Developmental Dysplasia of the Hip (DDH)

The normal growth of the acetabulum depends on normal epiphyseal growth of the triradiate cartilage and on the three ossification centers located within the acetabular portion of the pubis (os acetabulum), ilium (acetabular epiphysis), and ischium. Additionally, normal growth of the acetabulum depends on normal interstitial appositional growth within the acetabulum. The presence of the spherical femoral head within the acetabulum is critical for stimulating normal development of the acetabulum.

The anatomy of the dislocated hip, especially after several months, often includes formation of a ridge called the neolimbus. Closed reduction is often unsuccessful at a later date, secondary to various obstacles to reduction. These include adductor and psoas tendon contraction, ligamentous teres, a transverse acetabular ligament, and pulvinar and capsular constriction. With long-standing dislocations, interposition of the labrum can also interfere with reduction.

DDH involves abnormal growth of the hip. Ligamentous laxity is also believed to be associated with hip dysplasia, though this association is less clear. DDH is not part of the classic description of disorders that are associated with significant ligamentous laxity, such as Ehlers-Danlos syndrome or Marfan syndrome.

Children often have ligamentous laxity at birth, yet their hips are not usually unstable; in fact, it takes a great deal of effort to dislocate a child's hip. Therefore, more than just ligamentous laxity may be required to result in DDH. At birth, white children tend to have a shallow acetabulum.[7, 8] ; this may provide a susceptible period in which abnormal positioning or a brief period of ligamentous laxity may result in hip instability. However, this characteristic is not as true for children of African descent, who have a lower rate of DDH.[9]

The etiology of hip dysplasia is not clear, but this condition does appear to be related to a number of different factors.[10] One such factor is racial background: In Native Americans and Laplanders, the prevalence of hip dysplasia is much higher (nearly 25-50 cases per 1000 persons) than in other races, and the prevalence is very low in southern Chinese and black populations.[11, 12, 13, 9] An underlying genetic disposition also appears to exist, in that the frequency of hip dysplasia is 10 times higher in children whose parents had DDH than in those whose parents did not.[14]

Other factors possibly related to DDH include intrauterine positioning and sex, and some of these are interrelated. Female sex, being the first-born child, and breech positioning are all associated with an increased prevalence of DDH. An estimated 80% of persons with DDH are female,[15] and the rate of breech positioning in children with DDH is approximately 20% (compared with 2-4% in the general population).[16, 17] The prevalence of DDH in females born in breech position has been estimated to be as high as 1 case in 15 persons in some studies.[18]

Other musculoskeletal disorders of intrauterine malpositioning or crowding, such as metatarsus adductus and torticollis, have been reported to be associated with DDH.[19, 20] Oligohydramnios is also reported to be associated with an increased prevalence of DDH.[21] The left hip is more commonly associated with DDH than the right hip, possibly because of the common intrauterine position of the left hip against the mother's sacrum, which forces it into an adducted position.[21] Children in cultures in which the mother swaddles the baby, forcing the infant's hips to be adducted, also have a higher rate of hip dysplasia.[22]

Hip dysplasia can be associated with underlying neuromuscular disorders, such as cerebral palsy, myelomeningocele, arthrogryposis, and Larsen syndrome, though such cases are not usually considered DDH.

The overall frequency of DDH is usually reported as approximately 1 case per 1000 individuals, though Barlow believed that the incidence of hip instability during newborn examinations was as high as 1 case per 60 newborns.[23] According to Barlow's study, more than 60% of newborns with hip instability became stable by age 1 week, and 88% became stable by age 2 months, leaving only 12% (of the 1 in 60 newborns, or 0.2% overall) with residual hip instability.

Overall, the prognosis for children treated for hip dysplasia is very good, especially if the dysplasia is managed with closed treatment. If closed treatment is unsuccessful and open reduction is needed, the outcome may be less favorable[24] , although the short-term outcome appears to be satisfactory. If secondary procedures are needed to obtain reduction, then the overall outcome is significantly worse.

Some authors believe that patients with bilateral hip dysplasia have a poorer prognosis because of frequent delays in diagnosis and greater treatment requirements.[25, 26] In a study comparing the outcomes of walking-age children with bilateral hip dislocations who underwent open reduction and pelvic osteotomy with or without femoral osteotomy with those of walking-age children with unilateral dislocated hips who underwent the same set of procedures, the radiographic outcomes were similar.[27]

In this study, the rate of osteonecrosis was higher in the bilateral group, but this difference was explained by older age at surgery and a greater degree of hip dislocation before surgery.[27] The authors concluded that the clinical outcomes after surgery of the children with bilateral hip dislocations were worse mainly because of asymmetric outcomes.

