Type
Alpha angle (α)
Beta angle (β)
Descriptions
I
> 60°
<55°
Ia
Normal hip (at any age). This grade is further divided into (Ia; β < 55°) and (Ib; β > 55°). The significance of this subdivision is not yet established. Patient does not need follow-up.
> 55°
Ib
II
50–59°
IIa
< 77°
If the child is <3 months. This may be physiological and does not need treatment; however, Follow up is required.
IIb
< 77°
> 3 months, delayed ossification.
43–49°
IIc
Stable
< 77°
Critical zone, labrum not everted. This is further divided into stable and unstable by provocation test.
Unstable
D
43–49°
> 77°
This is the first stage where the hip becomes decentred (subluxed).
III
<43°
IIIa
Dislocated femoral head with the cartilaginous acetabular roof is pushed upwards. This is further divided into IIIa and IIIb depending on the echogenicity of the hyaline cartilage of the acetabular roof (usually compared to the femoral head) which reflects the degenerative changes.
IIIb
IV
<43°
Dislocated femoral head with the cartilaginous acetabular roof is pushed downwards
Fig. 4.1
Hips ultrasound; Graf hip types
The principles of treating DDH can be summarised in 4 steps:
- 1.
Achieve a concentric reduction without excessive force.
- 2.
Maintain the concentric reduction for an optimum time.
- 3.
Promote the normal growth and development of the hip.
- 4.
Minimise complications.
Although surgeons around the world agree on the above principle, they adapted different approaches to achieve the above principles. In this chapter, we have explored the evidence behind these approaches concentrating on the common questions that clinicians face in the management of children with DDH.
The hip continues to develop throughout childhood and complications of treatment in the young child may not become evident for many years. Large prospective studies with long-term outcomes are scarce and the majority has used surrogates such as avascular necrosis (AVN), failure of treatment, re-dislocations or further surgery etc. AVN is a major cause of long-term disability and is directly related to the treatment – it does not occur in untreated DDH. There are several classifications of AVN including Kalamachi and MacEwen [8], Bucholz and Ogden [9] and Salter’s classification [10].
The Severin classification (Table 4.2) describes the radiographic appearance of the hip 5 years after treatment for dislocations. It has been used by many authors as a surrogate for clinical outcome [11]. Several studies have questioned the reliability [12, 13] and showed unacceptably low levels of inter-observer and intra-observer reliability and agreement.
Table 4.2
Severin classification
Group | Description |
|---|---|
I | Well developed hip joint |
II | Moderate deformity of the femoral head, neck or acetabulum in otherwise well developed joint |
III | Dysplasia, but not subluxation |
IV | Subluxation |
V | Femoral head is articulating with secondary acetabulum |
VI | Re-dislocation |
DDH Management in the Infant (0–6 months of Age)
Is Universal Ultrasound Screening Program to Detect Hip Dysplasia Necessary?
This question was addressed in Chap.3.
When Should Treatment Be Commenced for the Dislocatable Hip?
While it is accepted that treatment should not be delayed with the dislocated but reducible hip but should the same apply to the neonate with a dislocatable hip or a stable but dysplastic hip? When should ‘watchful waiting’ be applied? Treatment with abduction splintage is not entirely benign with a 2–3 % rate of avascular necrosis (AVN) reported in children treated at less than 2 months of age, compared with 1 % after 6 months [14, 15]. Furthermore, AVN has also been shown to occur in the contralateral hip [16, 17].
It has been established that many unstable hips at birth will stabilise without the need for abduction splintage [18]. The fixed flexion contracture of the neonatal hip spontaneously resolves and rapid resolution of acetabular dysplasia also occurs in the majority of cases [19]. Therefore, at what age should treatment be initiated to avoid over-treatment in the clinically unstable hip?
