Changes in hospital consumption of opioid and non-opioid analgesics after colorectal endometriosis surgery

The aim of this study was to analyze postoperative consumption of analgesics during hospitalization following colorectal surgery for endometriosis. We conducted a retrospective study at Tenon University Hospital, Paris, France from February 2019 to December 2021. One hundred sixty-two patients underwent colorectal surgery: eighty-nine (55%) by robotic and seventy-three (45%) by conventional laparoscopy. The type of procedure had an impact on acetaminophen and nefopam consumed per day: consumption for colorectal shaving, discoid resection, and segmental resection was, respectively, 2(0.5), 2.1(0.6), 2.4(0.6) g/day (p = 10–3), and 25(7), 30(14), 31(11) mg/day (p = 0.03). The total amount of tramadol consumed was greater following robotic surgery compared with conventional laparoscopy (322(222) mg vs 242(292) mg, p = 0.04). We observed a switch in analgesic consumption over the years: tramadol was used by 70% of patients in 2019 but only by 7.1% in 2021 (p < 10–3); conversely, ketoprofen was not used in 2019, but was consumed by 57% of patients in 2021 (p < 10–3). A history of abdominal surgery (OR = 0.37 (0.16–0.78, p = 0.011) and having surgery in 2020 rather than in 2019 (OR = 0.10 (0.04–0.24, p < 10–3)) and in 2021 than in 2019 (OR = 0.08 (0.03–0.20, p < 10–3)) were the only variables independently associated with the risk of opioid use. We found that neither clinical characteristics nor intraoperative findings had an impact on opioid consumption in this setting, and that it was possible to rapidly modify in-hospital analgesic consumption modalities by significantly reducing opioid consumption in favor of NSAIDS or nefopam.


Introduction
Endometriosis is a well-known cause of pelvic pain affecting around 10% of women of reproductive age which represents at least 190 million women worldwide [1].Endometriosis is also a risk factor of infertility, quality of life alteration [1], and chronic consumption of opioid and non-opioid analgesics [2].
Women with endometriosis are three times more likely to use opioids for uncontrolled pain than patients without endometriosis [3].Depending on the study, 42% [2] to 62% [4] of patients with endometriosis use opioids.Moreover, in the postoperative period following resection of endometriosis lesions, a threefold increased risk of postoperative opioid consumption has been observed [5].
On one hand, there is the risk that exposure to opioids during hospital stay, particularly after surgery, can lead to dependence afterward [6,7] and, on the other hand, persistent pain in the immediate postoperative period may lead to longer hospital stays, readmission, and longer recovery times [8].Robotic surgery improves the surgeon's vision, facilitates nerve sparing techniques, and had been shown to reduce opioid consumption in benign hysterectomy [9].However, no data are available in the specific context of colorectal resection for endometriosis.
Therefore, the aim of this study was to compare postoperative consumption of analgesics during the hospital stay of patients who had undergone colorectal resection for endometriosis by robotic or conventional laparoscopy.

Methods
We conducted a retrospective cohort study including patients who had undergone colorectal surgery for endometriosis at Tenon University Hospital, Expert Centre in Endometriosis (GRC-6, C3E), Paris, France, from February 2019 (corresponding to the endowment of a robot from our hospital) to December 2021.Institutional review board approval was obtained, and all patients gave their written informed consent to participate in the study (CEROG 2012-GYN-10-03).

Patients
We included symptomatic patients > 18 years of age scheduled for colorectal surgery after failure of medical treatment and/or with infertility.Colorectal endometriosis was assessed by transvaginal ultrasonography (TVUS), magnetic resonance imaging (MRI), and rectal echo-endoscopy (REE) using previously published criteria [10][11][12].The robotic and conventional laparoscopy approaches were exclusively performed depending on the availability of the robotic theater: none of the patients were allocated to robotic or conventional laparoscopy based on preoperative characteristics.
For robot-assisted surgery, 4 × 8 mm skin incisions were made horizontally at the level of the umbilicus and another of 5 mm in the left flank for a conventional laparoscopic trocar.For standard laparoscopy, one 10 mm incision was made in the umbilicus and 3 × 5 mm skin incisions were made horizontally in the iliac fossae.The intervention began with the evaluation of the endometriosis stage according to the ASRM classification.Bowel endometriosis surgery was performed as previously described [13] and associated procedures including adnexal surgery (ovarian cystectomy or salpingo-oophorectomy), resection of the uterosacral ligament, torus uterinus, or parametrium, partial colpectomy, hysterectomy, ureterolysis, and ureteral reimplantation were performed when required.
After surgery, the patients were hospitalized in our gynecological surgery department.Acetaminophen (Sanofi, Paris, France) was systematically administered every 6 h for the first 24 h according to our standardized analgesic protocol.Patients with resistant pain were given ketoprofen 50 mg (VIATRIS SANTE, Lyon, France) or nefopam (VIATRIS SANTE, Lyon, France) every 6 h if the pain was rated at below 7/10 on the visual analogue scale (VAS), or in combination if the pain was rated at over 7/10.Tramadol 50 mg (ARROW, Lyon, France) was given if the pain was still resistant.Finally, oral morphine sulfate 10 mg (ETHY-PHARM, Saint cloud, France) or intravenous morphine chlorhydrate 3 mg (LABORATOIRE RENAUDIN, Itxassou, France) was prescribed every 4 h for patients whose pain remained resistant to these treatments with a VAS above 7/10, with the objective of reducing the score to under 3/10.Data about analgesic consumption were collected through our medical prescription and delivery software: Orbis (Agfa Healthcare, Mortsel, Belgium).Morphine consumption was converted to oral morphine equivalents using the guidelines published by the Centers for Disease Control and Prevention [14].
Postoperative complications were classified according to the Clavien-Dindo classification system as minor (grade I-II) or major (grade IIIA and IIIB-IV) 22 [15].

