Clavien-Dindo, comprehensive complication index and classification of intraoperative adverse events: a uniform and holistic approach in adverse event registration for (deep) endometriosis surgery

Quality assessment and follow-up are important in surgery. Quality assessment and registration of healthcare, including patient-reported outcome measures, could improve healthcare outcomes ( Øvretveit et al. , 2017 ; Pop et al. , 2019 ). A vital part of this follow-up is the reporting of clinical outcomes and the registration of adverse events (AEs). Correct registration makes it possible to evaluate care at the individual patient level, but also facilitates data analysis at group level to gain insight and create the possibility to compare quality of care. Correct data analysis at group level can provide insight into acknowledging and recognizing deviating trends that can be reacted to in time, if necessary, to prevent future AEs. Different systems can be used for the AE registration; however, the Clavien–Dindo (CD) system is currently the most used, validated, and accepted standard to classify perioperative AEs ( Dindo et al. , 2004 ). The CD system consists of five grades of severity, whereby Grade I is low and Grade V is death. A strength of the CD system is that it is widely accepted in surgery, making data comparison possible. Limitations of the CD are the fact that the data are ordinal, making data analysis difficult between two treatment options (a summative interpretation is not possible). Furthermore, because a weighting system is lacking, it is not possible to compare the impact of multiple minor AEs against one major AE. Additionally, for the ease of handling in AE registration (e.g. annual reports), it is common to capture only the AE with the highest grade ( Eisenberg et al. , 2010 ; Kadlec et al. , 2013 ; Dumitra et al. , 2018 ). This automatically leads to data loss (incomplete overview of total AE burden) in patients with combinations of multiple AEs and, therefore, failure in presenting the actual morbidity of AEs. To overcome these shortcomings, the comprehensive complication index (CCI) ( Slankamenac et al. , 2013 ) was created in 2013. Based on the CD, the CCI can be used to calculate the impact of all cumulative AEs within a single patient using an algorithm. The overall burden of AE per patient is reflected in a single number on a scale from 0 (no AE) to 100 (death). This enables the comparison of patients with a single AE to patients with multiple AEs. Furthermore, the CCI provides a more accurate selection of those patients that need to be discussed in the morbidity and mortality meetings. Patients with CCI values of ≥33.7 are considered severe ( Slankamenac et al. , 2013 ), and this cutoff point could be used for the selection of severe cases. However, it should be noted that the CD and CCI systems are only for postoperative adverse events (poAEs). Intraoperative adverse events (ioAEs) that are solved during surgery cannot be classified by CD. ioAEs are also important to register and classify, as these events are associated with adverse outcomes in the postoperative course ( Kaafarani et al. , 2014 ; Ramly et al. , 2015 ; Bohnen et al. , 2017 ). Furthermore, with the implementation of new surgical techniques (e.g. robotic surgery), the documentation of ioAEs will provide insight into the safety of the newly introduced techniques: this insight will not emerge with only the documentation of poAEs. Surprisingly, only recently (2020), the group of Dell-Kuster developed and validated a classification system named ClassIntra for reporting ioAEs ( Table 1 ), which was developed after the CD system ( Dell-Kuster et al. , 2020 ). ClassIntra version 1.0 for reporting intraoperative adverse events. Without the need for any additional treatment or intervention Patient asymptomatic or mild symptoms Bleeding: bleeding above average from small-caliber vessel: self-limiting or definitively manageable without additional treatment than routine coagulation Injury: minimal serosal intestinal lesion, not requiring any additional treatment Cautery: small burn of the skin, no treatment necessary Arrhythmia: arrhythmia (e.g. extrasystoles) without relevance With the need for any additional minor treatment or intervention Patient with moderate symptoms, not life-threatening and not leading to permanent disability Bleeding: bleeding from medium caliber artery or vein, ligation; use of tranexamic acid Injury: non-transmural intestinal lesion requiring suture(s) Cautery: