- Email: [email protected]
Effectiveness of a standardized back school program for patients with chronic low back pain after implementation in routine rehabilitation care
Accepted Manuscript Title: Effectiveness of a standardized back school program for patients with chronic low back pain after implementation in routine rehabilitation care Authors: Karin Meng, Stefan Peters, Hermann Faller PII: DOI: Reference:
S0738-3991(17)30011-3 http://dx.doi.org/doi:10.1016/j.pec.2017.01.011 PEC 5564
To appear in:
Patient Education and Counseling
Received date: Revised date: Accepted date:
9-3-2016 22-12-2016 13-1-2017
Please cite this article as: Meng Karin, Peters Stefan, Faller Hermann.Effectiveness of a standardized back school program for patients with chronic low back pain after implementation in routine rehabilitation care.Patient Education and Counseling http://dx.doi.org/10.1016/j.pec.2017.01.011 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 Effectiveness of a standardized back school program for patients with chronic low back pain after implementation in routine rehabilitation care
Karin Meng, Stefan Peters, Hermann Faller
Department of Medical Psychology, Medical Sociology, and Rehabilitation Sciences, University of Würzburg, Würzburg, Germany
Corresponding author Dr. Karin Meng University of Würzburg, Department of Medical Psychology, Medical Sociology, and Rehabilitation Sciences, Klinikstr. 3, D-97070 Würzburg, Germany Tel: (+)49 931 3182074 E-Mail: [email protected]
2 Highlights
Back schools are an integral part of rehabilitation in chronic low back pain.
Implementation of standardized programs is a challenge for routine care.
The implementation fidelity of a biopsychosocial back school was heterogeneous.
Despite heterogeneous implementation, the back school seems to be effective.
Further implementation within inpatient orthopedic rehabilitation may be recommended.
3 Abstract Objective. To evaluate the effectiveness of a standardized, patient-oriented, biopsychosocial back school after implementation in inpatient orthopedic rehabilitation. Methods. A multi-center, quasi-experimental controlled study of patients with low back pain (n = 535) was conducted. Patients in the control group received the traditional back school before implementation of the new program (usual care); patients in the intervention group received the new standardized back school after implementation into routine care. Patients` illness knowledge and conduct of back exercises (primary outcomes) and secondary selfmanagement outcomes and treatment satisfaction were obtained at admission, discharge, and 6 and 12 months after rehabilitation. Results. We found a significant small between-group intervention effect on patients` illness knowledge in medium- to long term (6 months: η2 = 0.015; 12 months: η2 = 0.013). There were trends for effects on conduct of back exercises among men (6 and 12 months: η2 = 0.008 both). Furthermore, significant small effects were observed for treatment satisfaction at discharge and physical activity after 6 months. Conclusions. The standardized back school seems to be more effective in certain outcomes than a usual care program despite heterogeneous program implementation. Practice Implications. Further dissemination within orthopedic rehabilitation may be encouraged to foster self-management outcomes.
Key words Chronic back pain, patient education, implementation effectiveness, controlled trial, rehabilitation
4 1. Introduction Back pain is a major health problem worldwide and responsible for disability and socioeconomic costs [1,2]. It is a leading cause for inpatient medical rehabilitation in Germany. In this regard, multidisciplinary biopsychosocial rehabilitation seems to be more effective than usual care and physical treatment in reducing pain and disability in the longterm [3]. However, how these programs should be designed to further improve outcome is not entirely clear [4]. Patient education programs, particularly back schools, are a mandatory part of rehabilitation for patients with chronic low back pain in Germany [5]. According to evidence, they should be based on the biopsychosocial model [6] and conducted multidisciplinary [7,8]. Additionally, the incorporation of strategies to promote long-term adherence to physical activity is recommended [9-11]. However, seemingly many back school programs offered as part of German medical rehabilitation are still based on a biomechanical approach. Many also do not fulfill quality requirements, such as the use of manuals, patient-oriented didactics and evaluation of effectiveness [12]. Therefore, a new standardized, patient-oriented and biopsychosocial back school with a theory-based approach employed to foster long-term adherence to physical activity was developed. A previous randomized, controlled trial in one inpatient rehabilitation clinic demonstrated the superior efficacy of this program with regard to patients’ illness knowledge, determinants of physical activity, fear-avoidance beliefs, and treatment satisfaction in short-term, as well as self-management behaviors up to 1 year [13,14]. Implementation of this back school may be complex because of its manual-based, modular and interdisciplinary structure as well as its new conception regarding contents and methods, requiring both organizational and behavioral changes. Generally, the implementation of standardized patient education programs represents a challenge for routine rehabilitation practice and is often not entirely successful, i.e. programs
5 are only partially implemented or modified [15,16]. This may pose a threat to their effectiveness in routine care. International reviews confirm that program adaptations are common and vary by providers [e.g. 17,18], and there is empirical support from health promotion and prevention research to the assumption that the level of implementation quality affects outcomes [17]. Overall, there is a need for more systematic evaluation of the implementation of patient education programs within inpatient medical rehabilitation. Our study is an implementation effectiveness trial, designed to address both implementation and clinical intervention aims [19,20]. Presenting a next step after the above-mentioned evaluation of the back school under standardized conditions (efficacy evaluation), it uses a mixed-methods approach for both the process evaluation of implementation and the evaluation of effectiveness in routine care. Results of the process evaluation showed that structural adherence as based on the transfer of key elements according to the manual varied between rehabilitation clinics. Minor adaptations appeared to be necessary to overcome structural barriers, achieve acceptance and facilitate program implementation [21]. In this paper, we report results of the effectiveness trial. The main research objective is to evaluate the effectiveness of a new back school after implementation in inpatient orthopedic rehabilitation with regard to patient outcomes. We hypothesized that the implemented new back school would be superior to traditional care regarding illness knowledge and back exercises (primary outcomes). Additionally, we expected superior effectiveness of the implemented program regarding treatment satisfaction as well as several self-management behaviors, such as physical activity, back posture habits, and coping with pain (secondary outcomes). Because of the results of our prior trial [14], we explored moderator effects of gender for behavioral outcomes.
2. Methods
6 2.1 Design and procedure Effectiveness within routine care was evaluated in a multi-centre, quasi-experimental controlled study in 535 patients with low back pain in 4 orthopedic rehabilitation clinics in Germany. Data were collected at admission and discharge as well as after 6- and 12-months with patient-reported questionnaires. Control patients received the traditional back school program of the hospital as delivered before implementation of the new program (CG); intervention patients received the new standardized back school as delivered after implementation into routine-practice (IG). Patients with a primary diagnosis of chronic low back pain (ICD-10-GM: M51, M53, M54) were included in the trial. Exclusion criteria were inadequate German language ability, age below 18 or above 70 years, severe visual or hearing impairment, severe comorbid psychiatric disorder, poor health state, and submitted pension application. Patients were blind to the allocated study condition; blinding of clinicians was impossible due to study design, though. Sample size was powered to detect small to medium effects in the primary outcomes (cf [13,14]; knowledge: d = 0.48; back exercise: d = 0.29; 2sided α = 0.05, 1- = 0.8). Therefore, 376 persons were required. We included 540 persons, based on an estimated drop-out of about 30%. As power to detect smaller between-group effects on the secondary outcomes may be lacking, effect sizes were reported throughout. The study was approved by the Ethics Committee of the Faculty of Medicine, University of Würzburg, on 25. March 2011 (ref: 48/11) and registered on WHO International Clinical Trials (DRKS00004835). 2.2 Participants Patients were consecutively recruited between May 2011 and April 2013; one-year follow-up was finished in May 2014. Figure 1 shows the participant flow. A physician evaluated the inclusion criteria at admission. 687 patients eligible were asked to participate in the study and
7 557 participated (81%) and signed the informed consent form. 22 persons had to be excluded. Thus, 535 persons comprised the initial study sample. Follow-up rates exceeded 65% at all time points. Analyses were performed with all persons who participated in at least one follow-up assessment. Non-responder analyses revealed no differences between participants and non-participants in gender and age (p > 0.05). Drop-out analyses at follow-up points were performed for socio-demographic variables, main outcome variables and allocation to study group. Patients who had not provided data at the 6- and 12months follow-ups, respectively, were younger, more often unfit to work, reported higher pain intensity, and had lower illness knowledge at baseline. Drop-outs at 12-months follow-up reported lower emotional well-being at baseline than did completers. Furthermore, drop-out after 6 months and by trend after 12 months was higher in the IG than the CG. 2.3 Intervention All patients took part in the 3-week multidisciplinary inpatient rehabilitation. CG patients received the traditional back school programs of the rehabilitation clinics as delivered before implementation of the new program (usual care). IG patients received the new standardized back school as delivered after implementation into routine-practice. Table 1 summarizes characteristics of back school programs in the CG as well as fidelity of the standardized back school program in the IG (implementation of structural key elements, contents and methods according to the manual). Details regarding fidelity assessment have been previously described [21]. CG: Traditional non-standardized back school programs (i.e. no closed group format, not manual-based, less patient oriented didactics, no interdisciplinary training team) with biomechanical orientation. Contents of the lessons included correct back posture and movements as well as back exercises. Theory-based intervention techniques to promote
8 physical activity and psychosocial aspects of low back pain were not systematically addressed. IG: Standardized back school program [13,14] with 7 patient-oriented, interactive sessions of 55 minutes, held in small groups of a closed format. The program is manual-based and interdisciplinary with sessions led by a physiotherapist (five sessions), an orthopedist (one session) and a psychologist (one session), respectively. In each session, patients are actively involved in the educational process using a combination of didactic methods (short lectures, group discussion, small group work, practice/exercises, and individual work). Didactic materials included presentations, flipcharts, handouts and work sheets. Contents of the sessions included basic knowledge about back pain (e.g. development, symptoms, treatment) with regard to physical (e.g. body perception, positions for short-term relief) and psychosocial aspects (fear-avoidance-endurance model, coping behavior, risk factors). Spine-related exercises (strength training, active stabilization) and theory-based techniques to promote physical activity (e.g. motivation, self-regulation) were also part of the sessions. All sessions followed a common bio-psycho-social model. However, realization of the program varied between clinics. 2.4 Outcomes and measurements The primary outcomes were patients` illness knowledge and conduct of back exercises. Secondary outcomes included physical activity, back posture and movements, and pain coping strategies, as well as satisfaction with the back school. Outcomes were assessed by validated measures. This paper focuses on primary and self-management outcomes that were of main interest at follow-up [14]. Furthermore, we report effects on patient satisfaction as an element of patient-centered outcome in short-term. Illness knowledge. Patients were asked to judge 44 statements about back pain and its treatment as true or false. Item examples: “Depression is a risk factor for low back pain”,
9 “Physical activity should last longer than 15 minutes to have a positive impact”. Items are summed up to yield a score ranging from 0 to 44, with higher scores corresponding to higher knowledge; missing answers were counted as wrong answers. This questionnaire was developed within the previous evaluation study [13,14]. Back exercises. A 3-item modified version of a self-report questionnaire was administered [13,14,22]. Patients were asked about the amount of back exercises they did during the last week using three items; e.g. “Stretching”, “Strength training”. The scale ranged from 0 (seldom or never) to 4 (almost every day), with higher scores indicating more exercises. It showed good internal consistency (Cronbach’s α = 0.85). Back posture habits. A 7-item modified version of a validated self-report questionnaire was administered [23]. Test-retest reliability after 5-8 weeks is rtt = 0.77. The scale is in accordance with ratings based on behavioral monitoring. Patients were asked how often they performed adequate sitting, standing, walking, lifting and carrying in their daily routine during the last week; e.g. “Sitting with a back-friendly posture”, “Keeping the back in a straight and stable position when lifting heavy objects”. The scale ranged from 0 (seldom or never) to 4 (almost every day), with higher scores indicating more adequate back-related behavior. The internal consistency is α = 0.85. Physical activity. The Freiburger Questionnaire on Physical Activity (FfkA) is a reliable and valid German self-report measure using eight items [24]. Test-retest reliability is high (14 days: rtt = 0.95). Validity was established by correlations with maximum oxygen uptake (sports activities: partial r = 0.42). Patients were asked to indicate their health-related physical activities in daily routine, leisure time, and sports and to report for each activity the frequency and duration during the last week or month, respectively. All activity reports were transformed to produce a total activity score (in hours per week) as well as three sub-scores (basic activity, leisure time activity, sports activity).
