Gel instillation sonohysterography (GIS) and saline contrast sonohysterography (SCSH): comparison of two diagnostic techniques
Abstract
Objective
To compare gel instillation sonohysterography (GIS) with saline contrast sonohysterography (SCSH) as diagnostic methods for the evaluation of the uterine cavity.
Methods
A prospective cohort study was performed at the Department of Obstetrics and Gynecology of the VU University Medical Center, Amsterdam, between September 2007 and April 2008. We included 65 women suspected of having an intrauterine abnormality with an indication for SCSH/GIS. First SCSH and subsequently GIS were performed in all women. Distension of the uterine cavity, image quality, visualization of intrauterine abnormalities and pain experienced on a visual analog scale (VAS score) were recorded for both procedures.
Results
The mean distension with GIS was 9.0 mm and with SCSH it was 8.5 mm (P = 0.15). The mean image quality, on a scale from 0 to 5, for SCSH was 4.0 and for GIS it was 3.6 (P = 0.01). No difference was found for the visualization of intrauterine abnormalities, and the VAS scores for pain experienced on SCSH and GIS were 1.5 and 1.6, respectively (P = 0.62).
Conclusions
The image quality of SCSH is slightly better than that of GIS. This difference is likely to be attributable to the presence of air bubbles in the gel. The small difference in uterine cavity distension in favor of GIS and comparable stable distension during at least 4 min make GIS a suitable alternative for SCSH if air bubbles can be prevented. Copyright © 2010 ISUOG. Published by John Wiley & Sons, Ltd.
Introduction
Abnormal uterine bleeding is highly prevalent and an important factor in female health. Saline contrast sonohysterography (SCSH) is an appropriate technique for the detection of focally growing lesions such as polyps and fibroids1, 2. With a sensitivity of 0.95 and a specificity of 0.88 it is an accurate diagnostic tool in the evaluation of the uterine cavity in premenopausal and postmenopausal women3. The uterine cavity is distended with physiological saline, which serves as a contrast medium and enables visualization of the endometrial surface. A problem that occurs during SCSH is unstable filling and inadequate distension of the uterine cavity owing to backflow of saline. In addition, patients experience discomfort owing to fluid leakage. In trying to overcome these disadvantages, the SCSH technique can be modified by instilling gel instead of saline. In one study hydroxyethyl glycerin gel was used, showing good distension and stable filling without backflow problems4. Comparison of both procedures has not been published yet. The purpose of this study was to compare gel instillation sonohysterography (GIS) with SCSH as diagnostic methods for the evaluation of the uterine cavity, with the focus on distension of the cavity.
Methods
This study was performed at our department of obstetrics and gynecology between September 2007 and April 2008. Women attending the department and suspected of having an intrauterine abnormality were asked to participate in the study. The following were used as inclusion criteria: all premenopausal women with abnormal uterine bleeding; postmenopausal women with abnormal uterine bleeding and endometrial thickness > 4 mm; women with infertility in combination with irregular endometrium and/or endometrial thickness > 10 mm; women with a history of premature birth in combination with irregular endometrium and/or endometrial thickness > 10 mm. Because of logistic reasons inclusion was not possible on certain days, while on other days women who met the inclusion criteria were included consecutively. Exclusion criteria were risk of pelvic inflammatory disease, presence of cervical cancer, pregnancy or being premenopausal and in the luteal phase without use of contraception.
Gynecologists and sufficiently trained residents performed first SCSH and subsequently GIS with the Goldstein Sonohysterography Catheter, which is 26 cm in length and 2.4 mm in diameter (Cook Medical, Spencer, USA). Most patients were pretreated with a non-steroidal anti-inflammatory drug (500 mg naprosine 1 day and 1 h before the procedure). Before performing SCSH, sterile saline was flushed through the catheter to rid it of small amounts of air and the cervix was cleaned with povidone–iodine solution. The catheter was inserted through the cervical canal and the ‘acorn’ of the catheter was adjusted at the external cervical os to reduce leakage of instilled saline. The transvaginal transducer was introduced and real-time ultrasonographic imaging (Accuvix XQ, Medison, Korea) was performed with the simultaneous, continuous instillation of sterile saline solution (0.9% sodium chloride) into the uterine cavity. Distension was assessed by measuring the greatest distance between the anterior and posterior uterine walls in the midsagittal section. A score between 0 and 5 was given for image quality, with 0 expressing low image quality and 5 expressing high image quality with optimal visualization. The score was based on the following criteria: contrast, sharpness and brightness of the image; air bubbles and other artifacts; distension; visualization in case of an intrauterine abnormality (e.g. contrast around the abnormality, possibility of assessing the intracavitary protrusion). In case of an intrauterine abnormality, the sonographer recorded the nature (e.g. submucous fibroid, polyp, Müllerian duct anomaly or artifact). The women were asked to express the pain experienced during instillation of the fluid using a visual analog scale (VAS), which is a measure of pain intensity, with 0 equivalent to no pain and 10 equivalent to the worst pain imaginable.
