Résumé
This study was conducted prospectively from June 1992 to September 1993, we evaluated 115 patients attending our infertility clinic for the first time to determine the correlation between abdominal ultrasound and laparoscopy in the detection of pelvic pathology in the initial workup of subfertile women, depicting laparoscopy as the gold standard. Normal ultrasound scans were reported in 80 women (69.56% of total) 58 of whom had also normal pelvic findings at laparoscopy. The remaining 22 patients with normal ultrasound scans had abnormal laparoscopic findings in the pelvis giving a false negative rate of 27.50 per cent. These were primarily related to; endometriosis with adhesions (10), filmy adnexal adhesions (5), omental and bowl adhesions (5), and hydrosalpinx (2). Thirty five women (30.44% of the total) had abnormal scans. Pelvic pathologies were confirmed in 30 at laparoscopy. The remaining five had a normal pelvis (three retroverted uteri, and two functional ovarian cyst had disappeared at time of laparoscopy) giving a false positive rate of 14.29 per cent for ultrasound scans. Statistical analysis revealed that abdominal ultrasound scan in the detection of pelvic pathology in the initial workup of subfertile women has a sensitivity of 57.69 per cent, specificity of 92.06 per cent, positive predictive value of 85.71 per cent with negative predictive value of 72.50 per cent and accuracy rate of 76.52 per cent. We conclude that the use of abdominal ultrasound in the detection of pelvic pathology in the initial workup of subfertile women is not an appropriate routine screening test as it has low efficiency in detecting the pelvic pathologies many of which are the causes of infertility.(ABSTRACT TRUNCATED AT 250 WORDS)