Prediction of vesicouterine adhesions by transvaginal sonographic sliding sign technique: validation study.

We conducted a prospective observational double‐blind diagnostic test accuracy study at the Amsterdam University Medical Center (UMC), Amsterdam, The Netherlands. Ethical approval of the study was provided by the ethical board of the Amsterdam UMC with numbers 2018.099 and W20_274. The trial was registered with ClinicalTrials.gov ( NCT05862155 ). The study was conducted and reported in accordance with STARD guidelines (Appendix  S1 ) 17 . Patients who visited our outpatient clinic between January 2020 and December 2022 and were scheduled for gynecological laparoscopy for a benign disorder were included consecutively. All patients were 18 years or older and gave signed informed consent. Indications for scheduled laparoscopy were diagnostic laparoscopy, uterine niche resection, hysterectomy and myomectomy. In the preoperative analysis, all patients underwent TVS. Patients were excluded if a potential adhesion‐provoking event occurred between the index test (TVS) and the reference standard (surgery) (e.g. other surgery or infection). Patients were also excluded if videoclips were either unavailable or of insufficient quality (e.g. no midsagittal plane, limited movements induced by the sonographer) or if the uterus was not visible continuously during the dynamic examination. Five experienced gynecological sonologists performed all real‐time dynamic transvaginal sliding sign ultrasound examinations at a dedicated tertiary referral outpatient clinic as part of routine gynecological TVS, using the technique described by Hudelist et al . 14 . The sliding bladder sign was recorded as a uniform videoclip of 5–20 s, with the vaginal probe placed in the anterior vaginal fornix and the examiner's free hand on the patient's lower abdomen. Repeated mild pressure was applied on the uterus with the vaginal probe and/or with the free hand on the fundus of the uterus while scanning in the midsagittal plane. The videoclips were saved in the patient record. The sliding bladder sign is considered positive if the anterior uterine wall glides freely in relation to the bladder. The sliding bladder sign is documented as negative if the bladder appears fixed to the uterus over more than 1 cm between the vesicovaginal and the vesicouterine fold 13 (Figure  1 , Videoclip S1 ). Transvaginal sonographic image showing anterior compartment between uterus and bladder in longitudinal section. During assessment of sliding bladder sign, mild pressure is applied on uterus ( ) with vaginal probe and/or examiner's free hand on the lower abdomen to determine whether the anterior uterine wall glides freely in relation to the bladder ( ) between the vesicovaginal and vesicouterine folds ( ) or appears to be fixed. Offline assessment was performed by two experienced gynecologists (R.A.d.L. and N.B.B.), and one resident in obstetrics and gynecology (J.W.T.R.). All were part of our dedicated tertiary outpatient clinic for at least 6 months and could be classified as Level 3 (R.A.d.L. and N.B.B.) or Level 2 (J.W.T.R.) according to the European Federation of Societies for Ultrasound in Medicine and Biology (EFSUMB) criteria 18 . All assessors evaluated all of the records for the presence of the sliding bladder sign independent from each other and in a random order. The offline assessors were blinded to the patient's demographic characteristics, medical or surgical history and laparoscopic findings. Laparoscopic surgery was performed < 9 months after the ultrasound assessment. Videoclips of the surgery were assessed by an experienced gynecologist (R.A.d.L.) in a random order. The assessor was blinded to medical or surgical history and to the findings during TVS. Vesicouterine adhesions were scored according to an adjusted classification of Tulandi and Lyell 19 , which distinguishes between the size and density of adhesions (Table  1 , Figure  2 ). During laparoscopic assessment of the adhesions, the size of the adhesions was determined subjectively by comparison with the width of the uterus. We divided the adhesion scores into three groups: no adhesions (score = 0), mild adhesions (score > 0 to < 16) and severe adhesions (score ≥ 16); the latter we also called a ‘fixed bladder’. Adjusted classification * of intra‐abdominal adhesions according to size of adhesion Score = 0, no adhesions; score > 0 to < 16, mild adhesions; score ≥ 16, severe adhesions. Maximum score is 48. Adapted from Tulandi and Lyell 19 . Laparoscopic images showing categorization of vesicouterine adhesions. (a) No adhesions between uterus and bladder. (b) Mild adhesions: filmy adhesions between uterus and bladder. (c) Severe adhesions: firm and dense adhesions between uterus, bladder and abdominal wall. TVS was the index test in the assessment of the sliding bladder sign for the presence of vesicouterine adhesions. The sliding bladder sign was recorded as positive or negative. The findings during laparoscopy recordings, categorized into absent, mild or severe adhesions, based on the adjusted classification of Tulandi and Lyell 19 , were used as the reference standard. Demographic findings, symptoms, and gynecological and surgical history were registered at baseline. Variables of the patients that could influence the quality of the ultrasound images or the a‐priori risk of having vesicouterine adhesions were also registered. These included a full or empty bladder and the position of the uterus on the ultrasound image. Uterine features were assessed in accordance with Morphological Uterus Sonographic Assessment (MUSA) 20 , International Federation of Gynecology and Obstetrics (FIGO) 21 and Delphi consensus 22 guidelines. Data were analyzed using IBM SPSS Statistics version 28 (Armonk, NY, USA). The diagnostic accuracy of the sliding bladder sign in predicting mild or severe adhesions between the bladder and uterus was determined in terms of sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV), with 95% CI. A negative sliding bladder sign is considered a positive result of the diagnostic test and refers to the absence of sliding of the bladder with the sliding sign technique. The sensitivity, specificity, PPV and NPV of the sliding bladder sign were analyzed in different subgroups to determine potential influencing factors. A P ‐value < 0.05 was considered statistically significant. Interobserver variability was assessed between two experienced gynecologists (R.A.d.L. and N.B.B.), and between one experienced gynecologist (R.A.d.L.) and the resident (J.W.T.R.). To evaluate the intraobserver variability, one of the experienced gynecologists (R.A.d.L.) re‐evaluated the same ultrasound scans 2 months after the first ultrasound evaluation. Cohen's kappa was calculated for the agreement. A kappa value of < 0.20 indicates poor agreement, 0.21–0.40 indicates fair agreement, 0.41–0.60 moderate agreement, 0.61–0.80 substantial to good agreement and 0.81–1.00 almost perfect agreement 23 . To determine the clinical relevance of the sliding bladder sign, a minimal predictive value of 0.60 was predetermined in patients with a relatively high risk for adhesions as null hypothesis. Sensitivity and specificity values lower than 0.60 are not clinically relevant, and using an expected higher predictive value would only result in a smaller sample size. Therefore, we agreed on a minimal predetermined predictive value for a larger sample size. With an expected sensitivity/specificity of 0.85, a significance level of 0.05 and a power of at least 0.80, the sample size was 87 participants 24 , 25 . We anticipated that 15% would be excluded because of insufficient quality of the ultrasound or laparoscopy videos. Therefore, we calculated a sample size of 115 women.

Between January 2020 and December 2022, 146 women visited our outpatient clinic preoperatively and were scheduled for laparoscopic surgery. A total of 30 women were excluded, four because they did not meet the inclusion criteria, 20 for insufficient ultrasound quality, as predefined, and six because of unavailable laparoscopic footage for offline assessment (Figure  3 ). The remaining 116 women were included in the study. The mean age was 37.2 years, with a mean body mass index (BMI) of 27.0 kg/m 2 (Table  2 ). In total, 87 women had a history of abdominopelvic surgery and/or pelvic infection. Flowchart showing process of patient selection and assessment of sliding bladder sign. TVS, transvaginal ultrasound. Demographic and transvaginal ultrasound characteristics of study population ( n  = 116) Data are given as mean ± SD, n (%) or n / N (%). Data on uterine position on ultrasound scan were missing in one patient. ≤ 45°. Straight position, in between anteverted and retroverted. BMI, body mass index. All 116 women included received a TVS scan before their scheduled surgery. In all women, the sliding bladder sign was evaluated. The time between TVS and surgery varied, but all ultrasound examinations were performed < 9 months before surgery. In 49.1% ( n  = 57) of women with available laparoscopic footage, sliding of the bladder was reported to be absent during offline assessment (Table  2 ). The position of the uterus was anteverted in 65.2% of women. Fibroids were found in 23.3% of women and a uterine niche was present in 64.7%. In 88.8% of women, the fundus was visible on screen during the entire ultrasound scan, and 46.6% had a full bladder during the examination. All patients underwent laparoscopic surgery for various benign indications. Vesicouterine adhesions were found in 68.1% ( n  = 79) of patients. Among the patients with vesicouterine adhesions, 64.6% ( n  = 51) had mild adhesions with an adjusted score between > 0 and < 16 and 35.4% ( n  = 28) had severe adhesions with a score ≥ 16. In 31.9% ( n  = 37) of all participants, no adhesions were found. Table  3 presents the diagnostic accuracy of the TVS negative sliding bladder sign for predicting vesicouterine adhesions of different severities on laparoscopy. Absent sliding of the bladder (negative sliding bladder sign) on ultrasound examination had a PPV of 94.7% (95% CI, 88.9–100%) for predicting the presence of any vesicouterine adhesions on laparoscopy. When there were no adhesions seen on laparoscopy, the bladder was sliding on ultrasound examination with a specificity of 91.9% (95% CI, 83.1–100%). Diagnostic performance of negative transvaginal sliding bladder sign for prediction of vesicouterine adhesions during laparoscopy Data are given as % (95% CI). NPV, negative predictive value; PPV, positive predictive value. For the presence of severe adhesions on laparoscopy, the sensitivity of the negative sliding bladder sign was 89.3% (95% CI, 77.8–100%). When the bladder was sliding on ultrasound examination, the sliding bladder sign had a NPV of 94.9% (95% CI, 89.3–100%) for the presence of severe adhesions (Table  3 ). Comparing the two experienced gynecologists (R.A.d.L. and N.B.B.), we found an interobserver agreement of 0.78 (substantial to good) using Cohen's kappa ( P  < 0.001) 23 . The interobserver agreement between the experienced gynecologist (R.A.d.L.) and the resident (J.W.T.R.) was lower, but there was still good agreement (Cohen's kappa = 0.67; P  < 0.001). After 2 months, the experienced gynecologist (R.A.d.L.) interpreted the ultrasound videos of all 116 women again and a kappa of 0.69 ( P  < 0.001) was found for intraobserver agreement. Apart from the primary outcome, we found several factors that were of influence on the predictive value of the sliding bladder sign. The most influential factors were the position of the uterus, a full bladder during ultrasound examination, a Cesarean scar defect or fibroids and BMI. However, sample sizes of these subgroups were small (Table  S1 ). The biggest increase in specificity and PPV for predicting the presence of adhesions was seen when TVS was performed with a relatively empty bladder ( n  = 62) (specificity, 95.8% (95% CI, 87.8–100%); PPV, 96.0% (95% CI, 88.3–100%)) and when the patient had a BMI < 25 kg/m 2 ( n  = 38) (specificity, 100% (95% CI, 100–100%); PPV, 100% (95% CI, 100–100%)). The NPV increased (65.1% (95% CI, 50.9–79.4%)) when the uterus was in the anteverted position ( n  = 75). The diagnostic accuracy of the sliding bladder sign for predicting the presence of severe vesicouterine adhesions was higher when we excluded patients with a uterine niche, indicated by higher specificity (94.9% (95% CI, 87.9–100%)) and PPV (50.0% (95% CI, 1.0–99.0%)) ( n  = 41).

