DDH Treatment Options

Developmental dysplasia of the hip is the most common hip disorder in infants. Early diagnosis and age-appropriate treatment are essential to prevent long-term disability, AVN and early osteoarthritis.

Who. The incidence is 1 in 1000 live births, with a female predominance of 7:1 (relaxin sensitivity). The left hip is affected in 60% (fetal position) and both hips in 20%. DDH is commoner in Caucasian, Lapland and Native American populations.

Risk factors. Beyond female sex and ethnicity:

• Breech presentation - 20% of children with DDH were breech, against 2% of the population
• Family history - 12% if parents had DDH
• Firstborn
• Packaging disorders - torticollis, CTEV, metatarsus adductus
• Calcaneovalgus - 5% risk of associated DDH

Natural history. 1 in 60 is abnormal at birth, but 60% stabilise by one week and 88% by two months. Untreated DDH leads to abnormal gait, limb shortening and early osteoarthritis. The earlier treatment starts, the better the harness works; late treatment brings worse outcomes and a higher AVN risk, and may require multiple surgeries.

The epidemiology figures above (1:1000 incidence, 7:1 female ratio, left 60%/bilateral 20%, breech 20% vs 2%, family history 12%, calcaneovalgus 5%) are conventional teaching figures - none of the cited series measured them. The measured anchors on this page are Abdulla 2023 (83.4% normalisation), Pollet 2010 (Graf-dependent late reduction), Gardner 2014 (AVN 20-24%) and Bohm 2002 (90% revision-free at 35 yr); the natural-history figure (~90% of mild instability resolving by 8 weeks) follows Vaquero-Picado 2019.

The spectrum. DDH runs from mild acetabular dysplasia to complete dislocation. The acetabulum is shallow, the labrum hypertrophies and inverts, the capsule is lax, the ligamentum teres elongates, fibrofatty pulvinar develops in the acetabulum, and the femoral head changes secondarily. Understanding these changes guides the treatment approach: the labrum, capsule, ligamentum teres and pulvinar each block reduction.

The acetabulum. The acetabular index is 30° at birth and falls to 20° by 24 months; in DDH it is elevated and the acetabulum is shallow. The acetabulum can remodel up to 5 years of age. After that remodelling is limited, and an osteotomy may be needed.

The femoral head. A recently dislocated head may be normal. A chronically dislocated one flattens and loses sphericity. Ossification is delayed in the dislocated hip. Reduced early, it may remodel.

Graf ultrasound. Graf grades the infant hip on two angles: the alpha angle, bony acetabular coverage, normal over 60°, and the beta angle, cartilaginous coverage, normal under 60°. Ultrasound is the imaging modality of choice under 4-6 months, because the cartilaginous head is not visible on X-ray.

View larger

Radiographs. Once the femoral head ossifies, after 4-6 months, radiographs take over. Hilgenreiner's line runs horizontally through the triradiate cartilage and Perkin's line vertically through the lateral edge of the acetabulum.

The neonate. Two tests, both read by a clunk:

• Ortolani - abduction with gentle pressure over the greater trochanter; a clunk indicates reduction
• Barlow - adduction with posterior pressure; a clunk indicates subluxation

Both are reliable in the first three months and become unreliable after 2-3 months as the soft tissues tighten. The other signs are asymmetric creases, limited abduction (under 60°) and the Galeazzi sign, which works in unilateral cases only.

The older infant. After 3 months, limited abduction is the primary sign, and between 3 and 12 months abduction range and the Galeazzi sign are moderately reliable. A unilateral dislocation produces a limb length discrepancy. Asymmetric skin creases may be present but are not diagnostic.

The walking child. Over 12 months the child presents with a limp, limb length discrepancy and limited motion: a waddling gait when both hips are affected, a Trendelenburg gait when one is. Examination is less reliable at this age, and imaging is key.

Ultrasound. Most useful from 4-6 weeks until the femoral head ossifies. Beyond the static Graf angles, a dynamic assessment evaluates stability during movement.

