Open Access Highly Accessed Open Badges Research article

Screening of the spine in adolescents: inter- and intra-rater reliability and measurement error of commonly used clinical tests

Ellen Aartun12*, Anna Degerfalk3, Linn Kentsdotter4 and Lise Hestbaek12

Author Affiliations

1 Department of Sports Science and Clinical Biomechanics, Faculty of Health Sciences, University of Southern Denmark, Campusvej 55, DK-5230 Odense M, Denmark

2 Nordic Institute of Chiropractic and Clinical Biomechanics, Campusvej 55, DK-5230 Odense M, Denmark

3 Clinical practice, Ryggcentrum Lund, Botulfsgatan 2, SE-22350 Lund, Sweden

4 Clinical practice, Aktivera kiropraktik & rehab, Glimmervägen 14, SE-19131 Sollentuna, Sweden

For all author emails, please log on.

BMC Musculoskeletal Disorders 2014, 15:37  doi:10.1186/1471-2474-15-37

Published: 10 February 2014



Evidence on the reliability of clinical tests used for the spinal screening of children and adolescents is currently lacking. The aim of this study was to determine the inter- and intra-rater reliability and measurement error of clinical tests commonly used when screening young spines.


Two experienced chiropractors independently assessed 111 adolescents aged 12–14 years who were recruited from a primary school in Denmark. A standardised examination protocol was used to test inter-rater reliability including tests for scoliosis, hypermobility, general mobility, inter-segmental mobility and end range pain in the spine. Seventy-five of the 111 subjects were re-examined after one to four hours to test intra-rater reliability. Percentage agreement and Cohen’s Kappa were calculated for binary variables, and interclass correlation (ICC) and Bland-Altman plots with Limits of Agreement (LoA) were calculated for continuous measures.


Inter-rater percentage agreement for binary data ranged from 59.5% to 100%. Kappa ranged from 0.06-1.00. Kappa ≥ 0.40 was seen for elbow, thumb, fifth finger and trunk/hip flexion hypermobility, pain response in inter-segmental mobility and end range pain in lumbar flexion and extension. For continuous data, ICCs ranged from 0.40-0.95. Only forward flexion as measured by finger-to-floor distance reached an acceptable ICC(≥ 0.75). Overall, results for intra-rater reliability were better than for inter-rater reliability but for both components, the LoA were quite wide compared with the range of assessments.


Some clinical tests showed good, and some tests poor, reliability when applied in a spinal screening of adolescents. The results could probably be improved by additional training and further test standardization. This is the first step in evaluating the value of these tests for the spinal screening of adolescents. Future research should determine the association between these tests and current and/or future neck and back pain.

Reliability; Measurement error; Scoliosis; Hypermobility; Intersegmental mobility; Spine; Adolescents; Mobility