Inter- and intra-observer reliability of cobb angle measurement between orthopaedic residents and spine specialists

Abstract

Cobb angle measurement is essential for diagnosing and monitoring scoliosis; however, its accuracy varies with observer experience. In many healthcare systems, including Oman, initial measurements are commonly performed by residents and influence referral urgency, follow-up intervals, and bracing decisions, where even small inaccuracies may alter clinical pathways. This study evaluated the inter- and intra-observer reliability of Cobb angle measurement among orthopaedic residents and spine specialists. A prospective comparative reliability study was conducted including 18 OMSB orthopaedic residents (PGY2-5) and five spine specialists. Each participant measured Cobb angles on five standardised standing scoliosis radiographs on two occasions, four weeks apart. Inter- and intra-observer reliability were assessed using the intraclass correlation coefficient (ICC) with a two-way random-effects, single-measurement, absolute-agreement model [ICC(2,1)], and Bland-Altman plots were used to assess intra-observer reproducibility. Mean Cobb angles ranged from 45° to 56°. Inter-observer reliability was poor among residents (ICC = -0.219; 95% CI -1.41 to 0.485) and remained low among specialists (ICC = 0.120; 95% CI -2.28 to 0.898). Intra-observer variation averaged approximately 5°, with a Bland-Altman mean bias close to zero and limits of agreement of ±5%. Negative ICC values indicated that between-observer variability exceeded within- observer consistency. Residents demonstrated markedly poorer inter-observer reliability compared to specialists, highlighting the impact of clinical experience on accurate end-vertebra selection and angle measurement. Targeted training modules and structured exposure during residency may improve measurement consistency, referral accuracy, and overall scoliosis management pathways.