Patients and Methods
Between 1 January 1996 and 31 December 2005, a prospective, observational longitudinal, targeted (clinical instability and 'at risk') hip ultrasound and clinical screening programme was undertaken at Blackburn Royal Infirmary. Neonates referred with clinical instability (as defined by a positive Ortolani or Barlow manoeuvre) were assessed by ultrasound within 1–2 weeks of age while infants considered to be 'at-risk' were assessed between 6 and 9 weeks of age. Other cases could be referred at any time from the community, usually after the 6-week general practitioner 'hip check' for clinical instability or limited hip abduction. The majority of referrals were in a bimodal distribution (within 1–2 weeks and at between 6 and 9 weeks). 'At risk' factors included a strong family history, breech presentation, postural and fixed foot abnormalities, oligohydramnios and torticollis. Infants presenting later (usually referred from the general practitioner) with LHA were assessed clinically and with sonographic and/or radiographic hip joint imaging (depending on age of presentation).
During the 10-year study period, 2876 (7.66% of the birth population) neonatal instability and 'at-risk' patients were assessed by the senior author (RWP), clinically and ultrasonographically, from a total of 37 518 births from the surrounding districts of Blackburn, Hyndburn and the Ribble Valley. The clinical assessment of hip abduction was made with the patient supine with both hips flexed to 90°. The prone method was not used as the senior author (RWP) found it was difficult to fix the pelvis, making accurate assessment poor. The clinical examination was undertaken independently of the ultrasound hip scan. LHA was noted at this initial assessment, and any block to full abduction was noted from the horizontal and was considered clinically present if it was estimated to be equal to or more than 20° compared to the other hip, as less than this has been shown to be within the normal range. Clinically, it is difficult to detect a difference between both hips in unilateral LHA of less than 20°; however, no assessment was made regarding the absolute degree of limited abduction and dysplasia as this sign was considered as positive if >20° or negative if <20°. Inclusion criteria were cases of unilateral or bilateral limitation of hip abduction hip (excluding syndromal, neuromuscular and skeletal dysplasia cases).
Ultrasound examinations were undertaken with the patient in the lateral decubitus position, with the hips flexed and adducted, in order to potentially minimise errors produced by pelvic obliquity. Static and dynamic (incorporating the Barlow's dynamic manoeuvre to demonstrate instability) sonographic hip imaging methods were used and the Graf α angle, measuring the osseous development of the acetabular roof, was recorded. The Graf α angle is the angle subtended by two lines, a baseline running tangentially to the wall of the ilium and an acetabular roof line. A modification of these sonographic measurements was used (Figure 1).
(Enlarge Image)
Figure 1.
Illustration of the α angle denoting the slope of the bony acetabulum. A modified Graf classification used; Type I (normal) α angle >60°, Type II α angle 43–59°, Type III α angle <43°, Type IV subluxable, dislocated dislocatable hip.
'Pathological' sonographic DDH was considered to be Graf Type III, IV or hip joint irreducibility. As 90% of Graf Type II dysplasia spontaneously resolve, the majority of Graf Type II hips were considered to be 'physiological'. All Graf Type II hips were followed up until normal or treated if they deteriorated. If the sonographic hip dysplasia deteriorated, the worst classification was recorded in the data analysis (ie, Graf Type III rather than Graf Type II).
Any infant, who presented after the age of 4–6 months with LHA, was initially assessed with a plain pelvic radiograph (although the hip was in addition imaged sonographically if the child was below 40 weeks old). In these cases, the diagnosis of pathological hip dysplasia was made using the radiological maturation curve of Tonnis in which a pathological hip dysplasia was diagnosed if the acetabular index was below two SD of the mean for that age (the lowest 2.5% of the population).
The study was not completed until 4 years after 31 December 2005, in order to identify other cases of irreducible hip dislocation or 'pathological' dysplasia requiring surgery which were born within the 10 years of the study but which presented to the clinic at a later date (ie, 'late' cases).
Statistical analysis was conducted using XLstat comparing the three main groups of patients, namely unilateral LHA, bilateral LHA and no LHA to determine the sensitivity, specificity of the tests as well as the PPV and negative predictive values (NPV) of LHA.