Gardiner and Dunn [7] randomised 79 neonates with unstable (dislocatable) hips to receive either immediate splinting or ultrasound surveillance. At 2 weeks follow-up, 60 % of hips in the surveillance group had no further sonographic signs of hip instability. The remainder with persistent instability or acetabular dysplasia underwent abduction splinting at this stage. At 6 and 12 months follow-up, there was no difference in clinical or radiological appearance in the two initial groups. They recommended that a 2-week delay, prior to initiating treatment for the dislocatable hip is advisable to minimize over treatment. However, due to the small sample size, it would have required a major adverse outcome to be statistically significant.
Elbourne et al. [20] undertook a multi-centre randomised trial to determine whether ultrasonographic surveillance could reduce the number of children undergoing abduction splintage for hip instability, without resulting in a significant increase in the incidence of late treatment. Following a clinical diagnosis of hip instability, infants were randomised to receive treatment based on clinical examination alone (n = 315) or treatment decision based on a hip ultrasound at 2 weeks of age or older (n = 314). In the ultrasound group, those hips that were significantly displaced or unstable were splinted, while minor displacement or instability was monitored. If abnormality persisted until 8 weeks of age, splintage was initiated. In the clinical examination group, abduction splintage was prescribed based on clinical suspicion. Splintage was undertaken earlier in the clinical examination group (81 % by 2 weeks of age vs. 63 % in the ultrasound group) with more hips requiring treatment with clinical examination alone (50 vs. 40 % respectively). Radiographic parameters at 2 years of age and the need for surgical intervention were similar in both groups. The study concluded that ultrasound examination reduced the rate of abduction splintage by a third without increasing the need for late treatment.
While the study does not directly address the age at which treatment should be initiated in the unstable hip, it suggests that the decision on abduction splintage should be based on the ultrasound appearance, not just clinical examination. Furthermore, immediate abduction splintage (<2 weeks of age) was not deemed to be necessary.
These studies indicate that the decision on whether to treat the unstable hip should include ultrasound examination. Treatment prior to 2 weeks of age does not appear to be necessary as many hips will spontaneously stabilise. Grade of recommendation: B.
Should an Ultrasound-Confirmed Dysplastic but Stable Hip Be Treated?
Ultrasound examination allows us to quantify the degree of acetabular dysplasia, but should all neonates with stable dysplastic hips be treated? Neonates that have evidence of acetabular dysplasia at birth can demonstrate rapid remodeling over the first few weeks of life, without the need for abduction splinting [15]. There are four relevant studies:
Wood et al. [5] reported on a study with 44 infants aged 2–6 weeks with dysplastic stable hips (<40 % coverage on ultrasound examination). Random allocation was used to allocate patients to receive splintage or surveillance. At 3 months the acetabular coverage measured by ultrasound improved in both groups but the greatest improvement was found in those infants placed in an abduction splint (36.7–54.3 % in the splinted group and 32.8–48.6 % without splint, p < 0.03). However, at 3 months the acetabular index on plain x-ray was similar (24.79 vs. 24.28). There was a 69 % follow-up at 24 months – mean acetabular index 21.6° (splinted) and 23.5 (no splint). They concluded that in the 2–6 week old child with dysplastic but stable hips, abduction splinting confers no benefit. However, the sample size in the study was small.
Sucato et al. [21] investigated the predictive value of an abnormal ultrasound in an infant under 1 month of age. They performed a retrospective review of 112 infants (192 hips) less than one month of age with a normal hip examination but abnormal hip ultrasound (Graf IIa-III). Pavlik harness treatment was selected at the discretion of the treating physician allowing review of two groups: those treated (43 hips) and untreated (149 hips). The two groups had similar demographics and Graf classification, but the treated group demonstrated less femoral head coverage on stress maneuvers. At final follow-up (15.9 months), in the treated group, no hip had evidence of dysplasia while 2 hips (1.3 %) were considered dysplastic in the untreated group. These 2 hips were initially classified as Graf IIc and it was reported that other more dysplastic hips (Graf IId/III) in the untreated group had normal acetabular indices at final review. They concluded that Pavlik harness management did not appear to influence the incidence of acetabular dysplasia in the stable hip.
