Predictive value of sonographic examination to visualize retained placentadirectly after birth at 16 to 28 weeks

A prospective study was performed to assess the predictive value of an ultr asonographic examination directly after a spontaneous birth at 16 to 28 wee ks' gestation to exclude the possibility of retained placental tissue. The aim of this procedure is to prevent routine curettage, which can induce Ash erman's syndrome, uterine perforation, and anesthetic complications. Over a 2 year period the clinical course in 64 women, who had been delivered of t heir infants at 16 to 28 weeks' gestation, was followed through 6 weeks pos t partum. Sonographic examination was performed within 30 min after deliver y of the placenta independent of macroscopic judgment of completeness of pl acenta. The examination was classified into three categories ( with subsequ ent clinical interpretation): sharp lining of echogenic uterine wall with t ranslucent cavity (uterine cavity containing fluid blood), sharp lining of the wall with echogenic area in cavity not continuous with the wall (uterin e cavity with blood clot), and irregular lining with echogenic area continu ous with the uterine wall and extending into the cavity (uterine cavity con taining retained placental tissue). Women with sharp uterine lining without (n = 32) or with (n = 7) echogenicity in the cavity had no direct operativ e removal of placental tissue; 3 underwent curettage at a later stage (17, 18, and 34 days, respectively). A direct digital removal of placenta or cur ettage was performed on 25 women who revealed echogenicity continuous with the uterine wall. The 25 of 28 operatively obtained tissues were examined m icroscopically for trophoblasts. The sensitivity of the sonographic examina tion to find retained placental tissue was 85% (17 of 20) at 95% confidence intervals of 62 to 97%, the specificity was 88% (36 of 41) at 95% confiden ce intervals of 74 to 96%, and there were 25% (5 of 20) false positive judg ments and 88% (3 of 39) false negative judgments. The positive predictive v alue of ultrasonography to find retained placenta of 68% (17 of 22) at 95% confidence interval of 55 to 92% combined with the negative predictive valu e of 92% (36 of 39) is sufficient to strongly suggest that curettage should not be performed routinely in these pregnancies at high risk for retained placental tissue.