, advice to stay active, or care by a general practitioner). 12,16,22,23,26,86,92 In 6 trials, graded activity was compared with another form of exercise (motor control, high-intensity back school, general physical therapy exercises, active physical treatment, or physical therapy according to guidelines). 11,14,15,22,25,26 In 4 trials, graded activity was compared with graded exposure, 12,90,91,94 and in 2 trials, graded exposure was compared with a waiting list. 12,94 Graded Activity Versus a Minimal Intervention Seven trials with a total of 920 patients compared graded activity with a minimal intervention. 12,16,22,23,26,86,92 The methodological quality of these trials ranged from 4 to 9. Data for pooling were available for pain and disability in the short, intermediate, and long terms and for perceived effect in the short term. Data were pooled by use of a fixed-effects model for all outcomes except pain and disability in the long term and perceived effect in the short term, for which a random-effects model was used. The pooled results favored graded activity for both pain and disability at short- and medium-term follow-up evaluations, but the effects were small; for example, for pain in the short term, the weighted mean difference on a scale from 0 to 100 was 6.2 points (95% confidence interval [CI] 9.4 to 3.0). For both pain and disability at long-term follow-up evaluations, the effects were close to zero and not significant; for example, for pain the value was 0.1 (95% CI 10.4 to 10.2) (Fig. 2). Results for return to work were presented by 3 studies. 26,92,95 The results of 1 trial, that of Lindstrom et al, 92 were expressed in likelihood ratios and showed that return to work was faster for patients in the gradedactivity group than for patients in the control group. The median number of days to return to work for the graded-activity group was 35 days; that for the control group was 61 days. Heymans et al 26 used survival analysis to show that there were no significant differences between treatment groups at short-term follow-up evaluations (hazard ratio 1.0; 95% CI 0.8 to 1.4). The median number of days of sick leave for the gradedactivity group was 85; that for the usual-care group was 75. In contrast, the results of a survival analysis conducted by Streenstra et al 95 showed a significantly earlier return to work for the usual-care group (hazard ratio 0.52; 95% CI 0.32 to 864 f Physical Therapy Volume 90 Number 6 June 2010

6 Table 1. Details of Included Randomized Controlled Trials (RCTs) a Trial Participant Characteristics, Sample Size (N), and Duration of Complaint Recruited from referrals by specialist or primary care practitioners to hospital s physical therapy department Age: 18 y Main exclusions: prior spinal surgery, hematologic disease, and physical therapy in the preceding 6 mo N 143 Duration: 12 wk Interventions and Study Design Outcomes (Measures) and Time Points Included in This Review PEDro Score Critchley et al (2007) 11 de Jong et al (2005) 91 Heymans et al (2006) 26 Leeuw et al (2008) 90 Referred for outpatient behavioral rehabilitation and reporting substantial fear of movement Age: y Main exclusions: pregnancy and psychopathology N 6 Duration: 6 mo Patients who visited their occupational physicians Age: y Main exclusions: specific pathologies, pregnancy, and legal conflicts at work N 299 Duration: 7 wk (at least 1 mo of pain and 3 wk of sick leave) Recruited by physicians and from advertisements Age: y More than 3 points on the RMDQ-24 More than 33 points on the Tampa Scale for Kinesiophobia N 85 Duration: 3 mo Lindstrom and colleagues Workers in industrial ( blue-collar ) jobs, on sick (1992) 92,93 leave for 6 wk due to any low back pain diagnosis and not on sick leave for 12 wk before the current leave Main exclusions: specific pathologies, surgery, and psychiatric diagnosis N 103 Duration: 8 wk of sick leave Linton et al (2008) 94 Nicholas et al (1991) 15 Recruited from primary care facilities and advertisements in local newspaper More than 35 points on the Tampa Scale for Kinesiophobia Age: y N 46 Duration: 3 mo Referred to a pain clinic Age: y Main exclusion: no compensation claim within 12 mo N 62 Duration: 6 mo Graded activity vs motor control vs manual