Immediately after this procedure GIS was performed with the same catheter and with the use of Endosgel (Farco-Pharma, Köln, Germany) instead of saline solution. Endosgel is a sterile gel preparation, which has long been used by urologists for intraurethral instillation before cystoscopy, and contains chlorhexidine gluconate, sodium lactate, methyl hydroxybenzoate, propyl hydroxybenzoate, hydroxyethylcellulose and purified water. A syringe was connected to the base of the catheter, which was still located in the uterine cavity. A vacuum was created in the curette by withdrawing the plunger and negative pressure was applied until no fluid appeared in the catheter. A syringe was filled with Endosgel, connected to the base of the catheter and realtime ultrasonographic imaging was performed with the simultaneous instillation of gel. The instillation of gel was stopped when the patient felt slight menstrual-like cramps, backflow started or a maximum of 10 mL had been reached. The greatest distance between the anterior and posterior walls was measured, a score for the image quality and pain (VAS) during instillation of gel was given, and the presence of intrauterine abnormalities was recorded.
To evaluate the distension in time with GIS, distension curves were created. The distension in five consecutive women who were suspected of having an intrauterine abnormality was measured at different time points between 0 and 5 min after the instillation of gel and removal of the catheter.
Statistical analysis
The primary outcome was a difference in distension of the uterine cavity. A difference of 1.5 mm in distension was considered to be clinically relevant. Because of a lack of prior studies and in order to prevent an underpowered study, a standard deviation of 4.0 mm was chosen to perform a power calculation. To achieve 80% power and an alpha of 5%, 56 women would be needed to detect a difference of 1.5 mm in distension. Anticipating technical or procedural failures, we decided to include 65 patients.
The difference in distension between SCSH and GIS was evaluated with the paired samples t-test and the difference in visualization of intrauterine abnormalities was tested with the chi-square test. The Wilcoxon signed-rank test was used to evaluate the difference in quality of the ultrasound images and VAS score. All statistical analysis was performed two-sided using Statistical Packages for the Social Sciences software (SPSS, Inc., Chicago, IL, USA), and P of less than 0.05 was considered statistically significant.
Results
SCSH and GIS were performed in 65 women suspected of having an intrauterine abnormality, after they had given informed consent. Most women undergoing SCSH/GIS had complained of heavy and/or irregular menstrual bleeding (49%) or infertility (20%). The findings for SCSH and GIS are presented in Table 1. Distension after GIS was slightly higher than with SCSH (9.0 mm vs. 8.5 mm), but this difference did not reach statistical significance (P = 0.15). All relevant intrauterine abnormalities detected with SCSH were also identified using GIS (P = 1.00): 14 fibroids (22%), nine polyps (14%) and two Müllerian duct anomalies (3%). Four (6%) artifacts (not air bubbles) were seen with GIS, while three (5%) were seen with SCSH (P = 0.72). The mean image quality for SCSH and GIS was 4.0 and 3.6, respectively, on a scale from 0 to 5 (P = 0.01). The pain experienced was comparable for SCSH and GIS, with mean VAS scores of 1.5 and 1.6, respectively (P = 0.62). Distension curves were created for five women, and show that the distension was stable during at least 4–5 min (Figure 1).
Distension of the uterine cavity after instillation of gel and removal of the catheter in five different women.
| Parameter | SCSH | GIS | P |
|---|---|---|---|
| Distension of uterine cavity (mm, mean ± SD) (95% CI) | 8.5 ± 4.2 (7.4–9.6) | 9.0 ± 4.6 (7.8–10.1) | 0.15 |
| Women with uterine cavity abnormality (n (%)) | 21 (32) | 21 (32) | 1.00 |
| Women with submucous fibroid (n (%)) | 14 (22) | 14 (22) | 1.00 |
| Women with polyp (n (%)) | 9 (14) | 9 (14) | 1.00 |
| Women with Müllerian duct anomaly (n (%)) | 2 (3) | 2 (3) | 1.00 |
| Women with artifact (n (%)) | 3 (5) | 4 (6) | 0.72 |
| Image quality (mean ± SD) (95% CI)* | 4.0 ± 1.1 (3.7–4.3) | 3.6 ± 1.3 (3.3–3.9) | 0.01 |
| VAS score for pain (mean ± SD) (95% CI)† | 1.5 ± 1.7 (1.0–1.9) | 1.6 ± 2.1 (1.1–2.2) | 0.62 |
- * Quality assessed on a range from 0 (low quality) to 5 (high quality).
- † Pain assessed on a visual analog scale (0–10).
Discussion
Distension of the uterine cavity was 6% greater when using GIS than when using SCSH but this difference did not reach statistical significance. We expected more distension with GIS owing to the higher viscosity of the gel, but menstrual-like cramps and backflow may have limited the instilled gel volume. The distension of the cavity depends on the instilled volume, the instillation rate, the time of instillation, the backflow rate and the intracavitary pressure. Although instillation of saline solution and gel, and therefore distension, are partly operator dependent, we tried to standardize both procedures. Instillation of saline solution was always performed with the same pressure (80 mmHg) and instillation of gel was stopped when the patient felt slight menstrual-like cramps, backflow started or a maximum of 10 mL had been reached. Complete standardization of the two methods would have been impossible because of the difference in viscosity between the two fluids and the different instillation techniques used in clinical practice and therefore in this study (e.g. instillation of saline solution was with a continuous flow). The distension per milliliter is expected to be higher with gel since the instillation of saline solution was continuous, but we did not quantify this in the current study. The distension may also have been influenced by the study design, since SCSH was always performed first and we did not succeed in all cases in removing the saline solution from the cavity completely. Nevertheless, it seems likely that the remainder of the saline solution was expelled by the gel through the cervix or Fallopian tubes.
It would have been appropriate to report the proportion of cases reaching a certain threshold volume using each technique, but the disadvantage with this approach is that the appropriate threshold volume is different for every woman and this would be difficult to assess. Our previous experience has suggested that greater distension results in better image quality and we therefore decided to compare the maximum uterine distension observed during each of the two procedures.
The image quality for SCSH was better than that for GIS. It must be noted that the score was a subjective impression, and therefore observer bias cannot be excluded. An important reason for the difference in quality is the presence of air bubbles in the gel. The currently used GIS technique may not be optimal in terms of avoiding air bubbles, because the syringe was filled manually with Endosgel. In addition, the catheter was not removed from the uterine cavity between SCSH and GIS in order to reduce inconvenience for the patient, although filling the catheter with gel before entering the uterine cavity may have reduced the occurrence of air bubbles. Another hypothesis is that the presence of air bubbles may have been promoted by rising temperature, negative pressure while filling the syringe, turbulence during gel instillation and dissolved molecules.
For better performance of GIS, commercial products are available containing gel with a smaller amount of dissolved molecules, no need to transfer the gel manually from one syringe to another and with a special cervical applicator preventing cervical backflow or leakage. It may be questioned whether a difference of 10% in image quality (4.0 vs. 3.6 on a scale from 0 to 5) should be considered clinically relevant, since intrauterine abnormalities could be visualized by SCSH as well as GIS in all cases. On the other hand, performing GIS after SCSH in all women may have caused review bias, because visualization of an abnormality with the first procedure makes it easier to recognize the same abnormality using the second procedure. Although it would have been appropriate to use a cross-over design, it appeared that it would be impossible to clear all the gel from the uterus and perform SCSH immediately afterwards. Alternatively, we could have chosen to perform GIS and SCSH at different times by different examiners, blinded to the results of the other test. However, we did not use this design for two reasons: inconvenience to the patient at having to undergo the same procedure twice; and interobserver variation. Future studies may overcome this problem by recording both SCSH and GIS for retrospective random assessment by independent examiners blinded to the results of the tests.
The mean VAS score for pain during fluid instillation was low and practically the same for both SCSH and GIS, which demonstrates that both procedures are well tolerated. In our experience, the introduction of the speculum and catheter causes more discomfort than the instillation of saline solution or gel. Because SCSH was always performed first, the pain score of SCSH may have been influenced by insertion of the catheter. Therefore, a cross-over design would have been more appropriate.
GIS is an easy diagnostic method to perform and the higher viscosity of the gel causes less fluid leakage. Although backflow could not be prevented completely, Figure 1 demonstrates that GIS gives stable distension during at least 4–5 min, which is sufficient to complete the ultrasound evaluation. Small fluctuations of the distension in time may be caused by contractions of the uterine muscle.
The introduction of three-dimensional (3D) ultrasound enhances visualization of the uterine cavity, and improves diagnostic accuracy of SCSH or GIS5. Gel is considered to be advantageous when performing 3D ultrasound. More stable filling of the uterine cavity without continuous flow of fluid allows extension of the observation time, resulting in better 3D sonohysterographic images with fewer artifacts and moving tissue elements4.
We have noticed that some women experience vasovagal cramps after GIS. To prevent this discomfort, the amount of gel should be as small as possible. With an average of 4 mL, an optimal distension of the uterine cavity can be achieved4.
In conclusion, a small but significant difference in image quality between GIS and SCSH was found in favor of SCSH, which may be caused by the presence of air bubbles while performing GIS. This disadvantage of GIS may be less prominent if another technique is used (e.g. use of commercially available prefilled gel syringes). With a comparable low level of experienced pain, the slightly higher distension with GIS makes it a suitable alternative to SCSH if air bubbles can be prevented. Moreover, a beneficial effect of GIS is expected for 3D ultrasonography and it must be questioned if GIS should replace SCSH when using two-dimensional (2D) imaging. Additional prospective studies should be performed to determine the diagnostic accuracy of 2D and 3D GIS in comparison with SCSH, in which both procedures are assessed independently and preferably blinded.
