Schroth Exercises for Adolescent Idiopathic Scoliosis – Reliability, A Randomized Controlled Trial and Clinical Significance

  • Author / Creator
    Schreiber, Sanja
  • Background and objectives In America, care recommendations for adolescents with idiopathic scoliosis (AIS) include observation, bracing, and surgery. In Europe exercises are often prescribed. Schroth scoliosis-specific exercises have demonstrated promising results, but only in studies of suboptimal quality. Schroth exercise prescription is guided by curve classification. An algorithm for determining Schroth curve types was developed. Reporting statistical significance should be supplemented with clinical significance estimates, which is rare for research on conservative treatment for scoliosis. This thesis aimed: 1) to determine the reliability of Schroth therapists in classifying patients with AIS using the proposed algorithm; 2) to investigate the effect of Schroth exercises combined with standard of care on curve severity, qualityoflife (QOL), perceived appearance and back muscle endurance compared to standard of care; and 3) to determine the clinical significance of the effect of Schroth exercises for all outcomes. Methods For the reliability study, we recruited 44 participants with AIS and 10 consecutive volunteer Schroth certified therapists. Therapists rated video assessments presented randomly twice at least seven days apart blinded to identities. The reliability was determined using Gwet’s AC1 coefficients. For the RCT, we randomized 50 patients with AIS, aged 10-18 years, with curves 10°45° to standard of care (observation or bracing) or supervised Schroth exercises plus standard of care. After introducing Schroth exercises, a daily home program was adjusted during weekly supervised sessions for six months. Assessors and the statistician were blinded. The primary outcome was the change in the largest Cobb (LC) angle and the Sum of Cobb (SOC) angles. Secondary outcomes included Biering-Sorensen back muscle endurance test, Scoliosis Research Society (SRS22r) and Spinal Appearance Questionnaires’ (SAQ) scores. Per protocol and intention-to-treat linear mixed models analyses were reported. Clinical significance was determined using anchor and distribution-based methods. Numbers needed to treat (NNT), and proportion of improved, stable and deteriorated patients were reported. Results The overall intra-rater AC1 was 0.64 (95%CI 0.53-0.73), 0.70 (0.60-0.78) among well-trained raters, and 0.81 (0.77-0.85) in experienced raters. The weighted intra-rater AC1 averaged 0.75 (95%CI 0.63-0.84) overall, 0.82 (0.73-0.88) in well-trained raters, and 0.89 (0.80-0.94) in experienced raters. Inter-rater AC1 was 0.43 (95%CI 0.28-0.58) overall, 0.50 (0.38-0.61) for well-trained raters, and 0.67 (0.50-0.85) for experienced raters. The weighted inter-rater AC1 was 0.48 (95%CI 0.29-0.67) overall, 0.61 (0.49-0.72) among well-trained, and 0.79 (0.64-0.94) among experienced raters. After six months, Schroth group had by 3.5° (p<0.01) smaller LC in the per protocol analysis. The difference in the square root of the SOC also favored Schroth group (p<0.05) such that a patient with an average 51.2° SOC at baseline had a 49.3° at six months in the Schroth group,and 55° in the control group, while the difference increased with severity. Schroth exercises improved patients’ back muscle endurance, by 30 seconds (p=0.02). Intention-to-treat results were similar in direction and magnitude, but did not reach statistical significance. No statistically significant differences were observed between groups on QOL and perceived appearance outcomes. Some covariates (age, weight, height, self-efficacy, brace wear, and Schroth classification) had important main effects on outcomes in different models. Clinical significance was reached for the LC, LOC and SRS22r function score, with their respective cutoffs at 3.4°, 8.7°, and 0.11. Their NNT were 3 and 4, and 4 respectively. Proportions of improved or stable patients were significantly larger in the Schroth group for all outcomes. Biering-Sorensen test cutoff was 36.15 sec. High rates of ceiling effects were observed on questionnaires’ scores, which created problems in determining statistical and clinical significance. Treatment effects on the questionnaires’ scores were smaller than the measurement error and not statistically significant suggesting that the questionnaires were not responsive to change over time. Distribution-based methods produced cutoffs, which exceeded the magnitude of changes commonly seen in this population. Conclusions 1) Reliability of experienced and well-trained therapists was adequate when using the proposed algorithm. 2) Schroth exercises added to the standard of care led to statistically and clinically significant improvements in curve severity in patients completing the program. 3) Clinically significant improvement was observed for the Largest Cobb, Sum of Cobb angles, and SRS22r function scores. More responsive questionnaires may be needed to quantify treatment effects on QOL and perceived appearance.

  • Subjects / Keywords
  • Graduation date
    Spring 2015
  • Type of Item
  • Degree
    Doctor of Philosophy
  • DOI
  • License
    This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.
  • Language
  • Institution
    University of Alberta
  • Degree level
  • Department
  • Specialization
    • Rehabilitation Science
  • Supervisor / co-supervisor and their department(s)
  • Examining committee members and their departments
    • Kawchuk, Greg (Physical Therapy)
    • Vette, Albert (Mechanical Engineering)
    • Hedden, Douglas (Surgery)
    • Dolan, Lori (Department of Orthopaedics and Rehabilitation, University of Iowa)
    • Parent, Eric (Physical Therapy)