10 Pain coping strategies. Coping with pain was assessed using the self-report German Pain Management Questionnaire (FESV), consisting of 24 items with 6-point numerical scales (1 = don’t agree at all, 6 = fully agree) [25]. Items are divided into 2 domains of coping strategies, cognitive strategies and behavioral strategies, each comprising 3 scores: cognitive strategies include action-oriented coping (e.g., “If I have pain, I have a plan how to handle that pain”), cognitive restructuring (e.g., “If I have pain, I will not give up”), and subjective coping competence (e.g., “If I have pain, I remind myself that I can deal with it much better than in the past”); behavioral strategies include mental distraction (e.g., “If I have pain, I thumb through a magazine”), counter-activities (e.g., “If I have pain, I distract myself with activities around my home or garden”), and relaxation (e.g., “If I have pain, I concentrate on calm and steady breathing”). Scores may range from 4 to 24, with higher scores corresponding to stronger endorsement of the respective coping strategy. Cronbach’s α for the subscales was between 0.70 and 0.84. Test-retest reliability and validity have been demonstrated [25]. Treatment satisfaction. Patients were asked to judge the educational program using 11 items [13]. Each item comprised patient-centered educational criteria on a 6-point scale (1 = very good, 6 = very poor). Items were averaged into 3 subscores – content (topic selection, comprehensibility, alignment, usefulness, delivery), group/interaction (opportunity to make comments, exchange of experience, group atmosphere, size), and materials (slides, handouts). Scores showed good reliability (Cronbach’s α 0.79-0.85). Furthermore, one item inquired the satisfaction with the program in general. 2.5 Statistical analysis Statistical analyses were performed using SPSS 22.0 for Windows. For missing data in accordance with missing at random, five imputed data sets were created using the multiple imputation procedure implemented in SPSS. Missing values due to drop-out were analyzed by pair-wise deletion. Non-response analyses and drop-out analyses were carried out by
11 independent group comparisons using t tests for continuous variables and chi-square tests for categorical variables. Treatment effects (between-group effects) were evaluated separately for each follow-up time point using analysis of covariance (ANCOVA) adjusting for baseline values [26]. Statistical significance (p < 0.05, 2-sided) and effect sizes (η²) were reported for all between-group differences [27]. In addition, within-group effects, including standardized effect sizes (SES) and accompanying 95% confidence intervals (CIs) were reported for both study groups. Gender subgroup analysis was performed by including gender as an additional factor in the ANCOVA and examining interaction effects. When trends to significant interactions emerged (p < 0.10), simple effects analyses were performed by gender subgroups [28]. The significance level for interaction effects was not adjusted due to its exploratory nature.
3. Results 3.1 Patient characteristics Table 2 presents characteristics of participants by study group. The initial sample consisted of 535 rehabilitants with chronic low back pain. Patients` mean age was 50 years (SD = 8.3), 55% were female, and 78% reported to live together with a partner. The sample mostly consisted of employed persons (87%). The duration of sick leave due to back pain within the last 6 months averaged 34 days (SD = 51, Median = 10). Symptom duration was longer than 5 years for about half of the sample (53%). Subjective health status was reduced as compared to the general population [29]. There were no systematic differences between participants in the CG and IG. 3.2 Primary outcome Table 3 presents the intermediate- and long-term results on primary and secondary outcomes.
12 Participants of the IG, as compared to the CG, showed superior knowledge about chronic back pain and its treatment 6 and 12 months after rehabilitation (6 months: p = 0.018, η2 = 0.015; 12 months: p = 0.039, η2 = 0.013). There were no significant intervention effects on back exercises 6 and 12 months after rehabilitation. However, interaction effects of gender and intervention condition were observed as trends (p < 0.10; Figure 2). Simple effect analyses indicated trends for small intervention effects on exercise behavior among men, but not women (6 months: men: p = 0.073, η2 = 0.008. women: p = 0.570, η2 < 0.001; 12 months: men: p = 0.106, η2 = 0.008. women: p = 0.480, η2 = 0.001). 3.3 Intermediate- and long-term effects on self-management behavior 6-months after rehabilitation, both IG and CG showed significant, small to medium improvements in the self-management behaviors physical activity (only IG), back posture habits and coping with pain (except counter-activities in both groups). SESs indicate that the IG improved more than the CG regarding physical activity. Accordingly, ANCOVA resulted in a significant, medium effect in favor of the IG for physical activity. Differences are based on a medium effect on everyday activities (p < 0.001, η2 = 0.067) and a small effect in sports activities (p = 0.044, η2 = 0.010), respectively. After adjustment for baseline imbalance, estimated mean physical activity per week in the IG was about 135 minutes longer than in the CG, with 86 minutes pertaining to everyday activities, 24 minutes to leisure time activities, and 22 minutes to sports activities, respectively. No significant differences between the groups were observed for back posture habits as well as cognitive and behavioral strategies of coping with pain. No significant interaction effects with gender emerged (p > 0.10). At 12-months follow-up, significant increases in physical activity and back posture habits occurred among patients of both groups, while within-group changes in coping with pain were heterogeneous. SESs indicated that patients in the CG improved more with regard to back posture habits. By contrast, a significant, small increase in relaxation was evident only in the
13 IG, but not the CG. ANCOVAs showed significant between-group effects for both outcomes. There were no intervention effects on physical activity or other pain coping strategies. Moderator analysis indicated significant interactions with gender regarding certain dimensions of coping, but not for physical activity and back posture habits. Women in the IG reported lower mental distraction and counter-activities than women in the CG, whereas no such differences were found among men (mental distraction: women: p = 0.088, η2 = 0.008; men: p = 0.207, η2 = 0.005. counter activities: women: p = 0.107, η2 = 0.008; men: p = 0.331, η2 = 0.003). 3.4 Short-term effects on treatment satisfaction Participants of the IG as compared to the CG showed significant higher treatment satisfaction with regard to “content” (p = 0.05, d = 0.17) and “materials” (p = 0.013; d = 0.23). However, there were no significant effects on “group/interaction” (p = 0.23, d = 0.11) or the satisfaction with the back school in general (p = 0.26; d = 0.10).
4. Discussion and Conclusion 4.1 Discussion In this implementation effectiveness trial, we evaluated the medium- and long-term routine effectiveness of a new back school as delivered after implementation in inpatient orthopedic rehabilitation in comparison with the respective traditional back schools of the study clinics. With respect to the primary outcomes, we found a significant small treatment effect on patients` illness knowledge but no effect on back exercises in the medium- to long-term. However, there were trends for effects on conduct of back exercises for men. Additionally, we found significant small effects for treatment satisfaction at discharge, as well as the amount of physical activity carried out after 6 months. However, there were no additional effects of the new back school on back posture habits or coping with pain (except relaxation in the long-
14 term). Thus, our primary and secondary hypotheses regarding self-management outcomes were only partially supported. The lack of additional effects may be explained by the high intensity of the whole rehabilitation package as well as variations in both implementation fidelity and control conditions. First, we compared two active treatments that were only a minor part of a comprehensive rehabilitation treatment, which patients from both IG and CG received. Thus, significant changes occurred in both groups and the incremental effects of the educational program may have been attenuated. Several intervention studies in the German rehabilitation setting have shown that for such comparisons small treatment effects on proximal outcomes may be expected at most [e.g. 14,30,31]. Moreover, only small effects are reported for self-management interventions (like back schools) in chronic back pain [32]. Secondly, we intended to test the routine effectiveness of the back school program after implementation in several clinics. Hence, the implementation of the program varied between clinics with regard to key elements, i.e. structure, contents and delivery. The interactive, patient-centered approach of the new back school is supposed to foster learning effects regarding knowledge and self-management skills, and the use of behavior change techniques [9] should promote the uptake of physical activity and exercises in everyday life. Modifications made during implementation may interfere with these assumed treatment mechanisms of the program. Additionally, the traditional back school programs (usual care) being rather heterogeneous may have influenced results. Nevertheless, results are in line with our prior efficacy trial [14] regarding treatment effects on illness knowledge, physical activity, treatment satisfaction and, by trend, on exercises for men after rehabilitation. Effect sizes or time points may vary due to different study design, samples as well as treatment fidelity, though. However, even varying implementation of the back school components into routine practice resulted in positive self-management outcomes. Thus, the interactive small group setting might support gaining knowledge even when other
15 components of delivery are partly modified. Current reviews show that motivational and behavior change approaches are successful in increasing physical activity using varying methods of delivery [33,34]. Generally, the patient-centered approach allowing for adaptations to both therapists’ and patients’ needs may foster the therapeutic alliance that predicts outcome in low back pain patients [35]. Adaptations are not necessarily associated with poor delivery and might also reflect high levels of facilitator competence [36]. In contrast to our prior study, we observed a small effect on back posture habits in the longterm favoring the CG. The degraded importance of posture habits within modern back schools may explain this result. Only one session addressed back posture habits with a focus on individual and situational variability. The therapists encouraged patients to try out and individually select suitable postures and movements. This approach aims to counteract fearavoidance beliefs and foster self-efficacy [37]. Furthermore, there were no effects on pain coping skills. However, cognitive behavioral pain management constituted only a small part of this back school. Overall, it is necessary to study the effect of back schools and other selfmanagement interventions on the development of coping strategies and its association with improvements in pain and functional levels [38]. Some treatment effects were moderated by gender. The new back school worked better for men regarding exercise behavior and for women regarding some coping strategies, respectively. These moderator effects partly match with gender effects observed in our prior study. So far, clear evidence that gender is an essential moderator in self-management interventions in musculoskeletal pain is lacking. Further research on this issue is recommended [10]. Our study has certain strengths and limitations. A strength is the comprehensive evaluation of the implementation of a new back school based on a controlled effectiveness trial. Reports on implementation fidelity are often insufficient [39]. However, providing relevant data may
16 contribute to a better understanding of patient outcomes. Additionally, we strengthened generalizability of previous efficacy results for this back school program by using a multicenter approach, allocating a representative sample and evaluating routine application of the program. The following limitations need to be considered. Randomization was not possible in this implementation effectiveness trial. However, IG and CG patient samples were comparable with regard to socio-demographic and medical data. In all analyses, baseline scores of outcome parameters were adjusted for. Second, although drop-out occurred, its proportion was in the range of other studies. Factors associated with drop-out were consistent with previous research but the amount of drop-out after 6 months differed in the two study groups. Varying effects over time in secondary outcomes might be affected by differential drop-out. Third, the sample size was powered to detect small-to-medium effects, as was assumed for the primary outcome. Power to detect smaller effects in the secondary outcomes may have been too low, though. Finally, the study focused on self-management outcomes and not on distal health outcomes. However, Toomey and colleagues [40] argue, that self-report measures of pain and disability do not adequately assess success in self-management interventions for people with chronic pain conditions, as the aims are often not to cure pain but to enable the person to cope with their pain and regain control over their lives. Thus, programs that aim to promote self-management should assess outcomes such as knowledge and self-management skills, but often do not [41]. Indeed, there is a lack of consensus regarding the most appropriate outcome measures for self-management interventions. 4.2 Conclusions As to our knowledge, this is the first implementation effectiveness trial for a patient education program in German rehabilitation centers. Overall, despite heterogeneous program implementation in the participating clinics, the study showed that the standardized back school seems to be more effective in certain self-management and patient-centered outcomes
17 compared to usual care back school programs. Thus, its further implementation within the rehabilitation setting may be recommended, but expected as small improvements in selfmanagement outcomes of multidisciplinary medical rehabilitation. However, it should be noted that the efficacy trial with the standardized program showed greater effects. A further step within effectiveness studies would be evaluating the impact of implementation fidelity on patient outcomes. 4.3 Practice Implications Patient education programs are part of multidisciplinary rehabilitation for patients with chronic low back pain. Implementation of a standardized, patient-oriented, biopsychosocial back school might improve effects on certain self-management behaviors and patient-centered outcomes within orthopedic inpatient rehabilitation. Further dissemination of such programs may be encouraged.
