ABSTRACT
OBJECTIVE: Because the presence of echogenic fluid on transvaginal sonography has been shown to correlate well with hemoperitoneum in patients with possible ectopic pregnancy, the aim of this study was to compare echogenic fluid on sonography with the results of culdocentesis in predicting hemoperitoneum. MATERIALS AND METHODS: Free fluid on transvaginal sonography and the results of culdocentesis were correlated with the presence or absence of hemoperitoneum in 46 patients at surgery. Forty ectopic pregnancies and six nonectopic pregnancies were found. Echogenic fluid was the criterion used to establish hemoperitoneum on sonography. For statistical analysis, negative and nondiagnostic culdocentesis results were combined. The sensitivity, specificity, and positive and negative predictive values of each diagnostic technique were compared. RESULTS: In 40 of 46 patients with ectopic pregnancy, the sensitivity and specificity of echogenic fluid for establishing hemoperitoneum were 100% and 100%, respectively, compared with 66% and 80%, respectively, for culdocentesis. More important, the negative predictive value of a nondiagnostic culdocentesis was 25% compared with 100% for echogenic fluid in the ectopic subgroup of patients. In two patients with incomplete abortions, sonography failed to detect small amounts of hemoperitoneum at surgery performed 4 hr and 7 days after sonography. CONCLUSION: Sonography is more sensitive than culdocentesis in the detection of hemoperitoneum. Culdocentesis is invasive, and nondiagnostic results cannot be used to exclude hemoperitoneum. Culdocentesis should play no role in the evaluation of ectopic pregnancy except in the unusual circumstance in which high-resolution sonography cannot be readily performed.
