Developmental dysplasia of the hip (DDH) is the most common congenital anomaly of the lower extremity and represents a spectrum of disease ranging from an insufficiently developed, shallow acetabulum to degrees of subluxation and frank dislocation.1 Studies have demonstrated that rates of DDH vary from <1% to 13%,1–9 depending on the demographics, risk factors, and method of diagnosis.
Since its introduction, ultrasound (US) has become the imaging modality of choice for infantile DDH. The advantages of US for DDH are vast, with reports indicating increased accuracy of DDH surveillance, increased nonoperative treatment success rates, earlier treatment failure detection, decreased exposure to radiographs, decreased duration of nonoperative therapy, and lower rates of surgical intervention.6,10–14
Although US has become a routine diagnostic and management tool for DDH, the accepted standard for defining clinical hip instability are the Barlow and Ortolani signs. In reality, current practice for the management of DDH involves input from both clinical examination and ultrasonography (or radiography). However, relatively little data exist to correlate these 2 methods of assessment. Ultrasonographic classification of DDH relies on the pioneering works of Graf15, whose system was developed using data from 3500 hips, only 15 of which were luxated, and without mention of clinical examination finding. Though Graf made no formal mention of femoral head coverage in his seminal works, Graf type IV hips were termed “completely dislocated.” Graf15,16 noted subjectively that hips of this type show an “empty acetabulum” or the “femoral head lying in the soft tissue.” As a result, despite the widespread use of US, objective ultrasonographic definitions of hip dislocation remain somewhat ambiguous. This can create challenges in reporting differences between researchers who use only clinical examination to classify patients and those who rely on baseline imaging with variable criteria, all of which hinders the ability to generate high level, practice-changing evidence to improve the care of DDH.
Because of the lack of consistent, large scale multicenter US data on dislocated hips, there is confusion in the DDH literature about how best to objectively define a dislocated hip with US. The purpose of this paper is to provide primary data of US findings in clinically dislocated hips with the aim of creating standard sonographic criteria for a hip dislocation. We believe criteria of this sort will ultimately facilitate the development of evidence-based guidelines to improve patient care and outcomes of DDH treatment.
After obtaining Institutional Review Board (IRB) approval, the database of an international multicenter study group was queried. This database contains prospectively collected data from 9 participating institutions representing the United States, Mexico, Canada, Australia, and the United Kingdom. Only patients under 6 months of age with hips dislocated at rest, as defined by a positive Ortolani test on clinical examination, were included in the study. Patients were excluded from analysis if they were older than 6 months of age or did not have a baseline US before treatment. In addition, patients were excluded from analysis if there was insufficient documentation of femoral head coverage.
Femoral head coverage, alpha angle (α), and beta angle (β) were measured on pretreatment US by the treating surgeon, and were recorded directly into the database by each center. α and β were measured using the standard method on coronal images, as described by Graf.15,16 Percent coverage of the femoral head was calculated using the method described by Morin et al17 using the best available coronal flexion image (Fig. 1).
The distribution of US metric data was evaluated using the Shapiro-Wilk test to determine if it followed a normal distribution. Appropriate population statistics were then calculated based on the population distribution type; that is normal distribution or non-normal distribution. In addition to describing the data distribution, a Pearson product-moment correlation was calculated to evaluate the relationship between femoral head coverage and α, as well as β.
We identified 374 hips with a positive Ortolani sign, indicating that these hips were clinically dislocated at rest. Data were recorded from 304 patients, with 70 patients having bilateral dislocated hips. Female patients represented 86% of the population. Left hips accounted for 62% of the data. Eight hips were excluded from analysis because the patient was older than 6 months of age at time of US. Forty-one hips were excluded from analysis because of insufficient documentation of the femoral head coverage percentage. The final cohort, therefore, was composed of 325 dislocated hips from 267 patients. The average age at the time of initial US was 1.1±1.3 months.
For these clinically dislocated hips, pretreatment US demonstrated femoral head coverage data did not follow a normal distribution, but rather a right skewed distribution (P<0.001) (Fig. 2). Therefore, in lieu of reporting an average and SD, population statistics for non-normal distribution were calculated. The median femoral head coverage was 10%, with an absolute range of 0% to 60%. The interquartile range was 0% to 23%. In total, 126 hips (39%) were found to have 0% FHC. The 90th percentile of the data was found to be at 33% coverage.
α were reported from 264 hips. The median α was 43 degrees (range, 15 to 68 degrees) with an interquartile range from 37 to 49 degrees. The 90th percentile for the data was found to be 54.0 degrees (Fig. 3). β were reported from 164 hips. The median β was 66 degrees (range, 22 to 117 degrees), with an interquartile range from 57 degrees to 80 degrees. The 90th percentile was at 95 degrees (Fig. 4).