Rosendahl et al. [22] conducted a blinded RCT involving 128 neonates (less than 2 weeks of age) with mild hip dysplasia (alpha angle 43–49°). Patients were allocated to receive 6 weeks of abduction splinting or ultrasound surveillance alone. There was no loss to follow-up. 29 children (47 %) in the ‘active-surveillance’ group received abduction splinting at 6 weeks of age due to persistent dysplasia. At 12 months of age, the mean acetabular inclination was 24.2° in both groups. The authors concluded that in the dysplastic stable hip, an active surveillance policy reduces the need for treatment by 50 % compared with immediate abduction splintage.
In a prospective study of 8638 hips (4319 neonates assessed in the immediate postnatal period) from Israel [23], 8030 hips (93 %) were normal by clinical examination and ultrasound findings. These babies were discharged when they did not have risk factors. Babies with clinically stable hips but with either risk factors or with a Graf type IIa or IIc sonographic appearance were re-examined clinically and sonographically at 6 weeks of age. Those neonates with unstable hips or a stable hip with Graf type D appearance (or worse) were re-examined at 2 weeks of age. If the sonographic appearance showed no improvement of the unstable hips at 2 weeks, treatment with the Pavlik harness was commenced.
There were 479 hips (5.53 %) with abnormal sonographic findings placing them in Graf class IIa or worse; of these 81 hips (0.9 %) were unstable on clinical examination.
At the end of the established waiting periods, 90 % of the abnormal hips had become normal without treatment. Less than 3 % of the Graf IIa hips failed to normalise without treatment, whereas 17 % of Graf III hips and 80 % of Graf IV hips failed to normalise (Table 4.3).
Table 4.3
Summary of Bialik’s study findings
Type | Stable | Unstable | Total | Needed treatment (%) |
|---|---|---|---|---|
Type IIa | 255 | 6 | 261 | 7 (2.6 %) |
Type IIc | 93 | 19 | 112 | 12 (10.7 %) |
Type D | 24 | 28 | 52 | 12 (23 %) |
Type III | 4 | 13 | 17 | 3 (17.6 %) |
Type IV | 1 | 14 | 15 | 12 (80 %) |
These studies suggest that immediate abduction splintage in the dysplastic hip could result in some hips being unnecessarily treated. In the infant with a stable but dysplastic hip, the decision to treat may be delayed to 6 weeks of age without risk of unsatisfactory outcome. Grade of recommendation: B.
How Should a Dysplastic or Dislocated Hip Be Treated in Children Who Are Less Than 6 Months of Age?
The Pavlik harness is the most widely used method of treatment in this age group. The reported success rate ranged from 59 % to 97 % (Table 4.4). Likely reasons for such variation are the threshold for treatment, the length of treatment and definition of failure by reporting authors. As noted previously, 93 % of Graf IIa hips resolve spontaneously without any treatment. If these hips are included in a treatment protocol the reported success will be high.
Table 4.4
Summary of Pavlik Harness treatment
Study | Hips | Success rate (%) | AVN rate (%) | LOE | |
|---|---|---|---|---|---|
Pavlik [24] | 1912 | 84 | 0 | IV | |
Wada [25] | 2481 | 80 | 14.3 | IV | |
Walton [26] | 123 | 90 | 2.4 | IV | |
Cashman [27] | 546 | 97 | 1 | IV | |
Grill [28] | 3611 | 92 | 2.4 | IV | |
Johnson [29] | 91 | 90 | 0 | IV | |
Filipe [30] | 74 | NR | 5.4 | IV | |
Santos [31] | 159 | 93.7 | 16 | IV | |
Murnaghan [32] | 1218 | 94 | 9 | IV | |
Nakamura [33] | 130 | 81.6 | 12.3 | IV | |
van der Sluijs [34] | 62 | 60 | 16 | IV | |
Walton [26] | 123 | 95.2 | 2.3 | IV | |
Westacott [35] a | Groups A | 80 | 88 | 9 | III |
Group B | 48 | 71 | 4 | ||
Wilkinson [36] | 43 | 76.7 | 0 | III | |
A variety of other splints and braces have also been used to treat DDH. Comparable success rates have been reported in treating DDH with the Von Rosen splint (Table 4.5). Comparative studies between the above braces are contradictory and most are suboptimal.
