Early ultrasound-monitored treatment of developmental dysplasia of the hip (DDH) became widely accepted after the introduction of general ultrasound screening 1,2. Ultrasound screening has been shown to decrease the incidence of late presentation of hip dislocation and the necessity of surgical procedures 2,3. There is consensus that although normal values for the hip joint are achieved at the end of ultrasound-monitored treatment, radiological assessment is necessary until skeletal maturity. Because of endogenous factors, a physiological hip joint at the end of harness treatment can result in residual dysplasia requiring treatment. This is underlined by several reports showing late identification of residual dysplasia or even dislocation of the hip following normal ultrasound screening 4–7. To rule out residual dysplasia, the first anteroposterior pelvic radiograph is recommended at the age of 15–18 months. The acetabular index (AI) has been proven to be a reliable parameter for acetabular development.
A previous study reported a markedly high rate of residual DDHs at the time of the first radiographic follow-up 8. The aim of this examination was to evaluate to what extent progressive development of the hips can be expected after a follow-up at the age of 3 years.
Patients and methods
We retrospectively reviewed the findings of the second radiographic follow-up of 72 consecutive infants with residual DDH (54 females, 18 males, 144 hips). When at least one hip showed mild residual dysplasia at the first radiographic follow-up, the above-mentioned second radiographic follow-up was recommended. From the first to the second radiographic control, the patients did not receive specific treatment. All of the children were able to start walking timely. The mean age at the time of the second follow-up was 31.3 months (range 26–41 months). Anteroposterior pelvic radiograph was performed and the AI was measured. To assess the severity of residual dysplasia, the values were allocated according to the age-related classification of Tönnis (Tönnis 0=below 1 SD (normal), Tönnis 1=between 1 and 2 SD (mild residual dysplasia) and Tönnis 2=beyond 2 SD (severe residual dysplasia) (Table 1) 9.
The findings of the second radiological follow-up were evaluated in terms of a persisting residual dysplasia. The data of the first radiographic follow-up, which took place in an analogous manner at the mean age of 14.8 months (range 12–21 months), were related to the findings of the second follow-up. Infants presenting normal hips bilaterally at the first radiographic follow-up did not receive a second radiographic examination. All children in the study were initially referred to us after ultrasound screening according to the German Health Care System, which is mandated at the age of 4–6 weeks, and underwent ultrasound-monitored treatment in our hospital. This comprised abduction splinting (Fettweis plaster cast, Pavlik-harness or abduction splint) until physiological values were achieved in the regularly performed sonographic follow-up. In retrospect, to the time of the initial sonographic examination, 10 of our patients hips presented Graf type I, 74 presented Graf type II, 51 presented Graf type III and nine presented Graf type IV, respectively. The abduction treatment followed the clinical practice guidelines recommended by Graf. In particular, additional Vojta therapy was advised in the Graf III and IV hips. In our patients, the mean time of abduction therapy added up to 108 days (range 70–158 days).
The following exclusion criteria were defined: (a) evidence of relevant concomitant disease (e.g. neurological syndromes), (b) incomplete sonographic documentation, (c) insufficient radiographs (e.g. malrotation) and (d) open reduction.
Data were analysed using SAS statistical software, version 9.2 (SAS Institute, Cary, North Carolina, USA). The generalized estimating equation (GEE) was used to assess the statistical difference in the values of AI from the first to the second radiographic follow-up.
For the assessment of a difference in the evolution of AI in terms of sex (boys vs. girls) and laterality (right vs. left hip), a dependent t-test for paired samples was used. Before this, the normal distribution of the differences of AI from the first to the second follow-up was determined using the Kolmogorov–Smirnov test. Significance was expressed by P-values.
After allocation of the radiographic findings of the second follow-up according to the criteria described by Tönnis, the 144 hips were divided into 62 hips (43.0%) with normal AI (Tönnis 0), 56 hips (38.9%) with mild residual dysplasia (Tönnis I) and 26 hips (18.1%) with severe residual dysplasia (Tönnis II). In the preliminary first follow-up, the 144 reinvestigated hips initially showed the following distribution: 40 hips (27.8%) showed normal values and represented the normal hip in unilateral DDH, but were reinvestigated because of the dysplastic contralateral hip, 54 hips (37.5%) showed mild residual dysplasia (Tönnis 1) and 50 hips (34.7%) showed severe residual dysplasia (Tönnis 2).
In the normal hips (Tönnis 0) with the dysplastic contralateral hip at the first follow-up, the mean AI was 22.9° range 16–26°); in the second follow-up, the hips developed to a mean AI of 18.5° (range 12–25°).
