Surgical outcomes of laparoscopic hysterectomy with concomitant endometriosis without bowel or bladder dissection: a cohort analysis to define a case-mix variable

Background: Pelvic endometriosis is often mentioned as one of the variables influencing surgical outcomes of laparoscopic hysterectomy (LH). However, its additional surgical risks have not been well established. The aim of this study was to analyze to what extent concomitant endometriosis influences surgical outcomes of LH and to determine if it should be considered as case-mix variable.

A total of 2655 LH ’s were analyzed, of which 397 (15.0%) with concomitant endometriosis. For blood loss and operative time, no measurable association was found for stages I ( n = 106) and II ( n = 103) endometriosis compared to LH without endometriosis. LH with stages III ( n =9 3 )a n dI V(n = 95) endometriosis were associated with more intra- operative blood loss (p = < .001) and a prolonged operative time ( p = < .001) compared to LH without endometriosis. No significant association was found between endometriosis (all stages) and complications ( p =. 6 2 ) .

The findings of our study have provided numeric support for the influence of concomitant endometriosis on surgical outcomes of LH, without bowel or bladder dissection. Only stages III and IV were associated with a longer operative time and more blood loss and should thus be considered as case-mix variables in future quality measurement tools.

Concomitant endometriosis, Laparoscopic hysterectomy, Case-mix correction, Surgical outcome measures

Measuring surgical outcomes to improve quality of health care has received increasing attention over the past decades. Consequently, many national registra- tion systems have been developed to collect hospital data and compare surgical outcomes between hospi- tals or even surgeons. An important limitation of most of these registrations is the lack of correction of case-mix variables [ 1]. Case-mix variables are defined as specific patient characteristics that are known to independently influence surgical outcome measures and that are potentially explaining the differences in outcomes between hospitals and/or surgeons [ 1, 2]. Thus, for an appropriate and reliable interpretation of the surgical outcomes, a case-mix correction is abso- lutely mandatory. Specific for the laparoscopic hysterectomy (LH), studies have demonstrated that BMI, uterine weight, and previous procedures influence surgical outcome measures [ 1, 2]. As a result, a case-mix correction for these variables has been recommended when comparing outcomes of LH [ 1, 2]. In a recent sys- tematic review on this top ic, pelvic endometriosis * Correspondence: [email protected] 1Department of Gynecology, Leiden University Medical Centre, Leiden, the Netherlands 4Department BioMechanical Engineering, Delft University of Technology, Delft, the Netherlands Full list of author information is available at the end of the article Gynecological Surgery © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License ( http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Sandberg et al. Gynecological Surgery (2018) 15:8 https://doi.org/10.1186/s10397-018-1039-3 was also found to be a potential factor influencing surgical outcomes of LH [ 2]. However, no further

on concomita nt endometriosis could be drawn in this review as available evidence was lim- ited. Indeed, only three retrospective studies were found that demonstrated an association between endometriosis and prolonged operative time and in- creased complication risk of LH [ 3–5]. Furthermore, none of these studies applied a correction for the known case-mix variables, which may have poten- tially resulted in an over- or underestimation of the impact of endometriosis. The objective of this study was firstly to analyze to what extent concomitant endometriosis influences surgical outcomes of LH and secondly, to determine if concomitant endometriosis should be considered as case-mix variables in future quality tools.