Statistical analysis
The population was divided into two groups according to the surgical approach, i.e., robotic (robotic group) or conventional (conventional group) laparoscopy.Patients were also categorized according to the type of analgesics they consumed, i.e., opioids or non-opioid analgesics.
Univariate analysis was performed using the Student's t test for normally distributed data and the Chi-square test for categorical data.The Wilcoxon test was used for continuous variables and Fisher's exact test for categorical variables.All reported p values were two sided.Significant difference was denoted when p < 0.05.All statistical analysis was performed using commercially available software (RStudio Team (2020).RStudio: Integrated Development for R. RStudio, PBC, Boston, MA URL http:// www.rstud io.com).

Epidemiologic characteristics and surgical findings
The study population was composed of 162 patients undergoing colorectal surgery for bowel endometriosis between February 2019 and December 2021.Eighty-nine patients (55%) underwent robotic laparoscopy (robotic group), and seventy-three (45%) patients underwent conventional laparoscopy (conventional group).For 142 patients (142/162, 87%), pain was part of the surgical indication.The epidemiological characteristics of the population are depicted in Table 1.
At surgery, 137/162 (86%) patients had ASMR grade III-IV endometriosis.There was no difference in the ASRM score or the digestive procedure (rectal shaving, discoid, or segmental resection) between the robotic and conventional groups.Similarly, surgical procedures associated with colorectal resection did not differ between the groups except for a higher frequency of JJ stents use in the conventional group compared with the robotic group: 16% (12/73) vs. 4.5% (4/89), p = 0.011.Finally, the operating time was longer for the robotic group: 200 vs. 251 min (p = 0.003).Table 2 presents the surgical findings for both groups.
Table 4 shows the consumption of analgesics for the robotic and conventional groups.Among the tramadol users, more tramadol was consumed in the robotic group than in the conventional group (322 (222) mg vs. 242 (292) mg, p = 0.04)), with a tendency for longer use in the robotic group (3.7 (2.2) vs. 2.7 (1.9) days, p = 0.08)).There was no difference in the average daily dose received (p = 0.2).
Analgesic consumption varied significantly according to the year of surgery.Figure 1 shows the changes in consumption of the different analgesics according to the year.
Patients consuming ketoprofen were more likely to use nefopam and tramadol (p = 10 -3 ), whereas patients consuming tramadol were more likely to use morphine (p = 0.02).

Factors associated with an increased risk of opioid use
Univariable and multivariable linear regression analysis of the consumption of any opioid is presented in Table 5.A history of abdominal surgery (OR = 0.37 (0.16-0.78, p = 0.011)), and undergoing surgery in 2020 rather than in 2019 (OR = 0.10 (0.04-0.24, p < 10 -3 )), and in 2021 rather than in 2019 (OR = 0.08 (0.03-0.20, p < 10 -3 )) were the only variables independently associated with the risk of opioid use after colorectal surgery for endometriosis.