The mildly dysplastic hips (Tönnis 1) at the first follow-up reduced from a mean AI of 27.2° (range 23–31°) to a mean AI of 22.5° (range 9–29°).
The initially severely dysplastic hips (Tönnis 2) developed from a mean AI of 31.1° (range 27–36°) to a mean AI of 25.1° (range 15–32°) (Fig. 1).
P-values of GEE indicated for each group mentioned above (Tönnis 0, Tönnis 1 and Tönnis 2) showed significant differences for the values of AI (P<0.001).
The Kolmogorov–Smirnov test indicated a normal distribution of the differences of AI from the first to the second follow-up (P=0.0102).
In terms of laterality, the dependent t-test for paired samples indicated a highly significant improvement in the right hip as well as the left hip (P<0.001) (Fig. 2). The mean difference in AI from the first to the second follow-up was calculated to be 5.36° (range −4 to 14°) for the right hips and 5.34° (range −3 to 16°) for the left hips.
In terms of sex, the dependent t-test for paired samples indicated a highly significant improvement in boys’ hips as well as girls’ hips (P<0.001) (Fig. 3). The mean difference in AI from the first to the second follow-up was calculated to be 5.4° (range −4 to 14°) for the boys and 5.1° (range −1 to 14°) for the girls.
Although normal values were achieved at the end of ultrasound-monitored treatment, nine hips in seven infants presented a considerable residual dysplasia at the second follow-up. These infants were subjected to pelvic osteotomy. All except two (one hip Graf type 2a+, one hip Graf type 2c initially) of these hip joints were classified as unstable or even dislocated in the first ultrasound examination (one hip Graf type D, four hips Graf type 3a, two hips Graf type 4).
One hip showed a mild type I necrosis according to the criteria of Kalamchi and MacEwen 10 at the time of the first radiological control. We did not detect deterioration at the time of the second follow-up.
Since its introduction in the German Health Care System in 1996, general ultrasound screening undertaken during the first 6 weeks of life has been well established. The implementation of the neonatal ultrasound hip-screening programme led to a marked decrease in the incidence of late presentation of DDH and significantly reduced the number of surgical procedures 3. Nevertheless, to rule out residual dysplasia, radiological follow-up of every dysplastic hip is recommended.
Recent studies have proven that there remains a considerable number of hip joints that show residual dysplasia in the first radiographic follow-up, despite achievement of physiological values at the end of the ultrasound-monitored harness treatment 8,11–13. In particular, it has been pointed out that there is no correlation of the initial severity of DDH measured by ultrasound and the severity of residual dysplasia on the radiograph at the first follow-up. Even a physiological hip joint in initially unilateral dysplasia can have residual dysplasia in the radiographic control 8. According to these observations and clinical experience, acetabular development is not reliably predictable.
Radiographs at 12–15 months may be too early to draw conclusions for the long-term outcome of the hips with residual dysplasia. For this reason, regular radiographic control is necessary to monitor acetabular development 8,14. In our patient population, almost two-thirds of the hip joints showed at least mild residual dysplasia to the time of the first radiographic follow-up, which took place at a mean age of 14.8 months. The high number of residually dysplastic hip joints may be because of the rather early point in time of the first radiographic follow-up, shortly after the children started walking. Our results show that after a mean follow-up of 31.3 months, there was a significant improvement in residual dysplasia in all three groups. In particular, we detected significant regression of AI at the time of the second radiographic follow-up in the subgroup with initially severe dysplasia (Tönnis 2 hips) in the first follow-up (Fig. 1).
The comparison of the left and the right hip, in terms of the development from the first to the second follow-up, showed only marginal differences. Statistically, the right hip improved slightly better than the left hip (Fig. 2).
Similarly, t-values of sex-specific development showed a slightly better improvement in AI in the cohort of the boys compared with the girls (Fig. 3).
Although significant improvement in AI was observed in all three groups, nine hips in seven patients did not develop below a threshold of 30° AI at the age of 3 years; therefore, surgical treatment was indicated. However, at the same time, in terms of the development of Graf type IV hips in our cohort, six of the initially dislocated hips presented a physiological containment in the first radiographic control and an AI beyond the first SD according to the Tönnis classification. These findings indicate that the development of any once treated hip is not reliably predictable.
Interestingly, in the 40 normal hips with the contralateral dysplastic hip at the first radiographic follow-up, six hips turned out to be mildly dysplastic according to the Tönnis classification at the time of the second radiographic follow-up. Although the AI of these hips decreased, the above-mentioned six hips switched to mildly dysplastic because of the age dependence of the Tönnis classification.
