Abdominal wall blocks in robotic hysterectomy for endometrial cancer are associated with a modest reduction in the frequency of patients receiving post-operative intravenous opioids.

Ioana Bondre: Writing – original draft, Methodology, Formal analysis, Data curation, Conceptualization. H. Meryem Soylu: Writing – review & editing, Data curation. Francesca Corbetta: Writing – review & editing. Breana Hill: Writing – review & editing, Conceptualization. Elise Wilson: Writing – review & editing, Conceptualization. Rodney Gabriel: . Ramez Eskander: Writing – review & editing. Michael McHale: Writing – review & editing. Cheryl Saenz: Writing – review & editing. Pratibha Binder: . Steven Plaxe: Writing – review & editing.

We performed a retrospective cohort study of patients registered in the Vizient Clinical Database (CDB) who had robotic hysterectomy for EC between January 2019 and December 2022. Data from the Vizient Clinical Data Base™ was used by permission of Vizient, Inc., all rights reserved. The study was IRB exempt. Inpatient data from all hospitals in the database were generated in a grouped fashion such that no individual hospital, or protected health information was identified. Only hospitals with available clinical resource management (CRM) data were included. The following restrictions were applied to the database report builder to capture the target patient population: diagnosis of malignant neoplasm of the endometrium, procedures performed included hysterectomy and robotic surgery codes. We defined exposure as an abdominal wall block, identified in the database query with the procedure codes for “insertion of anesthetic agent into subcutaneous tissue, muscle, or peritoneal cavity – percutaneous approach”. We excluded all cases with laparotomy procedure codes (converted cases). The primary outcome was any in-hospital use of any opioid or opioid combination, which was identified by querying the database for pharmaceutical clinical resource utilization. We primarily analyzed the data excluding fentanyl from the outcome (frequency of intravenous opioid use) due to its common intraoperative administration among patients given general anesthesia. Intravenous ketorolac and oral acetaminophen use data were also collected to evaluate secondary outcomes. The quantitative assessment of each medication was generated in standardized resource units (SRU), unique to each medication. A medication SRU in Vizient is a metric used to quantify and standardize the consumption of medications across healthcare facilities. It is used to assist in evaluating the efficiency and effectiveness of medication management. The conversion of SRU to milligrams for each medication is proprietary. We compared length of stay (LOS) and direct cost and their indices (defined as ratios between the observed and expected values of LOS and direct cost), readmission rates (at 7, 14, and 30 days) and several comorbidities between the groups in our study population (those who received a block and who did not). Direct cost includes cost of labor, supplies and materials, and other services related to direct patient care such as imaging or laboratory tests. While we compared readmission rates, we did not include the cost of readmission in the analysis as these data were not available. The expected length of stay and expected cost is calculated for each patient with Vizient’s proprietary data analytics model accounting for all inpatient records, ICD-10 codes, procedure codes and demographic information ( Hughes et al., 2021 ). The comorbidities we considered in the present study were obesity, type 2 diabetes mellitus, hypertension, and chronic pain disorder, (defined as any diagnosis of endometriosis, fibromyalgia, or other chronic pain syndrome). The mean of continuous variables (LOS, direct cost) were compared using Student’s t -test. We performed post-hoc calculations for computing achieved power if no significant difference was noted. Associations between categorical variables were compared by calculating risk ratios (relative risks). Mean medication resource units used, defined as the number of billing unit increments the hospital charged to the patient, are reported for each agent. Resource units differ between individual agents, so comparisons of total quantities of all opioids used per patient is not feasible with the data supplied. Further, standard deviations for the mean value of medication resource units of each agent administered is not reported in the dataset; therefore, we were unable to evaluate the statistical significance of observed differences between the average doses used for each opioid. P ≤ 0.05 was used to define statistical significance. Statistical analysis was performed using RStudio version 2022.2.3.492 (Boston, MA).

A smaller fraction of patients who had an abdominal wall block used IV opioids; however, we did not find lower oral opioid use, cost, length of stay, or readmission rates among patients who had a block.

