Purpose: to calculate sensitivity, specificity and positive and negative predictive values for multiparameter ultrasound scores for Down's syndrome. Patients and Methods: sensitivity and specificity for Down syndrome were calculated for ultrasound scores in a prospective study of ultrasound signs from 16 to 24 weeks in a high-risk population who denied invasive procedures after genetic counselling. The signs and scores were: femur/foot length < 0,9 (1), nuchal fold > 5 mm (2), pyelocaliceal diameter > 5 mm (1), nasal bones < 6 mm (1), absent or hypoplastic fifth median phalanx (1) and major structural malformations (2). Complete follow-up was obtained in each case. Genetic amniocentesis was proposed in the case of score 2 or more. Results: a total of 963 patients were examined from October 93 to December 97 at a mean gestational age of 19.6 (range 16 -24) weeks. Women's age ranged from 35 to 47 years (mean 38.8) and 18 Down syndrome cases were observed (1.8%). Sensitivity was 94.5% (17/18) for score 1 and 73% (13/18) for score 2 (false positive rate of 9.8% for score 1 and 4.1% for score 2). Individual sign sensitivity and specificity were: femur/foot = 16.7% (3/18) and 96.8% (915/945); nasal bones = 22.2% (4/18) and 92.1% (870/945); nuchal fold = 44.4% (8/18) and 96.5% (912/945); pyelic diameter = 38.9% (7/18) and 94.3% (891/945); absent phalanx = 22.2% (4/18) and 98.5% (912/945); malformation = 22.2% (4/18) and 98.2% (928/945). Conclusions: the overall sensitivity for score 1 was high but false positive rates were also high. For score 2, sensibility was still good (73%) and false positive rate was acceptable (4.1%). Positive and negative predictive values can be calculated for each prevalence (women's age). More cases are needed to reach final conclusions about this screening method (specially in a low-risk population) although this system has been useful for high-risk patients who deny invasive procedures.
Down syndrome; Ultrasound; Prenatal diagnosis; Maternal age; Amniocentesis
