Learning Curve Analysis of Transvaginal Natural Orifice Transluminal Endoscopic Surgery in Treating Ovarian Cysts: A Retrospective Cohort Study

doi:10.21203/rs.3.rs-4152484/v1

Download PDF

Research Article

Learning Curve Analysis of Transvaginal Natural Orifice Transluminal Endoscopic Surgery in Treating Ovarian Cysts: A Retrospective Cohort Study

https://doi.org/10.21203/rs.3.rs-4152484/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 25 Jul, 2024

Read the published version in BMC Women's Health →

You are reading this latest preprint version

Background：Transvaginal Natural Orifice Transluminal Endoscopy (vNOTES) is regarded as a challenging surgical technique to learn but is promising in reducing perioperative pain and significantly improves the cosmetic outcomes. Previous studies on the learning curve analysis of vNOTES mainly focuses on the hysterectomy approach, while the vNOTES ovarian cystectomy’s learning curve was merely reported thought more frequently performed than vNOTES hysterectomy. Therefore, this study seeks to analyze the learning curve of three surgeons with varying levels of experience in performing endoscopic surgery for the treatment of ovarian cysts using vNOTES.

Methods: A total of 127 patients with ovarian cysts of a variety of pathological types were treated by vNOTES ovarian cystectomy performed by three surgeons of different levels of endoscopic surgical experience. Each surgeon’s learning curve was plotted using the Cumulative Sum method and divided into three or four phases of technique learning at the turning point of the learning curve. The sociodemographic and clinical features of patients in each phase were then compared and factors potentially associated with operation time were also screened.

Results: The learning curve was presented in four phases. The operation time (OT) was significantly shorter in phases II (53.66 ± 16.55 min) and IV (54.39 ± 23.45 min) as compared with phases I (68.74 ± 15.85) and III (75.93 ± 30.55) (p <0.001). More cases of serve pelvic adhesion and chocolate cyst were assigned in the later phases. The presence of pelvic adhesion [adjusted odds ratio (OR) 7.149 (0.506, 13.792), p = 0.035] and bilateral cyst [adjusted OR 16.996 (2.155, 31.837), p = 0.025], max diameter of cyst[adjusted OR 2.799 (0.174, 5.425), p = 0.037], and surgeon’s experience [adjusted OR -6.118 (-11.814, -0.423), p =0.035] were significantly associated with OT.

Conclusion: There learning curve of ovarian vNOTES has four phases. vNOTES ovarian cystectomy could be mastered after performing seven, nine, and 16 cases by surgeons with the most, average, and least experience in gynecologic endoscopic surgeries.

Trial registration: ChiCTR2200059282 (Registered on April 28th, 2022)

transvaginal natural orifice transluminal endoscopic surgery

ovarian cystectomy

learning curve

cumulative sum analysis

Ovarian cysts reportedly have an occurrence rate of 4–7% among premenopausal women and approximately 20% among postmenopausal women[1, 2, 3]. It has been recommended by the American College of Obstetricians and Gynecologists that simple cysts found in ultrasonographic examinations should be treated conservatively and followed-up safely even in postmenopausal women. However, for symptomatic or non-simple cysts, timely surgical intervention is indicated to avoid rupture and other adverse outcomes.

The transvaginal natural orifice transluminal endoscopic surgery (vNOTES) is an emerging minimally invasive surgical (MIS) technique that reportedly has faster postoperative recovery, no visible abdominal skin scar, and easier specimen removal compared with traditional and even transumbilical laparoscopic single site surgery(TU-LESS) [4, 5, 6]. Since the first report of vNOTES ovarian cyst surgery in 2012, multiple clinical studies regarding its application in treating ovarian cysts have shown its non-inferiority to laparoscopy in terms of surgical conversion and postoperative outcomes.[4, 6]

However, due to factors, including the totally different surgical approaches, opposite operating angles compared to traditional laparoscopy, narrow operating space, and chopstick effect (the instruments interfere with each other due to the narrow operating space in single port endoscopy), many specialists believe that the learning technique for vNOTES might be challenging, and the cost-effectiveness of learning the technique could be low[7, 8]. In this study, we presented the learning curve of three gynecologists with different levels of surgical experience in performing laparoendoscopy to evaluate the feasibility of learning the technique.