18 Acknowledgements The study was funded by the German Federal Ministry of Research and Education and the German Statutory Pension Insurance Scheme [grant number: 01GX1021] as part of the German grant program “Chronic illness and patient orientation”. We would like to thank the participating patients and all providers in the rehabilitation clinics who supported the study (Klinik Bad Wörishofen Deutsche Rentenversicherung Schwaben; Rehabilitations- & Präventionszentrum Bad Bocklet; Reha-Zentrum Schömberg, Klinik Schwarzwald; Sana Rheumazentrum Rheinland-Pfalz AG/ACURA Rheumazentrum Rheinland-Pfalz AG).
Declaration of interest The authors confirm that there is no conflict of interest.
"I confirm all patient/personal identifiers have been removed or disguised so the patient/person(s) described are not identifiable and cannot be identified through the details of the story."
19 References [1] T. Vos, A.D. Flaxman, M. Naghavi, R. Lozano, C. Michaud, M. Ezzat, Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010, Lancet 380 (2012) 2163-96. [2] C.M. Wenig, C.O. Schmidt, T. Kohlmann, B. Schweikert, Costs of back pain in Germany, Eur. J. Pain 13 (2009) 280-6. [3] S.J. Kamper, A.T. Apeldoorn, A. Chiarotto, R.J.E.M. Smeets, R.W. Ostelo, J. Guzman, M.W. van Tulder, Multidisciplinary biopsychosocial rehabilitation for chronic low back pain, Cochrane Database Syst. Rev. 9 (2014) CD000963. [4] F.P.C. Waterschoot, P.U. Dijkstra, N. Hollak, H.J. de Vries, J.H.B. Geertzen, M.F. Reneman, Dose or content? Effectiveness of pain rehabilitation programs for patients with chronic low back pain: A systematic review, Pain 155 (2014) 179-89. [5] Deutsche Rentenversicherung, [Reha-Therapiestandards Chronischer Rückenschmerz Leitlinie für die medizinische Rehabilitation der Rentenversicherung], DRV, Berlin, 2015. [6] G. Waddell, A.K. Burton, Concepts of rehabilitation for the management of low back pain, Best Pract. Res. Clin. Rheumatol. 19 (2005) 655-70. [7] G. Morone, T. Paolucci, M.R. Alcuri, M.C. Vulpiani, A. Matano, I. Bureca, S. Paolucci, V.M. Saraceni, Quality of life improved by multidisciplinary back school program in patients with chronic non-specific low back pain: a single blind randomized controlled trial, Eur. J. Phys. Rehabil. Med. 47 (2011) 533-41. [8] M. van Middelkoop, S.M. Rubinstein, A.P. Verhagen, R.W. Ostelo, B.W. Koes, M.W. van Tulder, Exercise therapy for chronic nonspecific low-back pain, Best Pract. Res. Clin. Rheumatol. 27 (2010) 193-204. [9] S. Michie, S. Ashford, F.F. Sniehotta, S.U. Dombrowski, A. Bishop, D.P. French, A refined taxomomy of behaviour change techniques to help people change their physical
20 activity and health eating behaviours: The CALO-RE taxonomy, Psychol. Health 26 (2011) 1479-98. [10] C.L. Miles, T. Pincus, D. Carnes, K.E. Homer, S.J.C. Taylor, A. Bremner, A. Rahman, M. Unterwood, Can we identify how programmes aimed at promoting self-management in musculoskeletal pain work and who benefits? A systematic review of sub-group analysis within RCTs, Eur. J. Pain 15 (2011) 775.e1-775.e11 [11] M. Weigl, A. Cieza, P. Cantista, JD Reinhardt, G. Stucki, Determinants of disability in chronic musculoskeletal health conditions: a literature review, Eur. J. Phys. Rehabil. Med. 44 (2008) 67-79. [12] A. Reusch, M. Schug, R. Küffner, H. Vogel, H. Faller, [Medical rehabilitation groupprogrammes concerning health promotion, patient education and psychoeducation - a 2010 national survey], Rehabilitation 52 (2013) 226-33. [13] K. Meng, B. Seekatz, H. Roßband, U. Worringen, H. Faller, H. Vogel, [Development of a standardized back school for in-patient orthopaedic rehabilitation], Rehabilitation 48 (2009) 335-44. [14] K. Meng, B. Seekatz, H. Roßband, U. Worringen, H. Vogel, H. Faller, Intermediate and long-term effects of a standardized back school for inpatient orthopedic rehabilitation on illness knowledge and self-management behaviors: A randomized controlled trial, Clin. J. Pain 27 (2011) 248-57. [15] I. Brandes, A. Bönisch, U. de Vries, I. Ehlebracht-König, F. Petermann, C. Krauth, [Exemplary implementation of patient education as rehabilitation programmes], Phys. Med. Rehab. Kuror. 18 (2008) 344-8. [16] U. de Vries, I. Brandes, C. Krauth, F. Petermann, [Patient education in pulmonary rehabilitation: results of an implementation study], Gesundheitswesen 70 (2008) 572-81.
21 [17] J.A. Durlak, E.P. DuPre, Implementation matters: a review of research on the influence of implementation on program outcomes and the factors affecting implementation, Am. J. Community Psychol. 41 (2008) 327-50. [18] L. Dusenbury, R. Brannigan, M. Falco, W.B. Hansen, A review of research on fidelity of implementation: implications for drug abuse prevention in school settings, Health Educ. Res. 18 (2003) 237–56. [19] G.M. Curran, M. Bauer, B. Mittmann, J.M. Pyne, C. Stetler, Effectivenessimplementation hybrid designs combining elements of clinical effectiveness and implementation research to enhance public health impact, Med. Care 50 (2012) 217-26. [20] B.R. Flay, Efficacy and effectiveness trials (and other phases of research) in the development of health promotion programs, Prev. Med. 15 (1986) 451-71. [21] K. Meng, S. Peters, A. Schultze, K. Pfeifer, H. Faller, [The Impact of 2 Interventions on Implementation Fidelity of a Standardized Back School Program in Inpatient Orthopedic Rehabilitation Facilities], Rehabilitation 54 (2015) 325-331.. [22] W. Göhner, [Prevention of pain chronification: adherence to physiotherapy], Logos, Berlin, 2003. [23] C. Herda, S. Keller, K. Ridder, H.-D. Basler, [Motivation zu rückenfreundlichem Verhalten], in: S. Keller (Ed.), [Motivation zur Verhaltensänderung], Lambertus, Freiburg i. B., 1999, pp. 213-28. [24] I. Frey, A. Berg, D. Gratwohl, J. Keul, [Freiburger questionnaire on physical activity – design, validation and application], Soz. Präventivmed. 44 (1999) 55-64. [25] E. Geissner, [Coping with chronic pain – Scales for the assessment of coping strategies and pain-related distress], Z. Klin. Psychol. 28 (1999) 280-90.
22 [26] A.J. Vickers, Analysis of variance is easily misapplied in the analysis of randomized trials: a critique and discussion of alternative statistical approaches, Psychosom. Med. 67 (2005) 652-5. [27] J. Cohen, Statistical power analysis for the behavioral sciences, Lawrence Erlbaum Associates, Hillsdale, New York, 1988. [28] A. Field, Discovering Statistics Using SPSS, SAGE Publications, London, 2005. [29] M. Bullinger, I. Kirchberger, [SF-36 Fragebogen zum Gesundheitszustand], Hogrefe, Göttingen, 1998. [30] M. Glattacker, K. Heyduck, C. Meffert, Illness beliefs, treatment beliefs and information needs as starting points for patient information - Evaluation of an intervention for patients with chronic back pain, Patient Educ. Couns. 86 (2012) 378-89. [31] K. Mattukat, D. Rennert, I. Brandes, I. Ehlebracht-König, K. Kluge, W. Mau, Short- and long-term effects of intensive training and motivational programme for continued physical activity in patients with inflammatory rheumatic disease, Eur. J. Phys. Rehabil. Med. 50 (2014) 395-409. [32] S. May, Self-management of chronic low back pain and osteoarthritis, Nat. Rev. Rheumatol. 6 (2010) 199-209. [33] N. McGrane, R. Galvin, T. Cusack, E. Stokes, Addition of motivational interventions to exercise and traditional physiotherapy: a review and meta-analysis, Psychotherapy 101 (2015) 1-12. [34] S. Michie, C. Abraham, C. Whittington, J. McAteer, S. Gupta, Effective techniques in healthy eating and physical activity interventions: a meta-regression, Health Psychol. 28 (2009) 690-701.
23 [35] P.H. Ferreira, M.L. Ferreira, C.G. Maher, K.M. Refshauge, J. Latimer, R.D. Adams, The therapeutic alliance between clinicians and patients predicts outcome in chronic low back pain, Phys. Ther. 93 (2013) 470-8. [36] T. Mars, D. Ellard, D. Carnes, K. Homer, M. Underwood, S.J.C. Taylor, Fidelity in complex behaviour change interventions: a standardised approach to evaluate intervention integrity, BMJ Open 3 (2013) e003555. [37] M. Leeuw, M.E.J.B. Goossens, S.J. Linton, G. Crombez, K. Boersma, J.W.S. Vlaeyen, The Fear-Avoidance Model of Musculoskeletal Pain: Current State of Scientific Evidence, J. Behav. Med. 30 (2007) 77-94. [38] E.J. Yang, W.-B. Park, H.-I. Shin, J.-Y. Lim, The effect of back school integrated with core strengthening in patients with chronic low-back pain, Am. J. Phys. Med. Rehabil. 89 (2010) 744-54. [39] E. Toomey, L. Currie-Murphy, J. Matthews, D.A. Hurley, Implementation fidelity of physiotherapist-delivered group education and exercise interventions to promote selfmanagement in people with osteoarthritis and chronic low back pain: A rapid review Part II, Man. Ther. 20 (2015) 287-94. [40] E. Toomey, L. Currie-Murphy, J. Matthews, D.A. Hurley, The effectiveness of physiotherapist-delivered group education and exercise interventions to promote selfmanagement for people with osteoarthritis and chronic low back pain: A rapid review Part I, Man. Ther. 20 (2014) 265–86. [41] F.P. Kroon, L.R. van der Burg, R. Buchbinder, R.H. Osborne, R.V. Johnston, V. Pitt, Selfmanagement education programmes for osteoarthritis, Cochrane Database Syst. Rev. 1 (2014) CD008963.