Early clinical manifestations of developmental dysplasia of the hip (DDH) are identified during examination of the newborn. The classic examination finding is revealed with the Ortolani maneuver, in which a palpable "clunk" is present when the hip is directed in and out of the acetabulum and over the neolimbus. A high-pitched "click" (as opposed to a clunk) in all likelihood has little association with acetabular pathology.[28, 29] Ortolani originally described this clunk as occurring with either subluxation or reduction of the hip (in or out of the acetabulum). More commonly, the Ortolani sign is referred to as a clunk felt when the hip reduces into the acetabulum, with the hip in abduction.

To perform this maneuver correctly, the patient must be relaxed. Only one hip is examined at a time. The examiner's thumb is placed over the patient's inner thigh, and the index finger is gently placed over the greater trochanter. The hip is abducted, and gentle pressure is placed over the greater trochanter. In the presence of DDH, a "clunk," similar to that noted in turning a light switch on or off, is felt when the hip is reduced. The Ortolani maneuver should be performed gently, in such a way that the fingertips do not blanch.[30]

Barlow described another test for DDH that is performed with the hips in an adducted position, in which slight gentle posterior pressure is applied to the hips. A "clunk" should be felt as the hip subluxates out of the acetabulum.[23]

The clinical examination for late DDH (age 3-6 months) is quite different. At this point, the hip, if dislocated, is often dislocated in a fixed position.[14] The Galeazzi sign is a classic identifier of unilateral hip dislocation (see the image below). This is performed with the patient lying supine and the hips and knees flexed. The examination should demonstrate that one leg appears shorter than the other. Although this finding is usually due to hip dislocation, it is important to realize that any limb-length discrepancy results in a positive Galeazzi sign.

Additional physical examination findings for late dislocation include asymmetry of the gluteal thigh or labral skin folds, decreased abduction on the affected side, standing or walking with external rotation, and leg-length inequality.[31]

Bilateral dislocation of the hip, especially at a later age, can be quite difficult to diagnose. This condition often manifests as a waddling gait with hyperlordosis. Many of the aforementioned clues suggesting a unilateral dislocated hip are absent, such as the Galeazzi sign, asymmetric thigh and skin folds, or asymmetrically decreased abduction. Careful examination is needed, and a high level of suspicion is important.

Any limp in a child should be considered abnormal. The diagnosis can be quite variable, but an underlying etiology must always be pursued.

Of primary importance is making the diagnosis of hip dislocation or dysplasia. Once this diagnosis is made, the patient should be examined to make sure that there is no underlying medical or neuromuscular disorder. Proximal femoral focal deficiency can masquerade as hip dysplasia and often manifests similarly. Because the femoral head does not ossify, the radiographic appearance also may be deceiving. Other neuromuscular disorders can manifest as dysplasia later in life, such as Charcot-Marie-Tooth disease.

Using expected-value decision analysis, Mahan et al found that the screening strategy associated with the highest probability of having a nonarthritic hip at the age of 60 years was to screen all neonates for hip dysplasia with a physical examination and to use ultrasonography (US) selectively for high-risk infants.[32] The expected value of a favorable hip outcome was 0.9590 for screening all neonates with physical examination and selective use of US, 0.9586 for screening all neonates with physical examination and US, and 0.9578 for no screening.

Numerous possible complications can occur, including redislocation, stiffness of the hip, infection, blood loss, and, possibly the most devastating, avascular necrosis (AVN) of the femoral head. The rate of femoral head necrosis varies substantially; depending on the study, it may be anywhere from 0% to 73%. Numerous studies demonstrate that extreme abduction, especially when combined with extension and internal rotation, results in a higher rate of AVN.[33, 34, 35]

Ultrasonography (US) has been of substantial benefit in the assessment and treatment of children with developmental dysplasia of the hip (DDH).[3, 36, 37, 38] The benefit of screening all children with US is controversial.[39, 40] Even with US screening, children with hip dysplasia can be diagnosed late, and one concern with routine US evaluation of newborns is overdiagnosis of hip dysplasia (ie, increased false-positive results).[41]