Our results are supported by the current literature, which shows evidence of late dysplasia and dislocation of the hip following initially normal bilateral ultrasound screening. Gwynne Jones et al.6 have reported on seven late-presenting dislocations in five children with initially stable hips in ultrasound. At birth, four hips of three children were diagnosed as clinically instable (Barlow positive), but stabilized by the time of an ultrasound scan. Despite reduced and stable presentation at 2–6 weeks of age, subsequent late dislocation occurred.
These findings support the assumption of an endogenous factor of DDH 15. The presence of an endogenous factor may be responsible for variations in acetabular development. Depending on the penetration of the assumed endogenous factor, the hip might develop physiologically or remain severely dysplastic (Figs 4 and 5). This underlines the imperative of monitoring of any once treated hip and the contralateral side.
In summary, in most cases, we found a significant improvement in residual dysplasia at the second radiographic follow-up. Nevertheless, there is a risk for persistent residual dysplasia even after normal values are obtained at the end of ultrasound-monitored treatment. In nine of initially 144 hips, the indication of acetabuloplasty was observed because of severe residual dysplasia without recognizable improvement.
We draw the following conclusions: even after successful ultrasound-monitored treatment of DDH, there remains a risk of residual dysplasia. For this reason, radiographic follow-up of every once treated hip as well as the initially physiological contralateral hip is necessary. We recommend a follow-up regimen, with the first radiographic follow-up at the age of 18–24 months and a second radiographic control about the age of 3 years in infants with residual dysplasia in the first follow-up. As there is a high chance of improvement of residual dysplasia until the age of 3 years, at earlier ages, the indication for operative treatment should be assessed with caution.
Conflicts of interest
There are no conflicts of interest.
1. Graf R. Classification of hip joint dysplasia by means of sonography. Arch Orthop Trauma Surg. 1984;102:248–255
2. Wirth T, Stratmann L, Hinrichs F. Evolution of late presenting developmental dysplasia of the hip and associated surgical procedures after 14 years of neonatal ultrasound screening. J Bone Joint Surg Br. 2004;86:585–589
3. Von Kries R, Ihme N, Oberle D. Effect of ultrasound screening on the rate of the first operative procedures for developmental hip dysplasia in Germany. Lancet. 2003:1883–1887
4. Jaiswal A, Starks I, Kiely N. Late dislocation of the hip following normal neonatal clinical and ultrasound examination. J Bone Joint Surg Br. 2010;92-B:1449–1451
5. Rafique A, Set P, Berman L. Late presentation of developmental dysplasia of the hip following normal ultrasound examination. Clin Radiol. 2007;62:181–184
6. Gwynne Jones DP, Dunbar JD, Theis JC. Late presenting dislocation of sonographically stable hips. J Pediatr Orthop B. 2006;15:257–261
7. Ilfield FW, Westin GW, Makin M. Missed or developmental dislocation of the hip. Clin Orthop. 1986;203:276–281
8. Dornacher D, Cakir B, Reichel H, Nelitz M. Early radiological outcome
of ultrasound monitoring in infants with developmental dysplasia of the hips. J Pediatr Orthop B. 2010;19:27–31
9. Toennis D. Normal values of the hip joint for the evaluation of x-rays in children and adults. Clin Orthop. 1976;119:39–47
10. Kalamchi A, MacEwen D. Avascular necrosis following treatment of congenital dislocation of the hip. J Bone Joint Surg Am. 1980;62:876–888
11. Sucato D, Johnston CE, Birch JG, Herring JA, Mack P. Outcome of ultrasonographic hip abnormalities in clinically stable hips. J Pediatr Orthop. 1999;19:754–759
12. Cashman JP, Round J, Taylor G, Clarke NMP. The natural history of developmental dysplasia of the hip after early supervised treatment in the Pavlik harness. J Bone Joint Surg Br. 2002;84:418–425
13. Gwynne Jones DP, Vane AGS, Coulter G, Herbison P, Dunbar JD. Ultrasound measurements in the management of unstable hips treated with the Pavlik harness. J Pedatr Orthop. 2006;26:818–822
14. Jellicoe P, Aitken A, Wright K. Ultrasound screening in developmental hip dysplasia: do all scanned hips need to be followed up? J Pediatr Orthop B. 2007;16:192–195
15. Stevenson DA, Mineau G, Kerber RA, Viskochil DH, Schaefer C, Roach JW. Familiar predisposition to developmental dysplasia of the hip. J Pediatr Orthop. 2009;29:463–466