Validity and diagnostic bias in the clinical screening for congenital dysplasia of the hip.

To assess the validity of the clinical maneuvers and signs in the screening for congenital dysplasia of the hip and the presence of a diagnostic bias in this screening, a random sample of 261 newborns was studied at a tertiary hospital. All newborns were clinically and sonographically studied in the first 48 hours after birth. Hips were classified according to Graf's scheme. Sensitivity, specificity and predictive values were estimated. Nineteen dysplasias and 100 physiologically immature hips were detected. Sensitivity and specificity for the Ortolani-Barlow maneuver were 26% and 84% respectively. A higher sensitivity was observed when all clinical maneuvers/signs were considered; however, specificity decreased, so the positive predictive value reached a similar figure to that of Ortolani-Barlow, 5%. A higher positive rate with the Ortolani-Barlow procedure was observed when newborns had risk factors for congenital disease of the hip (family history and female sex) or belonged to groups that could be considered as high-risk (low Apgar score, younger mothers). Other procedures apart from clinical examination should also be recommended as an effective screening procedure to decrease the number of cases missed and to avoid diagnostic bias in future epidemiologic research on this condition.