RESUMEN
OBJECTIVES: The aims of this paper were: (i) to examine the intra-observer and inter-observer reliability of the shaft-condylar angle (SCA) and the lateral capitellohumeral angle (LCHA); (ii) to study the influence of experience level on the inter-observer and intra-observer reliability; and (iii) to determine the influence of the the age of the patients on reliability. METHOD: A retrospective cohort study was conducted. The study reviewed 81 elbow radiographs. The patients were aged between 2 and 13 years. All the images taken between 2000 and 2017 were independently measured by a senior pediatric orthopaedic surgeon, a pediatric orthopaedic surgeon, a pediatric orthopaedic fellow, an orthopaedic chief resident, a general practitioner, and a pediatric orthopaedic research assistant. Measurement was performed two times within a 2-week interval. Inexperienced observers (general practitioner and research assistant) were supervised by senior pediatric orthopaedic surgeons for at least 30 radiographs before performing the measurement. Inclusion criteria were as follows: (i) age 2-13 years; and (ii) no previous elbow fracture. EXCLUSION CRITERIA: elbow radiographs do not show true lateral view. The intraclass correlation coefficient (ICC) was used to calculate the reliability. RESULTS: The mean values of SCA and LCHA were 43° and 48°, respectively. For SCA, intra-observer reliability was excellent (ICC = 0.85) for one observer, good (range = 0.73-0.76) for three observers, and moderate (0.59) for one observer. Inter-observer reliability was moderate (0.48, 0.58), whereas the reliability categorized by age group showed excellent agreement (0.88-0.94). For LCHA, intra-observer reliability was excellent (0.84-0.89) for three observers and good (0.66-0.80) for two observers. Inter-observer reliability was moderate (0.44-0.45). Conversely, the reliability classified by age group showed excellent agreement (0.83-0.91). CONCLUSION: Intra-observer reliability for LCHA and SCA were excellent to good for most observers. Inter-observer reliability was moderate for LCHA and SCA. Reliability classified by age group showed excellent to good agreement. Reliability was influenced by the level of experience, especially for non-medical staff.