24 Figure 1. Study flow
Figure 2. Back exercises (means) at 6 and 12 months follow-up by treatment group (IG intervention group; CG control group) and gender. Analyses of covariance/interaction effects: 6 months: p = 0.094, η2 = 0.007; 12 months: p = 0.099, η2 = 0.008.
25
Table 1. Characteristics of the traditional back schools (control group) and fidelity of the implemented back school (intervention group) Clinic 1
Clinic 2
Clinic 3
Clinic 4
Modular concept
yes
yes
no
no
Number of modules delivered by exercise therapist
5
5
3
3
Written manual
no
in part
no
no
Maximum 15 participants
no
yes
no
yes
Closed group (modules delivered by exercise therapist)
yes
no
no
no
Patient handouts
yes
yes
yes
no
Lecture physiciana
yes
yes
yes
yes
Lecture psychologista
no
no
yes
no
Number, duration and frequency of modules
A
A
A
A
Sequence of modules
A
B
A
B
Closed group format (all modules)
A
B
A
C
Number of participants (maximum 15)
A
A
A
C
Health professionals (interdisciplinary implementation)
A
A
C
B
Therapist constancy (modules of exercise therapist)
A
A
C
C
Indication (allocated patients)
B
A
A
B
Seating arrangement/room setup
A
B
A
B
Slides/patient booklet
A
B
B
B
Contents physician/psychologist/exercise therapist
A/A/C
A/A/A
A/B/A
A/B/B
Methods physician/psychologist/exercise therapist
B/A/C
B/A/A
C/B/A-B
B-C/B/B-C
Control group - traditional back school
Intervention group - implemented back school Structural key elements
Control group: elements of the traditional back schools were assessed by semi-structured interviews and observations. Intervention group: structural key elements of the implemented back school were assessed by standardized interviews, realization of contents and methods by structured observations, respectively.
26 A = key elements according to manual and mandantory contents/methods were predominantly performed. B = key elements with minor adaptation in concordance with suggestions to overcome barriers and mandatory contents/methods were partly performed. C = key elements with major adaptation not suggested for barrier management and mandatory contents/methods were predominantly not performed. a
Lectures of physicians and/or psychologists that are not inherent part of the back school program.
27 Table 2. Baseline sociodemographic and clinical characteristics
28
Female, n (%) Age in years, mean (SD) Marital status, n (%)
Control group (n = 256)
Intervention group (n = 279)
p value
139 (54.3)
158 (56.6)
0.602
50.5 (8.1)
49.7 (8.6)
a
0.683
Single
34 (13.3)
44 (15.6)
Married/with partner
173 (67.8)
183 (65.8)
Divorced/separated
37 (14.5)
43 (15.5)
11 (4.3)
8 (2.9)
Widowed Education, n (%)
b
0.697
Less than junior (< 10 y; basic secondary school)
120 (47.1)
126 (45.5)
Junior (10 y; middle-level secondary school)
91 (35.6)
92 (33.2)
Senior (Qualification for university)
39 (15.3)
52 (18.8)
5 (2.0)
7 (2.5)
Other Working status, n (%)
a
0.394
Employed Unemployed Retired/other Occupation, n (%)
0.227
227 (89.0)
238 (85.6)
16 (6.3)
22 (7.9)
12 (4.7)
18 (6.5)
c
0.244
Worker
91 (36.3)
89 (32.5)
151 (60.2)
167 (60.9)
9 (3.6)
18 (6.5)
Dorsalgia (M54)
180 (70.3)
212 (76.0)
0.144
Other intervertebral disc disorders (M51)
114 (44.5)
162 (58.1)
0.002
69 (27.0)
69 (24.7)
0.621
Employee Other Main diagnosis (ICD-10), n (%)
d
Other dorsopathies, not elsewhere classified (M53) Pain duration, n (%)
e
0.235
< 1 year
30 (12.5)
42 (15.6)
< 3 years
61 (25.4)
49 (18.2)
< 5 years
26 (10.8)
32 (11.9)
≥ 5 years
123 (51.2)
146 (54.3)
Pain intensity (NRS), mean (SD)
5.8 (2.0)
5.8 (2.0)
0.680
Days of sick leave due to back pain in the last 6 months, mean (SD) SF-12 physical component
35.8 (51.4)
32.4 (51.1)
0.454
37.3 (8.8)
36.5 (8.5)
0.330
SF-12 mental component
42.4 (11.2)
41.8 (11.9)
0.568
NRS Numerical rating scale of pain intensity (0-10, with high scores indicating higher pain intensity); SF-12 Short form health survey (0 to 100, with high scores representing best health status). a
missing = 2. b missing: n = 3. c missing = 10. d Multiple comorbid diagnoses possible. e missing: n = 26. Rated
by the physician.
29 Table 3. Intermediate-term (6 months) and long-term (12 months) within-group and between-group effects on primary and secondary outcomes
Pre-treatment
mean (SD)
6-months
Within-group
follow-up
change
mean (SD)
SES (95% CI)
Illness knowledge
Between-group difference
Pre-treatment
ANCOVA p-value
η2
0.018
0.015
12-months follow-up
mean (SD)
mean (SD)
CG
25.43 (5.09)
26.35 (5.27)
0.17 (0.05 – 0.29)
25.39 (5.18)
26.36 (5.49)
IG
25.41 (5.19)
27.26 (5.35)
0.34 (0.22 – 0.46)
25.55 (5.33)
27.34 (5.54)
Back exercises
0.464
0.001
CG
2.25 (1.08)
2.85 (0.94)
0.56 (0.39 – 0.72)
2.22 (1.09)
2.75 (0.97)
IG
2.25 (1.06)
2.91 (0.89)
0.61 (0.44 – 0.78)
2.28 (1.06)
2.81 (0.92)
Physical activity
<0.001
0.069
CG
8.17 (5.73)
6.98 (5.51)
-0.22 (-0.35 – -0.09)
8.36 (5.81)
9.57 (6.59)
IG
7.61 (5.15)
9.00 (5.10)
0.25 (0.09 – 0.42)
7.75 (5.13)
9.42 (5.81)
Back posture habits
0.695
0.000
CG
2.70 (1.00)
3.30 (0.81)
0.61 (0.45 – 0.78)
2.69 (0.98)
3.32 (0.75)
IG
2.82 (0.95)
3.32 (0.80)
0.51 (0.38 – 0.64)
2.80 (0.93)
3.21 (0.82)
Action-oriented coping
0.488
0.001
CG
14.30 (5.02)
16.14 (5.31)
0.36 (0.21 – 0.51)
14.23 (5.00)
15.90 (5.64)
IG
14.89 (4.96)
16.72 (5.08)
0.36 (0.20 – 0.52)
14.91 (4.92)
16.07 (4.77)
Cognitive restructuring
0.634
0.001
CG
13.83 (4.69)
15.45 (8.08)
0.34 (0.19 – 0.50)
13.77 (4.66)
15.64 (5.03)
IG
13.86 (4.80)
15.25 (5.00)
0.30 (0.15 – 0.44)
13.75 (4.56)
15.10 (4.80)
Subjective coping competence
0.961
0.000
CG
15.51 (4.86)
16.32 (5.02)
0.18 (0.03 – 0.32)
15.62 (4.91)
16.91 (5.01)
IG
15.45 (4.50)
16.27 (4.86)
0.18 (0.02 – 0.33)
15.43 (4.35)
16.22 (4.56)
Mental distraction
0.878
0.000
CG
9.97 (4.50)
10.75 (5.11)
0.17 (0.03 – 0.31)
9.72 (4.28)
10.42 (5.06)
IG
9.74 (4.83)
10.57 (4.86)
0.18 (0.03 – 0.33)
9.61 (4.81)
10.14 (5.02)
Counter-activities
0.794
0.000
CG
11.97 (4.47)
12.42 (4.40)
0.10 (-0.03 – 0.23)
12.13 (4.71)
12.45 (4.64)
IG
12.15 (4.61)
12.41 (4.40)
0.06 (-0.09 – 0.20)
12.08 (4.63)
12.18 (4.33)
Relaxation
0.833
0.000
CG
11.64 (5.16)
13.10 (5.21)
0.28 (0.15 – 0.42)
11.88 (5.13)
12.52 (5.19)
IG
12.25 (5.14)
13.54 (5.29)
0.25 (0.12 – 0.38)
11.70 (5.01)
13.36 (5.13)
ANCOVA indicates analysis of covariance; CI, confidence interval; SES, standardized effect size (positive SES represents improvement). CG, control group; IG, intervention group. Effect size η² (small: η2 = 0.0099, medium: η2 = 0.0588, large: η2 = 0.137); SES (small = 0.20, medium = 0.50, large = 0.80). Bold figures indicate: a)
30 regarding within-group effects: SES of the CG is outside the 95% CI of SES of the IG; or b) regarding betweengroup effects: significant between-group effect (ANCOVA: p < 0.05). 6-months follow-up: n = 398 (CG: n = 205, IG: n = 193). 12-months follow-up: n = 348 (CG: n = 177, IG: n = 171).
Enrolment Allocation Follow-up Analysis
Control group
Intervention group
Assessed for eligibility/contacted for participation (n = 299)
Assessed for eligibility/contacted for participation (n = 388)
Refused to participate (n = 33)
Refused to participate (n = 97)
Agreed to participate (n = 266)
Agreed to participate (n = 291)
Withdrew consent (n = 4), did not complete baseline questionnaire t1 (n = 4), exclusion criterion (n = 2)
Withdrew consent (n = 4), did not complete baseline questionnaire t1 (n = 8)
Received allocated intervention/ completed t1 assessment (n = 256)
Received allocated intervention/ completed t1 assessment (n = 279)
Completed t2 assessment (n = 236)
Completed t2 assessment (n = 260)
Lost/reason: no reply, empty questionnaire (n = 20)
Completed t3 assessment (n = 205) Lost/reason: no reply (n = 48), moved (n = 3)
Lost/reason: no reply, empty questionnaire (n = 19)
Completed t3 assessment (n = 193) Lost/reason: no reply (n = 79), moved (n = 7)
Completed t4 assessment (n = 177)
Completed t4 assessment (n = 171)
Lost/reason: no reply (n = 75), moved (n = 4)
Lost/reason: no reply (n = 98), moved (n = 10)
Analyzed (n = 242) Data at t1 and at least at one of t2 - t4
Excluded (n = 14) Only data at t1
Analyzed (n = 268) Data at t1 and at least at one of t2 - t4
Excluded (n = 11) Only data at t1
4,0
Back exercises
3,5
IG Women CG Women IG Men CG Men
3,0
2,5
2,0
1,5
1,0 Pretreatment
6 months follow-up
4,0
Back exercises
3,5
IG Women CG Women IG Men CG Men
3,0
2,5
2,0
1,5
1,0 Pretreatment
12 months follow-up
S0738-3991(17)30011-3 http://dx.doi.org/doi:10.1016/j.pec.2017.01.011 PEC 5564
To appear in:
Patient Education and Counseling
Received date: Revised date: Accepted date:
9-3-2016 22-12-2016 13-1-2017
Please cite this article as: Meng Karin, Peters Stefan, Faller Hermann.Effectiveness of a standardized back school program for patients with chronic low back pain after implementation in routine rehabilitation care.Patient Education and Counseling http://dx.doi.org/10.1016/j.pec.2017.01.011 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 Effectiveness of a standardized back school program for patients with chronic low back pain after implementation in routine rehabilitation care
Karin Meng, Stefan Peters, Hermann Faller
Department of Medical Psychology, Medical Sociology, and Rehabilitation Sciences, University of Würzburg, Würzburg, Germany
Corresponding author Dr. Karin Meng University of Würzburg, Department of Medical Psychology, Medical Sociology, and Rehabilitation Sciences, Klinikstr. 3, D-97070 Würzburg, Germany Tel: (+)49 931 3182074 E-Mail: [email protected]
2 Highlights
Back schools are an integral part of rehabilitation in chronic low back pain.
