This single-center, prospective cohort study evaluated the data of patients managed via a standardized surgical algorithm at our clinic between December 1, 2023, and October 3, 2025. The study was designed and reported in adherence to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. Ethical approval was obtained from the Harran University Clinical Research Ethics Committee (Date: 28.08.2023; Approval number: HRÜ/23.08.37). All data were prospectively collected and analyzed following the completion of each surgical procedure, and the study was conducted in accordance with the principles of the Declaration of Helsinki.

During the study period, all patients scheduled for hysterectomy for benign gynecological indications underwent a standardized preoperative evaluation process.

For every patient in the study group, the procedure was initiated with the objective of completing a pure vaginal hysterectomy (VH). Depending on the intraoperative course, the following steps were pursued:

Step 1 (Primary Goal): Pure Vaginal Hysterectomy (VH).

Step 2 (Rescue Strategy): Transition to vNOTES was triggered based on specific intraoperative criteria: (I) Failure to progress, defined as the inability to safely identify or secure major pedicles within 30 min of colpotomy; (II) Inadequate visualization of the adnexa when salpingo-oophorectomy was indicated; or (III) Uncontrolled venous oozing requiring endoscopic bipolar vessel sealing.

Step 3 (Final Resort): Conversion to TLH or laparotomy if Step 2 was unsuccessful.

Patients were stratified into three groups for analysis based on the final surgical method completed: Group 1 (Pure VH), Group 2 (vNOTES-Assisted VH), and Group 3 (conversion to laparoscopy/laparotomy).

The primary outcome was the “Vaginal Completion Rate,” defined as the successful removal of the uterus via the vaginal route (either pure VH or vNOTES) without abdominal conversion. Secondary outcomes included operative time, estimated blood loss (EBL), length of hospital stay, and perioperative complications (Clavien–Dindo ≥2).

Demographic, preoperative, intraoperative, and postoperative data were collected from hospital records. The following clinical characteristics, particularly those considered to potentially influence surgical difficulty, were recorded for comparison between groups:

The primary surgical outcome, operative time, was defined as the total time elapsed from the initial vaginal incision to the placement of the final vaginal cuff suture. Postoperative complications were standardized according to the Clavien–Dindo Classification (Grade ≥2).

The standardized pure vaginal hysterectomy technique applied to all patients included the following steps:

All surgical procedures were performed by a single experienced surgeon who has over 5 years of clinical experience in vaginal surgery and had performed more than 200 vNOTES procedures prior to the commencement of this study to ensure technical standardization and an accurate evaluation of the learning curve. For vNOTES procedures, a V-PORT® Advanced Access Platform (Biomicro Medical, Izmir, Turkey) was utilized. Endoscopic visualization was achieved using a 5-mm, 30-degree laparoscope, and dissection was performed with standard laparoscopic graspers and a bipolar vessel sealing device (LigaSure™). Following the placement of the patient in the dorsal lithotomy position and preparation of the surgical field, the uterus was mobilized by placing tenaculums at the 6 and 12 o’clock positions of the cervix. A diluted vasopressin solution was injected submucosally into the cervical mucosa to ensure hemostasis and delineate dissection planes. The vaginal mucosa was detached via a circumferential incision around the cervix. Anterior colpotomy was completed by entering the vesicouterine space and peritoneal cavity following blunt and sharp dissection of the bladder from the cervix. Similarly, posterior colpotomy was performed by entering the pouch of Douglas and the peritoneal cavity after separating the rectum via blunt dissection following a mucosal incision in the posterior fornix. Subsequently, the hysterectomy proceeded using a vessel sealing device (LigaSure™, Medtronic) instead of the traditional clamp-cut-tie technique. The uterosacral and cardinal ligament complexes on both sides of the uterus were identified by digital palpation close to the cervix. These ligaments were sequentially coagulated and transected using the LigaSure™ device. This maneuver provided significant downward uterine mobility. Dissection continued from the caudal to cranial direction. The vascular pedicle containing the uterine arteries and veins was isolated using forceps and safely coagulated and transected with LigaSure™. Continuing dissection towards the uterine fundus, the round ligament and tubo-ovarian ligament (or infundibulopelvic ligament if salpingo-oophorectomy was planned) were sealed and transected using the same method, and the uterus was exteriorized vaginally. Following hemostasis control, the operation was terminated by closing the vaginal cuff with appropriate suture material.