moderate burn requiring non-invasive wound care Arrhythmia: arrhythmia requiring administration of antiarrhythmic drug, no hemodynamic effect With the need for any additional moderate treatment or intervention Patient with severe symptoms, potentially life-threatening and/or potentially leading to permanent disability Bleeding: bleeding from large caliber artery or vein with transient hemodynamic instability, ligation, or suture; blood transfusion Injury: transmural intestinal lesion requiring segmental resection Cautery: severe burn requiring surgical debridement Arrhythmia: arrhythmia requiring administration of antiarrhythmic drug, transient hemodynamic effect With the need for any additional major and urgent treatment or intervention Patient with life-threatening symptoms and/or leading to permanent disability Bleeding: life-threatening bleeding with splenectomy; massive blood transfusion; ICU stay Injury: injury of central artery or vein requiring extended intestinal resection Cautery: life-threatening burn injury by cautery leading to fire requiring ICU treatment Arrhythmia: arrhythmia requiring electroconversion, defibrillation, or admission to the ICU With intraoperative death of the patient For the classification of poAEs in endometriosis surgery, the Clavien–Dindo and use of the CCI have been recommended since 2016 ( Vanhie et al. , 2016 ); however, there are currently no publications on the use of the CCI. Also, no data could be found on the topic on how to report ioAEs in surgery for deep endometriosis (DE), something that is vital for a complete overview of the surgical quality assessment, especially because DE is particularly associated with high AE rates, owing to the complexity of the disease and surgery. However, it is challenging to gain proper insight into how severe some AEs are, especially in patients with multiple AEs. This study aims to assess the hypothesis that additional use of the CCI and ClassIntra classification systems will provide a better view of the total burden of AEs compared to using CD alone. The objective is to establish a standardized registration of AEs in (deep) endometriosis surgery, facilitating comparisons at both national and international levels.

A total of 870 cases met the inclusion criteria (85.9% DE cases, 14.1% endometriosis cases). The mean age of the women was 36.4 (SD 7.1) years, with a mean BMI of 24.9 (SD 4.5) kg/m 2 ( Table 2 ). Pain was the primary indication for surgery in 87.1% of the cases. The median (IQR) operating time was 75.0 (40–120) min, with a median (IQR) blood loss of 50 ml (0–50) (range 0–1700 ml). Baseline characteristics of patients with surgical (deep) endometriosis. 12.1% of the cases (n = 105) did not have a rASRM stage, while there was endometriosis outside the internal reproductive organs and peritoneum, not possible to classify with the rASRM. 61.6% of the cases (n = 536) did not have an EFI score, while the indication for surgery was not fertility. Cranial to sigmoid junction. FU involvement is a combination of intrinsic and extrinsic endometriosis involvement. Surgical technique in one case not further specified. IQR: interquartile range; rASRM: revised American Society for Reproductive Medicine; EFI: Endometriosis Fertility Index; poAE: postoperative adverse event; Compartment A: Enzian vagina; Compartment B: Enzian ligaments; Compartment C: Enzian rectum; FB: Enzian bladder; FA: Enzian adenomyosis; FU: Enian ureter; FI: Enzian intestine; FO: Enzian other. Independent Student’s T-test and the Mann–Whitney test were used for statistical analysis. Table 3 shows the total number of poAEs reported (n = 168 in 145 cases), among which Grade II was most often present (10.3%). A total of 125 (86.2%) patients had a single poAE, 17 (11.7%) patients had 2 poAEs, and there were 3 (2.1%) patients with 3 poAEs. Overall, a total of 870 procedures were registered, of which 145 procedures resulted in one or more poAEs, resulting in a poAE rate of 16.7% (145/870), of which 36 (4.1%) cases were classified as severe. Postoperative adverse events in women following surgery for (deep) endometriosis. Some patients have more than one postoperative AE. CCI ≥ 33.7. poAE: postoperative adverse event; CCI: comprehensive complication index; CD: Clavien–Dindo. n = 168 single adverse events in 145 cases. Table 4 presents the type of poAE in relation to the severity. The majority of poAEs consisted of infections at organ level (35.1%), of which the majority involved a cystitis (29.2%). Of the infection