Abstract
BACKGROUND: Wide variation exists in reported prevalence estimates and management standards of developmental dysplasia of the hip (DDH). Discrepancies in diagnosticians' opinions may explain some of this variation. QUESTIONS/PURPOSES: We sought to determine (1) the consistency with which pediatric orthopaedic surgeons rate the importance of diagnostic criteria for DDH, and (2) whether there were geographic differences in how the diagnostic criteria were rated by surgeons. METHODS: One hundred ninety-seven of 220 members of the European Paediatric Orthopaedic Society and 100 of 148 members of the British Society of Children's Orthopaedic Surgery treating children with DDH participated in this cross-sectional study across 35 countries (15 regions). Each rated 37 items in four domains that specialists previously had identified as the most important features associated with DDH in early infancy. We determined consistency using the intraclass correlation coefficient (ICC; two-way random-effects model) interpreted as poor (0-0.40), acceptable (0.41-0.74), or good (≥ 0.75). RESULTS: Poor consistency among surgeons was found in rating the 37 diagnostic criteria (ICC, 0.33; 95% CI, 0.24-0.45). Consistency was poor for three domains (patient characteristics/history: ICC, 0.29; 95% CI, 0.16-0.58; ultrasound: ICC, 0.26; 95% CI, 0.14-0.52; radiography: ICC, 0.34; 95% CI, 0.12-0.95) and acceptable for one (clinical examination: ICC, 0.50; 95% CI, 0.33-0.73). Surgeons in particular regions appeared to have a concept of DDH diagnosis that distinguished them from specialists of other regions; consistency in eight regions was greater (ICC ≥ 0.40) than consistency among all 15 regions. CONCLUSIONS: The consistency of specialists in rating diagnostic criteria for DDH was lower than expected, and there was considerable geographic variation in terms of how specialists assigned importance ratings of the diagnostic criteria; these findings are somewhat counterintuitive, given the frequency with which this condition is diagnosed. These inconsistencies could explain, partly, the widely differing prevalence estimates and management standards of DDH.