Study design

The present study reviewed and analyzed the sociodemographic—age, body mass index (BMI), parity and gravity, previous delivery mode, etc.—and operation-related clinical features—surgical type, surgeons’ experience, pathological type, operation time (OT), estimated intraoperative blood loss, hemoglobin (Hb) decrease, cyst size, etc.—of 127 cases of vNOTES ovarian cystectomy, which were performed between February 2019 and March 2023 by three gynecologic surgeons with different levels of experience of gynecological laparoscopies at Chengdu Women’s and Children’s Central Hospital. We also analyzed the learning curves of each surgeon. Among the three surgeons, the most experienced one is an expert in laparoendoscopy, who has 20 years of experience in gynecology and performed approximately 500 cases of TU-LESS; the youngest surgeon in this study is a new attending physician who has 10 years of working experience and conducted more than 200 cases of TU-LESS; the other gynecologist has 15 years of working experience and successfully performed approximately 300 cases of TU-LESS. Briefly, the most experienced surgeon was designated as surgeon #1; the surgeon with middle-level experience was designated as surgeon #2; and the least experienced surgeon was designated as surgeon #3. Among the clinical characteristics included in the present study, we mainly focused on the OT which reflected the competency of surgeons in performing ovarian vNOTES. We applied the cumulative sum (CUSUM) methodology on the OT to plot the learning curve of each surgeon and divided their vNOTES technique learning process into three or four phases at the turning points of the CUSUM curve, namely the exploration phase (Phase I), competence-acquiring phase (Phase II), challenge phase (Phase III), and proficiency phase (Phase IV, for the most highly experienced surgeon only). In the first and second phases, relatively easier surgeries were assigned. Since Phase III, more technically challenging surgeries, mainly chocolate cystectomy and cases with relatively severe pelvic adhesion were assigned, especially to surgeon #1. The sociodemographic and clinicopathological characteristics of the cases performed in the different phases were also compared subsequently.

Patient selection

The patients were included based on the following criteria:

(1) has ovarian cysts which require surgical intervention;

(2) has low possibility of malignancy according to imaging features and tumor markers;

(3) shows preference for vNOTES over other surgical options.

The patients were excluded based on the following criteria:

has never had sexual intercourse;

has suspected or confirmed rectovaginal endometriosis; and

has confirmed severe pelvic adhesion.

Operating procedure of vNOTES ovarian cystectomy

This study applied similar surgical methods and equipment as described in previous publications [6, 9, 10]. The detailed surgical procedures were as follows:

(1) Patients were placed in the Trendelenburg position and treated under general anesthesia following endotracheal intubation and insertion of Foley catheter for urinary drainage. After disinfection and draping, the cervix and vagina were exposed, especially the posterior fornix, which were disinfected three times.

(2) The posterior labium of the cervix was pulled toward the upper and exterior direction using a cervical clamp to expose the posterior fornix of the vagina (Supplementary Figure. 1A). A 2–2.5-cm long posterior colpotomy incision was made at 0.5 cm below the cervical vaginal junction (Supplementary Figure. 1B) to get access to the abdominal cavity after incising the peritoneum (Supplementary Figure. 1C and 1D).

(3) A disposable retractor (Figure. 1A) was placed into the abdominal cavity through the posterior culdotomy incision. The surgical platform (Figure. 1B and 1C) was established after establishing the pneumoperitoneum. The conventional laparoendoscopic equipment was applied in the subsequent surgical procedures.

(4) The abdominopelvic cavity was then carefully investigated to confirm the location, size, and adjacent organs of the ovarian cysts. Any pelvic adhesion, if existed, would be separated.

(5) The cortex of the ovarian cyst was incised using scissor or monopolar electrotome. The cyst was then separated completely between the cyst wall and the rest of the ovarian stroma (Figure. 1D).

(6) The excised cyst was incorporated into the bag, then the excised samples and bag were removed through the vaginal incision.

(7) The ovary was sutured using an absorbable suture to stop bleeding and allow for reshaping (Figure. 1E).

(8) Finally, the pelvic cavity was washed with sterile saline at body temperature, followed by the suction of CO₂, removal of the retractor, and serial suturing of the vaginal incision using a 2/0 absorbable suture line.

Learning curve analysis

To analyze the changes in the surgeons’ proficiency in performing ovarian vNOTES during the study period, their two-dimensional learning curves were plotted using the following two parameters for each surgeon: X-value, which represents the number of ovarian vNOTES the corresponding surgeon has performed, which were ordered chronologically from the earliest to the latest, and Y-value, which indicates CUSUM_OT. The CUSUM_OT of each case were calculated using the formula:

$$CUSUM OTn={\sum }_{i=1}^{n}\begin{array}{c}\\ \left(\text{x}\text{i}-\mu \right)\\ \end{array}$$

CUSUM_OT is the running total of differences between the individual case’s OT and mean OT of all cases. Hence, it could be conducted recursively. The OT of a certain case is designated as xi, and the mean OT of all cases is designated as µ. For instance, the CUSUM_OT1 of case no.1 was the difference between the OT for the first case and µ. The CUSUM_OTn of case no. n is case no. (n – 1)'s CUSUM_OT(n−1) added to the difference between the OT for case no. n and µ. The calculation process was repeated until the final case was calculated, and CUSUM_{OT final} reaches zero[9, 11] .