The final study population consisted of 9,062 patients that underwent robotic hysterectomy for EC, in which 320 (3.5 %) received an abdominal wall block. Blocks were performed at 66 (20.4 %) of the 323 of hospitals contributing CRM data. There was no significant difference in frequency of patient comorbidities, including obesity, diabetes, hypertension, and chronic pain disorders between the study cohorts ( Table 1 ). Table 1 Abdominal wall block (N = 320) No abdominal wall block (N = 8742) Difference p-value Number of hospitals (n = 323) 66 (20.4 %) 319 (98.8 %) RR 0.21 [0.17–0.27] <0.001 LOS index 0.87 0.81 LOS (days ± SD) 2.88 ± 6.42 2.44 ± 4.16 0.44 [-0.27–1.15] 0.11 Direct cost index 0.95 0.96 Direct cost (USD ± SD) $11,119 ± 14,264 $10,655 ± 8,723 $464 [-1,109 – 2,037] 0.53 

 Readmissions (%) 7 days 0 81 (0.9 %) RR 0.17 [0.01–2.68] 0.20 14 days 4 (1.2 %) 131 (1.5 %) RR 0.83 [0.31–2.24] 0.72 30 days 5 (1.6 %) 159 (1.8 %) RR 0.86 [0.35–2.08] 0.73 

 Comorbidities Obesity (all BMI > 30) 163 (50.9 %) 4,040 (46.2 %) RR 1.1 [0.99–1.23] 0.08 Morbid obesity (BMI > 40) 80 (25.0 %) 2,121 (24.3 %) RR 1.03 [0.85–1.25] 0.76 Type 2 DM 108 (33.8 %) 2,844 (32.5 %) RR 1.04 [0.88–1.21] 0.64 Hypertension 218 (68.1 %) 5,849 (66.9 %) RR 1.02 [0.94–1.10] 0.64 Chronic pain 34 (10.6 %) 920 (10.5 %) RR 1.01 [0.73–1.39] 0.95 endometriosis 12 (3.8 %) 447 (5.1 %) RR 0.73 [0.42–1.28] 0.28 fibromyalgia 6 (1.9 %) 161 (1.8 %) RR 1.02 [0.45–2.28] 0.96 Other chronic pain 16 (5.0 %) 312 (3.6 %) RR 1.40 [0.86–2.29] 0.18 Table 2 Cases in which analgesics were used. Abdominal wall block (N = 320) No abdominal wall block (N = 8742) Difference p-value Any IV and/or oral opioid (fentanyl excluded) 212 (66.3 %) 5,662 (64.8 %) RR 1.02 [0.94–1.10] 0.57 Any IV opioid (fentanyl excluded) 184 (57.5 %) 5,890 (67.4 %) RR 0.85 [0.77–0.94] 0.001 Any oral opioid 203 (63.4 %) 5,365 (61.4 %) RR 1.03 [0.95–1.12] 0.44 Any non-opioid pain medicine 303 (94.7 %) 7,882 (90.2 %) RR 1.05 [1.02–1.08] <0.001 Only non-opioid pain medicine (except fentanyl) 98 (30.6 %) 2,421 (27.7 %) RR 1.10 [0.93–1.31] 0.24 No pain medicine used at all (only Fentanyl used) 10 (3.1 %) 659 (7.5 %) RR 0.39 [0.19–0.65] 0.0009 0 (0 %) 59 (0.6 %) RR 0.23 [0.01–3.75] 0.3044 Cases in which analgesics were used. Receipt of a block was associated with decreased intravenous opioid use (57.5 % vs 67.4 %, RR 0.85, 95 % CI 0.77–0.94, P = 0.001). Hydromorphone was the most used intravenous opioid in both groups of patients. Patients in the block group were 17 % less likely to receive hydromorphone compared to those who did not receive blocks (52.5 % vs 64.8 %, RR 0.81, 95 % CI 0.73–0.89, p-value < 0.001). We did not find a significant difference between use of intravenous morphine or meperidine between groups. 203 (63.4 %) patients in the block group received oral opioids post-operatively, compared to 5365 (61.4 %) in the no block group (RR 1.03, 95 % CI 0.95–1.12, p-value 0.44). Patients in the block group were more likely to receive oral hydromorphone (6.9 % vs 3.9 %, RR 1.73, 95 % CI 1.14–2.62, p-value 0.009) and hydrocodone/acetaminophen (11.3 % vs 0.88 %, RR 12.77, 95 % CI 8.73–18.67, p-value < 0.001). There were no significant differences in rates of use of other oral opioids ( Table 3 , Table 4 ). Table 3 Use of individual analgesic agents. Abdominal wall block (N = 320) No abdominal wall block (N = 8742) Difference p-value Intravenous opioids Fentanyl 304 (95.0 %) 7,468 (85.4 %) RR 1.11 [1.08–1.14] <0.001 Hydromorphone (IV) 168 (52.5 %) 5,667 (64.8 %) RR 0.81 [0.73–0.89] <0.001 Morphine (IV) 35 (10.9 %) 787 (9.0 %) RR 1.21 [0.88–1.67] 0.21 Meperidine (IV) 5 (1.6 %) 234 (2.7 %) RR 0.58[0.24–1.41] 0.23 