therapy home exercises (RCT) Graded activity education vs graded exposure education (crossover study) Graded activity (highintensity back school) usual care vs low-intensity back school (one third education, two thirds exercise) usual care vs usual care (RCT) Graded activity vs graded exposure (RCT) Graded activity vs waiting list (RCT) Graded exposure usual care vs waiting list usual care (RCT) Cognitive treatment vs behavioral treatment (graded activity) vs cognitive treatment relaxation vs behavioral treatment relaxation vs attention control condition vs physical therapy only (RCT) Pain (VAS) Disability (RMDQ-24) Quality of life (EuroQol Questionnaire) 6- and 12-mo follow-up evaluations Disability (RMDQ-24) Crossover in short term only, with 6- or 8-wk follow-up evaluations Pain (VAS) Disability (RMDQ-24) Global perceived effect (0 5) Sick leave (d) and return to work (d) 3- and 6-mo follow-up evaluations Pain (McGill VAS) Disability (QBPDS and PSFS) Posttreatment, 6-mo, and 12-mo follow-up evaluations Return to work (d), sick leave (d), and % of workers returning to work 12-mo follow-up evaluations Pain intensity (from Orebro) Disability (QBPDS) Posttreatment follow-up evaluations Pain rating chart 5-wk, 6-mo, and 12-mo follow-up evaluations Nicholas et al (1992) 14 Referred from a pain clinic and by specialist and general medical practitioners Age: y Main exclusion: compensation claim due for settlement within 12 mo N 20 Duration: 6 mo Cognitive behavioral treatment physical therapy (graded activity) vs physical therapy attention control condition (RCT) Pain rating chart 5-wk and 6-mo follow-up evaluations 6 (Continued) June 2010 Volume 90 Number 6 Physical Therapy f 865

7 Table 1. Continued Trial Participant Characteristics, Sample Size (N), and Duration of Complaint Recruited by health care professionals and from hospital waiting lists Age: y Main exclusions: spinal surgery, specific pathologies, and contraindications to exercise N 260 Duration: 6 wkbut 12 wk Interventions and Study Design Outcomes (Measures) and Time Points Included in This Review PEDro Score Pengel et al (2007) 23 Smeets et al (2006) 22 Smeets et al (2008) 24 Staal et al (2004) 16 Hlobil et al (2005) 98 Steenstra et al (2006) 95 Anema et al (2007) 96 van der Roer et al (2008) 25 Vlaeyen et al (2002) 97 Woods and Asmundson (2008) 12 Recruited by general practitioners and medical specialists Age: y More than 3 points on the RMDQ-24 Ability to walk at least 100 m Main exclusions: specific pathologies, comorbidities, and clear treatment preference N 227 Duration: 3 mo Workers employed by a major Dutch airline Main exclusions: signs of nerve root compression and cardiovascular problems N 88 Duration: 6 wk (for patients included in this review) Recruited by occupational physicians Patients in return-to-work program at 2 6 wk of sick leave Age: y Main exclusions: specific low back pain, cardiovascular and psychiatric contraindications, pregnancy, and sick leave granted for low back pain 1 mo N 112 Duration: 8 wk of sick leave Age: y New episode of nonspecific back pain Inability to resume daily activities in the last 3 wk N 114 Duration: 12 wk Referred for outpatient behavioral rehabilitation and reporting substantial fear of movement Age: y Main exclusions: pregnancy and psychopathology N 6 Duration: 6 mo Recruited from newspapers, advertisements, and posters Age: y More than 38 points on the Tampa Scale for Kinesiophobia Main exclusions: pending medical investigations and back surgery N 83 Duration: 6 mo Graded activity advice vs sham exercises advice vs graded activity sham advice vs sham graded activity sham exercises (RCT factorial design) Graded activity problem solving vs active physical treatment vs graded activity problem solving active physical treatment vs waiting list (RCT) Graded activity vs usual care (RCT) Graded activity vs usual care (advice to stay active care by general practitioners) (RCT) Graded activity vs guidelines (physical therapy according to clinical guidelines) (RCT) Graded activity vs graded exposure (crossover study) Graded activity vs graded exposure vs waiting list (RCT) Pain (VAS) Disability (PSFS) Global perceived effect ( 5 to5) 6-wk, 4-mo, and 12-mo follow-up evaluations Pain (McGill