Comparative review of diagnostic accuracy in tubal pregnancy: a 14-year survey of 1040 cases.

This paper summarizes our experience from 1972-1985 with the changing tools used in the diagnosis of ectopic pregnancy. The study was divided into three time periods based on the diagnostic methods used in each. In the first period (1972-1975), diagnosis was by culdocentesis and the urine pregnancy test. In the second period (1976-1978), diagnostic laparoscopy was added. Finally, in the third period (1979-1985), serum beta-human chorionic gonadotropin (beta-hCG) assay and pelvic ultrasonography were added. A rising trend in diagnostic accuracy was observed from the first period to the third. The predictability and the negative predictive value of diagnostic tests in the third period were 96.3 and 94.9%, respectively--significantly higher than in the first and second periods. The incidence of unruptured tubal pregnancy increased significantly, from 27.6% in the first and the second periods to 42.5% in the third period. Laparoscopy, ultrasonography, and serum beta-hCG assay each revealed a positive rate of more than 90% in the diagnosis of tubal pregnancy. Laparoscopy had the highest positive predictive value as a single test in the third period. Among suspected cases, however, laparoscopy was found to have a low accuracy in discriminating between tubal pregnancy and infection. The combination of serum beta-hCG testing and ultrasonography produced a 99% positive predictive value, and revealed a higher degree of accuracy in the diagnosis of tubal pregnancy than any other single test or combination of tests.