Disclosures: Dr. Van den Bosch said he had no financial conflicts of interest. Dr. Emanuel reported being a stockholder and receiving royalties from Giskit BV.
HAMBURG, GERMANY — Diagnostic accuracy tends to be higher when the uterine cavity is distended with gel rather than saline during sonohysterography in women with abnormal bleeding, research results showed.
Use of gel instillation sonohysterography (GIS) also may resolve cervical leakage during saline infusion sonohysterography (SIS), thereby making the procedure easier on both the patient and the clinician.
“We had a large need for towels and even sometimes boots,” Dr. Mark Emanuel said at the 19th World Congress on Ultrasound in Obstetrics and Gynecology.
He presented data from the first randomized trial to compare the two techniques, demonstrating a diagnostic accuracy of 89% for GIS and 73% for SIS in 103 women with abnormal uterine bleeding and an abnormal or inconclusive transvaginal ultrasound. The between-group difference failed to reach significance, but statistical calculations showed it would have if an additional 50 patients were added to the cohort. The sensitivity for GIS was 97% vs. 85% for SIS, and the specificity was 82% vs. 77%.
Gel instillation cuts investigation time and can be performed by one clinician, whereas SIS requires two persons to help refill the uterine cavity with saline, said Dr. Emanuel, a gynecologist at Spaarne Hospital, Hoofddorp, the Netherlands.
Cavity distention was up to 3 minutes in most cases after approximately 4 mL of gel were placed in the cervix. That compares with about 10-15 minutes with saline, which had to be reinfused through a catheter in the uterine cavity after about 20 seconds.
“It saves an enormous amount of time,” Dr. Emanuel said in an interview. “To instill a few milliliters of gel is much easier and can be done by almost anyone who can perform transvaginal scanning.”
Scores on visual analog scales measuring GIS and SIS inconvenience with 0 being very inconvenient were not significantly different, however, among clinicians (7.9 vs. 6.4) or patients (7.6 vs. 7.8).
Dr. Emanuel said gel instillation is a safe alternative to SIS and provides clinicians with a more stable image. “If you want to do 3-D scanning it's mandatory that you use gel,” he said. “For 3-D scanning, you don't want any moving fragments because they create artifacts. Saline infusion with 3-D is very difficult.”
During the same session, Dr. Thierry Van den Bosch, who has worked extensively with gel instillation, said it produces lower failure rates and at least similar diagnostic accuracy. In an observational study of two consecutive cohorts of 402 women, the technical failure rate for GIS was 1.9%, compared with 5% for SIS. Failure due to inadequate uterine cavity distention was 0.3% vs. 1.5%.
The sensitivity for GIS was 85% vs. 77.8% for SIS, and the specificity was 78.1% vs. 80.7%. Negative predictive values were “comparable” at 88.6% for GIS and 79.1% for SIS, said Dr. Van den Bosch of the University Hospitals, Catholic University of Leuven (Belgium).
He also presented data from a study in which six gynecologists were asked to predict intracavity uterine pathology in 111 consecutive women from the hospital's bleeding clinic based on 3-D volumes, including unenhanced ultrasound as well as GIS with and without power Doppler. The only clinical information they had was patients' age (mean 50.4 years) and menopausal status (63% were premenopausal).
Although the examiners were able to accurately differentiate normal from abnormal endometrium in the majority of cases, interobserver variability for these “off-line” assessments was substantial. The best agreement between three examiners was a Schouten kappa of 0.69, with an overall kappa of 0.51.
“Certainly interobserver variability should be improved, and I think it illustrates how difficult it is to give your opinion [when you are] not seeing the patient or doing the ultrasound yourself,” said Dr. Van den Bosch. “Perhaps if we had more strict [diagnostic sonographic] criteria we would have done better.”
The first experience with GIS was reported in 2007 (Fertil. Steril. 2007;87:152-5), but concerns were raised at the time about the safety of the gels (Fertil. Steril. 2007;88:536-7).
Dr. Van den Bosch said a study he presented at last year's congress showed that the addition of lidocaine to the gel did not alter patients' pain and should not be used because of the potential for anaphylactic reactions.
It's been suggested that gels containing chlorhexidine may reduce the risk of infection, but concerns have been raised about embryotoxicity in younger women.
An audience member questioned whether pregnancy has been an issue after gel exposure. Several patients have gotten pregnant without difficulty, said Dr. Van den Bosch.