 Oral opioids Oxycodone (Oral) 152 (47.5 %) 3,754 (42.9 %) RR 1.11 [0.98–1.24] 0.09 Oxycodone/Acetaminophen (Oral) 17 (5.3 %) 380 (4.8 %) RR 1.22 [0.76–1.96] 0.40 Hydromorphone (Oral) 22 (6.9 %) 345 (3.9 %) RR 1.73 [ 1.73–2.62] 0.009 Morphine (Oral) 3 (0.9 %) 47 (0.5 %) RR 1.74 [0.54–5.57] 0.34 Hydrocodone/Acetaminophen (Oral) 36 (11.3 %) 77 (0.88 %) RR 12.77 [8.73–18.67] <0.001 

 Non-opioid pain medication Ketorolac (IV) 138 (43.1 %) 3,322 (38.0 %) RR 1.13 [0.99–1.29] 0.05 Ketorolac (Oral) 1 (0.3 %) 7 (0.08 %) RR 3.90 [0.48–31.62] 0.20 Acetaminophen (IV) 109 (34.1 %) 2,693 (30.8 %) RR 1.10 [0.94–1.29] 0.21 Acetaminophen (Oral) 230 (71.9 %) 6.302 (72.1 %) RR 0.99 [0.93–1.07] 0.08 Ibuprofen (Oral) 105 (32.8 %) 2,388 (27.3 %) RR 1.20 [1.02–1.41] 0.025 Tramadol (Oral) 27 (8.4 %) 959 (11.0 %) RR 0.77 [0.53–1.11] 0.16 Table 4 Standardized resource units of individual agents utilized and duration of use. Abdominal wall block No abdominal wall block % Difference (block vs. no block) (N = 320) (N = 8,742) Opioid utilization per case (mean resource units) Morphine (IV) 4.9 2.7 45 % Hydromorphone (IV) 3.3 3.1 6 % Meperidine (IV) 1.2 1.2 0 % Oxycodone (oral) 6.5 5.7 12 % Oxycodone/Acetaminophen (oral) 9.8 5.3 46 % Hydromorphone (oral) 2 4.1 −105 % Hydrocodone/Acetaminophen (oral) 4.5 4.8 −7% Morphine (oral) 1 13 −1200 % 

 Opioid utilization per case per day (resource units) Morphine (IV) 2.5 1.8 28 % Hydromorphone (IV) 2.5 2.4 4 % Meperidine (IV) 1.2 1.2 0 % Oxycodone (oral) 2.8 2.7 4 % Oxycodone/Acetaminophen (oral) 3.2 2.4 25 % Hydromorphone (oral) 1.6 2.3 −44 % Hydrocodone/Acetaminophen (oral) 2.2 2.3 −5% Morphine (oral) 1 3.4 −240 % 

 Days (mean) any opioid used Morphine (IV) 2 1.5 25 % Hydromorphone (IV) 1.3 1.3 0 % Meperidine (IV) 1 1 0 % Oxycodone (oral) 2.4 2.2 8 % Oxycodone/Acetaminophen (oral) 3.1 2.2 29 % Hydromorphone (oral) 1.2 1.8 −50 % Hydrocodone/Acetaminophen (oral) 2.1 2.1 0 % Morphine (oral) 1 3.9 −290 % 

