Diagnostic accuracy of the sonographic sliding sign for predicting pelvic organ adhesions in gynecologic endoscopic surgery

This research was designed as a retrospective cohort and observational study. A single-blind design was implemented, whereby the physician conducting the initial assessment of the participants was blinded to the surgical and postoperative outcomes, which were observed by a different physician. The study commenced after receiving approval from the local ethics committee (Approval ID:2024/64) according The STARD reporting guideline . A total of 283 non-virgin women over the age of 18 who underwent endoscopic surgery for gynecologic pathology at our tertiary care center during the last six months of 2024 were included. Participants were determined consecutively from surgical scheduling list. Twenty-one patients with body mass index (BMI) over 35 were excluded due to inadequate sonographic visibility. All initial examinations were performed by the same physician with 10 years of clinical experience. Demographic data for the 262 included patients—such as age, BMI, gravidity, parity, history of prior abdominal surgery or cesarean section, and chronic illnesses—were recorded on individual case forms. The presence of endometriosis or related symptoms (dysmenorrhea, chronic pelvic pain, dyspareunia, dysuria, dyschezia, chronic constipation), as well as obstructive or restrictive pulmonary diseases that could limit deep inspiration, were documented. Confirmed endometriosis was defined as the intraoperative identification of typical endometriotic lesions by an experienced surgeon. During inspection, previous incision scars were noted and categorized (hypotrophic, keloid or flat), and vaginal examination included an assessment for obliteration of the Douglas pouch. All patients underwent both transabdominal and transvaginal ultrasonographic evaluation by the same physician. Voluson S8 (GE Healthcare Technologies, Inc., Chicago, Illinois, USA) ultrasound branded device was used with 5Mhz convex probe and endocavitary probe. This evaluation included trocar entry sites, areas beneath previous incisions, and the relationship of pelvic and abdominal organs to adjacent structures and the anterior abdominal wall. The operator had access to the clinical history. Preoperative ultrasonographic evaluation was completed within one day before surgery. The sonographer was not involved in the operations. During transabdominal ultrasonographic examination, the evaluation of visceral organ mobility was conducted by instructing the patient to perform maximal inspiratory effort. Under normal physiological conditions, diaphragmatic excursion during deep inspiration induces caudal displacement of abdominal viscera, resulting in observable gliding movement between adjacent organ surfaces and the parietal peritoneum. The presence of unimpeded sliding motion between visceral and parietal layers indicates preserved serosal integrity and the absence of pathological adherence. The transvaginal sonographic component of the examination employed dynamic assessment through the application of graduated mechanical pressure using the endocavitary transducer directed toward the anterior and posterior vaginal fornices. This maneuver facilitates evaluation of the mobility of pelvic structures, particularly the relationship between the uterus, adnexa, rectosigmoid colon, and adjacent peritoneal surfaces. Under normal circumstances, gentle probe manipulation should elicit free sliding movement between these anatomical structures. The absence of expected organ sliding—designated as a “negative sliding sign”—was interpreted as sonographic evidence suggestive of intra-abdominal or pelvic adhesions. This finding indicates potential obliteration of the normal potential space between visceral surfaces, consistent with fibrous adhesion formation secondary to prior surgical intervention, inflammatory processes, endometriosis, or other pathological conditions resulting in peritoneal scarring and subsequent restriction of organ mobility. Operative and postoperative data were recorded by a different physician who was blinded to the preoperative sonographic assessment results, including the presence or absence of the sliding sign. These data included presence of true intra-abdominal adhesions, endometriozis, hemorrhage, bladder injury, bowel injury, duration of surgery, and length of postoperative hospital stay. The presence of true intra-abdominal adhesions was confirmed intraoperatively through direct visualization and manual assessment of fibrous bands connecting peritoneal surfaces or visceral organs. The Peritoneal Adhesion Index (PAI) is a widely used scoring system for evaluating the extent and severity of intra-abdominal adhesions according to predefined abdominal regions (right upper, right lower, right flank, left upper, left lower, left flank, epigastrium, pelvis, central and bowel to bowel). In this assessment, the abdominal cavity was divided into nine equal regions. Adhesions were graded as follows: filmy adhesions requiring blunt dissection were assigned 1 point; strong adhesions requiring sharp dissection were assigned 2 points; and very strong, vascularized adhesions in which sharp dissection with a high risk of unavoidable tissue damage was required were assigned 3 points. Adhesions were considered present if a score exceeding zero was recorded in at least one anatomical region [ 11 ]. Hemorrhage was defined as intraoperative aspiration of approximately 500 ml or a decrease in hemoglobin ≥ 2 g/dL at first day of surgery. Bladder injury and bowel injury were evaluated as injuries requiring repair in the organs, including the entire organ wall, or serosal injuries. The duration of surgery was calculated in minutes from the moment the surgeon started the surgery until the incision was completely sutured, and the postoperative hospital stay was calculated in days. For the index test (sliding sign), examinations in which the sliding sign could not be adequately assessed due to technical limitations (e.g., poor image quality, patient-related factors, or limited organ mobility) were classified as indeterminate and excluded from the primary diagnostic accuracy analysis. These cases were not reclassified as positive or negative to avoid misclassification bias. Missing data for the reference standard were handled by excluding patients who did not undergo surgical exploration, as intraoperative findings were required for definitive confirmation of adhesions. No imputation methods were applied, given the relatively small proportion of missing data and the risk of introducing systematic bias. A complete-case analysis was therefore performed. The number of excluded cases and the reasons for exclusion were explicitly reported in the flow diagram and the Methods section to ensure transparency (Fig. 1 ). Fig. 1 Flowchart showing process of patient selection and assessment of sliding bladder sign Flowchart showing process of patient selection and assessment of sliding bladder sign All demographic variables were analyzed using descriptive statistics including median, mean, and standard deviation. Associations between independent categorical variables were examined using Pearson’s Chi-Square test and Fisher’s Exact test. The distribution of continuous variables was tested with the Kolmogorov-Smirnov test, and non-parametric methods were used accordingly. Relationships between continuous and categorical variables were analyzed using the Mann-Whitney U test. The predictive performance of the sliding sign in detecting intra-abdominal adhesions was assessed using sensitivity and specificity calculations, followed by construction of ROC curves. 95% confidence intervals were calculated using the Wilson score method. A p -value of < 0.05 was considered statistically significant. Statistical analyses were performed using SPSS version 22.0 (SPSS Inc., Chicago, IL, USA). The sample size was determined based on expected diagnostic accuracy parameters for the sliding sign reported in previous studies. Assuming an expected sensitivity of approximately 73% and specificity of 86%, with an estimated prevalence of clinically significant intra-abdominal adhesions of 35%, a minimum sample size of approximately 210 patients was required to estimate sensitivity and specificity with a 95% confidence interval width of ± 10% (α = 0.05, power = 80%). The final study population consisted of 262 patients, exceeding the minimum required sample size.

The demographic data of all 262 patients included in the study were analyzed. The mean age was 45.1 (± 10.3) years, mean BMI was 27.4 (± 4.5), mean gravidity was 2.4 (± 1.3), and mean parity was 2.3 (± 1.3). Only 20 patients (7.6%) were nulligravid. The mean hospital stay for patients undergoing gynecologic endoscopic surgery was 1.9 (± 0.7) days, and the mean operative time was 132.7 (± 61.4) minutes (Table  1 ). Table 1 Demographic Data (mean ± SD, range) Age 45.1 ± 10.3, range (18–80) BMI 27.4 ± 4.5, range (17–34) Hospital Stay (day) 1.9 ± 0.7, range (1–6) Operation Duration (min) 132.7 ± 61.4, range (17–298) Gravidity 2.4 ± 1.3, range (0–7) Parity 2.3 ± 1.3, range (0–7) Demographic Data (mean ± SD, range) A history of previous abdominal surgery was present in 115 patients (43.9%). Adhesions were identified in 48 of these 115 patients. Of the 56 patients found to have adhesions, 8 had no prior abdominal surgery. A statistically significant relationship was found between a history of abdominal surgery and the presence of intra-abdominal adhesions ( p  < 0.001) (Table  2 ). PAI scores among the 56 patients with adhesions were: 1 ( n  = 28, 50%), 2 ( n  = 21, 37.5%), and 3 ( n  = 7, 12.5%). Table 2 Number of Patients According to Adhesions and History of Previous Surgeries ( n (%)) Surgical History Present, n (%) Yes No Adhesion Present 48 (18.3%) 8 (3.1%) 56 (21.4%) Absent 67 (25.6%) 139 (53%) 206 (78.6%) 115 (43.9%) 147 (56.1%) Number of Patients According to Adhesions and History of Previous Surgeries ( n (%)) Endometriosis was diagnosed in 7 patients (2.7%). A significant relationship was found between a confirmed diagnosis of endometriosis and the presence of adhesions ( p  = 0.01). A total of 124 patients (47.3%) had either a confirmed or suspected diagnosis of endometriosis or presented with symptoms potentially associated with intra-abdominal adhesions, such as dysmenorrhea, pelvic pain, dysuria, dyspareunia, chronic constipation, or dyschezia. A statistically significant relationship was observed between these symptoms and the presence of intra-abdominal adhesions ( p  = 0.004). Findings related to endometriosis are presented as exploratory and hypothesis-generating rather than confirmatory. Obstructive pulmonary disease was present in 15 patients (5.7%). Among these, only one patient had adhesions, and in that case, the sliding sign was negative. Regarding scar assessment, 5 patients (1.9%) had keloid scars, 19 (7.3%) had hypotrophic scars, and 91 (34.7%) had well-healed scars(flat). The remaining 147 patients (56.1%) had no visible scars. Of those with hypotrophic scars, 73.7% had adhesions. Among those with keloid scars, 60% had adhesions. There was a strong statistically significant association between poorly healed scars (hypotrophic or keloid) and the presence of intra-abdominal adhesions ( p  < 0.001). In gynecological examination, the Douglas pouch was obliterated in 8 patients (3.1%). A strong statistical correlation was found between obliteration of the Douglas pouch and the presence of intra-abdominal adhesions ( p  < 0.001). The mean operative time in patients with adhesions was 142.5 (± 66.8) minutes compared to 130.1 (± 59.8) minutes in those without adhesions; however, this difference was not statistically significant ( p  = 0.275). The mean length of postoperative hospital stay was 2.1 (± 0.8) days for patients with adhesions and 1.8 (± 0.7) days for those without. The presence of intra-abdominal adhesions significantly increased hospital stay duration ( p  = 0.006). Among the 56 patients with adhesions, 11 (19.6%) experienced hemorrhage and 3 (5.4%) had bladder injuries, while no bowel injuries were reported. In the group without adhesions, these complications were observed in 12 (5.8%), 1 (0.5%), and 1 (0.5%) patients, respectively. A statistically significant relationship was found between adhesions and the occurrence of hemorrhage ( p  = 0.003) and bladder injury ( p  = 0.03), but not with bowel injury ( p  = 1.0). The relationships between adhesions and the aforementioned findings are summarized in Table  3 . Table 3 Association Between Adhesions and Clinical Variables: Analysis Results Adhesion No Adhesion P Value Previous Surgeries ( n ) 48/56 8/206 p  < 0.001 α Sliding Sign ( n ) 54/56 0/206 p  < 0.001 α Endometriosis (n) 5/56 2/206 p  = 0.01 β Adhesion-related complaints (n) 36/56 88/206 p  = 0.004 α Douglas Obliteration (n) 7/56 1/206 p  < 0.001 α Poorly Healed Scar (Hypotrophic/ Keloid) (n) 17/48 7/67 p  < 0.001 β