Statistical analysis

IBM SPSS Statistics for Windows, version 25 (IBM Corp., Armonk, N.Y., USA) and Prism for Window, version 9.0 (GraphPad Software Inc., San Diego, CA, USA) were used for statistical analysis. Categorical variables are presented as numbers and percentages and were analyzed using Chi-squared test or Fisher Exact test when appropriate. As for continuous data, the normally distributed ones are shown as average ± standard deviation and compared using one-way analysis of variance. The non-normally distributed continuous variables are compared using the Kruskal–Wallis test. The learning curves of each surgeon were plotted using the aforementioned CUSUM methods. To screen out the parameters which were significantly associated with the OT, we established a multivariable linear regression model for OT and included all the relevant factors, such as age, BMI, previous pelvic surgery, pelvic adhesion, learning curve phase, estimated blood loss, uni- or bilaterality of cyst(s), occurrence complications, maximum cyst diameter, surgeon experience, parity, and pathologic types (chocolate or non-chocolate cyst). A two-sided p-value lower than 0.05 was considered statistically significant.

Overall profile of participants

Table 1 listed the general sociodemographic and perioperative information of all participants in our study. There were 127 patients with an average age of 35.52 ± 11.32 years and BMI of 21.83 ± 3.12 (kg/m2). The maximum diameter of their ovarian cysts was 5.33 ± 1.79 cm. Approximately 30% had undergone pelvic surgery before vNOTES ovarian cystectomy. Approximately one-quarter (27 cases) experienced cesarean sections, and 51 cases (40.2%) had delivered vaginally. Approximately 30% of patients had pelvic adhesion, among which 22 cases (17.3%) were mild, 9 cases (7.1%) were average, and the remaining 12 cases (9.4%) were complicated with severe pelvic adhesion. Only 4 cases (3.1%) had endometriosis (except ovarian chocolate cysts). Thirteen cases (10.2%) had bilateral cysts. The postoperative pathological diagnosis confirmed that there were 52 cases (40.9%) of teratoma, 39 cases (30.7%) of chocolate cyst, 20 cases (15.7%) of simple cyst, 16 cases (12.6%) of cystadenoma among all the cases. The patients had a postoperative hospitalization of 2.91 ± 0.93 days and intraoperatively lost 62.72 ± 93.93 mL blood.

Table 1

Description of the patients' characteristics(N = 127)
Variables*
Age (year)	35.52 ± 11.32
BMI (kg/m2)	21.83 ± 3.12
Max diameter of cyst (cm)	5.33 ± 1.79
History of pelvic surgery
0	91(71.7)
1	25(19.7)
2	11(8.7)
Previous delivery mode
Cesarean section	27(21.3)
Both	1(0.8)
Vaginal delivery	51(40.2)
None	48(37.8)
Presence of pelvic adhesion
None	84(66.1)
Mild	22(17.3)
Middle	9(7.1)
severe	12(9.4)
Presence of endometriosis**	4(3.1)
Laterality of cyst(s)
Unilateral	114(89.8)
Bilateral	13(10.2)
Postoperative pathological type
Teratoma	52(40.9)
Cystadenoma	16(12.6)
Simple cyst	20(15.7)
Chocolate cyst	39(30.7)
Postoperative hospitalization	2.91 ± 0.93
Intraoperative blood loss(ml)	62.72 ± 93.93
Postoperative pain score***
Day 0	2.87 ± 0.34
Day 1	2.10 ± 0.72
Day 2	1.26 ± 0.63
Day 3****	0.70 ± 0.60
Perioperative Infection	3(14.3)
Intraoperative complications*****	3(14.3)
* presented as Mean ± standard deviation or number of case with percentage
** Ovarian chocolate cysts were not included
***Visual Analogue Scale
****Only 29, 37, 29 and 18 patients in Phase I, II, III, IV were respectively included in the Day 3 VAS analysis due to the discharge of other patients.
***** 1 blood transfusion, 1 Paralytic ileus, 1 pelvic infection.

Learning curves of surgeons

The line graphs of OT and CUSUM value along individual case’s number (in chronologic order) for each surgeon were plotted in Fig. 2A and 2B (surgeon #1), Fig. 2C and 2D (surgeon #2), Fig. 2E and 2F (surgeon #3). As shown in Fig. 2A, 2C and 2E, their average OT of surgeon #1–3 were 64.44min, 67.35 and 57.17 min relatively. Generally, the surgeons spent longer than average OT in the exploration stage (Phase I) and shorter than average OT in the competency stage (Phase II). Consistently, we noticed that the CUMSUM curve of these three surgeons all presented an upward slope in Phase I and a downward slope in Phase II. A steep climbing trend in Phase I was observed in surgeon #1 (Fig. 2B), who performed 7 cases in the technique exploration stage (phase I). The surgeon #2 and #3 performed 9 and 16 cases respectively in the initial technique learning stage (Fig. 2D and 2F). As the surgeon and surgical team got more familiar with the vNOTES ovarian cystectomy procedure, their operation time started to decrease since the beginning of Phase II and their CUSUM curve also showed a downward slope. Surgeon #1 performed 11 cases (case 8–18) in Phase II; while surgeon #2 and #3 both performed 18 cases (case 10–27) in Phase II. Since Phase III, more technically challenging cases were assigned to these surgeons, especially the most experienced one, making their learning curve regained the upward trend, though some fluctuation could also be observed in the Phase III of surgeon #2 and #3’s CUSUM-Case-number curve. As a result, surgeon #1 conducted 13 cases (case 18–30) in Phase III; while surgeon #2 conducted 11 cases (case 27–37); surgeon #3 performed 10 cases in the challenge phase (case 33–42). As more surgeries performed, a downward slope re-appeared in final stage of surgeon #1’s learning curve, “demarcating” a fourth phase (18 cases, case 31–48), during which surgeon #1 gradually gained surgical proficiency of vNOTES ovarian cystectomy despite higher technical difficulty.