In this study, all hysterectomies registered, between April 2014 and September 2016, in the web-based ap- plication of QUSUM ( https://www.qusum.org )w e r e included [ 1]. Detailed information on this database has been described elsewhere [ 1], but briefly, data were obtained by asking gynecologists performing lap- aroscopic surgery to register anonymously their con- secutive LH ’s for benign indication and low-grade malignancy (e.g., cervical and endometrial dysplasia). All Dutch gynecologists performing LH ’s were re- cruited via a personal e-mail invitation, and gynecolo- gists from abroad were asked to participate in the study at international meetings and conferences. The Institutional Review Board (IRB) of Leiden University Medical Centre, Leiden, the Netherlands reviewed the study and exempted it from IRB approval. To evaluate the association between endometriosis and surgical outcomes, we retrospectively compared surgical outcomes of patients undergoing LH with and without concomitant endometriosis. Only proce- d u r e sw i t hh y s t e r e c t o m i e sa so n l ys u r g e r yw e r ec o n - sidered, including LH ’s with concomitant adnexal surgery. Hysterectomies including bowel or bladder dissection of endometriosis were excluded. Indeed, it did not seem fair to compare these operations to ‘simple ’ LH’s for benign indication or low-grade malignancy. We considered as primary outcomes the following three surgical outcome measures: intra-operative blood loss (ml), operative time (listed as time from incision to closure), and complications (up to 6 weeks postoperatively). Complications were defined accord- ing to the internationally accepted classification of the Dutch Society of Obstetrics and Gynecology (Appendix 1 )[ 6]. The severity of the complications was graded according to the following classification: level A, recovery without reoperation; level B, reop- eration indicated; level C, permanent injury and loss of function; level D, death [ 6]. Only reactive conver- sions to laparotomy, as defined by Blikkendaal et al., were included [ 7]. To adequately compare LH ’sw i t h and without endometriosis, a correction of the three known case-mix variables BMI (kg/m 2), uterine weight (grams), and previous abdominal procedures (previous laparotomy and therapeutic laparoscopy) was applied [ 2]. For each LH registered in the QUSUM database, we abstracted the data of the three primary outcomes, the three case-mix variables, and the following base- line characteristics: the presence and stage of con- comitant endometriosis, the age of the patients at procedure (years), the type of hysterectomy (total lap- aroscopic hysterectomy (TLH)), supracervical laparo- scopic hysterectomy (SLH), laparoscopic-assisted vaginal hysterectomy (LAVH) and robotic hysterec- tomy (RH)), and the experience of the surgeons (years of experience, number of LH ’s per year, and total number of LH ’s performed). Gynecologists were asked to enter the information on experience at initial regis- t r a t i o ni nt h ew e ba p p l i c a t i o n[1]. Endometriosis was classified into four stages, ac- cording to the revised definition of the American Society for Reproductive Medicine [ 8]. Stage I was defined as minimal endometriosis with only superfi- cial lesions and a few filmy adhesions; stage II, the mild variant, included additional deep lesions in the Douglas cavity; stage III was a moderate stage where, in addition to the previous stages, endometriomas on the ovary are observed together with more dense adhesions. Finally, stage IV , the severe endometriosis, included large endometriomas and extensive adhesions. Statistics For the statistical analysis, SPSS version 23.0 (SPSS Inc., Chicago, IL, USA) and R statistical software ver- sion 3.3.1 were used [ 9]. Categorical data were pre- sented as frequency with percentages (%) and continuous data were presented as mean with stand- ard deviation (SD) or as geometric mean with geo- metric standard deviation (GSD), if data were skewed. To assess significant differences between the baseline characteristics of the group with endometri- osis and the one without, independent sample t tests, Fisher ’s tests, and chi-square with trends were used as appropriate. Statistical significance was defined as a p value < .05. A (generalized) linear mixed model regression (univariable and multivariable analysis) was Sandberg et al. Gynecological Surgery (2018) 15:8 Page 2 of 9 performed to define the association between endo- metriosis and surgical outcomes. The case-mix vari- ables BMI, uterine weight, and previous procedures were included as co-variables [ 1, 10]. The characteris- tic “previous procedures ” was dichotomized into no previous procedures or at least one. Data with a skewed distribution were transformed for analysis into log values. We found t hat blood loss, operative time, and uterine weight needed a log transformation, which resulted in normalization of the distribution. In 55 cases (including five patients with endometri- osis) surgeons reported no blood loss after procedure (0 ml). Blood loss was ther efore transformed into log(x +1 ) . To correct for the fact that surgeons entered various procedures, analyses were performed with a random intercept for the performing surgeon. For blood loss and operative time, findings of the regression analyses were presented as percentage increase or decrease with 95% confidence interval (95% CI). The results for complica- tion risk were reported as odds ratio with 95% CI. To make data easier to interpret, an index patient was used to demonstrate the additional effect of the different stages of endometriosis on surgical outcomes. The other case-mix characteristics of this index patient were fixed and based on the mean values of the entire cohort.