The employment of US only in high-risk infants has not yet been shown to reduce the prevalence of late diagnosis of hip dysplasia.[42] However, most authors agree that it is an excellent tool for assessing children with suspected hip instability and a useful aid in the treatment of children with DDH, especially in monitoring reduction by closed methods.[43]

An ultrasound evaluation is typically performed either by assessing the alpha and beta angles or by performing a dynamic evaluation.[36, 38, 44] An alpha angle outlines the slope of the superior aspect of the bony acetabulum, with an angle greater than 60º considered normal. The beta angle, which is considered normal if less than 55º, depicts the cartilaginous component of the acetabulum. Many institutions now use a dynamic form of US, as heralded by Harcke.[43]

Standard radiographic views for DDH include a standing anteroposterior (AP) view of the pelvis, with the hips in neutral position, and a false profile view in which the patient is standing angled at 65º from the x-ray plate. The radiograph is then taken, profiling the anterior aspect of the acetabulum. If any evidence of hip subluxation is present, an abducted internal rotation view can help determine if the hip reduces and better determines the true neck-shaft angle of the proximal femur.

Radiographic evaluation is typically carried out as follows (see the image below). From an AP radiograph of the hips, a horizontal line (Hilgenreiner line) is drawn between each triradiate cartilage. Next, lines perpendicular to the Hilgenreiner line are drawn through the superolateral edge of the acetabulum (Perkin lines), dividing the hip into four quadrants. The proximal medial femur should be in the lower medial quadrant, or the ossific nucleus of the femoral head, if present (usually observed in patients aged 4-7 months), should be in the lower medial quadrant.

Additionally, the acetabular indices can be measured. These refer to the angle between the Hilgenreiner line and a line drawn from the triradiate cartilage to the lateral edge of the acetabulum. Typically, the angle decreases with age and should measure less than 20º by the time the child is aged 2 years.[45, 46]

The Shenton line—a line drawn from the medial aspect of the femoral neck to the inferior boarder of the pubic rami—can also be evaluated. This line should create a smooth arc that is not disrupted. Disruption of the Shenton line indicates the presence of some degree of hip subluxation.

Computed tomography (CT) can also be helpful in determining femoral anteversion and in determining the extent of posterior acetabular coverage. Three-dimensional (3D) images are also quite popular and can be beneficial in visualizing the overall shape of the acetabulum.

Magnetic resonance imaging (MRI) can be beneficial in identifying the underlying bony and soft-tissue anatomy. One study evaluated MRI findings in pediatric orthopedic patients who showed residual subluxation after reduction of DDH.[47] Twenty-two subjects were followed conservatively, and 14 subjects underwent corrective surgery.

The subjects in the surgery arm of the study showed the presence of a high-signal intensity area (HSIA) within the weight-bearing portion of the acetabular cartilage preoperatively, which decreased or disappeared after the surgical procedure.[47] In the conservative arm, those with HSIAs demonstrated poor acetabular growth and those without HISAs showed acetabular growth. The researchers concluded that HSIAs on MRI may be a marker for poor acetabular growth, which would make these areas valuable findings in corrective surgery decision-making.

A retrospective review compared CT with MRI in the evaluation of hip reduction in patients younger than 13 months with hip dysplasia.[48] The results indicated that whereas MRI was a viable alternative to CT, CT required significantly less scan time than MRI did and cost less. However, CT was slightly less specific than MRI was.[48]

Arthrography is a dynamic study, performed by injecting radiopaque dye into the hip joint and then carrying out a fluoroscopic examination, usually with the patient under anesthesia. Although it can be performed independently, it is routinely performed in conjunction with a closed reduction. Arthrography can be helpful in determining the underlying cartilaginous profile and dynamic stability of the hip.[37] It has also been used in conjunction with a hip MRI study to facilitate demonstration of labral tears.

When arthrography is performed in combination with a closed reduction, the adequacy of the reduction can be assessed. Increased medial joint space, as demonstrated by medial pooling of the dye and a rounded or interposing limbus, may be indicative of poor long-term results. After closed reduction and immobilization in a hip spica cast, a limited CT scan in the transverse plane is obtained to ensure the hip is not subluxated or dislocated posteriorly.