Clinical characteristics of cases in the four stages

Table 2 shows the comparison of the overall demographic and clinical features of participants in the four phases of vNOTES ovarian cystectomy. The comparisons of their age, BMI, gravidity, parity, delivery mode, previous history of abdominopelvic surgery, uni- or bilaterality of ovarian cysts, estimated intraoperative blood loss, Hb decrease at the 72nd postoperative hours, and postoperative pain score did not show any statistical significance. Meanwhile, we noticed a significantly shorter OT in phase II (53.66 ± 16.55 min) and IV (54.39 ± 23.45 min) as compared with phase I (68.74 ± 15.85) and III (75.93 ± 30.55) (p < 0.001). There were more cases with severe pelvic adhesion in later phases though not statistically significant [0 case in Phase I, 3 (6.4%) cases in Phase II, 4 cases (12.9%) in Phase III, and 5 cases (27.8%) in Phase IV, p = 0.591]. Similarly, there were significantly more cases with endometriosis (3 cases, 16.7%) in phase 4 than in phases I (1 case, 3.2%), II (0 case) and III (0 case) (p = 0.004). Moreover, the maximum cyst diameter significantly increased in later phases, with 4.79 ± 1.39 cm in phase I, 5.11 ± 1.70 cm and 5.93 ± 1.96 cm in Phase III, and 5.78 ± 2.06 cm in phase IV (p = 0.041). Similarly, the incidence of postoperative infection and other complications also increased remarkably in later phases, with 1 case (3.2%) of infection and blood transfusion occurring in phase III and two cases (11.1%) of infection, 1 case of paralytic ileus, and 1 pelvic infection in phase IV. Infection and other complications did not occur in phases I and II (p = 0.047).

Table 2

Characteristics and perioperative data of four phases
	Phase I(N = 31)	Phase II(N = 47)	Phase III(N = 31)	Phase Ⅳ(N = 18)	P-value
Age (year)	33.58 ± 9.97	36.17 ± 11.31	34.55 ± 7.92	38.83 ± 17.20	0.422^a
BMI (kg/m²)	21.44 ± 2.39	21༎62 ± 3.01	22.66 ± 3.65	21.65 ± 3.52	0.408^a
Gravidity					0.090^b
None	9(29.0)	7(14.9)	11(35.4)	6(33.3)
1	3(9.7)	18(38.3)	10(32.3)	3(16.7)
2	7(22.6)	12(25.5)	5(16.1)	2(11.1)
≥ 3	12(38.7)	10(21.3)	15(48.4)	7(38.9)
Parity					0.277^b
None	11(35.5)	14(29.8)	16(51.6)	7(38.9)
1	15(48.4)	26(55.3)	13(41.9)	6(33.3)
≥ 2	5(16.1)	7(14.9)	2(6.5)	5(27.8)
Delivery mode					0.582^b
None	11(35.5)	14(29.8)	16(51.6)	7(38.9)
Vaginal delivery	13(41.9)	22(46.8)	9(29.0)	7(38.9)
Cesarean section	7(22.6)	11(23.4)	5(16.1)	4(22.2)
Both	0	0	1(3.2)	0
History of abdominopelvic surgery					0.738^b
None	24(77.4)	31(66.0)	24(77.4)	12(66.7)
1	5(16.1)	12(25.5)	5(16.1)	3(16.7)
2	2(6.5)	4(8.5)	2(6.5)	3(16.7)
Presence of pelvic adhesion					0.591^c
None	25(80.6)	30(63.8)	18(58.1)	11(61.1)
Mild	5(16.2)	1021.3)	6(19.4)	1(3.2)
Middle	1(3.2)	4(8.5)	3(9.7)	1(3.2)
Severe	0	3(6.4)	4(12.9)	5(27.8)
Presence of endometriosis	1(3.2)	0	0	3(16.7)	0.004^c
Max diameter of cyst (cm)	4.79 ± 1.39	5.11 ± 1.70	5.93 ± 1.96	5.78 ± 2.06	0.041^a
Uni- or bilaterality of ovarian cysts					0.457^c
Unilateral	26(83.9)	44(93.6)	27(87.1)	17(94.4)
Bilateral	5(16.1)	3(6.4)	4(12.9)	1(5.6)
Operation time (min)	68.74 ± 15.85	53.66 ± 16.55	75.93 ± 30.55	54.39 ± 23.45	<0.001^c
Estimated blood loss (ml)	44.51 ± 35.58	52.87 ± 53.92	90.97 ± 149.76	71.11 ± 115.40	0.202^c
Postoperative hospitalization	2.87 ± 0.67	2༎81 ± 0.97	3.00 ± 0.52	3.11 ± 1.57	0.629^a
Postoperative pain score^*
Day 0	2.90 ± 0.30	2.89 ± 0.31	2.80 ± 0.40	2.83 ± 0.38	0.624^a
Day 1	2.16 ± 0.64	2.11 ± 0.73	2.13 ± 0.85	1.94 ± 0.64	0.778^a
Day 2	1.35 ± 0.61	1.22 ± 0.66	1.23 ± 0.67	1.28 ± 0.57	0.805^a
Day3^**	0.79 ± 0.56	0.76 ± 0.68	0.59 ± 0.50	0.61 ± 0.60	0.480^a
Hb decrease at postoperative 72 h (g/L)	20.71 ± 8.27	19.98 ± 9.00	22.68 ± 9.09	21.78 ± 10.30	0.614^a
Perioperative Infection	0	0	1(3.2)	2(11.1)	0.047^c
Intraoperative complications	0	0	1(3.2)^***	2(11.1)^****	0.047^c
^aAverage and standard deviation. One-way Analysis of Variance
^bNumber (percentage). Chi-squared Test
^cNumber (percentage). Fisher Exact Test
^dAverage and standard deviation. Kruskal-Wallis Test
*Visual Analogue Scale
**Only 29, 37, 29 and 18 patients in Phase I, II, III, IV were respectively included in the Day 3 VAS analysis due to the discharge of other patients.
***1 blood transfusion
****1 Paralytic ileus; 1 pelvic infection