During the study period, a total of 2655 LH ’sw e r e analyzed of which 397 cases (15.0%) with concomitant endometriosis. Table 1 gives an overview of the base- line patient characteristics and surgical outcomes. Pa- tients in the endometriosis group were younger (43.5 Table 1 Baseline patient characteristics and surgical outcomes of LH with and without endometriosis LH with endometriosis n = 397 LH without endometriosis n = 2258 p value Patient characteristics Age, years, mean ± SD 43.5 ± 7.7 49.8 ± 11.9 < .001 BMI, kg/m2, mean ± SD 27.9 ± 5.5 28.6 ± 6.4 .04 Uterine weight, gram, geometric mean; GSD* 149.31 ± 1.8 162.5 ± 2.1 < .001 Previous abdominal procedures, n (%) Two 57 (14.4) 142 (6.3) Stage of endometriosis, n (%) Stage I 106 (26.7) – Stage II 103 (25.9) Stage III 93 (23.4) Stage IV 95 (23.9) Procedure type, n (%) .03 TLH 376 (94.7) 2036 (90.2) SLH 13 (3.3) 109 (4.8) LAVH 6 (1.5) 88 (3.9) Robotic 2 (0.5) 25 (1.1) Surgical outcomes Blood loss (ml), geometric mean; GSD* 80 ± 3.1 67 ± 3.4 ɸ Operative time (min), geometric mean; GSD* 94 ± 1.5 91 ± 1.5 ɸ Complications, n (%) 24 in 22 patients (5.5%) 175 in 153 patients (6.8%) ɸ Lesion Bladder 3 (0.8) 27 (1.2) Ureter 1 (0.3) 9 (0.4) Vessel – 1 ( 1000 mL intra-operative 5 (1.3) 13 (0.6) Wound dehiscence – 6 (0.3) Thrombosis – 3 (0.1) Dysfunction ileus – 2 (0.1) Infections 1 (0.3) 50 (2.2) Dysfunction incontinence – 10 (0.4) Other 2 (0.5) 11 (0.5) Conversion 10 (2.5) 8 (0.4) .16 Table 1 Baseline patient characteristics and surgical outcomes of LH with and without endometriosis (Continued) LH with endometriosis n = 397 LH without endometriosis n = 2258 p value Severity of complications, n (%) Level A–recovery 18 (4.5) 113 (5.0) Level B–reoperation 3 (0.8) 39 (1.7) Level C–permanent injury 1 (0.3) 0 Level D–death 0 1 (< 0.1) Data are presented as number (percentage) or as mean ± standard deviation or as geometric mean (GM) with geometric standard deviation (GSD) *Interpretation of the GSD: 95% of the data are expected to lie in the range of GM/(GSD)2 to GM*(GSD) 2 Statistics: independent t test for continuous data; Fisher ’s test for categorical data; chi-square with trend for number of previous procedures ɸThis means that for every tenfold increase in uterine weight, blood loss and operative time increased by the stated percentages and odds of complication increased by the stated OR SD standard deviation, GSD geometric standard deviation, LH laparoscopic hysterectomy, TLH total laparoscopic hysterectomy, SLH supracervical laparoscopic hysterectomy, LAVH laparoscopic-assisted vaginal hysterectomy Sandberg et al. Gynecological Surgery (2018) 15:8 Page 3 of 9 (7.7) versus 49.5 (11.9), p < .001), had lower uterine weight (149.3 (1.8) gram versus 162.5 (2.1), p < .001), lower BMI (27.9 (5.5) versus 28.6 (6.4), p = .04), and had had more previous abdominal surgeries (61.6 ver- sus 42.9%, p < .001). Most of the hysterectomies performed were TLH (94.7% in the endometriosis group versus 90.2%). The four stages of endometriosis were almost equally divided in the group with endometriosis (stage I: n = 106 (26.7%); stage II: n = 103 (25.9%); stage III: n =9 3 (23.4%); stage IV: n = 95 (23.9%)). A total of 199 compli- cations occurred in 175 patients (6.6%), including 22 patients with endometriosis. Regarding