Treat, follow-up, or discharge? An aid in sonographic decisions for the borderline centered neonatal/infantile hips.

Aim: To investigate whether linear measurements or ratios on the Graf's "standard plane" ultrasound images of the neonatal/infantile hip, can support the clinically important differentiation between type I and type II hips.

Material and methods: A total of 60 Graf type II hips and 124 randomly selected Graf type I hips, matched to the gestational age at delivery, birth weight, delivery mode, and age at the time of the examination, were identified through our hip screening service, during a period of two years. The images were diagnostically suitable, following anatomical identification and usability check, according to Graf. Anatomical landmarks including the lower limb of the os ilium, the bony rim, the silhouette of the os ilium, the labrum and the femoral head's borders, were used to determine the measurements and ratios which quantified their inter-relationships.

Results: The indices which differed significantly between type I and type II hips included: (a) the width of the "bony roof" (cut-off value 5.91 mm, sensitivity: 75%, specificity: 70%), (b) the ratio of the width of the "bony roof" to the femoral head's width (cut-off value 0.40, sensitivity 83%, specificity 71%), and (c) the ratio of the cartilaginous acetabular roof's width (including the labrum), to the width of the femoral head (cut-off value 0.450, sensitivity 82%, specificity 67%).

Conclusions: Newly introduced measurements and calculated ratios on "standard plane" ultrasound images can be used as additional indices in the differentiation between Graf's types of "centered hips", thus increasing the diagnostic certainty of the examiner in borderline cases and limiting unnecessary re-examinations or treatment.