Abstract
This review sought to identify, critically appraise, compare, and summarize the literature on the reliability, discriminative validity and responsiveness of the flexion relaxation ratio (FRR) in adults (≥ 18 years old) with or without spine pain (any duration), in either a clinical or research context. The review protocol was registered on Open Science Framework ( https://doi.org/10.17605/OSF.IO/27EDF ) and follows COSMIN, PRISMA, and PRESS guidelines. Six databases were searched from inception to June 1, 2022. The search string was developed by content experts and a health services librarian. Two pairs of reviewers independently completed titles/abstracts and full text screening for inclusion, data extraction, and risk of bias assessment (COSMIN RoB Toolkit). At all stages, discrepancies were resolved through consensus meetings. Data were pooled where possible with a three-level random effects meta-analyses and a modified GRADE assessment was used for the summary of findings. Following duplicate removal, 728 titles/abstracts and 219 full texts were screened with 23 included in this review. We found, with moderate certainty of evidence, that the cervical FRR has high test-retest reliability and lumbar FRR has moderate to high test-retest reliability, and with high certainty of evidence that the cervical and lumbar FRR can discriminate between healthy and clinical groups (standardized mean difference - 1.16 [95% CI - 2.00, - 0.32] and - 1.21 [- 1.84, - 0.58] respectively). There was not enough evidence to summarize findings for thoracic FRR discriminative validity or the standard error of measurement for the FRR. Several studies used FRR assuming responsiveness, but no studies were designed in a way that could confirm responsiveness. The evidence supports adequate reliability of FRR for the cervical and lumbar spine, and discriminative validity for the cervical and lumbar spine only. Improvements in study design and reporting are needed to strengthen the evidence base to determine the remaining measurement properties of this outcome.