at organ level, most of the severity was scored as a Grade II severity. Local infections were reported in 16.7% of cases, mostly surgical site infections (13.3%). Anastomosis leakage occurred in 5.4% of the poAEs (1.0% in total population), and all of these poAEs were scored with severity IIIB. The highest poAE was a Grade IVB and was a bowel injury. Acute kidney failure was reported in 3.6% (n = 6). Type of postoperative adverse events and severity score by the Clavien–Dindo classification. poAE: postoperative adverse event. Table 3 shows the median CCI of 20.9 (IQR 20.9–31.7) for all poAEs. Looking at the patients with a severe poAE, a median CCI of 33.7 was reported (IQR 33.7–39.7). Figure 1 shows the patients with the highest reported CD poAE (black bars) in relation to the CCI (pink line). There are 36 cases (24.8%) with a CCI ≥33.7, and these are considered as severe poAEs (see reference line in Fig. 1 ). The area under the CCI curve (pink) shows the increase in CCI of patients with multiple poAEs compared to only reporting the most severe AE (as would apply if only the CD system was used). To illustrate, the case with the most severe poAE has a CCI of 54.4. The highest CD is a Grade IIIb poAE, which would correspond to a CCI of 33.7. The gap between 33.7 and 54.4 is explained by multiple other poAEs that are not captured if only the CD system is used. In 20 patients (13.8%), the CCI was higher than the CD because of multiple poAEs. As shown in Table 5 , cases with multiple AEs increase in CCI compared to only reporting the highest CD. Case #1 increases in the CCI from 33.7 to 44.9, Case #2 increases from 33.7 to 47.6, and Case #3 remains the same while it is only a single AE. All patients with endometriosis with postoperative adverse events in relation to the most severe Clavien–Dindo classification and comprehensive complication index. On the left axis, the severity of the Clavien–Dindo (CD) is shown, while on the right axis, the comprehensive complication index (CCI) is shown. The pink line presents the CCI and the black vertical lines present the highest CD grade. N = 145. The area under the CCI curve (pink) shows the increase in CCI of patients with multiple postoperative adverse events (poAEs) compared to only reporting the most severe AE (as would apply if only the CD system was used). A CCI ≥33.7 is considered as a severe poAE. To illustrate, the case with the most severe poAE has a CCI of 54.4. The highest CD is a Grade IIIb poAE, which would correspond to a CCI of 33.7. The gap between 33.7 and 54.4 is explained by multiple other poAEs that are not captured if only the CD system is used. The three cases with major surgery and the corresponding CCI. #1 Endometriosis classification: Enzian A2B3C3, FA, FI Surgical procedure: hysterectomy + BSO, low rectal resection, discoid resection sigmoid Compartment A: Enzian vagina; Compartment B: Enzian ligaments; Compartment C: Enzian rectum; FB: Enzian bladder; FA: Enzian adenomyosis; FU: Enian ureter; FI: Enzian intestine; BSO: bilateral salpingo-oophorectomy; CCI: comprehensive complication index. The Enzian classification, surgical procedure, type of postoperative adverse event, most severe Clavien–Dindo (CD) score, and total CCI are shown. In total, 11 (11/870, 1.3%) ioAEs arose and were corrected directly during surgery ( Table 6 ). These included serosa injury (n = 6 of 870 (0.7%)), blood vessel injury (n = 2), uterus perforation (n = 1), rectal perforation (n = 1), and perforation of the sigmoid (n = 1). Of these, 10 (90.9%) ioAEs were staged as Grade II and 1 (9.1%) ioAE as Grade III because of 1700 ml blood loss. The most frequent ioAE was a directly sutured intestinal serosa lesion in 6 (54.5%) cases, followed by 2 (18.2%) cases with a bleeding. When summing up poAEs and ioAEs, this results in a total AE rate of 17.6% (n = 153). Eight patients (0.9%) only had an ioAE, 142 (16.3%) patients only had a poAE, and 3 (0.3%) patients had both an ioAE and a poAE. Of the three patients with both an ioAE and poAE, one patient showed a relation between the ioAE and the poAE; this was serosa injury of the rectum that was sutured during the surgery, and this patient developed a peritonitis based on a perforation that was furthermore complicated with respiratory distress. Intraoperative adverse events, classified by ClassIntra. Because of rounding there is a difference of 0.1% between total percentage and subgroups. ioAE: intraoperative adverse event; poAE: postoperative adverse event. N = 870 surgeries.

