The aim was to assess the value of the GP 6–8 week hip examination. In a 15-year prospective observational longitudinal cohort study, every infant referred by the GP with suspected pathological developmental dysplasia of the hip (DDH) had their hip joints clinically and sonographically examined in a specialist hip screening clinic. Graf Type IV and dislocated hips were classified as pathological. Screening failures were defined as those who had not been identified by the 6–8 week check and presented with late instability. Secondary univariate and multivariable analysis was performed to determine which clinical findings are predictive of instability. 64,518 infants underwent the 6–8 week GP check. Of 176 referrals, 5 had pathological hips. 13 screening failures, presented between the ages of 17 and 80 weeks. The 6–8 week check has a sensitivity of 28% and a specificity of 99.7%. Univariate analysis revealed positive Ortolani tests and patients referred as ‘unstable hip’ to be significant predictors of hip pathology. Clicky hips, asymmetric skin creases, and leg length inequality were not predictive of pathological hips. A multivariable model showed a positive Ortolani test to be the sole independent predictor of instability at 6–8 weeks. This is the first attempt to test the validity of the 6–8 week GP clinical hip check. A low rate of hip pathology was identified. The high rate of false negatives raises questions about the value of screening at this age. At 6–8 weeks, clinical signs of hip instability are unreliable as hips become irreducible and stiff. Based on our findings, we recommend that at 6–8 weeks, referrals are only made if the Ortolani test is positive. We advocate the reintroduction of the 8-month check, including an assessment for limited hip abduction, which may improve the detection rate of those missed by initial screening.
Over a 15-year prospective period, 201 infants
with a clinically unstable hip at neonatal screening were subsequently
reviewed in a ‘one stop’ clinic where they were assessed clinically
and sonographically. Their mean age was 1.62 weeks (95% confidence
interval (CI) 1.35 to 1.89). Clinical neonatal hip screening revealed
a sensitivity of 62% (mean, 62.6 95%CI 50.9 to 74.3), specificity
of 99.8% (mean, 99.8, 95% CI 99.7 to 99.8) and positive predictive value
(PPV) of 24% (mean, 26.2, 95% CI 19.3 to 33.0). Static and dynamic
sonography for Graf type IV dysplastic hips had a 15-year sensitivity
of 77% (mean, 75.8 95% CI 66.9 to 84.6), specificity of 99.8% (mean,
99.8, 95% CI 99.8 to 99.8) and a PPV of 49% (mean, 55.1, 95% CI
41.6 to 68.5). There were 36 infants with an irreducible dislocation
of the hip (0.57 per 1000 live births), including six that failed
to resolve with neonatal splintage. Most clinically unstable hips referred to a specialist clinic
are female and stabilise spontaneously. Most irreducible dislocations
are not identified from this neonatal instability group. There may
be a small subgroup of females with instability of the hip which
may be at risk of progression to irreducibility despite early treatment
in a Pavlik harness. A controlled study is required to assess the value of neonatal
clinical screening programmes. Cite this article:
