Comparison of transvaginal sonography, saline infusion sonography, and hysteroscopy in premenopausal women with abnormal uterine bleeding

Abstract

Purpose Saline infusion sonography (SIS) is a relatively new technique in the evaluation of abnormal uterine bleeding. We compared the diagnostic accuracy of SIS with that of transvaginal sonography (TVS) in the detection of intracavitary abnormalities in premenopausal women with abnormal uterine bleeding. Methods In this prospective study, consecutive premenopausal women who underwent a hysteroscopy for abnormal uterine bleeding also underwent TVS and SIS. The findings at TVS and SIS were compared with the hysteroscopic and histologic findings. Sensitivity, specificity, and likelihood ratios were calculated. Receiver operating characteristic curves were constructed to assess the performance of endometrial thickness measured using TVS. Results Sixty‐two patients were included in the study. TVS demonstrated 60% sensitivity in directly visualizing intracavitary abnormalities and 93% specificity. The likelihood ratio of the presence of an intracavitary abnormality was 8, and the likelihood ratio of the absence of an intracavitary abnormality was 0.43. Defining an abnormality at TVS as direct visualization of an intracavitary abnormality or an endometrial thickness greater than 5 mm, TVS had an 85% sensitivity and a 21% specificity, with corresponding likelihood ratios of 1.1 and 0.71, respectively. For SIS, the sensitivity, specificity, and likelihood ratios of the presence and absence of intracavitary abnormalities were 88%, 95%, 10, and 0.13, respectively. Conclusions SIS is more accurate in the diagnosis of intracavitary abnormalities in premenopausal women than is TVS. An approach using endometrial thickness measurement by TVS and reserving SIS for patients who have an endometrial thickness greater than 5 mm or an intracavitary abnormality visualized by TVS would be the most effective method to reduce the number of hysteroscopies. © 2000 John Wiley & Sons, Inc. J Clin Ultrasound 28:217–223, 2000.

Accuracy of saline contrast sonohysterography in detection of endometrial polyps and submucosal leiomyomas in women of reproductive age with abnormal uterine bleeding: systematic review and meta‐analysis

“…The five studies included in the review were effectively cross-sectional without sample size calculation, each conducted at a single institution, including a total of 543 women [14][15][16][17][18] . Average patient age was 40 years, with the mean age among studies ranging from 30.6 to 44 years.…”

“…Quality assessment showed one study with a low risk of bias for all four items 18 , two studies with a high risk of bias for two of the items (description of index test and reference standard) 14,15 , while four studies had an unclear risk of bias for the description of flow and timing [14][15][16][17] . Three studies had low risk for the three items of applicability [16][17][18] .…”

“…For detection of endometrial polyps, sensitivity and specificity of 2D-SCSH ranged from 88% to 98% and from 61% to 100% among the five studies, respectively ( Figure 2). Pooled analysis gave sensitivity and specificity of 93% (95% CI, 89-96%) and 81% (95% CI, 76-86%), respectively [14][15][16][17][18] . Pooled LR+ was 5.41 (95% CI, 2.60-11.28) and LR-was 0.10 (95% CI, 0.06-0.17).…”

“…Pooled LR+ was 5.41 (95% CI, 2.60-11.28) and LR-was 0.10 (95% CI, 0.06-0.17). 17 The Netherlands 62 0 44 (36-54) Proliferative 2D-SCSH Hysteroscopy Soguktas (2012) 18 Turkey 89 NA 43.1 ± 2.9 NA 2D-SCSH Hysteroscopy…”

Submucous myomas: diagnosis and therapy

“…In comparative studies between various methods of intra-uterine diagnostic procedures, SIS appears to be more reliable than transvaginal ultrasound alone and as predictive as hysteroscopy. The reliability of transvaginal ultrasound increases with the addition of saline solution as a contrast medium [17,[33][34][35][36]. In a meta-analysis by de Kroon et al the following data were established in this respect: sensitivity 0.95 (95% CI 0.93 to 0.97), specificity 0.88 (95% CI 0.85 to 0.92), likelihood ratios 8.23 (95% CI 6.2 to 11) and 0.06 (95% CI 0.04 to 0.09) [37].…”