These studies suggest that the dislocated or dysplastic hip can be treated with a Pavlik harness or a Von Rosen splint with a high success rate and low incidence of AVN. Grade of recommendation: B.
Does Weaning the Pavlik Harness Treatment Provide an Advantage Over Immediate Discontinuation?
Pavlik harness treatment is usually discontinued when ultrasound examination confirms a normal hip morphology and the clinical examination is negative. However, some authors prefer to wean the infant from the harness with a period of part-time wear that occurs over several weeks or months.
Westacott et al. [35] compared the outcome of Pavlik harness treatment between two centres. Eighty children in Centre A underwent staged weaning of the Pavlik harness once three consecutive weekly ultrasounds demonstrated a Graf Grade I hip. Forty-eight children were treated in Centre B where Pavlik harness treatment discontinued immediately with no weaning period. No statistically significant difference was found in the success rate (88 % vs. 71 %) although there was a non-significant trend towards higher intervention when the harness was immediately stopped. While there was also no significant difference in AVN rate, there was a trend towards a lower AVN rate with the immediate cessation (9 % vs. 4 %).
This single study suggests that there is no obvious advantage of weaning the Pavlik harness treatment over the immediate cessation and there may be an associated higher AVN rate. Grade of recommendation: C.
How Should Femoral Nerve Palsy During Pavlik Harness Treatment Be Managed?
Femoral nerve palsy is an uncommon complication of Pavlik harness treatment. In a study [32] of 1218 patients treated by Pavlik harness, 30 cases of femoral nerve palsy were identified (incidence of 2.5 %). 87 % presented within one week of application of the harness. Femoral nerve palsy was more likely in older, larger patients in whom the developmental dysplasia of the hip was of higher severity. Nineteen patients were treated with temporary suspension of harness treatment and subsequent reapplication when femoral nerve function returned, 6 were treated with adjustment of harness to reduce hip flexion, 5 were managed with complete abandonment of the harness, with 4 requiring subsequent closed or open reduction of the hip. There was no correlation between the method of management of the femoral nerve palsy and the success of treatment. Of those 19 patients who did have reinstitution of harness therapy, only 3 developed a recurrence of the palsy. Pavlik harness treatment was abandoned in all patients who demonstrated recurrence of femoral nerve palsy. All patients had eventual complete return of full quadriceps function, with no clinically evident long-term motor or sensory deficit. Patients whose femoral nerve palsy resolved within 3 days had a 70 % chance of having successful treatment with harness, whereas those who had not recovered by 10 days had a 70 % chance of having treatment failure. Notably, the success rate associated with treatment with a Pavlik harness was 94 % in the control group and 47 % in the palsy group.
The published evidence to guide on the best treatment for femoral nerve palsy associated with Pavlik harness is limited. The above study suggested that it is reasonable to temporarily stop the Pavlik harness treatment until the nerve recovers or alternatively reduce the amount of hip flexion. The presence of femoral nerve palsy is associated with a higher failure rate. Grade of recommendation: C.
What Is the Next Step in the Hip That Fails to Reduce and Stabilise in a Pavlik Harness?
Several factors have been associated with failure of Pavlik harness management: breech presentation, bilateral dislocation, age at application, incorrect application, poor parental compliance and an initially irreducible hip [42–46].
The clinically irreducible hip can be successfully treated in a Pavlik harness, but close monitoring is essential. The harness can gradually reduce the abduction contracture and with ongoing active motion and flexion, the femoral head can relocate. It is the most commonly used device in the early management of DDH [27, 47, 48]. However, prolonged positioning of the dislocated hip in adduction and flexion can potentiate femoral head and posterolateral acetabular dysplasia causing greater difficulty in successfully obtaining stability with future closed or open reduction [49].