Indications for surgical treatment of developmental dysplasia of the hip (DDH) are met if the results of such treatment would be better than the results of the natural progression of the disease.[49] The natural history of DDH depends, in part, on the severity of the disease, bilaterality, and whether or not a false acetabulum is formed.[10, 50, 51]

Unilateral dislocations result in significant leg-length inequality, with a gait disturbance and possibly associated hip and knee pain. In addition, hip pain commonly manifests as knee or anterior thigh pain as a consequence of the innervation of the hip joint (obturator and femoral nerve distribution). Typically, true hip pain is identified as groin pain.

The development of a false acetabulum is associated with a poor outcome in approximately 75% of patients. Bilateral hip dislocation in a patient without false acetabula has a better overall prognosis. In fact, a case was reported of a 74-year-old man with no history of hip or thigh pain whose dislocated hips were only discovered shortly before his death.[52]

Indications for treatment depend on the patient's age and the success of previous techniques. Children younger than 6 months with instability upon examination are treated with a form of bracing, usually a Pavlik harness. If this is not effective or if the hip instability or dislocation is noted when the child is older than 6 months, closed reduction is typically recommended, with or without the administration of traction before the reduction.

When the child is older than 2 years or with failure of the previous treatment, open reduction is considered. If the patient is older than 3 years, femoral shortening is performed instead of traction, with additional varus applied to the femur, if necessary. A patient with residual acetabular dysplasia who is older than 4 years should be treated with an acetabular procedure.

Treatment of DDH diagnosed when the patient is a young adult can be considered for residual DDH. Unfortunately, radiographic characterization of DDH that is severe enough to lead to early osteoarthrosis is difficult. A center-edge angle smaller than 16º often has been used to predict early osteoarthrosis,[53] but other authors have found this measurement to be less reliable.[54, 55] Subluxation, defined as a break in the Shenton line (see Workup), has been shown to be associated with osteoarthrosis and decreased function.[54]

Relative contraindications for surgery include older age (>8 years for a unilateral hip dislocation or >4-6 years for bilateral hip dislocation, especially if a false acetabulum is not present). Other contraindications for surgery include a neuromuscular disorder, such as a high myelomeningocele or spinal cord injury, or cerebral palsy in a patient who has had a hip dislocation for longer than 1 year.

Treatment of DDH begins with careful examination of the newborn. If evidence of instability is present, a Pavlik harness should be considered and, if used, fitted appropriately.[56, 57, 58, 59, 60]

The Pavlik harness should be placed so that the chest strap is at the nipple line, with 2 fingerbreadths of space between chest and strap. The anterior strap is at the midaxillary line and should be set so that the hips are flexed to 100-110º; excessive flexion can lead to femoral nerve compression and inferior dislocations. Quadriceps function should be determined at all clinic visits.

The posterior abduction strap should be at the level of the child's scapula and adjusted to allow comfortable abduction. This should prevent the hips from adducting to the point where the hips dislocate. Excessive abduction should be avoided out of concern regarding the potential development of avascular necrosis (AVN). The fitting of the harness should then be checked clinically within the first week and then at regular intervals thereafter. The patient must be carefully monitored to ensure that the harness fits and the hips are reduced.

Ultrasonography (US) is an excellent means of documenting the reduction of the hip in the Pavlik harness and should be performed early in the course of treatment.[61] If the hip is posteriorly subluxated, then the Pavlik harness therapy should be discontinued.[62]

Using the Pavlik harness for guided reduction, which occurs when the hip does not completely reduce initially but is pointed toward the triradiate cartilage, is controversial. When a Pavlik harness is used for guided reduction, the physician should obtain a radiograph after the harness is placed to determine if the femoral heads are pointing toward the triradiate cartilage. An ultrasonogram should be obtained to determine the success, or lack thereof, of the guided reduction.

Several authors have expressed concern that use of the Pavlik harness in patients with bilateral involvement has an increased likelihood of failure.[58, 63] One group compared the success rates of Pavlik harness treatment for unilateral and bilateral frankly dislocated hips in otherwise normal children and found no significant difference.[64]

There is no consensus on the optimal overall duration of Pavlik harness therapy.[65, 66] If the hip is reduced satisfactorily in the harness, the author's practice is to maintain this treatment at least until the hip is considered to be stable both on clinical grounds and on the basis of US findings with the patient out of the brace. Abduction splinting is maintained thereafter if radiographic evidence of residual dysplasia is present.