the severity of the complications, no difference between the group with and without endometriosis was observed ( p = .16). Spe- cifically, no increased risk was observed in the endomet- riosis group for the different organ injuries (bladder, ureter, and bowel). No p values were calculated for the three primary surgical outcomes, as these data were used in the regression analysis (Table 3). As demonstrated in Table 2, a total of 93 surgeons reg- istered on average 33 (30.6) procedures during the study period, performed yearly 31 (17.5) LH ’s and had on aver- age 6 years (4.5) of experience. As the data were entered anonymously, it is unknown how many hospitals were involved. A significant difference was observed for the overall surgical experience and LH ’s performed with and without endometriosis ( p < .001). In 50.1% of the cases with endometriosis, procedures were performed by sur- geons with an experience of at least 200 LH ’s, compared to 34.8% in the group without endometriosis. Of note, no significant correla tion was observed between the number of LH ’s registered in the application and the overall experience of the surgeons (Spearman ’s rho = 0.145, p = .169, Appendix 2 ). Table 3 summarizes the association between the dif- ferent case-mix variables, the stages of endometriosis, and surgical outcomes, using the mixed model regres- sion analysis (univariate a nd multivariate analysis). For BMI, uterine weight and previous abdominal pro- cedures, significant associations were observed with blood loss and operative time. Regarding endometri- osis, we demonstrated that for blood loss and opera- tive time, no measurable association was found for stages I and II endometriosis compared to LH ’sw i t h - out endometriosis (geometric mean blood loss: no endometriosis: 67 (3.4) mL; stage I: 71.4 (3.2), p = .48; stage II: 79.0 (2.8), p = .10; operative time: no endo- metriosis 90.6 (1.5) min; stage I: 88.1 (1.5), p = .23; stage II: 81.9 (1.5), p = .68). Compared to LH ’sw i t h - out endometriosis, LH ’sw i t hs t a g eI I Ia n dI Ve n d o - metriosis were associated with more intra-operative blood loss (stage III: 70 (3.6), p =. 0 1 ) a n d s t a g e I V : 106.4 (2.9), p = < . 0 0 1 )a n dap r o l o n g e do p e r a t i v et i m e (stage III: 95.1 (1.5), p = <.001 and stage IV: 117.6 (1.6), p = <.001). No significant association was found between endometriosis (all stages) and complication rates ( p = .62). For an index patient with stage IV endometriosis, a mean intra-operative blood loss of 140.8 ml (109.4 – 180.9) was expected compared to a mean of 65.7 ml (57.2–75.6) for a patient with the same characteristics but without endometriosis. Also, an additional mean operative time of 47 min was demonstrated for an Table 2 Surgeon’s characteristics Surgeon characteristics* Total n = 93 surgeonsɸ Number of procedures registered in QUSUM 33 (30.6), (1 –130) Number of procedures per year 31 (17.5), (10 –140) Years of experience 6 (4.5), (0 –21) Overall experience 160 (158.9), (0 –800) Total surgical experience for the LH ’s with and without endometriosis** Total n =9 3 surgeonsɸ LH’s with endometriosis n = 397 LH’s without endometriosis n = 2258 LH p value Overall experience < .001 0–99 42 (45.2) 125 (31.5) 954 (43.4) 100–199 21 (22.6) 