VAS) Disability (RMDQ-24) Global perceived effect (1 7) 10-wk, 6-mo, and 12-mo follow-up evaluations Pain (VAS) Disability (RMDQ-24) Absence from work (d) 3-mo, 6-mo, and 12-mo follow-up evaluations Pain (VAS) Disability (RMDQ-24) Sick leave (d) and return to work (d) 3-mo, 6-mo, and 12-mo follow-up evaluations Pain (numerical rating scale) Disability (RMDQ-24) Global perceived effect (0 6) 6-wk, 6-mo, and 12-mo follow-up evaluations Pain (VAS) Disability (RMDQ-24) Crossover with 12-wk and 12- mo follow-up evaluations Pain Disability Index McGill Pain Questionnaire (short form) 4-wk follow-up evaluations a VAS visual analog scale, RMDQ Item Roland-Morris Disability Questionnaire, QBPDS Quebec Back Pain Disability Scale, PSFS Patient-Specific Functional Scale, Orebro Orebro Musculoskeletal Pain Questionnaire. 866 f Physical Therapy Volume 90 Number 6 June 2010

8 Table 2. Details of Graded-Activity (GA) and Graded-Exposure (GE) Interventions Graded Activity and Graded Exposure for Low Back Pain Study Critchley et al (2007) 11 de Jong et al (2005) 91 Heymans et al (2006) 26 Duration of GA and GE Interventions Maximum of 9 sessions of 90 min Graded activity for 32 h over 8 wk and graded exposure for 24hover6wk 16 sessions of 1 h over 8wk Leeuw et al (2008) 90 Graded activity in 26 sessions of 1-h treatments starting twice/wk and graded exposure in 16 sessions of 1-h treatments starting twice/wk Lindstrom et al (1992) 92 Linton et al (2008) 94 3 d/wk until return to work individual sessions of behavioral therapy and 8 10 sessions of exposure in vivo Expertise of the Caregiver and Integrity Check Physical therapists who had at least 2yof clinical experience and who agreed to treat according to the trial protocol were provided brief training. The hospitals had their own common internal teaching program for the leaders of the graded-activity program; otherwise, there was no further training. Adherence to the protocol was assessed by recording treatments from physical therapy notes after treatment allocation was revealed. Home Program Plans for Future Relapse Adherence Participants reported difficulty attending classes twice/wk Supervised physical therapist Physical therapists with at least 6 mo of experience were trained in both treatments and received a manual with guidelines for dealing with problems. Therapists attended collective supervision sessions 3 times/y. A sophisticated method for judging adherence to treatment and contamination was used. Thirty random samples from 265 recorded sessions were judged. Home exercises were given during the treatment period Most practice of exercises occurred at home, and progress was evaluated at sessions; home exercises were given only for the gradedactivity group Physical therapist No home exercises Therapy was performed by psychologists who were trained by Vlaeyen et al 97 and also received support from a physical therapist. A written treatment manual was developed for the study. Sessions ended with home assignment that incorporated movements into activities of daily living at home and work Patients attended an average of 13 sessions; 70% of workers completed all treatments, and 10% received no treatment 29% of patients receiving graded exposure and 33% of patients receiving graded activity did not finish the treatments (Continued) June 2010 Volume 90 Number 6 Physical Therapy f 867

9 Table 2. Continued Study Duration of GA and GE Interventions One 2-h session and one 1.5-h session/wk for 5 wk (only 5 sessions were cognitive behavioral therapy) Expertise of the Caregiver and Integrity Check Home Program Plans for Future Relapse Adherence Nicholas et al (1991) 15 Nicholas et al (1992) 14 Pengel et al (2007) 23 Smeets et al (2006) 22 Smeets et al (2008) 24 Staal et al (2004) 16 Hlobil et al (2005) 98 One 2-h session and one 1.5-h session/wk for5wk 12 sessions over 6 wk: 3 times/wk in weeks 1 and 2, twice/wk in weeks 3 and 4, and once/wk in weeks 5 and 6; participants also attended 3 sessions of advice 18 sessions over 10 wk, changing from 3 sessions/wk to 1 session/wk, for a total of 11.5 h of treatment 1-h exercise session twice/wk until return to work or a maximum of 3 mo Therapy was performed by a clinical psychologist with5yofexperience and