Wm. Ankum et al., TRANSVAGINAL SONOGRAPHY AND HUMAN CHORIONIC-GONADOTROPIN MEASUREMENTSIN SUSPECTED ECTOPIC PREGNANCY - A DETAILED ANALYSIS OF A DIAGNOSTIC-APPROACH, Human reproduction, 8(8), 1993, pp. 1307-1311
In this prospective study among 208 high-risk patients with suspected
ectopic pregnancy, the diagnostic value of transvaginal sonography and
serum human chorionic gonadotrophin (HCG) measurements were analysed
in detail. The absence of an intra-uterine gestational sac obviously w
as the most constant sonographic finding among patients with ectopic p
regnancy (n = 89), with a very high sensitivity (0.99) but a low speci
ficity (0.41). The application of different HCG cut-off levels improve
d specificity to 1.00 for values exceeding 4500 IU/l. Clinical utility
obviously decreased, as many patients presented with HCG values well
below this level. The additional effect of adnexal findings was analys
ed. Sonographic identification of an ectopic pregnancy was very specif
ic (0.99) but had low sensitivity (0.56) because many ectopics were no
t detected. The additional effect of HCG values on these results was m
inor. The low HCG cut-off levels advocated in recent studies are quest
ioned by the results of our analysis: whereas the combined use of sono
graphy and HCG measurements is shown to be of great benefit, the limit
ations are also documented, underlining the need for reevaluation at i
ntervals of patients with low HCG values. The question of which cut-of
f level should be used in practice, however, hinges on a difficult cho
ice between a certain specificity and clinical utility.