In cases where progress could not be made during pure vaginal hysterectomy or additional intervention was required (e.g., adnexal mass, salpingectomy), a transition to vNOTES was made at any stage of the procedure. Following the decision to transition, a V-PORT® Advanced Access Platform was placed into the abdominal cavity via the vaginal route. The abdominal cavity was insufflated with carbon dioxide at a pressure of 10–12 mmHg. The remaining surgical steps (e.g., adhesiolysis, salpingo-oophorectomy, completion of hysterectomy) were completed under endoscopic visualization using standard laparoscopic instruments.

Data analysis was performed using IBM SPSS Statistics (Version 25.0). Conformity to normal distribution was assessed using the Shapiro–Wilk test, and homogeneity of variances was evaluated using Levene’s test. For the comparison of baseline demographic and perioperative variables among the three surgical outcome groups (Pure VH, vNOTES, Conversion), ANOVA (with post-hoc Tukey) was used for normally distributed variables, and the Kruskal–Wallis test was used for non-normally distributed variables. Chi-square or Fisher’s Exact tests were used for categorical variables (Tables 1, 2). To avoid model overfitting given the number of failure events (n = 20), the multivariable logistic regression for ‘Pure VH Failure’ was restricted to the two strongest clinical predictors identified in univariate analysis: Uterine Weight and History of Cesarean Section. Furthermore, a Bonferroni correction was applied to multiple comparisons in perioperative outcomes to minimize Type I error. A post-hoc power analysis was conducted to confirm the statistical robustness of the primary outcomes. Learning curve analysis was performed by chronologically dividing the cohort into two halves (First 82 vs. Last 83 cases) and comparing results via Chi-square and T-tests. Missing data were managed via complete-case analysis. Statistical significance was set at p < 0.05 for all analyses.

During the study period, a total of 175 patients scheduled for hysterectomy for benign indications were evaluated at our clinic. Due to the primary contraindications specified in the Materials and methods section, 10 patients (5.7%) were deemed unsuitable for the standard vaginal-first algorithm and were operated on directly via laparoscopy (n = 7) or laparotomy (n = 3). These patients were excluded from the study. The remaining 165 patients constituted the main analysis cohort of the study and underwent the standard stepwise surgical algorithm (Figure 1).

Among the 165 patients managed via the standard algorithm, the operation was successfully completed via the vaginal route (either pure VH or vNOTES) without abdominal incision in 157 patients (95.2%). Post-hoc power analysis based on the primary outcome (vaginal completion rate) demonstrated a statistical power of 0.91, confirming the adequacy of the sample size.

The distribution of surgical methods was as follows:

The baseline demographic and clinical characteristics of the study cohort are summarized in Table 1. No statistically significant differences were found among the three surgical outcome groups (Pure VH, vNOTES, Conversion) in terms of age, body mass index (BMI), parity, menopausal status, number of previous cesarean sections, or history of abdominal surgery (p > 0.05 for all). Only final uterine weight was found to be statistically significantly higher in the vNOTES and Conversion groups compared to the Pure VH group (p = 0.031).

Key surgical and postoperative outcomes are compared between groups in Table 2. As expected, operative time, estimated blood loss, and length of hospital stay significantly increased from the Pure VH group to the Conversion group (p < 0.001 for all). The rates of severe intraoperative and postoperative complications were generally low and did not show a statistically significant difference between the groups (p > 0.05).