73 (18.4) 518 (21.8) 200–299 10 (10.8) 99 (24.9) 347 (15.4) ≥ 300 19 (20.4) 100 (25.2) 437 (19.4) *Data are presented as mean (SD), (minimum-maximum) **Data are presented as number (percentage) ɸFor one surgeon no information on experience was available Statistics: chi-square test for trend LH laparoscopic hysterectomy Number of procedures per year is the number of LH they perform an average on a yearly basis Years of experience is defined as the number of years of experience since they have finished residency Overall experience is defined as the total number of LH ’s performed by a gynecologist during his or her career as attending, including the teaching cases Sandberg et al. Gynecological Surgery (2018) 15:8 Page 4 of 9 Table 3 Influence of each covariate on blood loss, operative time and complications Crude analysis Adjusted analysis Blood loss (Geometric) mean + (G)SD§ or number (%) Percentage increase (95% CI) p value of univariate analysis Percentage increase (95% CI) Index patient* Data expressed in ml (95% CI) p value of multivariate analysis Patient characteristics – 10LOG uterine weight ɸ 153.6; 2.0 84.8 (74.5 –95.4) < .001 BMI (increase/1 kg/ m2) 28.5 ± 6.2 2.4 (1.8 –3.0) < .001 Previous procedures - Yes (versus no) 1213 (45.7) 13.1 (4.3 –22.6) < .001 Stage of endometriosis < .001 < .001 No endometriosis 67.3; 3.4 Reference Reference 65.7 (57.2 –75.6) Stage I 71.4; 3.2 4.9 ( −16.1–31.2) .68 7.6 ( − 12.3–31.9) 70.7 (55.7 –89.7) .48 Stage II 79.0; 2.8 16.0 ( − 7.3–44.9) .20 18.8 ( − 3.2–45.6) 78.0 (61.4 –99.2) .10 Stage III 70.0; 3.6 27.9 (1 –62.1) .04 34.0 (8.0 –66.4) 88.1 (68.7 –113.0) .01 Stage IV 106.4; 2.9 98.8 (56.7 –152.4) < .001 114.3 (72.1 –166.7) 140.8 (109.4 – 180.9) < .001 Operative time Number (%) Percentage increase (95% CI) P-value of univariate analysis Percentage increase (95% CI) Index patient* Data expressed in min (95% CI) p value of multivariate analysis Patient characteristics – 10LOG uterine weight ɸ 20.0 (18.1–21.8) < .001 BMI (increase/1 kg/ m2) 1.0 (0.8–1.2) < .001 Previous procedures – Yes (versus no) 3.8 (1.5 –6.1) <.001 Stage of endometriosis < 0.001 < .001 No endometriosis 90.6; 1.5 Reference Reference 92.0 (87.4 –96.9) Stage I 88.1; 1.5 − 4.4 (− 10.1–1.7) .16 − 3.3 (− 8.6–2.1) 88.9 (82.7 –95.7) .23 Stage II 81.9; 1.5 0.3 ( − 5.7–6.7) .93 1.2 ( − 4.3–6.9) 93.1 (86.5 –100.3) .68 Stage III 95.1; 1.5 15.3 (8.0 –23.1) < .001 17.0 (10.3 –24.0) 107.7 (99.8 –116.2) < .001 Stage IV 117.6; 1.6 47.6 (38.0 –57.6) < .001 51.0 (42.3 –60.3) 139.1 (128.8 – 150.1) < .001 Complications Number (%) Odds ratio (95% CI) p value of univariate analysis Odds ratio (95% CI) – p value of multivariate analysis Patient characteristics – 10LOG uterine weight ɸ 0.69 (0.56–0.82) < .001 BMI (increase/1 kg/ m2) 1.00 (0.98–1.03) .935 Previous procedures – Yes (versus no) 1.17 (0.87 –1.58) .305 Stage of endometriosis .72 .62 No endometriosis 153 (6.8) Reference Reference Stage I (versus no) 5 (2.9) 1.27 (0.55 –2.95) .57 1.27 (0.54 –2.98) .59 Stage II 7 (4.0) 0.69 (0.33 –1.43) .32 0.67 (0.32 –1.36) .26 Sandberg et al. Gynecological Surgery (2018) 15:8 Page 5 of 9 index patient with stage IV endometriosis compared to a patient with the same characteristics but without endometriosis (139.1 min (128.8 –150.1) versus 92 min (87.4–96.9)).

