349

Daniel Zewdneh Solomon. Ethiop Med J, 2019, Vol. 57, No. 4

ORIGINAL ARTICLE
HIP ULTRASOUND IN DEVELOPMENTAL HIP DYSPLASIA: AN INITIAL EXPERI-
ENCE AT A TERTIARY SPECIALIZED TEACHING HOSPITAL IN ADDIS ABABA
Daniel Zewdneh Solomon, MD, MHA, ScRAD, SScPRAD
ABSTRACT

Introduction: Developmental dysplasia of the hip is a congenital anomaly. Clinical assessment is not effective and
efficient as assumed previously, and the role of ultrasound imaging has increased over time and is now employed
for screening of neonates. The local literature lacks information. The purpose of this article was, therefore, to do a
preliminary ultrasound assessment of its status among infants.
Methods: A cross-sectional study was conducted in the pediatric unit of the Department of Radiology at Tikur An-
bessa Specialized Hospital from August 2018- to February 2019.Participants were selected based on recom-
mended age (4-24 weeks) Hip ultrasound was performed using the Graf static method to measure alpha and beta
angles of both hips and measurements were grouped according to the Graf classification. Mean values of alpha
and beta angles of both hip sides and types were analyzed. Independent sample T test was used for analysis where
appropriate.
Result: Over the specified period, 65 neonates within the age group 4-24 weeks were scanned. 40(61.5%) were
males and 25 (38.5%) females. Graf type II (immature and dysplastic) hips comprised 4.61% for the right hip and
10.7% for the left in males and 4.61% each for both hips in females).Type IV(dislocated) hips comprised 4.61% for
the right hip and 1.5% for the left in males and was not identified in females.
Conclusion: The proportion of abnormal hips identified in the study warrants the need for a much larger and more
comprehensive study and the use of ultrasound for a wider clinical application or a potential screening strategy is
recommended.
Key Words: developmental hip dysplasia, Ultrasound, hip angles

INTRODUCTION Vivek Gulet showed an overall incidence ranging

from 1.5- 2.5 per 1000 live births (1). Yet another
Developmental dysplasia of the hip (DDH) which was Australian study showed a rather higher incidence
previously addressed in the literature as congenital of 7.3 per 1000 live births (4). In a similar man-
dislocation of the hip does not have an established ner, the prevalence of DDH shows variations
etiology. The condition encompasses a wide spectrum across studies where a range of 0.15-4% was the
of conditions from mild, clinically insignificant reported prevalence in one study (5); while another
acetabular dysplasia in the adult to irreducible, total showed a 0.07-0.16% (6). The local literature does
dislocation of the hip in neonates (1). It was first re- not have any information regarding incidence in
ported as hip instability by Roser in 1879. Le Damny Ethiopia apart from sporadic clinical encounters in
and Seiget, were the first people to describe clinical routine practice. The only study found was one
tests for the condition in 1910. This was further devel- conducted in the Ethiopian Jewish community in
oped in 1937. Later on Palmer in 1961 and Barlow in Israel which showed an incidence of 0.44% com-
1962 expanded on these clinical tests for identifying pared to 5.9% in white Jewish neonates (7)
subluxation and dislocation (2).The incidence of the
disease, as described in the literature, tremendously Contemporary evidence in the literature shows that
shows a wide range of variations. DDH is associated with certain risk factors such as
breech presentation, positive family history, and
A systematic medical literature review by Randal T gender (female).Some human leucocyte antigens
Loder et al came up with incidence ranges of 0.06 per (HLA) A, B, and D gene types demonstrate an
1000 live births in Africans to 76.1 in native Ameri- increase in DDH.
cans showing a significant variability within racial
groups and across geographic locations (3).

1
Consultant pediatric radiologist , Department of Radiology , School of Medicine, College of Health Sciences ,
Addis Ababa University.
*Corresponding Author E-mail: [email protected]
 350