Implementation of standardized programs is a challenge for routine care.
The implementation fidelity of a biopsychosocial back school was heterogeneous.
Despite heterogeneous implementation, the back school seems to be effective.
Further implementation within inpatient orthopedic rehabilitation may be recommended.
3 Abstract Objective. To evaluate the effectiveness of a standardized, patient-oriented, biopsychosocial back school after implementation in inpatient orthopedic rehabilitation. Methods. A multi-center, quasi-experimental controlled study of patients with low back pain (n = 535) was conducted. Patients in the control group received the traditional back school before implementation of the new program (usual care); patients in the intervention group received the new standardized back school after implementation into routine care. Patients` illness knowledge and conduct of back exercises (primary outcomes) and secondary selfmanagement outcomes and treatment satisfaction were obtained at admission, discharge, and 6 and 12 months after rehabilitation. Results. We found a significant small between-group intervention effect on patients` illness knowledge in medium- to long term (6 months: η2 = 0.015; 12 months: η2 = 0.013). There were trends for effects on conduct of back exercises among men (6 and 12 months: η2 = 0.008 both). Furthermore, significant small effects were observed for treatment satisfaction at discharge and physical activity after 6 months. Conclusions. The standardized back school seems to be more effective in certain outcomes than a usual care program despite heterogeneous program implementation. Practice Implications. Further dissemination within orthopedic rehabilitation may be encouraged to foster self-management outcomes.
Key words Chronic back pain, patient education, implementation effectiveness, controlled trial, rehabilitation
4 1. Introduction Back pain is a major health problem worldwide and responsible for disability and socioeconomic costs [1,2]. It is a leading cause for inpatient medical rehabilitation in Germany. In this regard, multidisciplinary biopsychosocial rehabilitation seems to be more effective than usual care and physical treatment in reducing pain and disability in the longterm [3]. However, how these programs should be designed to further improve outcome is not entirely clear [4]. Patient education programs, particularly back schools, are a mandatory part of rehabilitation for patients with chronic low back pain in Germany [5]. According to evidence, they should be based on the biopsychosocial model [6] and conducted multidisciplinary [7,8]. Additionally, the incorporation of strategies to promote long-term adherence to physical activity is recommended [9-11]. However, seemingly many back school programs offered as part of German medical rehabilitation are still based on a biomechanical approach. Many also do not fulfill quality requirements, such as the use of manuals, patient-oriented didactics and evaluation of effectiveness [12]. Therefore, a new standardized, patient-oriented and biopsychosocial back school with a theory-based approach employed to foster long-term adherence to physical activity was developed. A previous randomized, controlled trial in one inpatient rehabilitation clinic demonstrated the superior efficacy of this program with regard to patients’ illness knowledge, determinants of physical activity, fear-avoidance beliefs, and treatment satisfaction in short-term, as well as self-management behaviors up to 1 year [13,14]. Implementation of this back school may be complex because of its manual-based, modular and interdisciplinary structure as well as its new conception regarding contents and methods, requiring both organizational and behavioral changes. Generally, the implementation of standardized patient education programs represents a challenge for routine rehabilitation practice and is often not entirely successful, i.e. programs
5 are only partially implemented or modified [15,16]. This may pose a threat to their effectiveness in routine care. International reviews confirm that program adaptations are common and vary by providers [e.g. 17,18], and there is empirical support from health promotion and prevention research to the assumption that the level of implementation quality affects outcomes [17]. Overall, there is a need for more systematic evaluation of the implementation of patient education programs within inpatient medical rehabilitation. Our study is an implementation effectiveness trial, designed to address both implementation and clinical intervention aims [19,20]. Presenting a next step after the above-mentioned evaluation of the back school under standardized conditions (efficacy evaluation), it uses a mixed-methods approach for both the process evaluation of implementation and the evaluation of effectiveness in routine care. Results of the process evaluation showed that structural adherence as based on the transfer of key elements according to the manual varied between rehabilitation clinics. Minor adaptations appeared to be necessary to overcome structural barriers, achieve acceptance and facilitate program implementation [21]. In this paper, we report results of the effectiveness trial. The main research objective is to evaluate the effectiveness of a new back school after implementation in inpatient orthopedic rehabilitation with regard to patient outcomes. We hypothesized that the implemented new back school would be superior to traditional care regarding illness knowledge and back exercises (primary outcomes). Additionally, we expected superior effectiveness of the implemented program regarding treatment satisfaction as well as several self-management behaviors, such as physical activity, back posture habits, and coping with pain (secondary outcomes). Because of the results of our prior trial [14], we explored moderator effects of gender for behavioral outcomes.
2. Methods
6 2.1 Design and procedure Effectiveness within routine care was evaluated in a multi-centre, quasi-experimental controlled study in 535 patients with low back pain in 4 orthopedic rehabilitation clinics in Germany. Data were collected at admission and discharge as well as after 6- and 12-months with patient-reported questionnaires. Control patients received the traditional back school program of the hospital as delivered before implementation of the new program (CG); intervention patients received the new standardized back school as delivered after implementation into routine-practice (IG). Patients with a primary diagnosis of chronic low back pain (ICD-10-GM: M51, M53, M54) were included in the trial. Exclusion criteria were inadequate German language ability, age below 18 or above 70 years, severe visual or hearing impairment, severe comorbid psychiatric disorder, poor health state, and submitted pension application. Patients were blind to the allocated study condition; blinding of clinicians was impossible due to study design, though. Sample size was powered to detect small to medium effects in the primary outcomes (cf [13,14]; knowledge: d = 0.48; back exercise: d = 0.29; 2sided α = 0.05, 1- = 0.8). Therefore, 376 persons were required. We included 540 persons, based on an estimated drop-out of about 30%. As power to detect smaller between-group effects on the secondary outcomes may be lacking, effect sizes were reported throughout. The study was approved by the Ethics Committee of the Faculty of Medicine, University of Würzburg, on 25. March 2011 (ref: 48/11) and registered on WHO International Clinical Trials (DRKS00004835). 2.2 Participants Patients were consecutively recruited between May 2011 and April 2013; one-year follow-up was finished in May 2014. Figure 1 shows the participant flow. A physician evaluated the inclusion criteria at admission. 687 patients eligible were asked to participate in the study and
7 557 participated (81%) and signed the informed consent form. 22 persons had to be excluded. Thus, 535 persons comprised the initial study sample. Follow-up rates exceeded 65% at all time points. Analyses were performed with all persons who participated in at least one follow-up assessment. Non-responder analyses revealed no differences between participants and non-participants in gender and age (p > 0.05). Drop-out analyses at follow-up points were performed for socio-demographic variables, main outcome variables and allocation to study group. Patients who had not provided data at the 6- and 12months follow-ups, respectively, were younger, more often unfit to work, reported higher pain intensity, and had lower illness knowledge at baseline. Drop-outs at 12-months follow-up reported lower emotional well-being at baseline than did completers. Furthermore, drop-out after 6 months and by trend after 12 months was higher in the IG than the CG. 2.3 Intervention All patients took part in the 3-week multidisciplinary inpatient rehabilitation. CG patients received the traditional back school programs of the rehabilitation clinics as delivered before implementation of the new program (usual care). IG patients received the new standardized back school as delivered after implementation into routine-practice. Table 1 summarizes characteristics of back school programs in the CG as well as fidelity of the standardized back school program in the IG (implementation of structural key elements, contents and methods according to the manual). Details regarding fidelity assessment have been previously described [21]. CG: Traditional non-standardized back school programs (i.e. no closed group format, not manual-based, less patient oriented didactics, no interdisciplinary training team) with biomechanical orientation. Contents of the lessons included correct back posture and movements as well as back exercises. Theory-based intervention techniques to promote
8 physical activity and psychosocial aspects of low back pain were not systematically addressed. IG: Standardized back school program [13,14] with 7 patient-oriented, interactive sessions of 55 minutes, held in small groups of a closed format. The program is manual-based and interdisciplinary with sessions led by a physiotherapist (five sessions), an orthopedist (one session) and a psychologist (one session), respectively. In each session, patients are actively involved in the educational process using a combination of didactic methods (short lectures, group discussion, small group work, practice/exercises, and individual work). Didactic materials included presentations, flipcharts, handouts and work sheets. Contents of the sessions included basic knowledge about back pain (e.g. development, symptoms, treatment) with regard to physical (e.g. body perception, positions for short-term relief) and psychosocial aspects (fear-avoidance-endurance model, coping behavior, risk factors). Spine-related exercises (strength training, active stabilization) and theory-based techniques to promote physical activity (e.g. motivation, self-regulation) were also part of the sessions. All sessions followed a common bio-psycho-social model. However, realization of the program varied between clinics. 2.4 Outcomes and measurements The primary outcomes were patients` illness knowledge and conduct of back exercises. Secondary outcomes included physical activity, back posture and movements, and pain coping strategies, as well as satisfaction with the back school. Outcomes were assessed by validated measures. This paper focuses on primary and self-management outcomes that were of main interest at follow-up [14]. Furthermore, we report effects on patient satisfaction as an element of patient-centered outcome in short-term. Illness knowledge. Patients were asked to judge 44 statements about back pain and its treatment as true or false. Item examples: “Depression is a risk factor for low back pain”,
9 “Physical activity should last longer than 15 minutes to have a positive impact”. Items are summed up to yield a score ranging from 0 to 44, with higher scores corresponding to higher knowledge; missing answers were counted as wrong answers. This questionnaire was developed within the previous evaluation study [13,14]. Back exercises. A 3-item modified version of a self-report questionnaire was administered [13,14,22]. Patients were asked about the amount of back exercises they did during the last week using three items; e.g. “Stretching”, “Strength training”. The scale ranged from 0 (seldom or never) to 4 (almost every day), with higher scores indicating more exercises. It showed good internal consistency (Cronbach’s α = 0.85). Back posture habits. A 7-item modified version of a validated self-report questionnaire was administered [23]. Test-retest reliability after 5-8 weeks is rtt = 0.77. The scale is in accordance with ratings based on behavioral monitoring. Patients were asked how often they performed adequate sitting, standing, walking, lifting and carrying in their daily routine during the last week; e.g. “Sitting with a back-friendly posture”, “Keeping the back in a straight and stable position when lifting heavy objects”. The scale ranged from 0 (seldom or never) to 4 (almost every day), with higher scores indicating more adequate back-related behavior. The internal consistency is α = 0.85. Physical activity. The Freiburger Questionnaire on Physical Activity (FfkA) is a reliable and valid German self-report measure using eight items [24]. Test-retest reliability is high (14 days: rtt = 0.95). Validity was established by correlations with maximum oxygen uptake (sports activities: partial r = 0.42). Patients were asked to indicate their health-related physical activities in daily routine, leisure time, and sports and to report for each activity the frequency and duration during the last week or month, respectively. All activity reports were transformed to produce a total activity score (in hours per week) as well as three sub-scores (basic activity, leisure time activity, sports activity).