Pearson product-moment correlation was utilized to elucidate the relationship between FHC and α, as well as β. It demonstrated that the variation in FHC explains about 6.6% of the variation in α (r2=0.0662; P<0.001). For the relationship between FHC and β, the variation in FHC accounted for 15.4% of the variation in the β (r2=0.1543; P<0.001) (Fig. 5).
US has become a key modality for the diagnosis of children with risk factors for DDH, dysplastic growth, or grossly abnormal hips, with some health care systems even utilizing it for routine screening of all newborns.18,19 As a management tool, US is useful because it allows for repeated qualitative and quantitative assessment of the development of the immature acetabulum without radiation. Of all of the information provided by hip US, DDH literature has primarily focused on the use of FHC to guide management, as this represents an objective metric for the amount of femoral head displacement. Holen and colleagues utilized FHC, in addition to clinical examination, to select patients for harness therapy. Using FHC, patients, who based on clinical examination findings would have been treated, were spared unnecessary harness therapy, while still ensuring appropriate acetabular development.14 In addition, Lerman et al,20 White et al,21 and Novais et al22 have shown that lower percent FHC represents a more severe form of disease, and therefore can be used to predict treatment failure with Pavlik harness. Similarly, Alexiev et al23 found that dynamic coverage index, a modified form of FHC, was a predictor of residual dysplasia in patients successfully treated with a Pavlik harness.
Despite the importance of both the clinical examination and US in the evaluation of infants with DDH, there is need for improved understanding of the relationship between these 2 methods of assessment. Currently, the most widely accepted definition of a dislocated hip is one with a positive Ortolani sign, which indicates that the hip is clinically dislocated at rest but reducible. Certainly, hips may be dislocated and irreducible (and therefore Ortolani negative), in which case the clinical examination does not provide definitive evidence for the dislocation and imaging verification is necessary to establish the diagnosis. However, for these cases there is no clear definition for what constitutes a hip dislocation on US. This information gap also creates challenges when conducting rigorous multicenter research where it is advantageous to use objective standards to group and compare cohorts rather than relying on subjective data such as an investigator’s clinical examination finding.
At first glance, it may seem intuitive to define a hip dislocation on US simply as a percent FHC of zero, as was suggested by Graf for type IV hips.15 However, several smaller studies evaluating the boney coverage of dislocated hips, using US, have contradicted this expectation. Terjesen et al,9 reported an average of 35.9% bony rim percentage in 7 patients with Ortolani positive hips. Holen et al6,14 have reported ranges from 33% to 37% average femoral head coverage in Ortolani positive hips. Finally, Novais et al22 documented a median FHC of 23% in 78 hips with a positive Ortolani test, though it was unclear if this measurement was done before or after harness initiation. In an effort to take the first step towards an evidence-based ultrasonographic definition of a hip dislocation, we conducted the current study to report the baseline US findings of a large, multicenter cohort of clinically dislocated hips. Our data demonstrates a median FHC of 10% in our cohort. Further, our data reveal that 39% of Ortolani positive hips had zero FHC, and 90% had a FHC of ≤33%. These findings generally support those of the previous smaller series.6,9,14,21,22
We also report on the potential relationship between FHC and both α and β. Our Pearson correlation values reveal that FHC is only very weakly correlated with these 2 metrics. As a result, measures of FHC can be considered to be independent of α and β for all practical purposes.
There are several limitations to this study. First, although all patients in our database had pretreatment US, we relied on measurements that were entered directly from each individual site. As a result, our data may be susceptible to interrater variability as each center performed its own measurements. Secondly, infantile hip US is a dynamic test requiring specific skills, and as a result, ultrasonographic measurements may be affected by the technique and level of experience of the person performing the US. Small variations in US protocols and positions at different institutions may also contribute to differences in measured US indices. This is reflected in the fact that 8 hips (2.2%) were found to have ≥50% FHC despite being classified as clinically dislocated. These may represent hips that relocated when the baby was positioned in lateral decubitus position and gently manipulated during the US process. Finally, we acknowledge that this is an observational study of all Ortolani positive hips in our cohort, and thus does not reflect the US findings in normal children, Barlow positive infants (reduced but unstable hips), or those with irreducible dislocations. As such, the current study represents only the first step toward an evidence-based US definition of a dislocation, and will require further work to validate the findings across different disease severities.
