Multiple Regression Analysis of Ultrasound and Clinical Features for Quantitative Evaluation of Tubal Pregnancy Rupture.

Abstract

The aim is to search for quantitative indicators of ultrasound and clinical features that suggest tubal pregnancy rupture, and to identify independent risk factors for tubal pregnancy rupture through multiple regression analysis. Retrospective analysis of 166 cases of tubal pregnancy was confirmed by laparoscopy, including 97 cases of unruptured type and 69 cases of ruptured type. Compare the ultrasound and clinical features of the 2 groups to identify quantitative indicators of tubal pregnancy rupture. Ultrasound features include: uterine position (anterior or posterior) and endometrial thickness, location of ectopic pregnancy (right or left), size, morphology, internal echoes, boundaries, and pelvic fluid accumulation. Clinical features include: age, number of days of menopause, abdominal pain and vaginal bleeding, intrauterine device, history of ectopic pregnancy and pelvic inflammatory disease, number of pregnancies and miscarriages, and preoperative β-human chorionic gonadotropin (β-HCG) value. Establish receiver operating characteristic (ROC) curves to determine the diagnostic efficacy of various ultrasound and clinical features for tubal pregnancy rupture and the optimal threshold for predicting the cause of rupture. Through multiple logistic regression analysis, identify the risk factors for tubal pregnancy rupture. The ultrasound features of unruptured tubal pregnancy (UNRTP) were: in 20 out of 97 cases, a mixed echo like a gestational sac could be seen in the attachment area, with clear boundaries and partial presence of yolk sac and embryo inside. In 50 out of 97 cases, there was not much pelvic fluid accumulation (<15 mm by ultrasound). Ruptured tubal pregnancy (RTP) ultrasound features were: large mixed echo mass in the attachment area, without obvious boundaries, with chaotic internal echoes, and a large amount of pelvic fluid accumulation. Univariate analysis showed that there was no difference in terms of uterine position, endometrial thickness, and mass location between the 2 groups (P > .05), but the RTP group had a larger maximum mass diameter, unclear boundaries, and more pelvic fluid accumulation (P < .05). Clinical characteristics: There was no difference in terms of age, vaginal bleeding, intrauterine device, history of ectopic pregnancy, number of pregnancies, history of miscarriage, and surgical methods (P > .05), but in the RTP group, there were more cases of abdominal pain, pelvic inflammatory disease, high preoperative β-HCG (P < .05). The ROC curve showed that the maximum diameter of the mass, unclear boundaries, pelvic fluid accumulation, abdominal pain, preoperative β-HCG, AUC are 0.741, 0.726, 0.752, 0.897, 0.585, 0.631 (all P < .05), which could be used to evaluate tubal pregnancy rupture. If the AUC of pelvic inflammatory disease is 0.585 (P > .05), it could not be used to evaluate tubal rupture. The cut-off values showed that the maximum diameter of the mass was >36.5 mm, the pelvic fluid volume measured by ultrasound was >34.5 mm, preoperative HCG > 3094.5 U/L, indicating the possibility of tube pregnancy rupture. Multiple logistic regression analysis showed that the accumulation of pelvic fluid measured by ultrasound and preoperative β-HCG were independent risk factors for tubal rupture (P < .05). The accumulation of pelvic fluid measured by ultrasound and preoperative β-HCG was independent risk factors for ruptured tubal pregnancy.