As shown in Table 3, among all the 127 cases, there were 31 cases of vNOTES ovarian cystectomies performed in phase I, 47 cases performed in phase II, 37 cases in phase III, and 18 cases in phase IV. We further divided the cases in each phase into four subgroups including the teratoma, cystadenoma, simple cyst, and chocolate cyst according to their pathologic types. The Fisher exact test showed that there were markedly less teratoma (2 cases, 11.1%) and more cholate cysts (10 cases, 55.6%) conducted in phase IV. Conversely, there was an opposite trend in ovarian vNOTES in treating more challenging pathologic types, including chocolate cyst, the percentage of which climbed from 12.9% (4 cases) in phase I to 29.8% (14 cases) and 35.5% (11 cases) in phases II and III, respectively, to 55.6% (10 cases) in phase IV (p = 0.039).

Table 3

Description of the pathologic type of four phases
	Teratoma(N = 52)	Cystadenoma(N = 16)	Simple cyst(N = 20)	Chocolate cyst(N = 39)
Phase I (N = 31)	20(64.5)	3(9.7)	4(12.9)	4(12.9)
Phase II (N = 47)	17(36.2)	6(12.8)	10(21.3)	14(29.8)
Phase III (N = 31)	13(41.9)	4(12.9)	3(9.7)	11(35.5)
Phase Ⅳ (N = 18)	2(11.1)	3(16.7)	3(16.7)	10(55.6)
P value	0.039^a
^aNumber (percentage). Fisher Exact Test

Table 4 presents the OT of vNOTES in treating four different pathologic types of ovarian cysts in four distinct phases. The comparison in OT in treating different pathologic types within each phase did not show any statistical significance. The comparison of OT for specific pathologic types in different phases showed that there was a remarkable decrease in OT in treating teratoma and cystadenoma in the competency phase and a noticeable surge of OT in challenge phase. The mean ± standard deviation of vNOTES in treating teratoma dropped from 70.20 ± 14.81 min in phase I to 46.06 ± 11.33 min in phase II, rebounded to 74.38 ± 33.75 min in phase III, and eventually decreased to 42.40 ± 6.58 min in phase IV (p<0.001). A similar trend was also observed in vNOTES cystadenoma cystectomy, the OT of which decreased from 88.33 ± 14.43 min in phase I to 50.83 ± 9.37 min in phase II and surged again to 69.00 ± 39.56 min in phase III and 71.67 ± 20.82 min in phase IV (p = 0 .040). vNOTES ovarian cystectomies for treating simple cysts had similar average OT in phases I (51.75 ± 3.50 min) and II (55.80 ± 19.46 min), but remarkably longer OT in phase III (69.33 ± 12.90 min) and shorter OT in phase IV (38.00 ± 1.73 min). Such comparison of vNOTES chocolate cystectomy did not reveal any statistical difference, though a longer OT in the 3rd phase (82.09 ± 29.20 min) was observed (p = 0.206).