The most frequent reason for transitioning to vNOTES support was the need for intervention in adnexal pathology (50.0%) or inadequate uterine mobility/visualization (33.3%). In these cases, vNOTES served as a successful rescue strategy. For the conversion group, the predominant cause was dense pelvic adhesions (62.5%) making both vaginal and endoscopic progression impossible. As detailed in Table 3, some patients presented with multiple intraoperative challenges, such as uncontrollable bleeding and difficulty in anatomical orientation, necessitating an upgrade in surgical tier.

The transparency of intraoperative decisions to transition to a higher-tier surgery is critical for understanding the clinical utility of the algorithm. Table 3 details the underlying reasons for the need for vNOTES support (Step 2) and conversion to laparotomy/laparoscopy (Step 3). The most frequent reason for transitioning to vNOTES was the need to perform a safe adnexal intervention, such as salpingo-oophorectomy or the evaluation of an adnexal mass (41.7%). In these instances, the endoscopic capability of vNOTES allowed for the safe completion of the procedure without an abdominal incision. Conversely, the most frequent cause for full conversion (Step 3) was dense pelvic adhesions (62.5%) that rendered both vaginal and endoscopic dissection impossible. It is important to note that intraoperative decisions were often multifaceted; for instance, some patients required transition due to a combination of inadequate mobility and difficulty in hemorrhage control, as reflected in the overlapping categories in Table 3.

To test the robustness of our algorithm, surgical success rates were examined in high-risk patient subgroups traditionally considered to present challenges for vaginal hysterectomy (Table 4). Even in the presence of obesity (BMI ≥ 30 kg/m²), a history of two or more cesarean sections, and a uterine weight exceeding 250 grams, pure vaginal hysterectomy and total vaginal completion rates remained remarkably high.

Multivariable logistic regression analysis, performed to determine which patients failed pure VH and required vNOTES or conversion (n = 20), demonstrated that uterine weight was the sole independent and statistically significant predictor of failure (Table 5). Every 100-gram increase in uterine weight increased the probability of pure VH failure (i.e., the need for vNOTES or conversion) by 2.1 times [Adjusted Odds Ratio (aOR): 2.11; 95% confidence interval (CI): 1.34–3.32; p = 0.001]. Although BMI (aOR: 1.04; 95% CI: 0.91–1.19; p = 0.560) and history of cesarean section (aOR: 1.12 for 1 C/S; aOR: 1.35 for ≥2 C/S; p > 0.05) increased the risk of pure VH failure, these increases were not statistically significant.

Comparison of the ‘First 82’ and ‘Last 83’ groups showed no significant differences in baseline demographics, including mean age (48.1 vs. 48.3; p = 0.81), BMI (30.1 vs. 29.5; p = 0.42), and uterine weight (195 g vs. 188 g; p = 0.55), confirming that improved outcomes were related to surgical experience rather than case selection. To evaluate the effect of surgical experience and the standardization of the algorithm on outcomes, the patient cohort was chronologically divided. As shown in Table 6, as surgical experience increased over time: the pure vaginal hysterectomy rate increased significantly (from 81.7 to 93.9%; p = 0.018); the rate of transition to vNOTES (the need for the “safety valve”) decreased significantly (from 11.0 to 3.6%; p = 0.045); and the mean operative time for cases completed via pure VH was significantly shortened (from 75 min to 61 min; p < 0.001). The full conversion rate showed a decreasing trend from 7.3 to 2.4% (p = 0.134).

These findings strongly support the hypothesis that vNOTES plays the role of an “educational bridge,” enhancing the surgeon’s competence in vaginal surgery, and becomes less necessary over time. Comparison of baseline demographic and clinical characteristics (age, BMI, parity, uterine weight) between the two periods revealed no statistically significant differences (p > 0.05 for all), indicating that the improvement in outcomes was attributable to increased surgical experience rather than a change in the patient population.