10 Pain coping strategies. Coping with pain was assessed using the self-report German Pain Management Questionnaire (FESV), consisting of 24 items with 6-point numerical scales (1 = don’t agree at all, 6 = fully agree) [25]. Items are divided into 2 domains of coping strategies, cognitive strategies and behavioral strategies, each comprising 3 scores: cognitive strategies include action-oriented coping (e.g., “If I have pain, I have a plan how to handle that pain”), cognitive restructuring (e.g., “If I have pain, I will not give up”), and subjective coping competence (e.g., “If I have pain, I remind myself that I can deal with it much better than in the past”); behavioral strategies include mental distraction (e.g., “If I have pain, I thumb through a magazine”), counter-activities (e.g., “If I have pain, I distract myself with activities around my home or garden”), and relaxation (e.g., “If I have pain, I concentrate on calm and steady breathing”). Scores may range from 4 to 24, with higher scores corresponding to stronger endorsement of the respective coping strategy. Cronbach’s α for the subscales was between 0.70 and 0.84. Test-retest reliability and validity have been demonstrated [25]. Treatment satisfaction. Patients were asked to judge the educational program using 11 items [13]. Each item comprised patient-centered educational criteria on a 6-point scale (1 = very good, 6 = very poor). Items were averaged into 3 subscores – content (topic selection, comprehensibility, alignment, usefulness, delivery), group/interaction (opportunity to make comments, exchange of experience, group atmosphere, size), and materials (slides, handouts). Scores showed good reliability (Cronbach’s α 0.79-0.85). Furthermore, one item inquired the satisfaction with the program in general. 2.5 Statistical analysis Statistical analyses were performed using SPSS 22.0 for Windows. For missing data in accordance with missing at random, five imputed data sets were created using the multiple imputation procedure implemented in SPSS. Missing values due to drop-out were analyzed by pair-wise deletion. Non-response analyses and drop-out analyses were carried out by
11 independent group comparisons using t tests for continuous variables and chi-square tests for categorical variables. Treatment effects (between-group effects) were evaluated separately for each follow-up time point using analysis of covariance (ANCOVA) adjusting for baseline values [26]. Statistical significance (p < 0.05, 2-sided) and effect sizes (η²) were reported for all between-group differences [27]. In addition, within-group effects, including standardized effect sizes (SES) and accompanying 95% confidence intervals (CIs) were reported for both study groups. Gender subgroup analysis was performed by including gender as an additional factor in the ANCOVA and examining interaction effects. When trends to significant interactions emerged (p < 0.10), simple effects analyses were performed by gender subgroups [28]. The significance level for interaction effects was not adjusted due to its exploratory nature.
3. Results 3.1 Patient characteristics Table 2 presents characteristics of participants by study group. The initial sample consisted of 535 rehabilitants with chronic low back pain. Patients` mean age was 50 years (SD = 8.3), 55% were female, and 78% reported to live together with a partner. The sample mostly consisted of employed persons (87%). The duration of sick leave due to back pain within the last 6 months averaged 34 days (SD = 51, Median = 10). Symptom duration was longer than 5 years for about half of the sample (53%). Subjective health status was reduced as compared to the general population [29]. There were no systematic differences between participants in the CG and IG. 3.2 Primary outcome Table 3 presents the intermediate- and long-term results on primary and secondary outcomes.
12 Participants of the IG, as compared to the CG, showed superior knowledge about chronic back pain and its treatment 6 and 12 months after rehabilitation (6 months: p = 0.018, η2 = 0.015; 12 months: p = 0.039, η2 = 0.013). There were no significant intervention effects on back exercises 6 and 12 months after rehabilitation. However, interaction effects of gender and intervention condition were observed as trends (p < 0.10; Figure 2). Simple effect analyses indicated trends for small intervention effects on exercise behavior among men, but not women (6 months: men: p = 0.073, η2 = 0.008. women: p = 0.570, η2 < 0.001; 12 months: men: p = 0.106, η2 = 0.008. women: p = 0.480, η2 = 0.001). 3.3 Intermediate- and long-term effects on self-management behavior 6-months after rehabilitation, both IG and CG showed significant, small to medium improvements in the self-management behaviors physical activity (only IG), back posture habits and coping with pain (except counter-activities in both groups). SESs indicate that the IG improved more than the CG regarding physical activity. Accordingly, ANCOVA resulted in a significant, medium effect in favor of the IG for physical activity. Differences are based on a medium effect on everyday activities (p < 0.001, η2 = 0.067) and a small effect in sports activities (p = 0.044, η2 = 0.010), respectively. After adjustment for baseline imbalance, estimated mean physical activity per week in the IG was about 135 minutes longer than in the CG, with 86 minutes pertaining to everyday activities, 24 minutes to leisure time activities, and 22 minutes to sports activities, respectively. No significant differences between the groups were observed for back posture habits as well as cognitive and behavioral strategies of coping with pain. No significant interaction effects with gender emerged (p > 0.10). At 12-months follow-up, significant increases in physical activity and back posture habits occurred among patients of both groups, while within-group changes in coping with pain were heterogeneous. SESs indicated that patients in the CG improved more with regard to back posture habits. By contrast, a significant, small increase in relaxation was evident only in the
13 IG, but not the CG. ANCOVAs showed significant between-group effects for both outcomes. There were no intervention effects on physical activity or other pain coping strategies. Moderator analysis indicated significant interactions with gender regarding certain dimensions of coping, but not for physical activity and back posture habits. Women in the IG reported lower mental distraction and counter-activities than women in the CG, whereas no such differences were found among men (mental distraction: women: p = 0.088, η2 = 0.008; men: p = 0.207, η2 = 0.005. counter activities: women: p = 0.107, η2 = 0.008; men: p = 0.331, η2 = 0.003). 3.4 Short-term effects on treatment satisfaction Participants of the IG as compared to the CG showed significant higher treatment satisfaction with regard to “content” (p = 0.05, d = 0.17) and “materials” (p = 0.013; d = 0.23). However, there were no significant effects on “group/interaction” (p = 0.23, d = 0.11) or the satisfaction with the back school in general (p = 0.26; d = 0.10).
4. Discussion and Conclusion 4.1 Discussion In this implementation effectiveness trial, we evaluated the medium- and long-term routine effectiveness of a new back school as delivered after implementation in inpatient orthopedic rehabilitation in comparison with the respective traditional back schools of the study clinics. With respect to the primary outcomes, we found a significant small treatment effect on patients` illness knowledge but no effect on back exercises in the medium- to long-term. However, there were trends for effects on conduct of back exercises for men. Additionally, we found significant small effects for treatment satisfaction at discharge, as well as the amount of physical activity carried out after 6 months. However, there were no additional effects of the new back school on back posture habits or coping with pain (except relaxation in the long-
14 term). Thus, our primary and secondary hypotheses regarding self-management outcomes were only partially supported. The lack of additional effects may be explained by the high intensity of the whole rehabilitation package as well as variations in both implementation fidelity and control conditions. First, we compared two active treatments that were only a minor part of a comprehensive rehabilitation treatment, which patients from both IG and CG received. Thus, significant changes occurred in both groups and the incremental effects of the educational program may have been attenuated. Several intervention studies in the German rehabilitation setting have shown that for such comparisons small treatment effects on proximal outcomes may be expected at most [e.g. 14,30,31]. Moreover, only small effects are reported for self-management interventions (like back schools) in chronic back pain [32]. Secondly, we intended to test the routine effectiveness of the back school program after implementation in several clinics. Hence, the implementation of the program varied between clinics with regard to key elements, i.e. structure, contents and delivery. The interactive, patient-centered approach of the new back school is supposed to foster learning effects regarding knowledge and self-management skills, and the use of behavior change techniques [9] should promote the uptake of physical activity and exercises in everyday life. Modifications made during implementation may interfere with these assumed treatment mechanisms of the program. Additionally, the traditional back school programs (usual care) being rather heterogeneous may have influenced results. Nevertheless, results are in line with our prior efficacy trial [14] regarding treatment effects on illness knowledge, physical activity, treatment satisfaction and, by trend, on exercises for men after rehabilitation. Effect sizes or time points may vary due to different study design, samples as well as treatment fidelity, though. However, even varying implementation of the back school components into routine practice resulted in positive self-management outcomes. Thus, the interactive small group setting might support gaining knowledge even when other
15 components of delivery are partly modified. Current reviews show that motivational and behavior change approaches are successful in increasing physical activity using varying methods of delivery [33,34]. Generally, the patient-centered approach allowing for adaptations to both therapists’ and patients’ needs may foster the therapeutic alliance that predicts outcome in low back pain patients [35]. Adaptations are not necessarily associated with poor delivery and might also reflect high levels of facilitator competence [36]. In contrast to our prior study, we observed a small effect on back posture habits in the longterm favoring the CG. The degraded importance of posture habits within modern back schools may explain this result. Only one session addressed back posture habits with a focus on individual and situational variability. The therapists encouraged patients to try out and individually select suitable postures and movements. This approach aims to counteract fearavoidance beliefs and foster self-efficacy [37]. Furthermore, there were no effects on pain coping skills. However, cognitive behavioral pain management constituted only a small part of this back school. Overall, it is necessary to study the effect of back schools and other selfmanagement interventions on the development of coping strategies and its association with improvements in pain and functional levels [38]. Some treatment effects were moderated by gender. The new back school worked better for men regarding exercise behavior and for women regarding some coping strategies, respectively. These moderator effects partly match with gender effects observed in our prior study. So far, clear evidence that gender is an essential moderator in self-management interventions in musculoskeletal pain is lacking. Further research on this issue is recommended [10]. Our study has certain strengths and limitations. A strength is the comprehensive evaluation of the implementation of a new back school based on a controlled effectiveness trial. Reports on implementation fidelity are often insufficient [39]. However, providing relevant data may
16 contribute to a better understanding of patient outcomes. Additionally, we strengthened generalizability of previous efficacy results for this back school program by using a multicenter approach, allocating a representative sample and evaluating routine application of the program. The following limitations need to be considered. Randomization was not possible in this implementation effectiveness trial. However, IG and CG patient samples were comparable with regard to socio-demographic and medical data. In all analyses, baseline scores of outcome parameters were adjusted for. Second, although drop-out occurred, its proportion was in the range of other studies. Factors associated with drop-out were consistent with previous research but the amount of drop-out after 6 months differed in the two study groups. Varying effects over time in secondary outcomes might be affected by differential drop-out. Third, the sample size was powered to detect small-to-medium effects, as was assumed for the primary outcome. Power to detect smaller effects in the secondary outcomes may have been too low, though. Finally, the study focused on self-management outcomes and not on distal health outcomes. However, Toomey and colleagues [40] argue, that self-report measures of pain and disability do not adequately assess success in self-management interventions for people with chronic pain conditions, as the aims are often not to cure pain but to enable the person to cope with their pain and regain control over their lives. Thus, programs that aim to promote self-management should assess outcomes such as knowledge and self-management skills, but often do not [41]. Indeed, there is a lack of consensus regarding the most appropriate outcome measures for self-management interventions. 4.2 Conclusions As to our knowledge, this is the first implementation effectiveness trial for a patient education program in German rehabilitation centers. Overall, despite heterogeneous program implementation in the participating clinics, the study showed that the standardized back school seems to be more effective in certain self-management and patient-centered outcomes
17 compared to usual care back school programs. Thus, its further implementation within the rehabilitation setting may be recommended, but expected as small improvements in selfmanagement outcomes of multidisciplinary medical rehabilitation. However, it should be noted that the efficacy trial with the standardized program showed greater effects. A further step within effectiveness studies would be evaluating the impact of implementation fidelity on patient outcomes. 4.3 Practice Implications Patient education programs are part of multidisciplinary rehabilitation for patients with chronic low back pain. Implementation of a standardized, patient-oriented, biopsychosocial back school might improve effects on certain self-management behaviors and patient-centered outcomes within orthopedic inpatient rehabilitation. Further dissemination of such programs may be encouraged.