The TVS sliding bladder sign technique is a non‐invasive, well‐tolerated and accurate method to detect postoperative vesicouterine adhesions. This study shows that, when the sliding bladder sign is negative, the chances of severe vesicouterine adhesions are very high, with a sensitivity of 89.3%. When the bladder is sliding, the chances are very high that there are no vesicouterine adhesions present, with a specificity of 91.9%. Mild vesicouterine adhesions are difficult to predict. A positive sliding bladder sign rules out severe adhesions, but mild adhesions can still be present. However, the clinical relevance of mild adhesions can be debated. We have shown that the interobserver variability is sufficient to good, even among observers with different experience levels. Most discrepancies in interpretation occurred in scans of women with a full bladder or a high BMI. These findings indicate that the reasonably low intra‐ and interobserver agreement can be improved when patients are scanned with an empty bladder. Also, we need to be aware that the assessment of the sliding bladder sign can be less accurate in patients with a high BMI. Further study of false‐negative and false‐positive results would be interesting to optimize and equalize our technique and evaluation skills for the assessment of vesicouterine adhesions. Many studies have investigated the role of TVS in diagnosing deep endometriosis 26 , including the validation of the predictive value of the sliding sign for the obliteration of the pouch of Douglas, with Reid et al . reporting a sensitivity of 83.3%, specificity of 97.1%, PPV of 92.6% and NPV of 93.2% 16 . As a result, the International Deep Endometriosis Analysis (IDEA) group has published a systematic approach to ultrasound examinations in patients with suspected endometriosis 27 . Use of the ultrasound‐based endometriosis staging system (UBESS) 28 has been proposed and, most recently, Leonardi and Condous have introduced a more practical approach to performing ultrasound examinations to evaluate patients with suspected endometriosis 29 . Thanks to these developments, the efficacy of the sliding sign in predicting deep‐endometriosis‐related adhesions is now indisputable. A few studies have shown that the sliding sign of the uterus can predict the presence of intra‐abdominal adhesions in pregnant women with previous Cesarean delivery, performed by transabdominal evaluation in the third trimester 13 , 30 , 31 . Only Ayachi et al . looked at the prediction of pelvic adhesions using the TVS sliding sign in women with previous abdominopelvic surgery. They found a sensitivity of 96.7% and specificity of 70.6% for predicting adhesions in the vesicouterine pouch 12 . However, they assessed vesicouterine adhesions as part of pelvic adhesions in a subgroup analysis, with a small sample size of 15 women. In addition, neither inter‐ nor intraobserver variability was analyzed. To our knowledge, this is the first study to validate the TVS sliding bladder sign with laparoscopy as the reference standard, including inter‐ and intraobserver variability analysis. We succeeded in blinding ultrasound interpretation from the laparoscopic adhesion scores. The ultrasound images were recorded by experienced gynecologists according to standard outpatient tertiary evaluation procedures 14 , 32 ; therefore, there was no specific training beforehand, nor was an adjustment for the learning curve made. All sonographers were well trained by their experience in general practice. It is possible that the learning curve and lack of clear instructions may have compromised our results. Moreover, it is possible that the predictive value of the sliding sign may improve when Bayesian diagnostic principles are considered, in which the sonographer is made aware of the patient's history and therefore is more alert to possible vesicouterine adhesions 33 . A similar interpretation bias in our study was the presence of a uterine niche in a subgroup of patients, which could have caused the ultrasound assessor to interpret a sliding bladder sign as negative. The results could also have been influenced by our inclusion of cases with a full bladder during the ultrasound examination. Another limitation of this study was that the intra‐ and interobserver agreement was based on videoclips only. Overall, we presented results on validation of the sliding bladder sign; we did not investigate the relationship between the presence of adhesions and clinical symptoms, such as pain. Despite these limitations, our dataset represents patients seen routinely in clinical practice and, as the number of Cesarean sections increases dramatically, the sliding bladder sign will provide valuable information in an increasing number of cases. This study shows that the sliding bladder sign is an accurate technique to predict the absence or presence of severe vesicouterine adhesions. By incorporating this