ABSTRACT
OBJECTIVE: To evaluate the accuracy of combined transvaginal ultrasound and beta-hCG discriminatory zone for diagnosing intrauterine pregnancy, abortion, and ectopic pregnancy in early abnormal pregnancy. METHODS: Initial ultrasound findings and beta-hCG level were compared with final pregnancy outcome in 164 early pregnant women who visit our hospital with vaginal bleeding or abdominal pain. The sensitivity, specificity, and predictive value by the combination of two diagnostic tools were calculated. Statistic analysis of collected data used x2 of SPSS(9.0). RESULTS: Of 90 women with normal outcome, 64(71.1%) had a gestational sac 5mm, and in 47 cases, the hCG level was above 1,800 mIU/ml. Transvaginal ultrasound was non-diagnostic in 47(28.6%) of 164 women, and especially, 43(53.1%) of 81 cases with beta-hCG levels below 1,800 mIU/ml. The portion of accurate ultrasound diagnosis was significantly higher in women above 1,800 mIU/ml (85.5%, 71 of 83 cases) compard with levels below 1,800 mIU/ml (37.0%, 30 of 81 cases) : P < 0.001; Relative Risk(RR) 2.31; CI 95%. Sensitivity of transvaginal ultrasound diagnosis of intrauterine pregnancy, abortion, and ectopic pregnancy was 90.2%, 79.3%, and 66.7% in women who presented with beta-hCG levels above 1,800 mIU/ml, and 41.0%, 23.5%, and 75% below 1,800 mIU/ml, respectively. And, negative predictive value was 83.9%, 89.7%, and 98.8% in each of intrauterine pregnancy, abortion, and ectopic pregnancy above 1,800 mIU/ml, and 64.6%, 60%, and 97.3% below 1,800 mIU/ml, respectively. CONCLUSIONS: The sensitivity, specificity, and predictive value of transvaginal ultrasound for diagnosing abnormal early pregnancy were poor except cases of ectopic pregnancy when beta-hCG levels were low than discriminatory zone. Ultrasound impressions were well related with beta-hCG levels.