Complications (n)  Hemorrhage 11/56 12/206 p  = 0.003 β  Bladder Injury 3/56 1/206 p = 0.03 β  Bowel Injury 0/56 1/206 p = 1 β Duration of Surgery (Mean ± SD) 142.5 ± 66.8 130.1 ± 59.8 p  = 0.275 µ Length of Hospital Stay (Mean ± SD) 1 2.1 ± 0.8 1.8 ± 0.7 p  = 0.006 µ 1 α : Pearson-Chi-Square ; β : Fisher’s Exact Test ; µ : Mann Whitney U Association Between Adhesions and Clinical Variables: Analysis Results 1 α : Pearson-Chi-Square ; β : Fisher’s Exact Test ; µ : Mann Whitney U The diagnostic value of a negative sliding sign on transvaginal ultrasound (Table 4 ) for detecting adhesions was found to be: sensitivity 77% (95% CI: 64.2–85.9), specificity 100% (95% CI: 98.2–100.0), positive predictive value (PPV) 100% (95% CI: 91.8–100.0), and negative predictive value (NPV) 94% (95% CI: 90.1–96.5). The receiver operating characteristic (ROC) curve for this analysis is shown in Fig.  2 . ROC analysis demonstrated a steep initial rise in sensitivity, indicating that the test correctly identified a substantial proportion of true positives even at low false-positive rates. The ROC curve rapidly approached the upper-left corner of the plot, reflecting strong discriminatory ability. The area under the curve (AUC) was 0.884 (95% CI, 0.82–0.95), indicating highly predictive overall diagnostic performance. This high AUC value suggests that the transvaginal ultrasound has a robust capacity to distinguish intraabdominal adhesions. Table 4 Number of Patients with Negative Sliding Sign (Transvaginal Ultrasound) and Intra-abdominal Adhesions ( n (%)) Adhesion, n (%) Present Absent Sliding Sign  Positive 13 (23.2%) 206 (100%)  Negative 43 (76.8%) 0 (0%) Number of Patients with Negative Sliding Sign (Transvaginal Ultrasound) and Intra-abdominal Adhesions ( n (%)) Fig. 2 Receiver operating characteristic (ROC) curve for transvaginal ultrasonography in predicting intra-abdominal adhesions. The area under the curve (AUC) was 0.884 (95% CI: 0.82–0.95), indicating highly predictive diagnostic performance Receiver operating characteristic (ROC) curve for transvaginal ultrasonography in predicting intra-abdominal adhesions. The area under the curve (AUC) was 0.884 (95% CI: 0.82–0.95), indicating highly predictive diagnostic performance For abdominal ultrasound, the diagnostic values of a negative sliding sign were: sensitivity 96%, specificity 100%, PPV 100%, and NPV 99%. The ROC curve for this analysis is presented in Fig.  3 . ROC analysis demonstrated an demonstrating substantial predictive ability of the test to discriminate intraabdominal adhesions. The ROC curve showed a near-perfect trajectory approaching the upper-left corner of the plot. The area under the curve (AUC) was 0.982 (95% CI, 0.95–1.00), indicating outstanding diagnostic performance. This exceptionally high AUC value suggests that the abdominal ultrasound provides a highly reliable distinction between intraabdominal adhesions. Fig. 3 Receiver operating characteristic (ROC) curve for transabdominal ultrasonography in predicting intra-abdominal adhesions. The area under the curve (AUC) was 0.982 (95% CI: 0.95–1.00), indicating outstanding diagnostic performance Receiver operating characteristic (ROC) curve for transabdominal ultrasonography in predicting intra-abdominal adhesions. The area under the curve (AUC) was 0.982 (95% CI: 0.95–1.00), indicating outstanding diagnostic performance A statistically significant association was found between the presence or absence of the sliding sign (assessed during deep inspiration via both abdominal and transvaginal ultrasound) and the presence of true intra-abdominal adhesions ( p  < 0.001). Table  5 summarizes the number of patients with and without the sliding sign and confirmed adhesions. Table 5 Number of Patients with Negative Sliding Sign (Both Transabdominal and Transvaginal Ultrasound) and Intra-abdominal Adhesions ( n (%)) Adhesion, n (%) Present Absent Sliding Sign  Positive 2 (0.8%) 206 (78.6%)  Negative 54 (20.6%) 0 (0%) Number of Patients with Negative Sliding Sign (Both Transabdominal and Transvaginal Ultrasound) and Intra-abdominal Adhesions ( n (%)) When both abdominal and transvaginal ultrasonographic evaluations were combined preoperatively to detect true intra-abdominal adhesions, the diagnostic values of the sliding sign were identical to those of abdominal ultrasound alone: sensitivity 96%, specificity 100%, PPV 100%, and NPV 99%. No adverse events related to the index test were observed, as ultrasonographic imaging is a non-invasive observational modality.

