The aim of our study was to evaluate the incidence of DDH in breech patients referred to our institution for screening ultrasound and to determine what, if any, radiographic follow-up is necessary. Breech presentation is a well-known risk factor for DDH, and rates in the literature vary based on the definition of dysplasia and the method of its determination. Suzuki and Yamamuro [9
] reported an overall incidence of DDH among 6,559 infants as 0.7%, but 2% in footling presentation and 20% in single-breech presentation. DDH in their study was diagnosed clinically, with a positive Barlow examination. Holen et al. [10
] reported a 6.1% incidence of neonatal hip instability in breech patients determined by clinical examination and confirmed by dynamic ultrasonography. An additional 3.7% had normal clinical examinations, but abnormal dynamic ultrasounds. In our current study, 9% of patients had a positive Barlow and/or Ortolani examination at the time of initial evaluation. When the diagnosis was based on ultrasonographic criteria, in the setting of abnormal hip abduction, we found that 27% of referred patients were considered to be dysplastic and were treated with a Pavlik harness. For those patients with an initially normal clinical examination and screening ultrasound, we found that 29% were considered to be dysplastic based on plain radiographs and clinical abduction. Other studies have reported frequencies of neonatal hip instability in breech newborns from 12 to 24% [9
Ultrasound is known to be a safe and effective screening tool for hip dysplasia in infants, especially before the appearance of the femoral head ossific nucleus, and has been advocated to decrease the incidence of late-detected hip dysplasia. However, the recommendations regarding the radiographic follow-up of hips deemed normal by screening ultrasound are less clear. Several authors have suggested that patients can be discharged following a normal screening ultrasound. Jellicoe et al. [3
] identified 226 ‘at-risk’ infants (click, family history, breech, abnormal skin creases, clubfoot, and unstable hips) referred to their institution over a 3-year period for screening. They classified their patients based on ultrasound into four groups: normal, shallow but stable, shallow and unstable, and dislocated. One hundred and thirty-nine patients were considered to be normal and were reviewed by X-ray at 12 months. Forty failed to follow-up or refused X-ray. All of the remaining patients were considered normal at 12 months. Of the 69 hips considered to be shallow but stable, 61 improved with observation and serial ultrasounds; seven did not improve and were placed into a Pavlik harness. One was lost to follow-up for 3 months and was noted to have developed a dislocated hip upon return, requiring closed reduction; 12% of their ‘shallow but stable’ patients required treatment. Our results were quite similar in our number of patients with normal screening ultrasounds being treated as well. However, despite 12% of patients in their study receiving treatment, and even one patient requiring surgery, the authors recommended discharge following a normal ultrasound examination [3
Arumilli et al. [2
] evaluated 100 patients with a family history of DDH referred for screening; all had normal clinical examinations and normal screening static ultrasounds. Of the 89 available for final review, 71 had normal radiographs at 6–8 months and the remainder had normal radiographs by 24 months of age. They also recommend the discharge of patients with a family history of DDH and normal screening static ultrasound. Their results and recommendations correspond with those reported by Osarumwense et al. [4
] on 181 children with a family history of DDH and normal screening ultrasounds. Two of those patients had abnormal radiographs at 9–12 months of age, but both resolved with observation.
However, hip dysplasia is known to be an evolving condition that often improves, but may also worsen, over time. Rafique et al. [14
] published a case report of a female patient born breech at 35 weeks that had a normal clinical examination at 4 days of age and a normal static ultrasound. She was deemed ‘normal’ at 8 months, but was re-referred by her pediatrician at 12 months and was found to have a dislocated right hip. Jones et al. [15
] reported seven dislocations at follow-up radiographs in five patients with initially normal dynamic and static ultrasound; they recommended a pelvic radiograph at 4–6 months in patients with a normal ultrasound to prevent these late misses.
In our review, we identified 266 patients with the risk factor of breech presentation and stable clinical examination; 27% of these patients had abnormal screening ultrasounds and were subsequently treated with a Pavlik harness. We prefer to use the Harcke dynamic method for infant hip ultrasound. It has been shown that the interobserver and intraobserver agreement ratios of the Graf classification are only 51% and 65%, respectively [16
] and that dynamic assessment is a more reliable predictor of radiographic outcome [17
]. Of the 131 infants with normal screening ultrasound who returned at an average of 5 months follow-up, 29% had abnormal findings and were treated. Although the decision to treat was made in the setting of both clinical and radiographic evaluation, the percentage of patients with normal ultrasounds who were treated at the 4–6-month mark that had acetabular indices at least two standard deviations greater than age-adjusted normal values [7
] was 62% versus 26% of those patients ultimately discharged. Consistent, non-rotated pelvic radiographs can be difficult to obtain in this patient population, and the measurement error of acetabular indices ranges from 2° to 6° [18
]. We, therefore, take both anteroposterior (AP) and frog-leg pelvis X-rays in an attempt to capture the best possible view of the hips, which we find especially helpful in evaluating the quality of the sourcil and measuring the acetabular index. The decision to treat is based on clinical abduction, as well as radiographic parameters, at our institution; upon retrospective review, we found that, at the visit where the decision to treat was made, 27% of patients placed in a Pavlik harness had a greater than 10° side-to-side difference in abduction and 15% had symmetric abduction of less than 60°. In summary, we do not base treatment decisions exclusively on X-ray measurements, but, rather, on a more global evaluation of the patient’s physical and radiographic examination.
There are several limitations to our study. It is retrospective and has a limited follow-up of only 10 months, on average, an age when the hip joint is still developing. There is the possibility that we are over-treating, and some of these hips may have ultimately normalized. However, the treatment of mild dysplasia at a young age with a harness is fairly simple and has relatively low morbidity in comparison to the surgical treatment of an older child with missed or persistent dysplasia. Therefore, at our institution, we treat patients with clinically stable but subtly radiographically dysplastic hips with a Pavlik harness based on the belief that every attempt should be made to ensure the development of a normal hip joint that will last a lifetime. Weinstein has written, “the only guarantee of a lifetime of normal hip function is a completely normal radiographic appearance of the hip” [21
]. As many authors have shown, subtle dysplasia not detected in adolescence and early adulthood can lead to significant morbidity in later life [22
]. It remains to be evaluated by further study what the optimal detection and treatment of these patients in infancy may be, and, if early, aggressive bracing treatment will decrease the development of later dysplastic acetabulae requiring pelvic osteotomies in young adulthood.
The findings of our study support those of others that breech presentation is the most important environmental risk factor for hip dysplasia [28
] and that these patients require more vigilant evaluation. Our current protocol for breech patients, in addition to a careful newborn physical examination, is to screen with physical examination and dynamic ultrasound at 6 weeks of age and, if normal, to follow-up at 6 months for an evaluation of abduction, as well as AP pelvis and frog-leg lateral radiographs. This was developed after noting a small, but not insignificant, rate of acetabular dysplasia in those patients that followed up. At this point, we feel that the risk of subsequent dysplasia is too high to discharge patients after normal ultrasound, but further longitudinal studies will need to be done to validate any one protocol in this patient population.