This study used prospectively collected data on surgical cases with endometriosis or DE from a Dutch DE expertise center (Endometriose in Balans, Haaglanden Medical Centre, the Hague, the Netherlands). Data inclusion took place from February 2019 through December 2021 by the EQUSUM application ( Metzemaekers et al. , 2020 ). For this study, we used the definition of DE described by the international working group of American Association of Gynecologic Laparoscopists, European Society for Gynaecological Endoscopy, ESHRE, and World Endometriosis Society in 2021 ( Tomassetti et al. , 2021 ), which is classified with the Enzian-based system on clinical expertise combined with radiological and intraoperative findings. Patients who underwent a diagnostic laparoscopy were excluded from the study. For adenomyosis, we used the Morphological Uterus Sonographic Assessment (MUSA) for classification ( Van den Bosch et al. , 2019 ). Registered data included: general data on patients’ characteristics, surgical indication, surgical procedure, previous abdominal surgery, and accurate localization of DE lesions and adhesion scoring. The EQUSUM ( Metzemaekers et al. , 2020 ) application automatically generated the following classifications: revised American Society for Reproductive Medicine (rASRM) ( American Society for Reproductive Medicine, 1997 ), Enzian ( Keckstein et al. , 2003 ), and Endometriosis Fertility Index (EFI) scores ( Adamson and Pasta, 2010 ). The Enzian scores were assigned according to the original manuscript ( Tuttlies et al. , 2005 ). AEs were documented until 42 days post-procedure according to the CD classification ( Dindo et al. , 2004 ). Surgical poAEs were defined as any event that represents a deviation in the expected postoperative course ( Dindo and Clavien, 2008 ). The validated CCI ( Slankamenac et al. , 2013 ) was calculated with the following: CCI formula CCI ®  = √ (wC1 + wC2 + wC x )/2 where wC stands for weight of complication. CD Grade I = wC 300, Grade II = wC 1750, Grade IIIa = wC 2750, Grade IIIb = wC 4550, Grade IVa = wC 7200, and Grade IVb = wC 8550. CD Grade V always results in CCI ® 100. ioAEs were classified by the validated ClassIntra reporting system and defined as ‘any deviation from the ideal intraoperative course between skin incision and skin closure, and includes events related to surgery and anaesthesia’ ( Dell-Kuster et al. , 2020 ). Table 1 shows the classification of ioAEs, ranging from Grades I to V. The Statistical Package for Social Sciences (SPSS, Chicago, IL, USA), version 28.0 was used for analysis. Categorical variables were described as frequencies with percentages. For normally distributed or skewed data, data were presented as mean with SD or median with interquartile range (IQR), respectively. The Shapiro–Wilk test was used to determine the distribution of the data. For group comparisons with parametric data, an independent Student’s T-test was performed, while for non-parametric data, the Mann–Whitney test was performed. A two-tailed P -value of <0.05 was considered as statistically significant. Ethical approval was given by the Medical Ethics Committee of the Leiden University Medical Centre for the use of the anonymous data in the EQUSUM database (LUMC) (G20.019).

This study showcases the advantages of integrating two supplementary systems—the CCI and Classintra system—in the reporting of AEs that occur during endometriosis surgery. Although the Clavien–Dindo classification is a commonly used tool for evaluating postoperative complications, it has limitations in capturing the entirety of AEs, particularly those that arise during surgery. In order to address this limitation, the CCI and Classintra system were developed, providing standardized methods for categorizing and reporting multiple and ioAEs. By utilizing these systems, a more inclusive and thorough summary of all AEs can be generated, leading to a more comprehensive understanding of patient outcomes and the potential to lessen the overall burden of AEs associated with surgery.

