INTER-TESTER RELIABILITY IN A MANUAL STRATEGY CLINICAL CLASSIFICATION FOR TRAINING IN A NON-SPECIFIC LOW BACK PAIN WITH MALADAPTIVE MOTOR CONTROL

Abstract

 Study design: partially nested, multifaceted observational study of categories (C), subjects (S) and observers (O). Objectives: to analyze inter-examiner reliability in a manual strategy clinical classification for training in segmental stabilization exercises (SSE) in a chronic non-specific low back pain (NLBP) with maladaptive motor control. Back-ground: twenty-four subjects (mean age 38 and SD 11.8 years) with NLBP were included in the study. Methods and measures: three physiotherapists performed a clinical procedure for a new four-stage (category) classification system during which an assessment form was tested with verbal commands and cognitive-behavioral and manual techniques. Inter-tester concordance was calculated using generalizability theory. Results: the greatest percentage of variance (63 %) was associated with the facet «category». The designs on 288 observations yielded high indices of reliability and generalizability, with minimum error rates (0.009 and 0.079). Design CS/O showed satisfactory in-dices of reliability and generalizability (intra-class correlation: ICC), the value being 0.96. Conclusions: the study showed that the inter-examiner reliability of categorization for this learning strategy was acceptable. As there is considerable variability in a strategy for teaching SSE, it is necessary to develop protocols of inter-tester reliability and validity for clinical outcomes.