Neonatal clinical examination and selective ultrasound screening are not reliable for the early diagnosis of hip dysplasia: A retrospective cohort study

Aim: To compare the incidence, clinical features and timing of initial treatment of early- and later-presenting developmental dysplasia of the hip (DDH) in infancy in the context of a risk-based selective ultrasound screening programme.

Methods: We conducted a retrospective observational study of infants born at a tertiary perinatal centre in Melbourne, Australia, between July 2016 and June 2020 and treated for DDH up to 18 months of age at a tertiary paediatric centre. The incidence of DDH was calculated for early- and later-presenting cases. Clinical features and ultrasound indications were analysed using multivariate logistic regression.

Results: There were 192 infants with DDH requiring treatment, 100 (52%) of which were not detected by universal neonatal hip examination and selective ultrasound screening based on risk factors (overall incidence 6.3 per 1000 livebirths). The median age at which treatment was commenced was delayed by 8 weeks for the later-presenting group compared to those detected through screening (16.7 vs. 7.9 weeks, P < 0.001). Skin crease asymmetry was associated with later presentation (adjusted odds ratio (aOR) 44.03, 95% confidence interval (CI) 9.99-333), whilst breech presentation (aOR 0.08, 95% CI 0.02-0.26) and Barlow/Ortolani test positivity (aOR 0.06, 95% CI 0.01-0.25) were associated with early presentation. Only 39 (1.45%) infants receiving a screening ultrasound had DDH.

Conclusion: Approximately half of all infants with DDH were not detected by combining risk-based selective ultrasound screening with neonatal clinical findings, highlighting the failure of this screening method to reliably detect a significant proportion of dysplastic infant hips.