If the hip is reducible but fails to stabilise in a Pavlik harness, many specialists will elect to proceed with closed reduction ± adductor tenotomy and spica cast. However, there are increasing concerns on the effect of general anaesthesia on the developing brain [50, 51] and a viable alternative to a procedure under anaesthesia is appealing. Furthermore, closed reduction has been associated with a highly variable rate of AVN (see next section). Semi-rigid abduction bracing is an alternative and several authors have reported their experience in using the orthosis in those hips that fail to stabilise in a Pavlik harness. In particular the Ilfield orthosis is gaining in popularity. The Ilfield orthosis holds the hip in less flexion than in a Pavlik harness and in 50° abduction. It reduces the amount of hip motion compared to the Pavlik harness. Sankar et al. [47] postulate that the semi-rigid device may be particularly effective in those hips that are reducible and lie inferiorly or are excessively lax. Parental compliance due to ease of re-application may also be a factor.
Sankar et al. [47] reviewed two retrospective cohorts of patients who had failed Pavlik harness management at the same institution. Nineteen infants who had undergone Ilfield bracing following persistent hip instability following Pavlik harness management were compared to a consecutive retrospective cohort of 16 infants who had a closed reduction and spica cast application. The groups were comparable prior to the secondary intervention. The hips stabilized in 82 %vs 91 % of cases respectively and radiographic appearance was similar at one year. Notably, hips that were dislocated and irreducible were excluded from the study. Three hips in the closed reduction cohort had evidence of AVN at 12 months follow-up.
Hedequist et al. [52] reviewed their experience of using an abduction orthosis after failed Pavlik management in 14 infants. At the time of brace application, 12 of the hips were dislocated but reducible and 2 hips were unstable. There were no irreducible hips. 12/14 hips successfully stabilized in the abduction orthosis and 2 hips failed, requiring closed reduction. One of these developed radiographic evidence of grade 1 AVN at 3-year follow-up. The mean time for the hip stabilisation was 24 days (14–63) with duration in the splint of 46 days (18–91).
Swaroop et al. [53] performed a retrospective review of their experience in managing the dislocated but reducible hip. In the cohort 41/44 hips were successfully stabilised in a Pavlik harness and of the remaining 3 hips who failed Pavlik harness, 2 (67 %) were successfully treated in an abduction orthosis.
Ibrahim et al. [42] performed a retrospective review of 7 patients at a single institution who had failed Pavlik harness management and were then treated in an Ilfield abduction brace. In contrast to the experience of the previous authors, all hips failed to stabilize with the abduction splint and thus closed or open reduction was then required. Of note, three of the patients had a dislocated irreducible hip at initial presentation (unchanged post-Pavlik management) and one patient could not tolerate the brace and it was discontinued after three days. The remaining hips were unstable on commencement of the brace.
The experience of these authors suggest that while an abduction orthosis will not successfully stabilize the irreducible dislocated hip, it can be considered for the dislocated but reducible or unstable hip that fails Pavlik harness management. The evidence thus far are level 3 and 4 studies. Prospective trials are needed to further clarify the indications. Grade of recommendation: B/C.
Failure of Early Treatment or Late Presentation Between 6–18 Months of Age
When Should a Closed Reduction Be Considered?
Closed reduction of the dislocated hip with adductor +/− psoas tenotomy and spica cast is an accepted technique in the management of DDH. It can be performed as the initial procedure in the child over 6 months of age, or following failure post Pavlik harness/abduction brace management. However, it is not a benign procedure with wide variation in AVN rates reported from 4 % to 60 % [54–60]. Case selection, surgical technique, pre-operative traction and the presence of the ossific nucleus are thought to be the contributive factors. Excessive hip abduction in the post-operative spica is likely to be a cause of AVN, due to vascular occlusion and diminished blood supply to the femoral epiphysis – care should be taken to keep the degree of hip abduction within the ‘safe-zone’, as described by Ramsey [61].
Senaran et al. [62] hypothesised that reducing dislocated hips which fail Pavlik harness treatment within 3 months of age will result in a lower incidence of AVN. To support their hypothesis, they reviewed 21 consecutive cases (35 hips) that failed Pavlik harness treatment and underwent closed reduction before the age of three months. Successful closed reduction was achieved in 33 (94 %) of 35 hips, and open reduction required in 2 (6 %) of 35 hips. At latest follow-up, one (3 %) of 35 hips had AVN – it should be noted that follow-up duration was just 36 months. At the time of reporting, 1 (3 %) of the 35 hips has required an additional procedure (Pemberton osteotomy) for residual dysplasia. They concluded that the study supports their hypothesis that an early closed reduction following Pavlik harness management minimizes the rate of AVN.