Table 4

Operation time (in minutes) among the four phases stratified by pathologic disease
	Teratoma (N = 52)	Cystadenoma (N = 16)	Simple cyst (N = 20)	Chocolate cyst (N = 39)	P-value
Phase I (N = 31)	70.20 ± 14.81	88.33 ± 14.43	51.75 ± 3.50	63.75 ± 12.50	0.134
Phase II (N = 47)	46.06 ± 11.33	50.83 ± 9.37	55.80 ± 19.46	62.57 ± 18.64	0.188
Phase III (N = 31)	74.38 ± 33.75	69.00 ± 39.56	69.33 ± 12.90	82.09 ± 29.20	0.857
Phase Ⅳ (N = 18)	42.40 ± 6.58	71.67 ± 20.82	38.00 ± 1.73	62.57 ± 30.04	0.148
P-value	<0.001	0.040	0.013	0.206
^aAverage and standard deviation. One-way Analysis of Variance

Multivariate regression analysis for operation time of vNOTES ovarian cystectomy

To further evaluate the potential significant association between OT and other factors, a multivariable linear regression model was established using the aforementioned methods. The results (Fig. 3) revealed that there are significant associations between OT and the following variables: (1) presence of pelvic adhesion [adjusted odds ratio (OR) 7.149 (0.506, 13.792), p = 0.035], (2) presence of bilateral cyst [adjusted OR 16.996 (2.155, 31.837), p = 0.025], (3) max diameter of cyst[adjusted OR 2.799 (0.174, 5.425), p = 0.037], and (4) surgeon’s experience [adjusted OR -6.118 (-11.814, -0.423), p = 0.035].

vNOTES is quite different from LESS and conventional laparoendoscopic surgery in multiple aspects. The establishment of a surgical path might be the most challenging one, due to the risks of injuring neighboring organs, including the rectum and urinary bladder [12, 13]. More than 1,000 cases of ovarian vNOTES were conducted in our institute between December 2018 and October 2022 which were all performed through culdotomy in the posterior vaginal fornix[10]. After the initial technical exploration via approximately 30 cases of tentative vNOTES operation by two experts with more than 2 decades of experience in vaginal and laparoendoscopic surgeries, we established a standard operating procedure for conducting vNOTES and achieved better surgical outcomes and cosmetic satisfaction, faster recovery, and reduced postoperative pain compared with LESS or conventional laparoendoscopy[6, 9]. In this Standard Operating Procedure (SOP), we divested the cyst in the same manner as in LESS cystectomy after establishing the surgical platform. Our clinical practice of ovarian-vNOTES also supported the viewpoint that vNOTES is easier to perform compared to LESS because of the closer proximity to the ovaries via the culdotomy entrance, less severe chopstick effects, and easier specimen removal[12] .

According to the international consensus of vNOTES experts, beginners should start from learning complete hysterectomy via vNOTES since the surgical path established therefrom would be much easier to access than the culdotomy in the posterior or anterior vaginal fornix[14]. Others state that vNOTES has a long learning process and may be quite challenging in inexperienced hands[13]. However, in our department, we observed that ovarian cystectomy via vNOTES is much more frequently conducted even by less experienced hands due to the high incidence of ovarian and tubal diseases. Moreover, posterior vaginal fornix culdotomy is the dominantly preferred entrance for vNOTES surgery. Through a reasonably designed learning process (performing technically difficult cases in later stage or assigning them to experienced hands), our data showed that, even without experiencing vNOTES hysterectomy, surgeons with different levels of experience in performing laparoendoscopy could also master ovarian cystectomy via vNOTES in relatively less cases. The experienced surgeons only needed seven and nine cases, respectively, in the technique-acquiring phase, while the least experienced surgeons performed only 16 cases to reach competency, which supported the feasibility and promising prospect of promoting ovarian vNOTES. For young surgeons lacking experiences in conducting vaginal surgeries, they could also complete ovarian cystectomy via vNOTES after standardized training.

Our study assigned relatively diverse pathologic types in the surgeons’ different learning phases, among which the chocolate cyst almost takes up one-third of the total cases. Some previous studies reported that endometriosis might cause adhesion in the pouch of Douglas, and that vNOTES ovarian cystectomy of chocolate cysts should be avoided.[15] Probably due to this reason, very limited cases of vNOTES endometriotic cystectomy has been reported[12]. Moreover, a previous study of the learning curve of adnexal vNOTES did not include chocolate cysts[16]. In our study, we noticed that vNOTES may still be used to treat ovarian endometrioma with detailed preoperative examination and assessment to exclude cases with severe adhesion in the pouch of Douglas since it was not significantly related to the OT and surgical conversion did not occur; however, several cases of endometriosis and severe pelvic adhesion were included.