This study was performed to advocate uniform registration in the field of AEs in endometriosis surgery, and especially in DE, making use of the CD and CCI as well as the ClassIntra systems. This is one of the largest and most detailed studies on this topic, which could act as a first step toward benchmarking AEs in DE surgery. A high percentage of poAEs were registered, namely 16.7%. Easy discrimination between severe AEs is possible with the CCI (CCI ≥ 33.7). We found only 4.1% of patients with severe AEs, which are interesting to discuss in, for example, morbidity and mortality meetings. Furthermore, with the CCI, it is possible to calculate the total burden of AEs, instead of only reporting the most severe AE (as with CD). If we only used the most severe AE (which is common with the CD), 23 poAEs would not have been used in the total AE calculation (23/168, 13.7%). This is a significant part of the total AE burden. As far as we know, there are no endometriosis studies where the CCI is used for the registration of poAEs. In our study, the median CCI was 20.9 (IQR 20.9–31.7) in the total cohort, and 33.7 (IQR 33.7–39.7) in the patients experiencing severe AEs. This is comparable with non-endometriosis literature: Tamini et al. (2021) established a median CCI of 20.9 (IQR 20.9–36.2) in a cohort who underwent surgery for colon cancer. Furthermore, in a recent publication by Haas et al. (2021) concerning bladder surgery, the median 30-day CCI was 22.6 (IQR 8.8–39.7). The CCI also makes it easy to compare poAE rates over the years, something that is more challenging with an ordinal scale (CD). It was easy to show that the median CCI over the years did not significantly change. In contrast, with the CD alone, it was only possible to detect frequencies instead of comparing the total AE burden. In our cohort, 11 ioAEs occurred, which were treated intraoperatively. These 11 ioAEs would not have been classified if only the CD system was used. For a complete overview in the quality assessment of surgical audits, these 11 ioAEs are also a significant part of the AEs that need to be classified. This is especially important when new surgical techniques are introduced. The ClassIntra could be a good tool in the debriefing period at the end of a surgical procedure. In this way, the registration is directly performed and the whole team could instantly reflect on any ioAE that occurred. Furthermore, the team can discuss strategies aimed at preventing any poAE that may occur as a result of the ioAE, such as increased monitoring for peritonitis in cases with intraoperative intestinal injuries. To highlight the literature gap on ioAE reporting, a critical appraisal of 46 randomized controlled surgical trials on AEs concluded that only 41% of the studies reported ioAEs, with 13% providing a definition and only 9% using a classification for these ioAEs ( Rosenthal et al. , 2015 ). This shows that consensus is lacking, while there is a need for a widely applicable classification system. From our data, three cases showed both an ioAE and a poAE, and one case had a relation to both AEs (serosa injury and peritonitis). Others have illustrated that these types of ioAE are associated with adverse outcomes in the postoperative course ( Bohnen et al. , 2017 ). This also shows how one event could trigger a cascade of AEs, either directly or more indirectly owing to increased patient vulnerability and case complexity ( de Vos et al. , 2019 ). A limitation of the ClassIntra system is, for example, the fact that the system relies heavily on the individual surgeon to identify and report complications, which can lead to under-reporting. However, the question about the subject ‘what is an adverse event’ does not only apply to ioAEs but also to poAEs. This question can be answered with Delphi methods, but that is beyond the scope of this research, though of great importance in the standard reporting of AEs. This is the first AE study in endometriosis care that uses a total burden of all AEs (intra, post, single, and multiple AEs), which is a strength of our study. With this study, we hope to encourage other centers to also adopt these classification systems. Not only will this provide a more accurate and complete overview of AEs but also it will enable data comparison on multiple levels. A limitation of this study is the fact that it was performed in a single center, which limits the number of cases. However, even with a smaller number of cases, this study is able to illustrate how the addition of the CCI and ClassIntra systems would lead to a more comprehensive image of the complications that may arise in surgical endometriosis cases. In AE reporting, we recommend to use the Clavien–Dindo for poAEs, which aligns with international recommendations ( Vanhie et al. , 2016 ). The usage of the CCI is also recommended by these same guidelines ( Vanhie et al. , 2016 ); however, to our knowledge, no publications are available on this topic in relation to DE surgery. Therefore, we would like to promote the use of the CCI alongside the CD. Owing to several instrument-specific advantages, the CCI appears to be a useful addition to the CD system because: it presents the total burden of poAEs per patient (cumulative formula); it uses a linear scale instead of an ordinal scale, which is beneficial for research/statistical purposes; there is easy discrimination of severe poAEs when using the cutoff point of CCI ≥33.7; we can compare patients with one poAE to patients with multiple poAEs; and it creates easy and interpretable data for benchmarking. We acknowledge that implementing our recommendations will necessitate additional registration time for doctors, who already devote a significant amount of time to computer-based tasks rather than direct patient care. However, if the IT is properly designed and the registration is performed accurately, it will eventually save time and energy. For example, with the EQUSUM application, it is easy to perform statistics, while the syntax for all the calculations is already there. The CCI will automatically be generated after running the syntax and all other calculations for AE percentages. Proper use of the current and future technologies does have the potential to decrease workload instead of increasing it, even with more registration. Effort should be put into making it as easy as possible for doctors, and ‘reward’ them with easy and automatically generated monthly and annual reports for the morbidity and mortality meetings.