View larger

Radiographs. After 4-6 months:

• AP pelvis - acetabular index, Shenton's line, head position
• Frog-leg lateral - femoral head coverage, version

The acetabular index is normally under 30° at 1 year; the centre-edge angle is measured after age 5.

Arthrogram. Performed during the examination under anaesthesia for closed reduction, it shows the quality of the reduction and determines the safe zone, the abduction range that maintains reduction without AVN risk. A normal study shows a sharp labrum (the rose thorn sign) and minimal medial pooling. An abnormal one shows a widened medial joint space (pooling over 5mm), a blunted labrum and an hourglass constriction.

DDH itself is reducible or dislocatable, with a shallow acetabulum and no fixed deformity.

Age decides the treatment. The first-line treatment by age band:

• 0-6 months: Pavlik harness - first-line for dislocatable and dislocated hips
• 6-18 months: closed reduction - examination under anaesthesia, arthrogram and spica
• 18 months-3 years: open reduction - when closed reduction fails, with femoral shortening if the hip is tight
• Over 3 years: open reduction with osteotomy - open reduction is always required, with a femoral varus and shortening osteotomy and a pelvic osteotomy for residual dysplasia; more complex, with a higher complication risk

The Graf type decides the harness's odds. Abdulla's 83.4% sonographic normalisation spans all Graf types; the rate is far higher in Graf II and far lower in Graf III/IV. An earlier start favours success, mean age 6.7 weeks in successes against 8.8 weeks in failures. The conventional age ceiling for the harness is 6-7 months, but Pollet trialled a progressively applied harness in late-diagnosed dislocated hips aged 6-24 months, and the results again split by Graf type (table).

What the evidence does not give you. No series reports a success rate for closed or open reduction. For closed reduction the endpoints are intraoperative, an adequate safe zone and a concentric arthrographic reduction; for open reduction the reported harm is AVN. Over 3 years there is no comparative series; Bohm's Salter osteotomies supply the long-term figures.

Avascular necrosis. The most devastating complication, and it always results from treatment, not the disease. Preventing it is the most important principle in DDH treatment.

• Pavlik harness - 0% of reduced hips (Pollet 2010); 5-10% conventional range
• Closed reduction - ~20% conventional (Salter type 1)
• Open reduction - medial approach 20% measured, rising to 24% at skeletal maturity (Gardner 2014, 734 hips); anterior approach 10-20% conventional
• Open reduction with osteotomy - 15-25% conventional

What causes it. Forced or extreme abduction: immobilisation at 60° or more measurably raised AVN in Gardner's review. Extreme abduction combined with internal rotation and extension, multiple reduction attempts and early rigid casting add to the risk. Prevention follows: a gentle reduction technique, avoiding multiple reduction attempts, the human position in the spica, and a harness discontinued when the hip has not reduced.

View larger

Grading. Kalamchi-MacEwen grades the damage by the part of the head and physis involved:

• Class I - nucleus only; will be normal
• Class II - lateral physis; coxa valga from lateral physeal fusion
• Class III - central; moderate deformity
• Class IV - whole physis; severe deformity, poor outcomes

Established AVN. Managed with containment procedures, and may need later salvage. Monitor for growth disturbance, coxa magna, coxa breva and coxa vara.

View larger

View larger

Redislocation. Conventionally quoted at 5-10% of closed reductions (no measured rate in the cited series). An inadequate safe zone, poor positioning and cast problems are the risk factors, and it is managed by repeat closed reduction or open reduction.

Stiffness. May occur after open reduction. It usually improves with time and physiotherapy, and may require manipulation or arthrolysis if severe.

Limb length discrepancy. May occur with unilateral DDH. It is usually mild (under 2cm) and may require later equalisation if significant.

Early osteoarthritis. Even successfully treated DDH carries an increased risk, and it may require total hip arthroplasty in adulthood.