Pelvic adhesions may develop due to prior surgeries, endometriosis, malignancies, pelvic radiation, or infections [ 1 ]. These adhesions occur as a result of fibrin accumulation and insufficient fibrinolysis during the healing process of peritoneal surface injuries in pelvic organs. Increased concentrations of leukotriene B4 and prostaglandin E2 in peritoneal fluid, along with decreased plasminogen activity, play a key role in this process [ 2 ]. Additionally, other biochemical and molecular mechanisms such as cytokines, growth factors, and cell adhesion molecules also contribute to adhesion formation [ 3 ]. Such adhesions can lead to small bowel obstruction, female infertility, chronic pelvic pain, chronic abdominal pain, and increased rates of reoperation and surgical complications [ 4 ]. By increasing complication rates, they extend both operative times and hospital stays. The 10-year readmission rate due to adhesions is approximately 5.8%, while the rate of reoperation is 3.8% [ 5 ]. In patients with adhesions, laparotomy via a midline incision may take up to 18 min longer on average [ 6 ]. The increasing frequency of cesarean deliveries in Turkiye—motivated by maternal preference and efforts to ensure neonatal safety—has unfortunately led to a higher incidence of pelvic adhesions in women [ 7 ]. Given this burden, predicting adhesions preoperatively can assist surgeons in selecting patients in line with their expertise. When performed by experienced hands, minimally invasive gynecologic surgeries can improve surgical success and reduce both complication rates and the duration of hospitalization and surgery. Patient history plays a critical role in predicting pelvic adhesions. However, additional tools such as assessment of skin scar type and the presence or absence of a “sliding sign” seen in sonographic evaluation of intra-abdominal organ mobility during deep inspiration or compression are also useful [ 8 – 10 ]. Hypotrophic (sunken) scar tissue has been associated with intra-abdominal adhesions. In sonography, the presence of the “sliding sign” suggests the absence of adhesions, while its absence raises suspicion. This method is not widely used in routine practice, mainly due to a lack of awareness about its predictive value. This study aims to evaluate the diagnostic accuracy of the sliding sign observed via both abdominal and transvaginal ultrasonography in predicting intra-abdominal adhesions prior to endoscopic gynecologic surgeries.

Our findings support the sliding sign as a promising, accurate tool that may be considered as part of the preoperative assessment however external validation, multi-center data and assessment of interobserver variability are needed. Sliding sign via transvaginal ultrasound can help anticipate bladder and rectal injuries in V-NOTES procedures, while transabdominal ultrasound can predict potential trocar injuries in laparoscopy. In both cases, recognizing hemorrhage risk can allow for better preoperative blood preparation, thus ensuring safer surgical outcomes.