Main findings In this present study, we demonstrated that LH with concomitant stages I or II endometriosis had no measurable and clinical relevant associations with sur- gical outcome measures compared to LH without endometriosis. Stages III and IV endometriosis, how- ever, appeared to be of influence for the outcomes blood loss and operative time. Strengths and limitations A limitation of our study was that gynecologists reg- istered themselves the procedures and that these data could not be verified as entered anonymously. As a result, we cannot guarantee that the surgeons in- cluded all their consecutiv ec a s e s .Y e t ,t h i si ss i m i l a r to daily clinical practice where data are also entered by the surgeons. Moreover, previous studies have shown that these self-reported data are often accurate and reliable [ 10, 11]. Another limitation is the fact that information regarding the setting, the indication of surgery, or the participation of a fellow or resident during the procedure was missing. This might have biased the outcomes. Although these data would have been in- teresting, the baseline cha racteristics (BMI, uterine weight, and previous procedures), known to influence outcomes were analyzed and therefore we believe our comparison is reliable. Regarding the participation of at r a i n e ew eb e l i e v et h a ti ti st h er e s p o n s i b i l i t yo f the principle surgeon to judge if his or her involve- ment remains acceptable and therefore, no correction for this factor was applied. Additionally, it would have been interesting to know if surgeons knew pre- operatively of the presence of endometriosis as this might be relevant for the choice of the surgeon and the surgical planning. Although the outcome “previous surgeries ” has been demonstrated to be associated with worse sur- gical outcome [ 2, 10], collecting data on adhesions would have been interesting as well. The advantage of the registration we used was that the number of previous procedures was an objective measure, whereas adhesions and their grading would have been more at risk for intra-observer variations [ 12]. Finally, in our study, surgery with bowel and bladder dissection of endometriosis was excluded. This might have resulted in an underestimation of the impact of endometriosis on surgery in general. Our data should thus not be generalized to all endometriosis cases but are limited to LH with concomitant endometri- osis. Because the aim of our study was to determine the influence of concomitant endometriosis during LH, we believe that it would not be correct to com- pare these advanced cases to hysterectomies for be- nign indication or low-grade malignancy. Similar studies should be performed analyzing specifically surgical outcomes of procedures with bladder and/or bowel dissections. Strengths of this study included the relatively large database and number of cases with endometriosis. Also, the high number of sur- geons with varying experience adds to the generalizability of the data. Interpretation Although endometriosis is often mentioned as com- plicating factor during LH, the additional surgical risks associated with LH ’s with concomitant endo- metriosis have not been we ll established. Yet, for a reliable interpretation and comparison of surgical outcome measures between surgeons and/or hospi- tals, numeric support of the impact of concomitant e n d o m e t r i o s i sd u r i n gL Hi sn e c e s s a r y .T h i si sa l s o relevant for determining if concomitant endometri- osis should be considered a case-mix variable. In this study, we firstly demonstrated that stages III and IV Table 3 Influence of each covariate on blood loss, operative time and complications (Continued) Crude analysis Adjusted analysis Stage III 4 (2.3) 1.23 (0.48 –3.10)0.76 .67 1.16 (0.45 –2.96) .76 Stage IV 6 (3.4) (0.34 –1.65) .48 0.67 (0.30 –1.50) .34 §Interpretation of the geometric standard deviation (GSD): 95% of the data are expected to lie in the range of GM/(GSD) 2 to GM*(GSD) 2 *Index patient with BMI of 28.5, uterine weight of 161.3 g, and 50% previous procedures ɸThis means that for every tenfold increase in uterine weight, blood loss, and operative time increased by the stated percentages and odds of complicat ion increased by the stated OR Statistics: linear mixed model (for outcome blood loss and operative time) and generalized linear mixed model (with logistic link function) (for outcome complications) 95% CI 95% confidence interval BMI body mass index, GM geometric mean, GSD geometric standard deviation Sandberg et al. Gynecological Surgery (2018) 15:8 Page 6 of 9 endometriosis were associated with more intra- operative blood loss compared to the group without endometriosis. This finding has, to our knowledge, not been observed in previous studies [ 3–5, 10]. Al- though the amount of blood loss was twice as much in the group with stage IV endometriosis