registered physical therapists under the supervision of a senior physical therapist. Two masked raters independently and correctly identified 6 audiotapes randomly selected from cognitive behavioral, behavioral, and attention control conditions, providing evidence that the content of each condition was consistent with the treatment protocol. One hour of the first session each week was conducted by a physical therapist, and the other hour was conducted by a psychologist who had 5 y of experience since completing clinical qualifications. A physical therapist under the supervision of a senior physical therapist conducted the second session each week. Registered physical therapists received training from an experienced clinical psychologist. An investigator recorded and assessed a sample treatment session and visited each treatment site regularly to monitor delivery. All therapists received extensive training before the start of the trial and refresher courses. Social workers and psychologists had at least 5yofexperience. Three therapists provided treatment according to a protocol. Therapists were trained before the study in three 2-h sessions and practiced patient-therapist interactions. Sessions before and after the study were audiotaped and analyzed for content. Patients were verbally praised by the psychologist for practicing physical therapy exercises each week Patients were encouraged by physical therapists to practice exercises at home, but no check on the practice was made and no specific reinforcement for performing exercises was provided by physical therapists Individualized home treatment was regularly reviewed, and patients were encouraged to continue after the end of the treatment period Home assignment was given to allow patients to practice skills in their daily lives Means of 9.4 (SD 3.2) sessions of exercises and 2.9 (SD 1.1) sessions of advice Means of 14.3 sessions (maximum 20) for the cognitive behavioral therapy group and 11.9 sessions (maximum 19) for the combination group (Continued) 868 f Physical Therapy Volume 90 Number 6 June 2010

10 Table 2. Continued Study Steenstra et al (2006) 95 Anema et al (2007) 96 Duration of GA and GE Interventions 26 sessions of 1 h (twice/wk) over 13 wk or until return to work van der Roer et al 10 individual sessions (2008) 25 and 20 group sessions 0.86). The median number of days to return to work was 139 days for the graded-activity group and 111 days for the usual-care group. Graded Activity Versus Other Forms of Exercise Six trials (597 patients) with a methodological quality ranging from 4 to 8 compared graded activity with other forms of exercise. 11,14,15,22,25,26 Data for pooling were available for pain, disability, and perceived effect in the short, intermediate, and long term. Data were pooled by use of a fixed-effects model for all outcomes except perceived effect in the short Expertise of the Caregiver and Integrity Check Physical therapists (47 in 16 centers) were trained by the physical therapist participating in the study of Staal et al. 77 A protocol was used to standardize the intervention. Physical therapists had a 2-h feedback session every 3 mo. Physical therapists attended 2 intensive 6-h workshops and were trained according to the study protocol. Study forms were used to assess therapist adherence to the study protocol; 18% of patients did not receive adequate treatment. Home Program and intermediate term, for which a random-effects model was used. Pooled results revealed no statistically significant differences between graded activity and other forms of exercise for pain, disability, or global perceived effect at each time point (Fig. 3). A hazard ratio for the difference in days to return to work was calculated by Heymans et al 26 ; the results revealed no significant differences between treatment groups (hazard ratio 1.4; 95% CI 1.0 to 1.9). The median number of days of sick leave was 85 days for the graded-activity Plans for Future Relapse Adherence 35% of patients did not adhere to the treatment; average frequency was 14.1 sessions Vlaeyen et al (2002) 97 3 wk of treatment Patients were given 1 wk to practice new skills in home situations after the end of the 3-wk treatment Woods and Asmundson (2008) 12 Eight 45-min sessions conducted on twiceweekly basis over 4wk Graded activity was conducted by a registered physical therapist, and graded exposure was conducted by a clinical psychologist graduate student trained in the therapy and supervised by a registered doctorallevel psychologist. group and 68 days for the exercise group. Graded Exposure Versus Graded Activity Two low-quality trials (PEDro scores of 3 and 4), with 6 patients each, compared graded activity with graded exposure. 