Preoperative sonographic evaluation of intra-abdominal adhesions is a highly accurate, cost-effective, reproducible, and valuable method that can help predict and potentially prevent perioperative complications. Transabdominal sliding sign examination alone is equal diagnostic value to transabdominal combined transvaginal sliding sign evaluation. In a meta-analysis of 890 cases assessing periumbilical bowel adhesions prior to laparoscopy, findings similar to our study were reported: the sensitivity and specificity of the sliding sign were 95.9% and 93.1%, respectively [ 12 ]. In transabdominal evaluations, the accuracy of detecting intra-abdominal adhesions—particularly bowel adhesions—is notably high. In a large-scale review by Alcazar et al. involving 963 cases, the sliding sign was evaluated via transvaginal ultrasound in patients with suspected endometriosis, with specific attention to obliteration of the pouch of Douglas. The sensitivity and specificity of the sliding sign for predicting obliteration of the Douglas pouch were 88% and 94%, respectively [ 13 ]. A Cochrane meta-analysis reported a sensitivity of 79% and a specificity of 94% for predicting deep endometriosis via transvaginal ultrasonography [ 14 ]. These literature findings support our results. Although the diagnostic accuracy of transvaginal ultrasound for predicting periumbilical adhesions may not be as high as that of transabdominal ultrasound, its specificity remains high. This may be due to the relatively limited mobility of pelvic organs compared to the intestines. A recent study of 146 patients used the sliding sign on transvaginal ultrasound to predict adhesions between the uterus and bladder. The sensitivity and specificity were 68.4% and 91.9%, respectively—findings consistent with ours [ 15 ]. The sliding sign is also utilized in obstetrics. In a prospective cohort study of 380 pregnant women undergoing transabdominal ultrasound prior to cesarean section, the sensitivity and specificity of the sliding sign were 60.6% and 91.9%, respectively [ 10 ]. In a similar study involving 120 pregnant women, the sensitivity for predicting intra-abdominal adhesions was 100%, and the specificity was 86.8% [ 16 ]. Conversely, a separate study of 112 pregnant women reported a sensitivity of 53.3% and a specificity of 80.4%, suggesting lower diagnostic value in this population [ 17 ]. We believe the lower performance of the sliding sign in pregnant women may be attributed to the large size of the term uterus, which restricts peritoneal movement during deep inspiration. The gestational age at which the ultrasound was performed may also affect diagnostic accuracy. From the perspective of gynecologists or radiologists, identifying the sliding sign to reduce perioperative complications and assess the relationship between Douglas pouch adhesions and infertility is highly important. Learning this technique is relatively simple. Maicas et al. trained an artificial intelligence model to evaluate the sliding sign using just 414 videoclips, and tested it on 196 cases. The AI (artificial intelligence) model achieved a sensitivity of 88.6% and a specificity of 90.0% for predicting obliteration of the Douglas pouch [ 18 ]. Furthermore, learning curve studies have shown that clinicians can consistently interpret pelvic adhesions after performing 200 or more sliding sign evaluations [ 19 ]. In addition to the sliding sign, scar assessment from prior surgeries is also used to predict intra-abdominal adhesions. These scars may be flat, hypertrophic, or hypotrophic. Patients with flat scars have significantly lower rates of intra-abdominal adhesions. Meta-analyses have demonstrated, in line with our findings, that patients with hypotrophic scars have a 2.79-fold increased risk of adhesions, and those with hypertrophic scars have a 1.61-fold increased risk [ 20 ]. The strengths of our study include being the first large-scale, single-blind, observational retrospective cohort study conducted in the Turkish population; the sonographic evaluations were all performed by a single experienced physician; and both transvaginal and transabdominal evaluations of the sliding sign were performed prior to gynecologic endoscopic surgery. Limitations of our study include the exclusion of patients with BMI > 35,which limits the generalizability of the results to an obese population and the low number of patients with active pulmonary disease, which may have contributed to higher predictive values for the sliding sign. Moreover, the use of a single experienced clinician for ultrasound evaluations prevented assessment of interobserver variability and single-operator imaging may have resulted in overestimation of diagnostic value. The limited sample size may also have contributed to the lack of a statistically significant association between adhesions and bowel injury.