18 Acknowledgements The study was funded by the German Federal Ministry of Research and Education and the German Statutory Pension Insurance Scheme [grant number: 01GX1021] as part of the German grant program “Chronic illness and patient orientation”. We would like to thank the participating patients and all providers in the rehabilitation clinics who supported the study (Klinik Bad Wörishofen Deutsche Rentenversicherung Schwaben; Rehabilitations- & Präventionszentrum Bad Bocklet; Reha-Zentrum Schömberg, Klinik Schwarzwald; Sana Rheumazentrum Rheinland-Pfalz AG/ACURA Rheumazentrum Rheinland-Pfalz AG).
Declaration of interest The authors confirm that there is no conflict of interest.
"I confirm all patient/personal identifiers have been removed or disguised so the patient/person(s) described are not identifiable and cannot be identified through the details of the story."
19 References [1] T. Vos, A.D. Flaxman, M. Naghavi, R. Lozano, C. Michaud, M. Ezzat, Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010, Lancet 380 (2012) 2163-96. [2] C.M. Wenig, C.O. Schmidt, T. Kohlmann, B. Schweikert, Costs of back pain in Germany, Eur. J. Pain 13 (2009) 280-6. [3] S.J. Kamper, A.T. Apeldoorn, A. Chiarotto, R.J.E.M. Smeets, R.W. Ostelo, J. Guzman, M.W. van Tulder, Multidisciplinary biopsychosocial rehabilitation for chronic low back pain, Cochrane Database Syst. Rev. 9 (2014) CD000963. [4] F.P.C. Waterschoot, P.U. Dijkstra, N. Hollak, H.J. de Vries, J.H.B. Geertzen, M.F. Reneman, Dose or content? Effectiveness of pain rehabilitation programs for patients with chronic low back pain: A systematic review, Pain 155 (2014) 179-89. [5] Deutsche Rentenversicherung, [Reha-Therapiestandards Chronischer Rückenschmerz Leitlinie für die medizinische Rehabilitation der Rentenversicherung], DRV, Berlin, 2015. [6] G. Waddell, A.K. Burton, Concepts of rehabilitation for the management of low back pain, Best Pract. Res. Clin. Rheumatol. 19 (2005) 655-70. [7] G. Morone, T. Paolucci, M.R. Alcuri, M.C. Vulpiani, A. Matano, I. Bureca, S. Paolucci, V.M. Saraceni, Quality of life improved by multidisciplinary back school program in patients with chronic non-specific low back pain: a single blind randomized controlled trial, Eur. J. Phys. Rehabil. Med. 47 (2011) 533-41. [8] M. van Middelkoop, S.M. Rubinstein, A.P. Verhagen, R.W. Ostelo, B.W. Koes, M.W. van Tulder, Exercise therapy for chronic nonspecific low-back pain, Best Pract. Res. Clin. Rheumatol. 27 (2010) 193-204. [9] S. Michie, S. Ashford, F.F. Sniehotta, S.U. Dombrowski, A. Bishop, D.P. French, A refined taxomomy of behaviour change techniques to help people change their physical
20 activity and health eating behaviours: The CALO-RE taxonomy, Psychol. Health 26 (2011) 1479-98. [10] C.L. Miles, T. Pincus, D. Carnes, K.E. Homer, S.J.C. Taylor, A. Bremner, A. Rahman, M. Unterwood, Can we identify how programmes aimed at promoting self-management in musculoskeletal pain work and who benefits? A systematic review of sub-group analysis within RCTs, Eur. J. Pain 15 (2011) 775.e1-775.e11 [11] M. Weigl, A. Cieza, P. Cantista, JD Reinhardt, G. Stucki, Determinants of disability in chronic musculoskeletal health conditions: a literature review, Eur. J. Phys. Rehabil. Med. 44 (2008) 67-79. [12] A. Reusch, M. Schug, R. Küffner, H. Vogel, H. Faller, [Medical rehabilitation groupprogrammes concerning health promotion, patient education and psychoeducation - a 2010 national survey], Rehabilitation 52 (2013) 226-33. [13] K. Meng, B. Seekatz, H. Roßband, U. Worringen, H. Faller, H. Vogel, [Development of a standardized back school for in-patient orthopaedic rehabilitation], Rehabilitation 48 (2009) 335-44. [14] K. Meng, B. Seekatz, H. Roßband, U. Worringen, H. Vogel, H. Faller, Intermediate and long-term effects of a standardized back school for inpatient orthopedic rehabilitation on illness knowledge and self-management behaviors: A randomized controlled trial, Clin. J. Pain 27 (2011) 248-57. [15] I. Brandes, A. Bönisch, U. de Vries, I. Ehlebracht-König, F. Petermann, C. Krauth, [Exemplary implementation of patient education as rehabilitation programmes], Phys. Med. Rehab. Kuror. 18 (2008) 344-8. [16] U. de Vries, I. Brandes, C. Krauth, F. Petermann, [Patient education in pulmonary rehabilitation: results of an implementation study], Gesundheitswesen 70 (2008) 572-81.
21 [17] J.A. Durlak, E.P. DuPre, Implementation matters: a review of research on the influence of implementation on program outcomes and the factors affecting implementation, Am. J. Community Psychol. 41 (2008) 327-50. [18] L. Dusenbury, R. Brannigan, M. Falco, W.B. Hansen, A review of research on fidelity of implementation: implications for drug abuse prevention in school settings, Health Educ. Res. 18 (2003) 237–56. [19] G.M. Curran, M. Bauer, B. Mittmann, J.M. Pyne, C. Stetler, Effectivenessimplementation hybrid designs combining elements of clinical effectiveness and implementation research to enhance public health impact, Med. Care 50 (2012) 217-26. [20] B.R. Flay, Efficacy and effectiveness trials (and other phases of research) in the development of health promotion programs, Prev. Med. 15 (1986) 451-71. [21] K. Meng, S. Peters, A. Schultze, K. Pfeifer, H. Faller, [The Impact of 2 Interventions on Implementation Fidelity of a Standardized Back School Program in Inpatient Orthopedic Rehabilitation Facilities], Rehabilitation 54 (2015) 325-331.. [22] W. Göhner, [Prevention of pain chronification: adherence to physiotherapy], Logos, Berlin, 2003. [23] C. Herda, S. Keller, K. Ridder, H.-D. Basler, [Motivation zu rückenfreundlichem Verhalten], in: S. Keller (Ed.), [Motivation zur Verhaltensänderung], Lambertus, Freiburg i. B., 1999, pp. 213-28. [24] I. Frey, A. Berg, D. Gratwohl, J. Keul, [Freiburger questionnaire on physical activity – design, validation and application], Soz. Präventivmed. 44 (1999) 55-64. [25] E. Geissner, [Coping with chronic pain – Scales for the assessment of coping strategies and pain-related distress], Z. Klin. Psychol. 28 (1999) 280-90.
22 [26] A.J. Vickers, Analysis of variance is easily misapplied in the analysis of randomized trials: a critique and discussion of alternative statistical approaches, Psychosom. Med. 67 (2005) 652-5. [27] J. Cohen, Statistical power analysis for the behavioral sciences, Lawrence Erlbaum Associates, Hillsdale, New York, 1988. [28] A. Field, Discovering Statistics Using SPSS, SAGE Publications, London, 2005. [29] M. Bullinger, I. Kirchberger, [SF-36 Fragebogen zum Gesundheitszustand], Hogrefe, Göttingen, 1998. [30] M. Glattacker, K. Heyduck, C. Meffert, Illness beliefs, treatment beliefs and information needs as starting points for patient information - Evaluation of an intervention for patients with chronic back pain, Patient Educ. Couns. 86 (2012) 378-89. [31] K. Mattukat, D. Rennert, I. Brandes, I. Ehlebracht-König, K. Kluge, W. Mau, Short- and long-term effects of intensive training and motivational programme for continued physical activity in patients with inflammatory rheumatic disease, Eur. J. Phys. Rehabil. Med. 50 (2014) 395-409. [32] S. May, Self-management of chronic low back pain and osteoarthritis, Nat. Rev. Rheumatol. 6 (2010) 199-209. [33] N. McGrane, R. Galvin, T. Cusack, E. Stokes, Addition of motivational interventions to exercise and traditional physiotherapy: a review and meta-analysis, Psychotherapy 101 (2015) 1-12. [34] S. Michie, C. Abraham, C. Whittington, J. McAteer, S. Gupta, Effective techniques in healthy eating and physical activity interventions: a meta-regression, Health Psychol. 28 (2009) 690-701.
23 [35] P.H. Ferreira, M.L. Ferreira, C.G. Maher, K.M. Refshauge, J. Latimer, R.D. Adams, The therapeutic alliance between clinicians and patients predicts outcome in chronic low back pain, Phys. Ther. 93 (2013) 470-8. [36] T. Mars, D. Ellard, D. Carnes, K. Homer, M. Underwood, S.J.C. Taylor, Fidelity in complex behaviour change interventions: a standardised approach to evaluate intervention integrity, BMJ Open 3 (2013) e003555. [37] M. Leeuw, M.E.J.B. Goossens, S.J. Linton, G. Crombez, K. Boersma, J.W.S. Vlaeyen, The Fear-Avoidance Model of Musculoskeletal Pain: Current State of Scientific Evidence, J. Behav. Med. 30 (2007) 77-94. [38] E.J. Yang, W.-B. Park, H.-I. Shin, J.-Y. Lim, The effect of back school integrated with core strengthening in patients with chronic low-back pain, Am. J. Phys. Med. Rehabil. 89 (2010) 744-54. [39] E. Toomey, L. Currie-Murphy, J. Matthews, D.A. Hurley, Implementation fidelity of physiotherapist-delivered group education and exercise interventions to promote selfmanagement in people with osteoarthritis and chronic low back pain: A rapid review Part II, Man. Ther. 20 (2015) 287-94. [40] E. Toomey, L. Currie-Murphy, J. Matthews, D.A. Hurley, The effectiveness of physiotherapist-delivered group education and exercise interventions to promote selfmanagement for people with osteoarthritis and chronic low back pain: A rapid review Part I, Man. Ther. 20 (2014) 265–86. [41] F.P. Kroon, L.R. van der Burg, R. Buchbinder, R.H. Osborne, R.V. Johnston, V. Pitt, Selfmanagement education programmes for osteoarthritis, Cochrane Database Syst. Rev. 1 (2014) CD008963.