Novais et al. [63] recently published a systematic review and meta-analysis (level III) which provided some answers to our question. They included 66 studies in the systematic review and 24 in the meta-analysis. Data on 481 hips treated by closed reduction and 584 hips treated by open reduction were available to evaluate the association between AVN and age. The association between AVN and operative approach was assessed using data on 364 hips treated by medial open reduction and 220 hips treated by anterior open reduction. Novais reported that the overall, adjusted incidence of AVN (≥ Grade II) was 8.0 % (95 % CI, 2.8 %–20.6 %) among patients who underwent closed reduction at or before 12 months of age and 8.4 % (95 % CI, 3.0 %–21.5 %) among those who had closed reduction after 12 months. The difference between the two age groups was not significant (OR, 1.1; 95 % CI, 0.4–3.2; p = 0.9).
It is of note that this meta-analysis was based on the results of observational studies and potential confounding variables such as failure of previous treatment, associated procedures including adductor tenotomy, length of immobilisation, and degree of abduction in a spica cast were not accounted for.
The authors conducted a systematic review of the literature to investigate the incidence and predictors of AVN and the radiographic outcome in children who had a closed reduction under 2 years of age (unpublished) [64]. The study included 7 papers that had a 5-year minimum follow-up. The review included 539 hips across the studies. At a mean follow-up of 7.6 years there was a 10 % rate of AVN.
Although these studies indicate that a closed reduction is not a benign procedure with 8–10 % rate of AVN, it remains an effective treatment for the hip that fails to stabilise in a Pavlik harness. While careful positioning of the hip in a spica is important to minimize AVN, age at reduction is not conclusively associated with AVN.
Is Preliminary Period of Traction Necessary Before Closed Reduction of a Dislocated Hip?
Proponents of preliminary traction claim that traction reduces the risk of AVN and the need for open reduction. To reduce the cost and inconvenience of hospital admission, portable home traction devices may be used. Opponents cite the increased cost for this additional step in the management and argue that the evidence for reducing the risk of AVN has not been substantiated in recent studies. Several comparative studies (level III) have been identified and their findings have been summarised in Table 4.6. Crude pooling of the AVN rate is slightly lower with traction. Some retrospective case series tried to address the value of the preliminary traction in treating DDH but their findings were inconclusive [56, 70].
The published evidence for the value of traction prior to closed reduction is inconclusive. The potential benefit should be weighed against the cost and inconvenience of the traction. Grade of recommendation: C.
Should Treatment Be Delayed Until the Ossific Nucleus Is Visible?
In his above mentioned study [65], Segal also investigated the effect of several factors that might influence the rate of AVN including the presence of the ossific nucleus (ON). The study included 49 children (57 dislocated hips) who were <12 months old. Eighteen hips developed AVN. There was no significant difference in the occurrence of AVN with respect to variables such as preliminary traction, closed versus open reduction, Pavlik harness use, and age at the time of operative intervention. However, the presence of the ossific nucleus before reduction, detected either by radiographs (p < 0.001) or sonography (p = 0.033) was statistically significant in predicting AVN; one (4 %) of 25 hips with an ossific nucleus developed AVN, whereas 17 (53 %) of 32 hips without an ossific nucleus before reduction developed AVN.
Related posts:
Evidence-Based Treatment of Flexible Flat Foot in Children
The Evidence Base for Botulinum Toxin Injection for the Treatment of Cerebral Palsy–Related Spasticity in the Lower Limb: The Long-Term Effects
What Is the Best Treatment for Perthes’ Disease?
Evidence-Based Management of Ankle Fractures in Children
Private: Physeal Injury, Epiphysiodesis and Guided Growth
Femoro-Acetabular Impingement in Children
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