91,97 Both studies had a crossover design, making it more difficult to reach any conclusion about the results because no evaluation of carryover effects was performed. Another 2 studies 12,90 had information that allowed pooling for pain and disability in the short term. These 2 trials had PEDro scores June 2010 Volume 90 Number 6 Physical Therapy f 869

11 Figure 2. Forest plot of results of randomized controlled trials comparing graded activity with a control (minimal intervention). Values represent effect sizes (weighted mean differences) and 95% confidence intervals. The pooled effect sizes were calculated by use of a fixed-effects model for all outcomes except pain and disability in the long term and perceived effect in the short term, for which a random-effects model was used. 870 f Physical Therapy Volume 90 Number 6 June 2010

12 Figure 3. Forest plot of results of randomized controlled trials comparing graded activity with other forms of exercise. Values represent effect sizes (weighted mean differences) and 95% confidence intervals. The pooled effect sizes were calculated by use of a fixed-effects model for all outcomes except perceived effect in the short and intermediate term, for which a random-effects model was used. June 2010 Volume 90 Number 6 Physical Therapy f 871

13 Figure 4. Forest plot of results of randomized controlled trials comparing graded activity with graded exposure. Values represent effect sizes (weighted mean differences) and 95% confidence intervals. The pooled effect sizes were calculated by use of a fixed-effects model. of 5 and 6 and included 146 patients. The results of the pooled analysis revealed no statistically significant differences between treatment groups for pain and disability in the short term (Fig. 4). Graded Exposure Versus a Minimal Intervention Two trials (104 patients) (PEDro scores of 4 and 5) compared graded exposure with a minimal intervention (waiting list or usual care). 12,94 Data for pooling were available for pain and disability in the short term. Pooled results revealed no statistically significant differences between treatment groups for pain in the short term (weighted mean difference on a scale of points [95% CI 12.3 to 4.9]) and disability in the short term (weighted mean difference on a scale of points [95% CI 19.4 to 12.3]) (Fig. 5). No data were available for quality of life or recurrence for any of the comparisons. Discussion In this systematic review, we compared the effects of graded activity versus a minimal intervention, graded activity versus other forms of exercise, graded activity versus graded exposure, and graded exposure versus a minimal intervention for patients with persistent LBP. Only 4 of the pooled effect sizes were statistically significant. These statistically significant results favored graded activity over a minimal intervention for pain and disability at short- and intermediate-term follow-up evaluations. However, the effect sizes were small, approximately 7 points on a 100-point scale; this finding may indicate that they are not clinically meaningful. Among the studies in which graded activity was compared with a minimal intervention, 2 trials had counterintuitive results: The minimal intervention provided better outcomes than graded activity. 12,95 The study of Woods and Asmundson 12 poten- Figure 5. Forest plot of results of randomized controlled trials comparing graded exposure with a control (waiting list or usual care). Values represent effect sizes (weighted mean differences) and 95% confidence intervals. The pooled effect sizes were calculated by use of a fixed-effects model for pain and a random-effects model for disability. 872 f Physical Therapy Volume 90 Number 6 June 2010