Chromosome17q21 is very much associated with the specified study period and generate information
DDH. Lax ligaments and abnormalities in collagen for a more focused large scale study which may
metabolism, estrogen metabolism, and pregnancy- help develop future national screening strategies
associated pelvic instability have associations with which are currently non-existent.
DDH. On the other hand, children born premature,
with low birth weights, or to multi-fetal pregnancies PATIENTS AND METHODS
are somewhat protected from DDH (8).
This was a cross-sectional study conducted in the
Clinical assessment of DDH is not always as effective pediatric unit of the Department of Radiology at
and efficient as assumed previously, and the role of Tikur Anbessa Specialized Hospital from August
ultrasound imaging has increased over the years and is 2018- to February 2019 after approval by the de-
now extensively employed for screening of neonates partmental research and ethics committee. The
with clinical evidence of hip instability and those at source population was all pediatric patients present-
risk (7,8). The wide spread use of ultrasound began ing at the teaching hospital in the specified study
after Graf developed and published a static method of period.
assessment, and later in 1985, Harcke et al introduced
a dynamic method of examination(9) A non-random convenience sampling method was
Currently, there is no consensus as to whether ultra- used. The sample size obtained was 59 with a 10%
sound should be used as a universal screening or in contingency added, totaling 65, and this was esti-
selective situations before the age 6 weeks to mini- mated, using a higher range prevalence of 4% taken
mize late detection of DDH. The age definition of late from the literature and cited in a joint Norwegian
cases ranges from 4 to 24 weeks. However, studies and British study (5) as a baseline, and employing a
have also shown that positive ultrasound examinations single population proportion formula with a level of
in the universal plan have resulted in insignificant significance at 5% a confidence interval (Z) of 95%
reduction in the rate of late complication of DDH and an absolute precision (margin of error) at 5%(ά-
while increasing the rate of intervention compared to 0.05). Infants with known hip pathology were ex-
the selective approach; although the difference was cluded.
not statistically significant.(5): while other studies
recommend universal screening of newborns(10). Hip US exam was performed after informed consent
was obtained from parents. Data were collected in a
The two methods of ultrasound exam- Graf’s static data collection format. A Sonoace US machine with
method of ά and β acetabular angle measurement and a 7 MHz linear probe was used. The Graf method of
Harcke’s dynamic method are mostly used in conjunc- static scan was employed to measure the alpha and
tion, but there are some authors who recommend beta angles of both hips. Measurements were
Graf’s method as a more reliable one and advocate its grouped according to the Graf angle classification
separate use (6). The author of this article has em- shown below on Table 1 (11,12). Subjects were
ployed Graf’s method (Figure 1) as it is easier to use stratified by gender and age (4-10 weeks, 11-17
since it does not employ clinical maneuvers such as weeks and 18-24 weeks of age). All collected data
Barlow’s dislocation test. (11, 12). were then checked for clarity and completeness and
analyzed using SPSS statistical software version 20
The purpose of this article is to do a preliminary as- at 5% precision level. Percentage, mean and stan-
sessment of the incidence of late DDH in infants be- dard deviation of alpha and beta angles of both hip
tween the ages of 4 and 24 weeks visiting our pediat- sides as well as hip types were analyzed using the
ric ultrasound unit within Graf classification. Associations between variables
were checked with independent samples T test at a
P value of <0.05 taken as statistically significant.

A B
Figure 1: Schematic demonstration of alpha and beta angle measurements used in this study (17)
 351

Table 1: Graf Classification of hip angles used - (12)

All types Alpha angle > 60 degrees

Type 1a Beta angle < 55 degrees
Type 1b Beta angle > 55 degrees

(Normal)

All type Ⅱ
Type 2a Alpha angle 50-59 degrees (< 3 months)
Type 2b Alpha angle 50-59 degrees (> 3 months)
Beta angle 55-77

Type 2c Alpha angle 43-49 degrees

Beta angle < 77 degrees

Type 2d Alpha angle 43-49 degrees

Beta angle > 77 degrees
( "about to decenter")
I mmature/Dysplatic
Type 3 Alpha angle < 43 degrees
( (Type 3a and 3b distinguished on
t he grounds of structural
alteration of the cartilaginous roof)
Deficient/Subluxed

Type 4 Alpha angle < 43 degrees

(dislocated with labruminterposed
between femoral head and acetabulum)

Dislocated, highly deficient

RESULTS Hip types were grouped according to a modified

Graf classification for both sexes. Type I hips (all
US scan of the hips was performed in 65 neonates of subtypes lumped together) were the predominant
whom 40(61.5%) were males and 25 (38.5%) females. ones in both gender groups and on both sides with
The mean age was 12 weeks. In the female babies, the 34 (85%) for the right hip, and 32(80%) out of the
mean alpha angles were 63.08 (±4.30) degrees for the 40 for the left hip in males and 22(88%) out of the
right hip and 64.20(±5.45) for the left; whereas in the 25 females each for the right and left hips, (Table
male babies, the mean alpha angles were 64.65(±4.35) 2).
for the right hip and 63.63(± 7.16) for the left (figure
2). :The mean beta angles were 44.4 (±10.7) for the Type II hips (with all subtypes included) were seen
right hip and 45.36(±11.0) for the left in females; in 3 (7.5%) of our cases for the right hip and 7
while the mean beta angles were47.05(± 14) for the (17.5%) for the left out of the 40 males. Out of the
right hip and 47.12(± 11.65) for the left in males 25 females scanned, 3(12%) showed type II hips
(figure 3). equally for both hips.(Table 2). There were no
Graf type III hips identified in both genders. Graf
Differences in both mean alpha and beta angles be- type IV hip was found in 3(7.5%) for the right hip
tween males and females were not statistically signifi- and 1(2.5%) for the left, of the 40 males. It was not
cant (t=1.421, df=63, P=.160 for right alpha angle, identified in females.
t= .465, df=63, P=-.644 for the left alpha angle and
t=.808 df= 63 P= .422 for right beta angle and
t= .606, df=63, P=.547 for the left).
 352

Figure 2: Mean ά angles by gender among infants Figure 3: Mean beta angles by gender among infants
scanned at Tikur Anbessa Specialized Hospital: August scanned at Tikur Anbessa Specialized Hospital: Au-
2018- February 2019. gust 2018- February 2019.