A concentric reduction is itself the main treatment for the acetabulum. Residual dysplasia may occur despite a successful reduction and requires monitoring with serial radiographs. Acetabular development depends on a concentrically reduced femoral head providing the growth stimulus, so achieving and holding a concentric reduction lets the majority of acetabular dysplasia remodel spontaneously. The trajectory of the acetabular index over serial radiographs is the best predictor of persistent dysplasia: an index steadily falling toward normal can be observed, whereas one that plateaus or stays high will not correct on its own.

How long to wait. Most acetabular improvement occurs in the first 18-24 months after reduction, with useful remodelling potential continuing to about 4-5 years. A reasonable rule is to observe with serial films while the index is improving. Consider a pelvic (acetabular) procedure once it is clear that spontaneous correction has stalled, with an acetabular index over 30° or a centre-edge angle under 20° persisting beyond the expected window (typically after about age 3-5), rather than operating on every high index early.

View larger

Which osteotomy. The triradiate cartilage and the deformity decide. With it open, in a young child, a reshaping acetabuloplasty (Pemberton, or Dega for posterior deficiency) or a redirectional Salter innominate osteotomy is used; once it closes, a redirectional periacetabular (Ganz) osteotomy is required. A coexisting femoral deformity (excess anteversion or valgus) may need a varus derotation femoral osteotomy alongside.

Concentric and AVN-free first. As the long-term Salter data show, durability depends on a concentric, AVN-free reduction. A pelvic osteotomy corrects persistent structural dysplasia; it cannot rescue a non-concentric or AVN-damaged hip, so fix the reduction before the osteotomy, or reserve or stage it.

View larger

After open reduction. A spica for 3 months, changed at 6 weeks, then physiotherapy once the cast is off. Serial radiographs watch for AVN, which may present late.

To skeletal maturity. Follow-up continues until the skeleton is mature, looking for AVN, residual dysplasia and redislocation, and following acetabular development and femoral head shape. The acetabular index should normalise by age 2-3 years, and the centre-edge angle is assessed after age 5.

• Radiographs at 6 months, 1 year and 2 years
• Then annually until skeletal maturity

What predicts a good hip. Early diagnosis, and a concentric reduction achieved without force or extreme abduction. Late diagnosis correlates with worse outcomes.

View larger

View larger

Universal or selective ultrasound. Universal newborn ultrasound treats infants earlier and makes surgery less likely, but it markedly increases harness treatment (overtreatment of physiologically immature hips), and in the cited cohort it did not reduce delayed presentation. Selective screening avoids overtreatment but depends on reliable clinical examination. No global consensus exists; practice tracks resources and historical programmes.

Static or dynamic devices. The Pavlik harness allows active motion within a safe range, whereas rigid abduction braces (von Rosen, Tübingen, Plastazote) provide more fixed control. Evidence does not clearly favour one device, and rigid braces are often used as second-line after Pavlik failure or for non-compliant families rather than as proven superiors.

View larger

When to abandon the harness. The classic teaching is to abandon it if the hip is not reduced within 3-4 weeks, to limit AVN and Pavlik disease. Some series tolerate up to 6 weeks with a gentle progressive protocol. The unifying principle is that a persistently dislocated hip inside a harness causes posterior acetabular wear and must not be left.

Concurrent or staged osteotomy. In older children, whether to perform the pelvic or femoral osteotomy at the same sitting as the open reduction or in a staged fashion is debated. Long-term Salter data suggest that performing open reduction first and reserving or staging the osteotomy improves durability, but single-stage surgery is widely practised to limit anaesthetics.

The painless bilateral dislocation in the older child. In a walking child with bilateral, painless, completely dislocated hips presenting late (e.g. over 4-6 years), the morbidity of bilateral open reduction and osteotomy (stiffness, AVN, redislocation) may exceed the natural history of well-tolerated bilateral dislocation for many years. Whether to operate, operate on one side, or observe remains genuinely controversial and is individualised to symptoms, function and family expectations.