14 tially provided biased estimates of treatment effects because of a lack of masking, failure to conceal allocation, and a loss to follow-up of approximately 50%. In contrast, the study of Steenstra et al 95 seemed to be well conducted, and we are unable to explain why the results for pain and disability in the long term favored a minimal intervention over graded activity. The results of the 3 trials that measured return to work after graded activity versus a minimal intervention were conflicting. In 1 of these 3 trials, 92 graded activity provided a faster return to work than a minimal intervention. In the second trial, 26 there were no significant differences between treatment groups. In the third study, 95 a minimal intervention provided a faster return to work than graded activity. The authors of the last study suggested that the longer delay in return to work for patients in the graded-activity group was attributable to a delay in the beginning of the treatment for this group. Despite the possible bias present in these studies, the results regarding return to work are inconclusive, and further research is needed. The results for comparisons of graded activity with other forms of exercise are in accordance with the recommendations of most clinical guidelines, which suggest that no form of exercise is more effective than another. 99 However, Smeets et al 84 did find that graded activity was more cost-effective than active physical treatment (exercises), drawing attention to the need for more costeffectiveness studies. In many cases, the interventions were implemented by trained physical therapists. Although adherence to the treatment protocols was assessed in 6 of the 15 studies, no results of this assessment were provided; therefore, it remains unclear whether the interventions were properly administered. Additionally, it is uncertain whether 2 or 3 sessions of training provided the clinicians with sufficient skills to effectively implement the treatment protocols. There is some evidence that trained physical therapists lack the skills required to effectively implement a psychologically based intervention. 100 One important feature of both treatments considered in this review is inclusion of a plan for managing relapses. This strategy not only helps patients deal with anxiety and fear about a flare-up but also may assist in maintaining the long-term effects of the intervention. Although none of the studies mentioned the use of this strategy, it is not clear whether it was omitted from the intervention or simply was not reported. There are some limitations to the conclusions of this review. These include the low quality of the studies, primarily those evaluating graded exposure, the use of various types of outcomes, and differences in the implementation of the interventions. The small number of trials testing graded exposure is also an important limitation of this review. Some authors 92,97 suggested that graded exposure may have larger effects than graded activity because the former intervention more specifically targets a patient s fears; however, the small number of trials included in this review for this comparison does not allow a reliable conclusion to be drawn. Future research should focus on the conduct of higher-quality trials evaluating issues such as return to work and compensation and the conduct of cost-effectiveness studies. Additionally, trials including patients identified as having greater fear behaviors should be carried out because these patients may respond better to both graded activity and graded exposure interventions. 97 Conclusion The results of this systematic review suggest that graded activity is slightly more effective than a minimal intervention but not more effective than other forms of exercise for pain, disability, and global perceived effect in the short and intermediate terms for patients with persistent nonspecific LBP. The results also suggest that graded exposure is no more effective than a minimal intervention or graded activity. Because of the poor reporting in many of the studies, it often was unclear precisely how the interventions were implemented. Additionally, the smaller number and lower quality of graded-exposure trials limits the conclusions that can be drawn regarding this