compared to the group without, the mean blood loss was still low (141 mL for stage IV). The clinical relevancy for patients is therefore questionable, yet, for the sur- geons, it is important to be aware of this potentially increased amount of blood loss. Furthermore, our finding demonstrated that concomitant endometriosis independently influence surgical outcome and there- fore should be taken into consideration when apply- ing a case-mix correction. In agreement with previous studies [ 4, 13], stages III and IV endometriosis was associated with a prolonged operative time, up to 47 min for stage IV endometriosis. This finding is important to consider for pre-operatively scheduling but also as prolonged operative time has been associated with increased morbidity [ 14, 15]. One study even demonstrated that every additional hour of surgery during LH increases the risk of postoperative complications by 22% [ 15]. Regarding complications, previous studies have demonstrated an increased risk for LH ’sw i t he n d o - metriosis compared to LH ’s without endometriosis. [3–5, 13]S p e c i f i c a l l yf o rL H’s with moderate-severe endometriosis (stages III and IV), one of the studies showed an almost fourfold increase in the risk of complications compared to controls [ 13]. Interest- ingly, in our study, no significant difference in complication risk was observed (5.5% for the endo- metriosis group versus 6.8% for the non- endometriosis group). This could be explained by the fact that in previous studies, patients undergoing bowel and bladder resection were included. Other ex- planations could be related to the overall low num- ber of complications in our study and hence the lack of power to demonstrate a significant difference and/ or the high surgical experience for LH with endomet- riosis [ 16]. Indeed, LH ’sw i t hs t a g e sI I Ia n dI Ve n d o - metriosis were significantly more often performed by surgeons with more experi ence, and this might have affected the outcomes. However, we explicitly did not correct for surgical experience in our model as we aimed to demonstrate how patient ’s characteristics independently influence su rgical outcomes. In daily clinical practice, surgical experience cannot be used either to justify worse surgical outcomes. We want to underline that it is the responsibility of the surgeon to know his individual limitations when counseling a patient. This pre-operative awareness was reflected in our study as most severe endometriosis cases were performed by the more experienced surgeons, and this selection most probably has improved overall sur- gical outcomes. However, it is important to keep in m i n dt h a tah i g hs u r g i c a lv o l u m ed o e sn o tn e c e s s a r i l y directly stand for better surgical outcomes [ 1]. Al- though high surgical experience is often associated with positive outcomes, it is not a guarantee. As such, we would recommend surgeons to monitor their indi- vidual surgical performan ces over time rather than to focus on the number of surgeries T able 4 [1].

For a reliable comparison of surgical outcomes between hospitals and/or surgeons, it is necessary to correct for the patient characteristics that are independently influencing these outcomes. For LH, previous studies have already demonstrated that a case-mix correction for BMI, uterine weight, and previous procedures is required [ 1, 2]. The findings of our study have provided numeric support for the influence of concomitant endometriosis on the surgi- cal outcomes of LH. We demonstrated that stages III and IV endometriosis were associated with a longer operative time and more blood loss. These specific stages should thus be considered as case-mix variable for these out- comes in future quality measurement tools. Appendix 1 Table 4 Complication classification according to the NVOG Main category Complication Infection - Local - Organ - Systemic Injury - Vascular - Bowel - Bladder - Ureter - Other Wound dehiscence – Hemorrhage - > 1000 mL - Post-operative bleeding Thrombo-embolism – Dysfunction - Urinary retention - Incontinence - Ileus - Liver - Kidney Systemic - Medication error - Adverse drug event - Other Technical - Failed procedure - Retained foreign body Reactive conversion – Other – Sandberg et al. Gynecological Surgery (2018) 15:8 Page 7 of 9 Appendix 2

