Performance of Doppler ultrasonography as a screening test in low risk pregnancies: results of a multicentric study

To assess the efficacy of the Doppler velocimetry of the uterine and umbilical arteries as a screening test for preeclampsia and fetal growth retardation, we studied 916 low risk pregnancies. The S/D ratios of the uterine and umbilical arteries were obtained at 19 to 24 weeks and at 26 to 31 weeks of gestational age. Mean values, receiver operator curves, and the diagnostic accuracy of the tests were calculated for the following end-points: (1) pregnancy-induced hypertension, (2) low birth weight for gestational age, (3) small for gestational age with abnormal outcome, (4) pregnancy-induced hypertension needing preterm delivery. The prevalences for these outcomes were 3.4%, 4.6%, 1%, and 0.7%, respectively. The study was blinded. The umbilical and uterine artery S/D ratios were significantly higher in the abnormal than in the normal outcome group. When uterine arteries were studied at 19 to 24 weeks, sensitivity was 59% in the detection of pregnancy-induced hypertension, 11% in the detection of small for gestational age fetuses, 33% in the detection of small for gestational age fetuses with abnormal outcome, and 83% in the detection of pregnancy induced hypertension needing preterm delivery; the corresponding values for specificity were 69%, 94%, 94%, and 68%. At 26 to 31 weeks the sensitivity values were respectively, 69%, 58%, 75%, and 100% and specificity values were 80%, 59%, 39%, and 79%. When umbilical arteries were studied at 19 to 24 weeks, sensitivity was 38% in the detection of pregnancy-induced hypertension, 46% in the detection of small for gestational age fetuses, 78% in the detection of small for gestational age fetuses with abnormal outcome, and 67% in the detection of pregnancy-induced hypertension needing preterm delivery. The corresponding values of specificity were 74% for all four groups. At 26 to 31 weeks the sensitivity values were 38%, 43%, 87%, and 67%, respectively, and specificity values were 80% for all four groups. We concluded that Doppler examinations of the uterine and umbilical arteries can detect, at midpregnancy, the severe forms of pregnancy-induced hypertension and small for gestational age fetuses but they cannot be used to screen a low risk population in which the prevalence of the disease is low, and hence the positive predictive value is low.

Can ultrasound be used to screen uterine malformations?

Eighty-nine patients were examined by ultrasound during the second phase of the menstrual cycle. The operators were not aware of the patients' history or results of previous ultrasonographic or radiologic investigations. The operators' aim was to diagnose uterine malformations graded as normal, suspected, or abnormal. Ultrasound identified 7 of 9 patients with a didelphic or complete septate uterus and 2 out of 3 with a bicornuate or partial septate uterus. Two other cases were suspected (one with a unicornuate, the other with a didelphic uterus). One false-positive case occurred, and 10 patients were not identified as having a malformed uterus (sensitivity, 42.9%; specificity, 97.8%). False-negatives involved mainly minor malformations.

Characteristics of women under 20 with cervical intraepithelial neoplasia

The relationship between risk of cervical neoplasia and various sociodemographic, reproductive and sexual characteristics was studied in 126 women (cases) with an abnormal Papanicolaou test finding and 1914 controls seen at the same clinic in which cases were detected but showing no evidence of cervical neoplasia. All the subjects of this study were under the age of 20. Number of sexual partners and clinical occurrence of genital warts emerged as the most important determinants of the risk of cervical intraepithelial neoplasia (CIN) in these very young women (Relative Risk (RR) for three or more sexual partners = 2.45, 95% Confidence Interval (CI) 1.63-3.70, RR for occurrence of genital warts, adjusted for number of sexual partners = 9.15, 95% CI: 5.06-16.26). Also having grown up in a 'problem' family seemed to increase the probability of developing CIN (RR adjusted for number of partners = 1.64, 95% CI 1.06-2.52). For all other investigated characteristics, cases and controls were remarkably similar. These include socioeconomic status, parity, smoking habits, use of various contraceptive methods and also indicators of sexual activity such as age of first having sexual intercourse, and duration and frequency of intercourse. Although this group of women under 20 only allows the study of less severe precursors of cervical cancer, it helps to highlight the earliest effect of purported risk factors for cervical neoplasia, chiefly sexual habits, at a time very close to their being established.