intervention. Ms Macedo, Dr Smeets, Dr Maher, and Dr Latimer provided concept/idea/research design. All authors provided writing and data collection. Ms Macedo and Dr Smeets provided data analysis. Ms Macedo provided project management. Dr Smeets, Dr Maher, and Dr Latimer provided facilities/equipment. Dr Smeets provided institutional liaisons. Dr Smeets, Dr Maher, Dr Latimer, and Dr McAuley provided consultation (including review of manuscript before submission). The authors of this systematic review thank the authors of the retrieved trials who provided additional information or data from their trials. Ms Macedo holds a PhD scholarship jointly funded by the University of Sydney and the Australian Government. Dr Maher s research fellowship is funded by Australia s National Health and Medical Research Council, and Dr Latimer s research fellowship is funded by the Australian Research Council. The results of this study were presented at the Australian Physiotherapy Association Conference; October 1 5, 2009; Sydney, Australia. This article was submitted September 14, 2009, and was accepted January 24, DOI: /ptj June 2010 Volume 90 Number 6 Physical Therapy f 873

15 References 1 Vlaeyen J, Linton S. Pain-related fear and its consequences in chronic musculoskeletal pain. In: Linton S, ed. New Avenues for the Prevention of Chronic Musculoskeletal Pain and Disability, Pain Research and Clinical Management. Amsterdam, the Netherlands: Elsevier; 2002: Fordyce WE. Behavioral Methods for Chronic Pain and Illness. St Louis, MO: Mosby; Asmundson GJG, Norton PJ, Norton GR. Beyond pain: the role of fear and avoidance in chronicity. Clin Psychol Rev. 1999;19: Rossignol M, Arsenault B, Dionne C, et al. Clinic on Low-Back Pain in Interdisciplinary Practice (CLIP) Guidelines. Montreal, Quebec, Canada: Direction de Santé Publique, Agence de la Santé et des Services Sociaux de Montréal; Airaksinen O, Brox JI, Cedraschi C, et al. European guidelines for the management of chronic nonspecific low back pain. Eur Spine J. 2006;15(suppl 2):S192 S Ostelo RW, van Tulder MW, Vlaeyen JW, et al. Behavioural treatment for chronic low-back pain. Cochrane Database Syst Rev. 2005;1:CD Bombardier C, Van Tulder M, Pennick V, et al; Cochrane Back Group. About the Cochrane Collaboration (Cochrane Review Groups [CRGs]). 2008, issue 3. 8 Johnson RE, Jones GT, Wiles NJ, et al. Active exercise, education, and cognitive behavioral therapy for persistent disabling low back pain: a randomized controlled trial [with consumer summary]. Spine. 2007;32: Turner JA. Comparison of group progressiverelaxation training and cognitive-behavioral group therapy for chronic low back pain. J Consult Clin Psychol. 1982;50: von Korff M, Balderson BH, Saunders K, et al. A trial of an activating intervention for chronic back pain in primary care and physical therapy settings. Pain. 2005; 113: Critchley DJ, Ratcliffe J, Noonan S, et al. Effectiveness and cost-effectiveness of three types of physiotherapy used to reduce chronic low back pain disability: a pragmatic randomized trial with economic evaluation. Spine. 2007;32: Woods MP, Asmundson GJG. Evaluating the efficacy of graded in vivo exposure for the treatment of fear in patients with chronic back pain: a randomized controlled clinical trial. Pain. 2008;136: Turner JA, Clancy S, McQuade KJ, et al. Effectiveness of behavioral therapy for chronic low back pain: a component analysis. J Consult Clin Psychol. 1990;58: Nicholas MK, Wilson PH, Goyen J. Comparison of cognitive-behavioral group treatment and an alternative nonpsychological treatment for chronic low back pain. Pain. 1992;48: Nicholas MK, Wilson PH, Goyen J. Operant-behavioural and cognitivebehavioural treatment for chronic low back pain. Behav Res Ther. 1991;29: Staal JB, Hlobil H, Twisk JW, et al. Graded activity for low back pain in occupational health care: a randomized, controlled trial. Ann Intern Med. 2004;140: Sanders SH. Operant conditioning with chronic pain: back to basics. In: Gatchel RJ, Turk DC, eds. Psychological Approaches to Pain Management: A Practitioner s Handbook. New York, NY: The Guilford Press; 1996: Vlaeyen JWS, de Jong J, Sieben JM, et al. Graded exposure in vivo for pain-related fear. In: Turk DC, Gatchel RJ, eds. Psychological Approaches to Pain Management: A Practitioner s Handbook. 