Early treatment is considered essential for developmental dysplasia of the hip (DDH), but the choice of screening strategy is debated. The investigators evaluated the effect of a selective ultrasound (US) screening programme.

All infants born in a defined region during 1991-2006 with increased risk of DDH, i.e. clinical hip instability, breech presentation, congenital foot deformities or a family history of DDH, were subjected to US screening at age one to three days. Severe sonographic dysplasia and/or dislocatable/dislocated hips were treated with abduction splints. Mild dysplasia and/or pathological instability, i.e. not dislocatable/dislocated hips were followed clinically and sonographically until spontaneous resolution, or until treatment became necessary. The minimum observation period was 5,5 years.

Of 81564 newborns, 11539 (14,1%) were identified as at risk, of which 11190 (58% girls) were included for further analyses. Of the 81564 infants, 2433 (3•0%) received early treatment; 1882 (2,3%) from birth and 551 (0,7%) after six weeks or more of clinical and sonographic surveillance. Another 2700 (3,3%) normalised spontaneously after watchful waiting from birth. Twenty-six infants (0,32 per 1000, 92% girls, two from the risk group) presented with late subluxated/dislocated hips (after one month of age). Another 126 (1,5 per 1000, 83% girls, one from the risk group) were treated after isolated late residual dysplasia. Thirty-one children (0,38 per 1000) had surgical treatment before age five years. Avascular necrosis was diagnosed in seven of all children treated (0.27%), four after early and three after late treatment.

Interpretation The first 16 years of a standardised selective US screening programme for DDH resulted in acceptable rates of early treatment and US follow-ups, and low rates of late subluxated/dislocated hips compared to similar studies.

Inclusion Criteria:

• born at Haukeland University hospital January 1991-December 2006

Exclusion Criteria:

• Children with DDH due to neuromuscular syndromes were excluded.

All infants born at the maternity unit at Haukeland University Hospital from January 1991 through December 2006 were included. The hospital provides the only delivery unit for the city and suburbs of Bergen and a large rural area within the Hordaland County. It serves a population of approximately 400 000 inhabitants, predominantly ethnic Norwegians. The annual birth rates varied from 4723 to 6010.

Minimum observation time was 5,5 years.