Discussion
The objective of this study was to present the postoperative analgesic consumption of patients operated on in an expert center for colorectal endometriosis.We observed that surgical procedure did not impact opioid consumption.However, during the period, a shift in analgesic use was noted with an increased consumption of anti-inflammatory drugs and nefopam while tramadol consumption was divided by 10 in 3 years.
To the best of our knowledge, this is the first study to evaluate the modalities of consumption of postoperative analgesics in a large cohort of women operated on for colorectal endometriosis.As previously described, this is a challenging setting because colorectal endometriosis represents the most severe form of endometriosis and is the source of the most severe postoperative complications [16,17].Our complication rate is in agreement with previous reports: 13% of Clavien-Dindo grade IIIA, 8% grade IIIB, and 5% of grade IV [18].This relatively high complication rate can be explained by the incidence of some surgical risk factors such as a high rate of colpectomy, temporary diverting stoma, intra-operative JJ stent use, and ureteral or bladder resection [19,20].Moreover, as previously published for bowel cancer surgery, the high incidence of drainage in our series was associated with a risk of postoperative complications underlining the need to avoid systematic drainage [21,22].In our study, the surgical approach-robotic or conventional laparoscopyhad no impact on the likelihood of being exposed to an opioid during hospitalization.However, in patients using  endometriosis.Time-to-diagnosis for patients with endometriosis can be as long as 12 years during which time they often experience considerable pain [1].They have often been subjected to medical wandering and multiple surgeries, and can suffer from anxiety and depressive disorders [24], all of which have a negative impact on social and familial relationships as well as causing stigma.This explains the high preoperative incidence of analgesic consumption in this population [25].Our data focus on the in-hospital postoperative period and only on drugs that were both prescribed and actually taken by the patients.The first 2 days after colorectal surgery are the most painful [26], and this, therefore, represents a crucial moment for analgesic consumption: the likelihood of chronic opioid use increases with each additional day of medication supplied starting with the third day after surgery [27].In our study, the average duration was 2.06 days (2.09) for morphine consumption and 3.22 (2.09) days for tramadol for a mean hospital stay of 1 week.This means that very few patients were using opioids by the end of their hospital stay, and that there is no need to prescribe them at discharge.This study also showed that it was possible to radically change the prescription habits of all the practitioners working in our center.The example of tramadol is the most striking.Tramadol is a strong contributor to the risk of future opioid addiction [27] and was the first opioid analgesic cited in a 2018 survey of problematic use both among substance abusers and in the general population for pain management conducted by the French National Drug Safety Agency and published in February 2019 [28].Problematic use included dependence with signs of withdrawal occurring even when taken at recommended doses and over a short period of time, and leading to persistent intake by patients who no longer experience pain.The report highlighted that tramadol was also the first analgesic involved in deaths related to the use of analgesics, ahead of morphine.Following this report and recommendations to restrict the indications for tramadol prescription, we have been able to divide its consumption in our hospital by a factor of 10 in 3 years.
Opioid treatments have been replaced by the non-steroid anti-inflammatory drugs (NSAID) ketoprofen and nefopam.By 2021, two out of three patients were receiving ketoprofen in postoperative care and three out of four nefopam.Nefopam has been shown to be effective postoperatively [29] and to reduce opioid consumption when combined with an NSAID [30].NSAIDs should be the weapon of choice in postoperative prescriptions: they are effective postoperatively and reduce opioid consumption [31], but they also reduce the rate of fistulas after colorectal surgery [32,33] and shorten the time required to resume transit [34,35].However, as we have seen, there is still room for improvement to increase the prescription of NSAIDs during the postoperative in-hospital period in our department.In our study, in addition to the year of surgery, prior surgery was also found to represent a risk of opioid consumption.This is probably because the patient was exposed to opioids during their previous hospitalization and wishes to benefit from their analgesic effects again.Special attention should, thus, be paid to patients with a history of surgery, as can often be the case in endometriosis.Finally, although neither the digestive procedure performed nor the severity of the disease impacts postoperative opioid consumption, preoperative consumption has been shown to be a major factor in its use postoperatively [36].This is the main limitation of the study: we did not collect the patients' preoperative consumption habits and a future study will have to take this bias into account.Similarly, we did not collect data about the duration and type of analgesics used after hospitalization, which is important information regarding the risk of chronic opioid use.Finally, the retrospective nature of the data always represents a bias although we used a software dedicated to the prescription of analgesics allowing a comprehensive evaluation of drug consumption.

Conclusion
In conclusion, our study presents rare data on the immediate postoperative consumption of patients having undergone surgery for colorectal endometriosis.We were able to demonstrate that neither clinical characteristics nor intraoperative findings had an impact on opioid consumption in these patients.Furthermore, our study shows that it is possible to rapidly modify in-hospital analgesic consumption by significantly reducing opioid consumption in favor of NSAIDs or nefopam.

Table 1
Characteristics of the study population BMI body mass index, SD standard deviation Statistics for continuous variable are presented by mean (SD)

Table 2
Surgical findings according to the surgical approach Significant p values are shown in bold ASRM American society of reproductive medicine, SD standard deviation Statistics for continuous variable are presented by mean (SD)

Table 3
Analgesic consumption in patients with or without postoperative complications

Table 4
Analgesic consumption in patients depending on the surgical route Significant p values are shown in bold MME morphine milligram equivalent, NS not significant Statistics for continuous variable are presented by mean (SD)

Table 5
Univariable and multivariable analysis on the consumption of any opioid Significant p values are shown in bold ASRM American society of reproductive medicine, SD standard deviation, BMI body mass index Statistics for continuous variable are presented by mean (SD)