Table 2: Distribution of hip types (as determined by the Graf classification) by gender with group percent age
among infants scanned at Tikur Anbessa Specialized Hospital: August 2018- February 2019.
Graf hip type Gender
Male Female
Right Left Right Left

(n) (%) (n) (%) (n) (%) (n) %

Type I 34 85 32 80 22 88 22
88
Type II 3 7.5 7 17.5 3 12 3
12
Type III 0 0 0 0 0 0 0
0
Type IV 3 7.5 1 2.5 0 0 0
0
Total 40 100 40 100 25 100 25
100

Table 3: Distribution of hip types (as determined by the Graf classification) by laterality with both sexes combined
among infants scanned at Tikur Anbessa Specialized Hospital: August 2018- February 2019.

Graf hip type Laterality

Right Left
(n) (%) (n) (%)
Type I 56 86.2 54 83.1
Type II 6 9.2 10 15.4

Type III 0 0 00 00
Type IV 3 4.6 1 1.6
Total 65 100 65 100
 353

Among the study subjects, 62 of them had cephalic Although various studies in the literature describe
presentation and 3 had breech at delivery. Of the 62 the occurrence of immature, dysplastic, subluxed
cephalic presentations, 54 (83%) had type I hips, 5 or dislocated hips more in females and on the left
(7.7%) had type II hips and 3(4.61%) had type IV side, this study did not show the same except for
hips. Of the 3 breech presentations, 2 had type I hips type II hip which was observed in 10 of the sub-
and 1 had type II. 57 babies were delivered vaginally, jects on the left side; but again the small sample
and 8 by Cesarean section. Forty nine of those deliv- size and the study design pose a significant limita-
ered vaginally had type I hips, 5 had type II and 3 had tion to provide a solid explanation, and a more
type IV. From the 8 Cesarean section deliveries, 7 had comprehensive study is needed to provide a rea-
type I and 1 had type II hips. There were 26 first born sonable answer in the future.
and 39 non-firstborn babies. Twenty two of the first
born babies had type I, 3 had type II, and 1 type IV The literature has documented studies which state
hips whereas 34 of the non-first born babies had type breech lie and presentation as common predispos-
I, 3 type II and 2 type IV hips (tables not shown). Two ing factors in the development of DDH (3, 4, and
of our subjects had evidence of torticollis and were in 8). There were only 3 breech presentations with 1
the type II hip group. dysplastic hip documented out the 65 study sub-
jects and as such; no significant inference could be
DISCUSSION made. By the same token, differences observed in
modes of delivery (vaginal Vs Cesarean section)
In this study, mean alpha and beta angles were com- were not worth reasonable inference owing to the
puted for both genders and both sides. The slight small number of Cesarean sections observed in the
variations observed in mean alpha angle values be- study. Dysplastic hips (type II) were equally seen
tween male and female groups (higher in males on the in both first-born and non -first-born babies; while
right side and higher in females on the left side) were only 1 dislocated hip (type IV) in the first-born
not statistically significant in contrast to the findings group and 2 in the non-first-born group were seen,
of other studies in the literature which showed higher respectively, but differences in observation were
values for males(6). Mean beta angles also showed not statistically significant. There were two cases
slightly higher values in males, but the difference of torticollis found with type II hip dysplasia in
again was not statistically significant as were differ- this study, but it is very difficult to assume whether
ences of values with regard to laterality These find- this observation was accidental or indicates asso-
ings were in agreement with those of other similar ciation; although such association between DDH
studies in the literature (6, 13). and torticollis has been established in the literature
(16).
Further analysis of data showed that type I (mature
and normal) hips were the most predominant group Variations in the values among different studies
occurring in 56(86.2%) regardless of gender. The might be affected by various factors such as study
same predominance of occurrence was corroborated in designs, population types, and experience and ob-
other studies (6, 14). Type II hips (immature and dys- server variation of ultrasound examination.
plastic) were seen in 16 (24.6%).This figure was
higher than values obtained by Sernic (12.4%) (13), The immature/dysplastic (type II) and dislocated/
and slightly lower than figures by Bruno (6). Yet an- deficient (type IV) hips identified in the study,
other study by Garedaghi (15) revealed a figure which portray a clear need for a more comprehensive
was around 49.9%. longitudinal study design across health institutions
nation-wide to investigate the incidence and preva-
With regard to laterality, type II hips occurred more lence of DDH and all the predisposing factors
on the left side compared to the right and this is in mentioned above. This study had limitations in
agreement with the findings of the previous studies (6, terms of study design, scarcity of adequate high
13). Type III hips (subluxed type) were not identified end ultrasound equipment and study subjects com-
in the study subjects; whereas type IV hips ing from a single hospital population.
(dislocated) were identified in only three males on
right hip and one on the left as compared to the much
lower figure by Gharedaghi(15).