None Funding Evelien M. Sandberg has received a research grant from the Bronovo Research Fund (Bronovo Hospital, the Hague, the Netherlands). The funding source had no involvement in the conduct of the study. Authors’ contributions All authors contributed to the work presented in this paper. EMS was responsible for the project development, the acquisition, analysis and interpretation of data, as well as the drafting and finalization of the manuscript. EATB and SRCD were also responsible for the acquisition of data and contributed to the analysis and interpretation of the data as well as the finalization of the manuscript. NvG contributed to the analysis and interpretation of data as well as the finalization of the manuscript. JPB, MJGHS, and JPTR contributed to the acquisition, the analysis and interpretation of data as well as the finalization of the manuscript. FWJ was responsible for the conception of the study and contributed to the acquisition and interpretation of data and the finalization of the manuscript. All authors take responsibility for this study and its findings. All authors read and approved the final manuscript. Ethics approval and consent to participate The Institutional Review Board (IRB) of Leiden University Medical Centre, Leiden, the Netherlands reviewed the study and exempted it from IRB approval. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. For this retrospective study, formal consent is not required. This article does not contain any studies with animals performed by any of the authors. Competing interests The authors declare that they have no competing interests. Publisher’sN o t e Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Author details 1Department of Gynecology, Leiden University Medical Centre, Leiden, the Netherlands. 2Department of Medical Statistics, Leiden University Medical Centre, Leiden, the Netherlands. 3Department of Gynecology, Haaglanden Medical Centre, the Hague, the Netherlands. 4Department BioMechanical Engineering, Delft University of Technology, Delft, the Netherlands. 5Department of Gynecology, Minimally Invasive Surgery, Leiden University Medical Centre, PO Box 9600, 2300, RC, Leiden, the Netherlands. Received: 10 December 2017 Accepted: 7 February 2018

1. Driessen SR, van Zwet EW, Haazebroek P et al (2016) A dynamic quality assessment tool for laparoscopic hysterectomy to measure surgical outcomes. Am J Obstet Gynecol 2016;215(6):754.e1-754.e8 2. Driessen SR, Sandberg EM, la Chapelle CF, Twijnstra AR, Rhemrev JP, Jansen FW (2016) Case-mix variables and predictors for outcomes of laparoscopic hysterectomy: a systematic review. J Minim Invasive Gynecol 23:317 –330 3. Brummer TH, Jalkanen J, Fraser J et al (2011) FINHYST, a prospective study of 5279 hysterectomies: complications and their risk factors. Hum Reprod 26:1741–1751 4. Song T, Kim TJ, Kang H et al (2012) Factors associated with complications and conversion to laparotomy in women undergoing laparoscopically assisted vaginal hysterectomy. Acta Obstet Gynecol Scand 91:620 –624 5. Patzkowsky KE, As-Sanie S, Smorgick N, Song AH, Advincula AP (2013) Perioperative outcomes of robotic versus laparoscopic hysterectomy for benign disease. JSLS 17:100 –106 6. Twijnstra AR, Zeeman GG, Jansen FW (2010) A novel approach to registration of adverse outcomes in obstetrics and gynaecology: a feasibility study. Qual Saf Health Care 19:132 –137 Fig. 1 Scatter plot presenting surgeon ’s experience versus the number of procedures registered during the study Sandberg et al. Gynecological Surgery (2018) 15:8 Page 8 of 9 7. Blikkendaal MD, Twijnstra AR, Stiggelbout AM, Beerlage HP, Bemelman WA, Jansen FW (2013) Achieving consensus on the definition of conversion to laparotomy: a Delphi study among general surgeons, gynecologists, and urologists. Surg Endosc 27:4631 –4639 8. American Society for Reproductive Medicine (1997) Revised American Society for Reproductive Medicine classification of endometriosis: 1996. Fertil Steril 67:817 –821 9. R Development Core Team (2008). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org 10. Twijnstra AR, Blikkendaal MD, van Zwet EW, van Kesteren PJ, de Kroon CD, Jansen FW (2012) Predictors of successful surgical outcome in laparoscopic hysterectomy. Obstet Gynecol 119:700 –708 11. Van Leersum NJ, Snijders HS, Henneman D et al (2013) The Dutch surgical colorectal audit. Eur J Surg Oncol 39:1063 –1070 12. Corson SL, Batzer FR, Gocial B, Kelly M, Gutmann JN, Maislin G (1995) Intra- observer and inter-observer variability in scoring laparoscopic diagnosis of pelvic adhesions. Hum Reprod 10:161 –164 13. Uccella S, Marconi N, Casarin J et al (2016) Impact of endometriosis on surgical outcomes and complications of total laparoscopic hysterectomy. Arch Gynecol Obstet 294:771 –778 14. Catanzarite T, Saha S, Pilecki MA, Kim JY, Milad MP (2015) Longer operative time during benign laparoscopic and robotic hysterectomy is associated with increased 30-day perioperative complications. J Minim Invasive Gynecol 22:1049–1058 15. Hanwright PJ, Mioton LM, Thomassee MS et al (2013) Risk profiles and outcomes of total laparoscopic hysterectomy compared with laparoscopically assisted vaginal hysterectomy. Obstet Gynecol 121:781 –787 16. Maruthappu M, Gilbert BJ, El-Harasis MA et al (2015) The influence of volume and experience on individual surgical performance: a systematic review. Ann Surg 261:642 –647 Sandberg et al. Gynecological Surgery (2018) 15:8 Page 9 of 9