Detection of cesarean scars by transvaginal ultrasound

OBJECTIVE: To assess the ability of transvaginal ultrasound to detect cesarean scars and their defects in the nonpregnant state. METHODS: Asymptomatic, parous volunteers underwent transvaginal ultrasound of the cervix, uterus, and adnexa. Uterine measurements, the presence or absence of a cesarean scar, and the presence of a scar defect, defined as fluid within the scar, were recorded. All subjects completed a self-report questionnaire regarding obstetric history. Sonographers and investigators were blinded to subject history. RESULTS: A total of 70 subjects were enrolled. Of these, 38 women had a prior vaginal delivery and 32 women a prior cesarean delivery. One woman with a bicornuate uterus and three cesarean deliveries was excluded from data analysis. Real-time transvaginal ultrasound proved 100% sensitive (exact 95% confidence interval [CI] 88.8, 100) and 100% specific (exact 95% CI 90.7, 100). Stored image review had a sensitivity of 87% (exact 95% CI 70.2, 96.4) and a specificity of 100% (exact 95% CI 90.7, 100). Fluid was visualized within the scars of 13 of 31 subjects (42%) with a prior cesarean delivery. All 13 were found among the 23 subjects (56%) who had labored prior to cesarean delivery. Moreover, women with cesarean scar defects had a greater number of cesarean deliveries (P < .04) than women without scar defects. CONCLUSION: Transvaginal ultrasound is highly accurate in detecting cesarean hysterotomy scars. Cesarean scar defect, defined by the presence of fluid within the incision site, was more common when labor preceded cesarean delivery and with multiple cesarean deliveries.