Despite its limitations, our study represents the largest cohort to date of clinically dislocated hips with baseline ultrasonographic data. Observation of the distribution of these raw data allows for the development of cutoff values that may be used to preliminarily define a sonographically dislocated hip. Based on our findings that 90% of Ortolani positive hips had a femoral head coverage of ≤33%, it seems reasonable to consider ≤33% FHC as an initial threshold for defining an ultrasonographically dislocated hip. One could then reasonably term 34% to 50% coverage as subluxated. We recognize that this definition was generated using a population of only dislocated, but reducible hips, and therefore does not include measurements from reduced, but unstable hips among others. Therefore, further study utilizing the full spectrum of infantile hips, both pathologic and normal, is necessary to validate our proposed thresholds. Nonetheless, we believe that characterization of only dislocated/reducible hips provides a useful starting point for research purposes. Furthermore, we believe that our sonographic data provide background on an imaging modality that has been widely adopted for clinical and research purposes, and fills a gap in the current DDH literature, providing a link between the clinical examination and US findings.
1. Baronciani D, Atti G, Andiloro F, et al. Screening for developmental dysplasia of the hip
: from theory to practice. Collaborative Group DDH Project. Pediatrics. 1997;99:E5.
2. Bialik V, Bialik G, Blazer S, et al. Developmental dysplasia of the hip
: a new approach to incidence. Pediatrics. 1999;103:93–99.
3. Dogruel H, Atalar H, Yavuz OY, et al. Clinical examination versus ultrasonography in detecting developmental dysplasia of the hip
. Int Orthop. 2008;32:415–419.
4. Exner GU. Ultrasound screening for hip dysplasia in neonates. J Pediatr Orthop. 1988;8:656–660.
5. Gunay C, Atalar H, Dogruel H, et al. Correlation of femoral head coverage and Graf alpha angle in infants being screened for developmental dysplasia of the hip
. Int Orthop. 2009;33:761–764.
6. Holen KJ, Tegnander A, Eik-Nes SH, et al. The use of ultrasound in determining the initiation of treatment in instability of the hip in neonates. J Bone Joint Surg Br. 1999;81:846–851.
7. Loder R■ ■. The Deformed Hip, in Paediatric Orthopaedic Diagnosis: Asking the Right Questions. Joseph B, Robb J, Loder RT, Torode I, Editors. Springer India: New Delhi; 2015:3–12.
8. Peled E, Eidelman M, Katzman A, et al. Neonatal incidence of hip dysplasia: ten years of experience. Clin Orthop Relat Res. 2008;466:771–775.
9. Terjesen T, Bredland T, Berg V. Ultrasound for hip assessment in the newborn. J Bone Joint Surg Br. 1989;71:767–773.
10. Swaroop VT, Mubarak SJ. Difficult-to-treat Ortolani-positive hip: improved success with new treatment protocol. J Pediatr Orthop. 2009;29:224–230.
11. Marks DS, Clegg J, al-Chalabi AN. Routine ultrasound screening for neonatal hip instability. Can it abolish late-presenting congenital dislocation of the hip? J Bone Joint Surg Br. 1994;76:534–538.
12. 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.
13. Hangen DH, Kasser JR, Emans JB, et al. The Pavlik harness and developmental dysplasia of the hip
: has ultrasound changed treatment patterns? J Pediatr Orthop. 1995;15:729–735.
14. Holen KJ, Tegnander A, Terjesen T, et al. Ultrasonography of clinically unstable hips. A prospective study of 143 neonates at birth and early follow-up. Acta Orthop Scand. 1997;68:527–532.
15. Graf R. Classification of hip joint dysplasia by means of sonography. Arch Orthop Trauma Surg. 1984;102:248–255.
16. Graf R. Fundamentals of sonographic diagnosis of infant hip dysplasia. J Pediatr Orthop. 1984;4:735–740.
17. Morin C, Harcke HT, MacEwen GD. The infant hip: real-time US assessment of acetabular development. Radiology. 1985;157:673–677.
18. Portinaro NM, Pelillo F, Cerutti P. The role of ultrasonography in the diagnosis of developmental dysplasia of the hip
. J Pediatr Orthop. 2007;27:247–250.
19. Clarke NM, Castaneda P. Strategies to improve nonoperative childhood management. Orthop Clin North Am. 2012;43:281–289.
20. Lerman JA, Emans JB, Millis MB, et al. Early failure of Pavlik harness treatment for developmental hip dysplasia: clinical and ultrasound predictors. J Pediatr Orthop. 2001;21:348–353.
21. White KK, Sucato DJ, Agrawal S, et al. Ultrasonographic findings in hips with a positive Ortolani sign and their relationship to Pavlik harness failure. J Bone Joint Surg Am. 2010;92:113–120.
22. Novais EN, Kestel LA, Carry PM, et al. Higher Pavlik harness treatment failure is seen in graf type IV Ortolani-positive hips in males. Clin Orthop Relat Res. 2016;474:1847–1854.
23. Alexiev VA, Harcke HT, Kumar SJ. Residual dysplasia after successful Pavlik harness treatment: early ultrasound predictors. J Pediatr Orthop. 2006;26:16–23.