Our multivariable logistic regression analysis confirmed that uterine weight was the sole independent risk factor for pure vaginal hysterectomy failure (transition to vNOTES or conversion) (aOR: 2.11; p = 0.001). As seen in Table 4, large uteri (particularly >250 g) presented a challenge for the vaginal approach. However, this finding does not imply that large uteri cannot be removed vaginally. Indeed, the fact that 87.8% of patients with uteri weighing >250 g had their operations completed via the vaginal route demonstrates the efficacy of our algorithm in this challenging group. Our study reveals that the majority of large uterus cases (75.6%) can actually be completed with the simpler pure vaginal hysterectomy without even requiring vNOTES. This suggests that while uterine size is a difficulty factor, it is not an absolute contraindication within an algorithmic approach, and positioning vNOTES as a strategic tool that intervenes only when necessary may be the most accurate approach in terms of clinical and cost-effectiveness (13).

It is crucial to emphasize that our proposed “vaginal-first” approach is not a universal solution, and its applicability depends on patient anatomy. Our study cohort did not include transgender patients. A significant study by Donmez et al. (14) examined hysterectomy outcomes in virgo intacta transgender male patients. In this patient group, due to the narrow and atrophic vagina resulting from nulliparity and long-term testosterone use, even the initial steps of traditional vaginal hysterectomy are extremely difficult or impossible. In this specific and anatomically challenging population, pure VH, the first step of our algorithm, is not a suitable option. Donmez et al. (14) compared totally endoscopic vaginal NOTES hysterectomy (TVNH) with conventional laparoscopy (TLH) and demonstrated that TVNH offered significant advantages such as less postoperative pain and shorter hospital stay. These findings indicate that in special cases where vaginal access is restricted, vNOTES, which we position as “supportive,” is actually the most appropriate primary minimally invasive method. Therefore, while the results of our study are valid for the broad non-descensus patient population with suitable vaginal anatomy, we acknowledge that a “vNOTES first” approach is a more accurate strategy in special groups such as transgender men. This reinforces the principle that the surgical approach must be individualized according to the patient.

In our cohort, the overwhelming majority (87.9%) of patients with non-descensus uteri had their operations successfully completed with the simpler and cost-effective pure vaginal hysterectomy (VH). This raises the question of whether starting every patient routinely with vNOTES might impose an unnecessary technological and cost burden in the majority of cases. While the model by Tekin et al. (10) initiates every case with vNOTES, our algorithm positions vNOTES not as a “primary” method, but as an invaluable “safety valve” that activates in selected cases where pure VH is obstructed or additional maneuvers such as adnexal intervention are required. Indeed, the decrease in our vNOTES utilization rate from 11.0 to 3.6% as surgical experience increased demonstrates that this tool serves as an “educational bridge,” and the need for this “valve” diminishes as the surgeon’s competence in pure vaginal surgery grows. Consequently, while acknowledging the validity of the “vNOTES first” strategy in specific contexts, we argue that our stepped approach presents a more refined model that remains truer to the philosophy of minimally invasive surgery, encouraging the development of surgical skills while ensuring rational resource use.

At this point, it must also be emphasized that the “let’s do vNOTES for every suitable patient” mentality is not a rational approach. Our findings show that vNOTES-assisted operations and conversions result in significantly longer operative times and greater blood loss compared to pure vaginal hysterectomy (Table 2). We acknowledge that surgery time in cases requiring transition to vNOTES or conversion reflects the intraoperative decision-making process and the added complexity of the rescue steps rather than the efficiency of a single technique. Consequently, our primary time-efficiency analysis focused on Pure VH cases, where a significant reduction from 75 to 61 min (p < 0.001) was observed as experience grew. This distinction ensures that the surgical efficiency of the primary goal (VH) is not obscured by the inherent time penalty of rescue maneuvers in more complex cases. Furthermore, due to the specialized ports and instruments required, vNOTES is not a cost-effective method for routine use. Therefore, the stepped algorithm presented in our study offers a model that optimizes both clinical efficacy and efficient resource utilization by leveraging the technological advantage of vNOTES only when truly needed. Regarding the economic implications, although some argue that vNOTES may facilitate earlier discharge in complex cases, our algorithm achieved a mean hospital stay of 1.2 days in the Pure VH group without the need for specialized equipment. By successfully managing 87.9% of cases with pure VH, we avoided the costs of disposable ports and endoscopic instruments in the vast majority of the cohort, suggesting that prioritizing the vaginal route is not only clinically sound but also highly cost-effective.