technique into standard TVS examinations, surgeons will be better prepared before entering the operating theater. This procedure will help to evaluate the risk of bladder injury or intraoperative difficulties, counsel patients accordingly, refer patients to more experienced clinics or plan surgical time more accurately when adhesions are identified 3 , 34 . Assessing the sliding bladder sign when performing TVS should become routine practice. It is important that the right technique is used, including applying pressure using the examiner's free hand and the vaginal probe. To master the technique of the sliding bladder sign, we have developed a free online eLearning resource based on the results of this study, which is available at www.gynaecologyonline.com . We recommend integrating this into further research and taking a possible learning curve of less experienced sonographers into account. Future research should investigate whether standardizing the sliding bladder sign in outpatient clinic evaluation improves preoperative planning and reduces short‐term and long‐term patient risks. It is also important to determine the clinical implications of adhesions to better understand the relationship between pain and urinary‐related symptoms, and the sliding bladder sign could be key in future research analyzing this. Based on the results of future research, in particular, the evaluation of discrepancies in interpretation of the sliding bladder sign, we will adjust our eLearning resource. The sliding bladder sign using TVS is a reliable, minimally invasive diagnostic technique for predicting the presence of severe vesicouterine adhesions, in which a negative sliding bladder sign predicts the presence of severe adhesions. A positive sliding bladder sign indicates that the presence of severe adhesions is unlikely. When interpreting the sliding bladder sign, a patient's surgical history, the presence of fibroids and the position of the uterus should be considered. Establishing vesicouterine adhesions by TVS may be useful for preoperative planning and for future research assessing the relationship between vesicouterine adhesions and symptoms and the effect of adhesion‐prevention interventions.

Vesicouterine adhesions are adhesions of the anterior compartment between the uterus, bladder and/or anterior abdominal wall. There is an association with chronic pelvic pain, dyspareunia, small bowel obstruction, increased operative difficulty in gynecological surgery and possible infertility 1 , 2 , 3 . The most common cause of these adhesions is a Cesarean section. Other known risk factors include endometriosis, pelvic inflammatory disease or previous abdominopelvic surgery for other indications 4 , 5 , 6 . Chronic pelvic pain, infertility and dyspareunia are known long‐term complications of Cesarean delivery. Multiple studies have suggested that these are related to adhesions of the anterior compartment 7 , 8 , 9 . Verberkt et al . showed that, at 3‐year follow‐up, 47.6% of women had de‐novo mild dysmenorrhea after Cesarean delivery, of which 11% had severe dysmenorrhea (visual analogue scale ≥ 7) 10 . Moro et al . suggested that postoperative development of dysmenorrhea and chronic pelvic pain is related to vesicouterine adhesions 11 . Besides the fact that diagnosing vesicouterine adhesions can be important for analyzing pain‐related symptoms, it may also be relevant to predict surgical difficulty and, consequently, improve the preoperative planning of surgery and counseling of the patients. Ultrasonography is a non‐invasive diagnostic tool to evaluate the presence of postoperative adhesions using the sliding sign 12 , 13 . The sliding sign is a standardized technique, described by Hudelist et al . in 2009, to determine whether a structure glides freely in relation to adjacent structures 14 . The transvaginal sonographic (TVS) sliding sign has been validated previously for evaluating adhesions around the ovaries and the posterior compartment between the uterus and bowels due to endometriosis and it shows a good predictive value of 93% 15 , 16 . This dynamic sonographic marker can also be used for the anterior compartment to evaluate the presence of vesicouterine adhesions, termed the ‘sliding bladder’ sign. Ayachi et al . investigated the accuracy of the TVS sliding sign for predicting pelvic adhesions in women with previous abdominopelvic surgery and found a sensitivity of 96.3% and specificity of 92.6% 12 . To our knowledge, no study has validated the transvaginal sliding bladder sign specifically. This study aimed to determine the predictive value and the intra‐ and interobserver variation for the sliding bladder sign using TVS for the presence of vesicouterine adhesions.

Appendix S1 STARD checklist Table S1 Predictive value of transvaginal ultrasound sliding bladder sign, with subgroup analysis Videoclip S1 Negative sliding bladder sign on grayscale ultrasound.