To date, there are many studies on the learning curve of vNOTES hysterectomy, while that of vNOTES ovarian cystectomy or adnexectomy was only reported once by Huang et al [9, 11, 16, 17]. Compared with their study, our study had a similar cohort size but included more surgeons with different levels of gynecologic endoscopy experience. Moreover, we divided the learning curves into more stages, which could be a better reference for more surgeons. We also noticed that the mass diameter in their study was not significantly associated with the OT. In our study, cyst size and bilaterality are both positively associated with the OT, which might be caused by the different pathologic types included in our study. Another mismatch between our studies was that they did not find any cut-off point for determining the volume of cases needed to achieve mastery of adnexal vNOTES. Conversely, we noticed that seven to 16 cases were enough for initial technique establishment. This might also be explained by the various pathologic and surgeon types included in our study, given that all the cases were conducted by a single high-volume surgeon in their study. Nevertheless, our results both suggested that it might be more appropriate to start learning the vNOTES technique from ovarian or adnexal vNOTES, and less cases are needed to learn the technique and achieve proficiency than the vNOTES hysterectomy.

There are also several limitations in present study. Firstly, the surgeons who performed the vNOTES ovarian surgeries in our study may have performed more laparoscopic surgeries than many other less populous regions and countries, which may impair the generality of our findings. Secondly, the preoperative assessment of surgical difficulty was mainly made by subjective standards rather than standardized stratification or scoring system, which, to some extent, may hinder the assignment of surgeries to different surgeons and phases.

Our learning curve analysis of ovarian vNOTES depicted four specific stages. vNOTES for ovarian cystectomy could be mastered after performing seven, nine, and 16 cases by surgeons with different endoscopic surgical experience. The presence of pelvic adhesion or bilateral cyst and cyst size were positively related to OT, while the surgeon’s experience was negatively correlated to OT.

vNOTES: transvaginal natural orifice transluminal endoscopic surgery

MIS: minimally invasive surgery

TU-LESS: transumbilical laparoscopic single site surgery

OT: operation time

BMI: body mass index

CUSUM: cumulative sum

Hb:hemoglobin

Ethics approval and consent to participate

This study was approved by the ethics committee of Chengdu Women's and Children's Central Hospital [No.WCCH-B2021(39)]. Before the surgery, All patients were well-informed of the potential risks and advantages of vNOTES and other alternative surgical treatments, including TU-LESS or multi-port laparoscopy, and expressed their preference for vNOTES. Written informed consent was obtained from all the participants during the preoperative consultation. All identifiable private information were deleted or not transcribed while extracting the patients’ pathological and clinical characteristics from the Hospital Information System to protect patients’ privacy.

Consent for publication

Not applicable.

Availability of data and materials

Data supporting the conclusion are provided within the manuscript or supplementary information files. The original data are available from the corresponding authors upon reasonable request.

Competing interests

The authors declare no conflict of interest.

Funding

Chengdu High-level Key Clinical Specialty Construction Project, Chengdu Municipal Health Commission Project (Grant number:2021215), Fifth Round of Chengdu Municipal Science and Technology Research Program (Grant number:2021-YF05-00627-SN), Chengdu Municipal Medical Research Project in 2023 (Grant number:2023352), and Japan China Sasakawa Medical Fellowship Program(Grant number:4408).

Author contribution

This study was designed by Dan Feng, Tianjiao Liu, and Xin Li. Dan Feng and Lu Huang completed the data collection. Tianjiao Liu, Dan Feng, and Xin Li drafted this manuscript. Tianjiao Liu and Xin Li performed the statistical analysis. Li Xiao, Li He and Yonghong Lin supervised this project and critically reviewed this manuscript. All authors participated in the revision of the manuscript.

Acknowledgements

We thank Dr.Qiannan Hou for performing the vNOTES ovarian cystectomies. We also feel grateful for the participation of all the patients.