Furthermore, the economic implications of our algorithm merit consideration. While a vNOTES port and endoscopic equipment add an estimated $350–$600 to the procedural cost, our VH-first approach utilized these resources in only 7.3% of cases. In contrast, a ‘vNOTES-first’ strategy for the entire cohort would have incurred significantly higher institutional costs without necessarily improving clinical outcomes for the 87.9% of patients who were successfully managed with pure VH. This underscores the role of vNOTES not as a routine necessity, but as a cost-effective ‘rescue tool’ that prevents more expensive abdominal conversions.

The results of our study hold important implications for non-descensus hysterectomy practice. In light of these findings, the following practical recommendations can be made: (1) In patients with non-descensus and benign pathology, barring a few absolute contraindications, the first surgical option should always be vaginal hysterectomy. (2) vNOTES should be integrated into resident and specialist training programs not as a primary procedure, but as an advanced skill and safety tool that enhances vaginal surgical competence. (3) In hospital resource planning, vNOTES equipment should be evaluated as a strategic resource to support the vaginal approach in selected cases, rather than for routine use in every case.

The primary strength of this study is its prospective design and the application of a standardized stepwise algorithm to all consecutive patients, providing high-quality real-world data and minimizing selection bias. The involvement of a single surgeon allowed for a controlled evaluation of the learning curve, removing inter-surgeon technical variability. While this ensures high internal validity, we acknowledge that it may limit the immediate generalizability of our findings to centers with different surgical volumes or expertise levels. The outcomes reported here reflect a high-volume setting; therefore, our results should be viewed as a proof-of-concept. Future multicenter studies involving surgeons with diverse experience levels are warranted to confirm the external validity and reproducibility of this stepped algorithm across different clinical environments. Furthermore, the inclusion of high-risk subgroup and learning curve analyses allowed for the evaluation of the algorithm’s effectiveness from different perspectives, demonstrating that the strategy remains robust even in challenging clinical scenarios. However, several limitations must be acknowledged. First, being a single-center study with a single surgeon may limit the generalizability of our findings to centers with different expertise levels. Second, although multivariable logistic regression showed uterine weight to be an independent predictor, unmeasured confounders such as “intraoperative surgeon decision” or “vaginal access difficulty” may have played a role in the transition between surgical tiers. Third, due to the low number of failure events (n = 20), the predictive model was intentionally restricted to avoid overfitting, meaning these results should be interpreted as exploratory rather than definitive.

In conclusion, this study provides robust evidence supporting the vaginal route as the primary standard of care for hysterectomy in patients with non-descensus uteri, barring specific absolute contraindications. Within this algorithmic framework, vNOTES is redefined not merely as a routine surgical method, but as a strategic “catalyst” in the surgeon’s professional development. It functions as an essential “educational bridge” that hones proficiency in traditional vaginal techniques and an invaluable “safety valve” that preserves the minimally invasive nature of the procedure during challenging intraoperative scenarios. By prioritizing the simplest and most cost-effective method pure vaginal hysterectomy and utilizing vNOTES support only when clinically indicated, this stepped approach optimizes hospital resource allocation while maximizing patient safety. Ultimately, our findings suggest that the intelligent integration of vNOTES into a “vaginal-first” algorithm can successfully reverse the global decline in vaginal surgery, proposing a new, rational paradigm for modern gynecological practice.