2nd ed. New York, NY: The Guilford Press; 2002: Vlaeyen JWS, de Jong J, Leeuw M, et al. Fear reduction in chronic pain: graded exposure in vivo with behavioural experiments. In: Asmundson GJ, Vlaeyen JWS, Crombez G, eds. Understanding and Treating Fear of Pain. New York, NY: Oxford University Press; 2004: Sherrington C, Herbert RD, Maher CG, et al. PEDro: a database of randomized trials and systematic reviews in physiotherapy. Man Ther. 2000;5: Higgins JPT, Green S, editors. Cochrane Handbook for Systematic Reviews of Interventions [updated September 2006]. In: The Cochrane Library, Issue 4, Chichester, UK: John Wiley & Sons, Ltd. 22 Smeets RJEM, Vlaeyen JWS, Hidding A, et al. Active rehabilitation for chronic low back pain: cognitive-behavioral, physical, or both? First direct posttreatment results from a randomized controlled trial. BMC Musculoskelet Disord. 2006;7:5. 23 Pengel LHM, Refshauge KM, Maher CG, et al. Physiotherapist-directed exercise, advice, or both for subacute low back pain: a randomized trial. Ann Intern Med. 2007;146: Smeets RJEM, Vlaeyen JWS, Hidding A, et al. Chronic low back pain: physical training, graded activity with problem solving training, or both? The one-year post-treatment results of a randomized controlled trial. Pain. 2008;134: van der Roer N, van Tulder M, Barendse J, et al. Intensive group training protocol versus guideline physiotherapy for patients with chronic low back pain: a randomised controlled trial. Eur Spine J. 2008;17: Heymans MW, de Vet HC, Bongers PM, et al. The effectiveness of high-intensity versus low-intensity back schools in an occupational setting: a pragmatic randomized controlled trial. Spine. 2006;31: Review Manager (RevMan) [computer program]. Version 5.0. Copenhagen, Denmark: The Nordic Cochrane Centre, The Cochrane Collaboration, Menzel NN, Robinson ME. Back pain in direct patient care providers: early intervention with cognitive behavioral therapy. Pain Manag Nurs. 2006;7: Alaranta H, Rytokoski U, Rissanen A, et al. Intensive physical and psychosocial training program for patients with chronic low back pain: a controlled clinical trial. Spine. 1994;19: Bendix T, Bendix A, Labriola M, et al. Functional restoration versus outpatient physical training in chronic low back pain: a randomized comparative study. Spine. 2000;25: Fairbank J, Frost H, Wilson-MacDonald J, et al. Randomised controlled trial to compare surgical stabilisation of the lumbar spine with an intensive rehabilitation programme for patients with chronic low back pain: the MRC Spine Stabilisation Trial. BMJ. 2005;330: Gohner W, Schlicht W. Preventing chronic back pain: evaluation of a theorybased cognitive-behavioural training programme for patients with subacute back pain. Patient Educ Couns. 2006;64: Harkapaa K, Jarvikoski A, Mellin G, et al. Health locus of control beliefs and psychological distress as predictors for treatment outcome in low-back pain patients: results of a 3-month follow-up of a controlled intervention study. Pain. 1991; 46: Jellema P, Van Der Roer N, Van Der Windt DAWM, et al. Low back pain in general practice: cost-effectiveness of a minimal psychosocial intervention versus usual care. Eur Spine J. 2007;16: Magnussen L, Rognsvag T, Tveito TH, et al. Effect of a brief cognitive training programme in patients with long-lasting back pain evaluated as unfit for surgery. J Health Psychol. 2005;10: Jousset N, Fanello S, Bontoux L, et al. Effects of functional restoration versus 3 hours per week physical therapy: a randomized controlled study. Spine. 2004; 29: Keller A, Brox JI, Gunderson R, et al. Trunk muscle strength, cross-sectional area, and density in patients with chronic low back pain randomized to lumbar fusion or cognitive intervention and exercises. Spine. 2003;29: Jensen IB, Bergstrom G, Ljungquist T, et al. A 3-year follow-up of a multidisciplinary rehabilitation programme for back and neck pain. Pain. 2005;115: Jensen IB, Bergstrom G, Ljungquist T, et al. A randomized controlled component analysis of a behavioral medicine rehabilitation program for chronic spinal pain: are the effects dependent on gender? Pain. 2001;91: f Physical Therapy Volume 90 Number 6 June 2010