Borgfeldt C, Andolf E. Transvaginal sonographic ovarian findings in a random sample of women 25–40 years old. Ultrasound Obstet Gynecol. 1999;13(5):345–50.
Bottomley C, Bourne T. Diagnosis and management of ovarian cyst accidents. Best Pract Res Clin Obstet Gynaecol. 2009;23(5):711–24. 10.1016/j.bpobgyn.2009.02.001.
Sadowski EA, Paroder V, Patel-Lippmann K, et al. Indeterminate Adnexal Cysts at US: Prevalence and Characteristics of Ovarian Cancer. Radiology. 2018;287(3):1041–9. 10.1148/radiol.2018172271.
Zhang C, Duan K, Fang F, et al. Comparison of Transvaginal and Transumbilical Laparoscopic Single-Site Surgery for Ovarian Cysts. JSLS. 2021;25(2). 10.4293/JSLS.2021.00019.
Housmans S, Noori N, Kapurubandara S, et al. Systematic Review and Meta-Analysis on Hysterectomy by Vaginal Natural Orifice Transluminal Endoscopic Surgery (vNOTES) Compared to Laparoscopic Hysterectomy for Benign Indications. J Clin Med. 2020;9(12). 10.3390/jcm9123959.
Huang L, Lin Y-H, Yang Y, et al. Comparative analysis of vaginal natural orifice transluminal endoscopic surgery versus transumbilical laparoendoscopic single-site surgery in ovarian cystectomy. J Obstet Gynaecol Res. 2021;47(2):757–64. 10.1111/jog.14603.
Yoshiki N. Review of transvaginal natural orifice transluminal endoscopic surgery in gynecology. Gynecol Minim Invasive Ther. 2017;6(1):1–5. 10.1016/j.gmit.2016.11.007.
Li C-B, Hua K-Q. Transvaginal natural orifice transluminal endoscopic surgery (vNOTES) in gynecologic surgeries: A systematic review. Asian J Surg. 2020;43(1):44–51. 10.1016/j.asjsur.2019.07.014.
Huang L, He L, Huang L, et al. Learning curve analysis of transvaginal natural orifice transluminal endoscopic hysterectomy combined under the standard operating procedure. Int J Gynaecol Obstet. 2022;159(3):689–95. 10.1002/ijgo.14238.
Huang L, Feng D, Gu D-X, et al. Transvaginal natural orifice transluminal endoscopic surgery in gynecological procedure: Experience of a Women's and Children's Medical Center from China. J Obstet Gynaecol Res. 2022;48(11):2926–34. 10.1111/jog.15402.
Park JS, Ahn HK, Na J, et al. Cumulative sum analysis of learning curve for video-assisted mini-laparotomy partial nephrectomy in renal cell carcinoma. Med (Baltim). 2019;98(17):e15367. 10.1097/MD.0000000000015367.
Li Y-C, Ku F-C, Kuo H-H, et al. Transvaginal endoscopic surgery-assisted versus conventional laparoscopic adnexectomy (TVEA vs. CLA): A propensity-matched study and literature review. Taiwan J Obstet Gynecol. 2017;56(3):336–41. 10.1016/j.tjog.2017.04.013.
Brandão P, Almeida A, Ramôa P. vNOTES for adnexal procedures. J Obstet Gynaecol. 2021;41(7):1134–8. 10.1080/01443615.2020.1841124.
Kapurubandara S, Lowenstein L, Salvay H, et al. Consensus on safe implementation of vaginal natural orifice transluminal endoscopic surgery (vNOTES). Eur J Obstet Gynecol Reprod Biol. 2021;263:216–22. 10.1016/j.ejogrb.2021.06.019.
Wang C-J, Wu P-Y, Kuo H-H, et al. Natural orifice transluminal endoscopic surgery-assisted versus laparoscopic ovarian cystectomy (NAOC vs. LOC): a case-matched study. Surg Endosc. 2016;30(3):1227–34. 10.1007/s00464-015-4315-6.
Huang Y-T, Yang L-Y, Pan Y-B, et al. Learning Curve Analysis of Transvaginal Natural Orifice Adnexal Surgery. J Minim Invasive Gynecol. 2020;27(2):489–97. 10.1016/j.jmig.2019.04.009.
Lauterbach R, Matanes E, Amit A, et al. Transvaginal Natural Orifice Transluminal Endoscopic (vNOTES) Hysterectomy Learning Curve: Feasibility in the Hands of Skilled Gynecologists. Isr Med Assoc J. 2020;22(1):13–6.

No competing interests reported.

Supplementarymaterial.doc

Download PDF

Journal Publication

published 25 Jul, 2024

Read the published version in BMC Women's Health →

Editorial decision: Revision requested
03 Jun, 2024
Reviews received at journal
12 May, 2024
Reviewers agreed at journal
02 May, 2024
Reviews received at journal
24 Apr, 2024
Reviewers agreed at journal
10 Apr, 2024
Reviewers invited by journal
09 Apr, 2024
Editor invited by journal
27 Mar, 2024
Submission checks completed at journal
25 Mar, 2024
Editor assigned by journal
25 Mar, 2024
First submitted to journal
22 Mar, 2024

You are reading this latest preprint version

Learning Curve Analysis of Transvaginal Natural Orifice Transluminal Endoscopic Surgery in Treating Ovarian Cysts: A Retrospective Cohort Study

Status:

Journal Publication

Version 1

Abstract

Figures

Background

Methods

Study design

Patient selection

Operating procedure of vNOTES ovarian cystectomy

Learning curve analysis

Statistical analysis

Results

Overall profile of participants

Learning curves of surgeons

Clinical characteristics of cases in the four stages

Multivariate regression analysis for operation time of vNOTES ovarian cystectomy

Discussion

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1