.
 354

Conclusion and recommendation ACKNOWLEDGEMENT

In this study, although the majority of our study sub-
jects had normal hip types, the proportion of abnormal The author would like to express his deepest grati-
hips that were picked by ultrasound herald the need tude to the staff and residents of the Department of
for a more comprehensive longitudinal study design Radiology, for their considerate assistance and to
addressing associated risk factors, and the effects of the parents of the study subjects for their willing-
early and late diagnosis of DDH and the role of ultra- ness to be a part of the study and most of all to the
sound with the recommended protocols including dy- senior radiology faculty for their valuable technical
namic methods should be considered for a wider clini- help and meaningful review.
cal application.

REFERENCES
1. Gulati V, Eseonu K, Sayani J, et al. Developmental dysplasia of the hip in the newborn: A systematic
review: World Journal of Orthppedics 2013:4(2):32-41.
2. Noordin N, Umer M, Hafeez K, Nawaz H Developmental dysplasia of the hip: Orthop Rev (Pavia).
2010; 2(2): e19.
3. Loder RT, Skopelja EN. epidemiology and demographics of hip dysplasia: ISRN Orthopedics
2010;2011:1-46.
4. Lowry CA, Donoghue VB, Murphy JF. Auditing hip ultrasound screening of infants at increased risk of
developmental dysplasia of the hip. Arch Dis Child 2005;90:579-581.
5. Laborie LB, Engesæter IØ, Lehmann TG, Eastwood DM, Engesæter LB, Rosendahl K. Screening strate-
gies for hip dysplasia: Long term outcomes of a randomized controlled trial: Pediatrics 2013;132(3):492-
501.
6. Jacobino B, Galvao M, Silva A, Castro C. Using the Graf method of ultra sound examination to classify
hip dysplasia in neonates: Autopsy and Case Reports: 2012:2(2):5-10.
7. Vialik B, Berant M: Immunity of Ethiopian Jews to developmental dysplasia of the hips: A preliminary
sonographic study: J Pediar. Orthop 1997:6(4):253-254.
8. Emrie M, Scott V. Is ultrasound screening for DDH in babies born breech sufficient. J Child Orthop:
2010:4:3-8.
9. Clarke NM, Harcke HT, McHugh P, Lee MS, Borns PF, MacEwen GD. Real time ultrasound in the di-
agnosis of congenital dislocation and hip dysplasia of the hip: J Bone Joint Surg Br 1985 May;67(3):406-
12
10. Sharpe P, Mulpuri K, Chan A, and Cundy PJ. Differences in risk factors between early and late diag-
nosed developmental dysplasia if the hip: Arch Dis Child Neonatal Ed: 2006:91:158-162.
11. Graf R. The diagnosis of congenital hi-joint dislocation by the ultrasonic compound treatment. Arch Or-
thop Trauma Surg 1980; 97(2):117-33.
12. Alexiev VA, Harcke HT, Kumar SJ. Residual dysplasia after successful Pavlik harness treatment: early
ultrasound predictors. J Pediatr Orthop 2006;26:16-23.
13. Sernik R, Ascencio JEB. Quadril. In: Cerri GG, Sernik R. Ultrassonografiadosistemamusculoesquelé-
tico: correlação com ressonânciamagnética. Rio de Janeiro: Revinter; 2009. p. 259-315. Portuguese.
14. Baronciani D, Atti G, Andiloro F, et al. Screening for developmental dysplasia of the hip: from theory to
practice. Pediatrics. 1997;99(2):e5. PMid.
15. Gharedaghi M, Mohammadzadeh A, Zandi B. Comparison of clinical and sonographic prevalence of
developmental dysplasia of the hip. Acta Med Iran. 2011;49(1):25-7.
16. Park HK, Kang EY, Lee SH, Kim KM, Jung AY, Nam DH. The utility of ultrasonography for the diag-
nosis of DDH of the hip joint in congenital muscular torticollis: Ann Rehabil Med.2013: 37(1):26-32.
17. Martin et al: Developmental dysplasia of the hip (DDH) - Review of major imaging techniques: Educa-
tional Exhibit : European Society of Radiology (ESR): 2017: Poster No: C-2053: DOI: 10.1594/ecr2017/
C-2053.