24 Figure 1. Study flow
Figure 2. Back exercises (means) at 6 and 12 months follow-up by treatment group (IG intervention group; CG control group) and gender. Analyses of covariance/interaction effects: 6 months: p = 0.094, η2 = 0.007; 12 months: p = 0.099, η2 = 0.008.
25
Table 1. Characteristics of the traditional back schools (control group) and fidelity of the implemented back school (intervention group) Clinic 1
Clinic 2
Clinic 3
Clinic 4
Modular concept
yes
yes
no
no
Number of modules delivered by exercise therapist
5
5
3
3
Written manual
no
in part
no
no
Maximum 15 participants
no
yes
no
yes
Closed group (modules delivered by exercise therapist)
yes
no
no
no
Patient handouts
yes
yes
yes
no
Lecture physiciana
yes
yes
yes
yes
Lecture psychologista
no
no
yes
no
Number, duration and frequency of modules
A
A
A
A
Sequence of modules
A
B
A
B
Closed group format (all modules)
A
B
A
C
Number of participants (maximum 15)
A
A
A
C
Health professionals (interdisciplinary implementation)
A
A
C
B
Therapist constancy (modules of exercise therapist)
A
A
C
C
Indication (allocated patients)
B
A
A
B
Seating arrangement/room setup
A
B
A
B
Slides/patient booklet
A
B
B
B
Contents physician/psychologist/exercise therapist
A/A/C
A/A/A
A/B/A
A/B/B
Methods physician/psychologist/exercise therapist
B/A/C
B/A/A
C/B/A-B
B-C/B/B-C
Control group - traditional back school
Intervention group - implemented back school Structural key elements
Control group: elements of the traditional back schools were assessed by semi-structured interviews and observations. Intervention group: structural key elements of the implemented back school were assessed by standardized interviews, realization of contents and methods by structured observations, respectively.
26 A = key elements according to manual and mandantory contents/methods were predominantly performed. B = key elements with minor adaptation in concordance with suggestions to overcome barriers and mandatory contents/methods were partly performed. C = key elements with major adaptation not suggested for barrier management and mandatory contents/methods were predominantly not performed. a
Lectures of physicians and/or psychologists that are not inherent part of the back school program.
27 Table 2. Baseline sociodemographic and clinical characteristics
28
Female, n (%) Age in years, mean (SD) Marital status, n (%)
Control group (n = 256)
Intervention group (n = 279)
p value
139 (54.3)
158 (56.6)
0.602
50.5 (8.1)
49.7 (8.6)
a
0.683
Single
34 (13.3)
44 (15.6)
Married/with partner
173 (67.8)
183 (65.8)
Divorced/separated
37 (14.5)
43 (15.5)
11 (4.3)
8 (2.9)
Widowed Education, n (%)
b
0.697
Less than junior (< 10 y; basic secondary school)
120 (47.1)
126 (45.5)
Junior (10 y; middle-level secondary school)
91 (35.6)
92 (33.2)
Senior (Qualification for university)
39 (15.3)
52 (18.8)
5 (2.0)
7 (2.5)
Other Working status, n (%)
a
0.394
Employed Unemployed Retired/other Occupation, n (%)
0.227
227 (89.0)
238 (85.6)
16 (6.3)
22 (7.9)
12 (4.7)
18 (6.5)
c
0.244
Worker
91 (36.3)
89 (32.5)
151 (60.2)
167 (60.9)
9 (3.6)
18 (6.5)
Dorsalgia (M54)
180 (70.3)
212 (76.0)
0.144
Other intervertebral disc disorders (M51)
114 (44.5)
162 (58.1)
0.002
69 (27.0)
69 (24.7)
0.621
Employee Other Main diagnosis (ICD-10), n (%)
d
Other dorsopathies, not elsewhere classified (M53) Pain duration, n (%)
e
0.235
< 1 year
30 (12.5)
42 (15.6)
< 3 years
61 (25.4)
49 (18.2)
< 5 years
26 (10.8)
32 (11.9)
≥ 5 years
123 (51.2)
146 (54.3)
Pain intensity (NRS), mean (SD)
5.8 (2.0)
5.8 (2.0)
0.680
Days of sick leave due to back pain in the last 6 months, mean (SD) SF-12 physical component
35.8 (51.4)
32.4 (51.1)
0.454
37.3 (8.8)
36.5 (8.5)
0.330
SF-12 mental component
42.4 (11.2)
41.8 (11.9)
0.568
NRS Numerical rating scale of pain intensity (0-10, with high scores indicating higher pain intensity); SF-12 Short form health survey (0 to 100, with high scores representing best health status). a
missing = 2. b missing: n = 3. c missing = 10. d Multiple comorbid diagnoses possible. e missing: n = 26. Rated
by the physician.
29 Table 3. Intermediate-term (6 months) and long-term (12 months) within-group and between-group effects on primary and secondary outcomes
Pre-treatment
mean (SD)
6-months
Within-group
follow-up
change
mean (SD)
SES (95% CI)
Illness knowledge
Between-group difference
Pre-treatment
ANCOVA p-value
η2
0.018
0.015
12-months follow-up
mean (SD)
mean (SD)
CG
25.43 (5.09)
26.35 (5.27)
0.17 (0.05 – 0.29)
25.39 (5.18)
26.36 (5.49)
IG
25.41 (5.19)
27.26 (5.35)
0.34 (0.22 – 0.46)
25.55 (5.33)
27.34 (5.54)
Back exercises
0.464
0.001
CG
2.25 (1.08)
2.85 (0.94)
0.56 (0.39 – 0.72)
2.22 (1.09)
2.75 (0.97)
IG
2.25 (1.06)
2.91 (0.89)
0.61 (0.44 – 0.78)
2.28 (1.06)
2.81 (0.92)
Physical activity
<0.001
0.069
CG
8.17 (5.73)
6.98 (5.51)
-0.22 (-0.35 – -0.09)
8.36 (5.81)
9.57 (6.59)
IG
7.61 (5.15)
9.00 (5.10)
0.25 (0.09 – 0.42)
7.75 (5.13)
9.42 (5.81)
Back posture habits
0.695
0.000
CG
2.70 (1.00)
3.30 (0.81)
0.61 (0.45 – 0.78)
2.69 (0.98)
3.32 (0.75)
IG
2.82 (0.95)
3.32 (0.80)
0.51 (0.38 – 0.64)
2.80 (0.93)
3.21 (0.82)
Action-oriented coping
0.488
0.001
CG
14.30 (5.02)
16.14 (5.31)
0.36 (0.21 – 0.51)
14.23 (5.00)
15.90 (5.64)
IG
14.89 (4.96)
16.72 (5.08)
0.36 (0.20 – 0.52)
14.91 (4.92)
16.07 (4.77)
Cognitive restructuring
0.634
0.001
CG
13.83 (4.69)
15.45 (8.08)
0.34 (0.19 – 0.50)
13.77 (4.66)
15.64 (5.03)
IG
13.86 (4.80)
15.25 (5.00)
0.30 (0.15 – 0.44)
13.75 (4.56)
15.10 (4.80)
Subjective coping competence
0.961
0.000
CG
15.51 (4.86)
16.32 (5.02)
0.18 (0.03 – 0.32)
15.62 (4.91)
16.91 (5.01)
IG
15.45 (4.50)
16.27 (4.86)
0.18 (0.02 – 0.33)
15.43 (4.35)
16.22 (4.56)
Mental distraction
0.878
0.000
CG
9.97 (4.50)
10.75 (5.11)
0.17 (0.03 – 0.31)
9.72 (4.28)
10.42 (5.06)
IG
9.74 (4.83)
10.57 (4.86)
0.18 (0.03 – 0.33)
9.61 (4.81)
10.14 (5.02)
Counter-activities
0.794
0.000
CG
11.97 (4.47)
12.42 (4.40)
0.10 (-0.03 – 0.23)
12.13 (4.71)
12.45 (4.64)
IG
12.15 (4.61)
12.41 (4.40)
0.06 (-0.09 – 0.20)
12.08 (4.63)
12.18 (4.33)
Relaxation
0.833
0.000
CG
11.64 (5.16)
13.10 (5.21)
0.28 (0.15 – 0.42)
11.88 (5.13)
12.52 (5.19)
IG
12.25 (5.14)
13.54 (5.29)
0.25 (0.12 – 0.38)
11.70 (5.01)
13.36 (5.13)
ANCOVA indicates analysis of covariance; CI, confidence interval; SES, standardized effect size (positive SES represents improvement). CG, control group; IG, intervention group. Effect size η² (small: η2 = 0.0099, medium: η2 = 0.0588, large: η2 = 0.137); SES (small = 0.20, medium = 0.50, large = 0.80). Bold figures indicate: a)
30 regarding within-group effects: SES of the CG is outside the 95% CI of SES of the IG; or b) regarding betweengroup effects: significant between-group effect (ANCOVA: p < 0.05). 6-months follow-up: n = 398 (CG: n = 205, IG: n = 193). 12-months follow-up: n = 348 (CG: n = 177, IG: n = 171).
Enrolment Allocation Follow-up Analysis
Control group
Intervention group
Assessed for eligibility/contacted for participation (n = 299)
Assessed for eligibility/contacted for participation (n = 388)
Refused to participate (n = 33)
Refused to participate (n = 97)
Agreed to participate (n = 266)
Agreed to participate (n = 291)
Withdrew consent (n = 4), did not complete baseline questionnaire t1 (n = 4), exclusion criterion (n = 2)
Withdrew consent (n = 4), did not complete baseline questionnaire t1 (n = 8)
Received allocated intervention/ completed t1 assessment (n = 256)
Received allocated intervention/ completed t1 assessment (n = 279)
Completed t2 assessment (n = 236)
Completed t2 assessment (n = 260)
Lost/reason: no reply, empty questionnaire (n = 20)
Completed t3 assessment (n = 205) Lost/reason: no reply (n = 48), moved (n = 3)
Lost/reason: no reply, empty questionnaire (n = 19)
Completed t3 assessment (n = 193) Lost/reason: no reply (n = 79), moved (n = 7)
Completed t4 assessment (n = 177)
Completed t4 assessment (n = 171)
Lost/reason: no reply (n = 75), moved (n = 4)
Lost/reason: no reply (n = 98), moved (n = 10)
Analyzed (n = 242) Data at t1 and at least at one of t2 - t4
Excluded (n = 14) Only data at t1
Analyzed (n = 268) Data at t1 and at least at one of t2 - t4
Excluded (n = 11) Only data at t1
4,0
Back exercises
3,5
IG Women CG Women IG Men CG Men
3,0
2,5
2,0
1,5
1,0 Pretreatment
6 months follow-up
4,0
Back exercises
3,5
IG Women CG Women IG Men CG Men
3,0
2,5
2,0
1,5
1,0 Pretreatment
12 months follow-up