The use of an abduction brace after a failure of the Pavlik harness has been suggested. In one study, 13 of 15 patients were treated successfully in this manner, and the remaining two patients had a successful closed reduction.[67]

In patients older than 6 months, the success rate with a Pavlik harness is less than 50%; therefore, this therapy should not be used in these patients.[6] If the child is diagnosed when older than 6 months or if the Pavlik harness is determined to be unsuccessful, a closed reduction is attempted.

Historically, traction was performed for a 2- to 3-week period before closed reduction was attempted.[68] Traction (usually skin traction) was performed either at home or in the hospital. This required careful monitoring to ensure the integrity of the skin. Although there remains considerable controversy regarding the overall benefit of traction, reports of long-term follow-up have shown satisfactory outcomes in a majority of patients.[69, 70]

Closed reduction is typically performed with the aid of arthrography, which is used to determine the adequacy of the reduction. A medial dye pool and an interposing limbus are both associated with a poor prognosis. If, on the other hand, a sharp or even a blunted limbus and no medial dye pooling are present, the prognosis is good.[71] In addition, the safe zone of Ramsey (ie, the angle between the maximum abduction and minimum abduction in which the hip remains reduced) should be at least 25º and can be increased with release of the adductor longus.

The cone of stability—a cone that involves hip flexion, abduction, and internal or external rotation—has also been defined. If this cone measures greater than 30º, it is considered satisfactory.[71]

A spica cast is placed, with care taken in molding over the posterior aspect of the greater trochanter of the ipsilateral limb. After this is done, computed tomography (CT) or magnetic resonance imaging (MRI) is performed to ensure that no evidence of posterior subluxation is present. The cast is typically worn for 6-12 weeks, at which time the hip is reexamined. If the hip is found to be stable, the patient is placed in an abduction brace. If the hip remains unstable, the patient is again placed in a spica cast.

Open reduction is the treatment of choice for DDH in children who are older than 2 years at the time of the initial diagnosis or in whom attempts at closed reduction have failed. In children with teratologic hips, with failure at a much younger age, open reduction can be performed via a medial approach. The medial approach has a number of advantages, as follows:

• Both hips can be reduced at the same time (in a patient with bilateral DDH)
• Obstacles to reduction (eg, psoas tendon) are easily identified
• The adductor longus can be sectioned through the same incision
• Because the hip abductor muscles are not at risk for injury, residual weakness is unlikely to occur
• The iliac apophysis is not at risk for injury
• The incision has a very good cosmetic result

Problems with this approach include the following:

• Potentially increased rates of AVN
• The possibility that the surgeon may be unfamiliar with this approach
• The inability to perform capsular plication (capsulorrhaphy) or a pelvic procedure through this incision

With the use of a medial approach, the cast plays a much more important role.

Most often, especially in older children, the standard anterolateral or Smith-Petersen approach is used. This can be combined with a capsule plication, if needed, with an acetabular procedure, or with both. In a child older than 3 years, femoral shortening osteotomy has been found to be safer and more effective than traction (see the image below).[72] At that time, if proximal femoral dysplasia is present, such as that observed with significant anteversion or coxa valga, this can also be corrected.

Even in children younger than 2 years, the derotational shortening femoral osteotomy has been safely performed with open reduction of the hip in selected hip dislocation patients.[73] Furthermore, a study of walking children with idiopathic hip dislocations suggested that performing an open reduction of the hip without concurrent femoral osteotomy strongly predicts the need for a secondary procedure.[74]

Pelvic osteotomy may be needed for residual hip dysplasia.[49, 75, 76, 77] The recommended timing for the pelvic osteotomy varies among surgeons. Some authors suggest pelvic osteotomy in children as young as 18-24 months, whereas others suggest waiting until the children are at least 4 years old.

If open reduction is performed in a child older than 4 years with significant hip dysplasia, an acetabular procedure should be considered at the time of open reduction. If a closed reduction is performed earlier, at least 12-18 months of acetabular remodeling should be allowed before an acetabular procedure is undertaken. At that time, if no evidence of acetabular modeling is noted, a pelvic osteotomy should be considered.

When open reduction is performed, the patient wears a hip spica cast for 6-12 weeks, then is placed in an abduction orthosis. The length of time for which a child remains in a hip orthosis is quite controversial and depends on the treating physician's experience and on the individual patient.