 Non-opioid analgesic utilization per case (resource units) Acetaminophen (Oral) 15.8 15.1 4 % Acetaminophen (IV) 151.5 137.5 9.2 % Ketorolac (IV) 4.1 4.5 −10 % Ketorolac (Oral) 1.0 1.7 42 % Ibuprofen (Oral) 4.5 3.8 16 % Tramadol (Oral) 6.9 3.9 43.4 % 

 Non-opioid analgesic utilization per case per day (resource units) Acetaminophen (Oral) 5.7 5.5 4 % Acetaminophen (IV) 92.0 98.9 −7.5 % Ketorolac (IV) 2.5 2.3 8 % Ketorolac (Oral) 1.0 1.2 −20 % Ibuprofen (Oral) 2.2 2.3 −5% Tramadol (Oral) 2.9 1.9 34 % 

 Days any non-opioid analgesic used Acetaminophen (Oral) 2.8 2.7 4 % Acetaminophen (IV) 1.6 1.4 12.5 % Ketorolac (IV) 1.6 2.0 −25 % Ketorolac (Oral) 1.0 1.4 −40 % Ibuprofen (Oral) 2.1 1.7 19 % Tramadol (Oral) 2.4 2.0 17 % Use of individual analgesic agents. Standardized resource units of individual agents utilized and duration of use. Most patients in our cohort received non-opioid analgesics. Those who received a block were modestly more likely to receive a non-opioid analgesic (94.7 % vs 90.2 %, RR 1.05, 95 % 1.02–1.08, p < 0.001). Patients with blocks were 20 % more likely to receive ibuprofen (32.8 % vs 27.3 %, RR 1.20, 95 % CI 1.02–1.41, p 0.025). However, we did not find a significant difference (30.6 % vs 27.7 %, RR 1.10, 95 % 0.93–1.31, p 0.24) in rates of patients receiving only non-opioid medications for post operative pain relief ( Table 2 ). Higher SRUs (higher amounts) of IV morphine, IV hydromorphone, oxycodone and oxycodone/acetaminophen were used in the block group, with the biggest differences noted in IV morphine and oxycodone/acetaminophen. The number of days opioids were used and the mean number of standardized resource units of each opioid (among patients who received opioids) used was similar between groups but trended toward higher usage in the block group ( Table 3 ). We could not evaluate statistical significance of the differences in resource units used between groups, due to the standard deviations not being available in the Vizient dataset. Readmission rates appeared similar between groups at all time-points ( Table 1 ). We did not find statistically significant differences between length of stay or direct cost of procedure between groups ( Table 1 ). While the direct cost was $464 higher and the LOS 0.44 days longer in the block group, these differences were not found to be statistically significant. Our population provided 80 % power to detect a difference of greater than or equal to $2240 in direct cost and a difference in LOS of 1.02 days, at an alpha of 0.05.

Abdominal wall blocks are infrequently performed in patients undergoing robotic hysterectomies for EC in the Vizient Clinical Database. Overall, there is no difference in opioid use dependent on whether a block is given; however, given the uncertainty about when the fentanyl is given, there might be a modest decrease. Despite less frequent IV opioid use, we found a suggestion of longer LOS and higher direct cost among patients who had abdominal wall blocks. Further studies are needed for a more precise assessment of opioid use in this patient population. Our findings suggest that additional investigation is warranted to determine the best use of abdominal wall blocks prior to robotic hysterectomy for endometrial cancer.

We have reported on the association between abdominal wall block and opioid use in individuals undergoing robotic hysterectomy for EC. Endometrial cancer patients were included in other studies reporting on surgical outcomes and regional perioperative analgesia; however, these studies also included patients undergoing hysterectomy for benign indications or patients undergoing both robotic and non-robotic laparoscopic hysterectomies ( Hughes et al., 2021 , Shin et al., 2020 , El Hachem et al., 2015 , Ghisi et al., 2016 , Hansen et al., 2021 , Hotujec et al., 2015 , Hutchins et al., 2015 , Jadon et al., 2021 , Lopez-Ruiz et al., 2022 , Mathew et al., 2019 , McDonald et al., 2022 , Torup et al., 2015 , Yap et al., 2018 ). To the best of our knowledge, this is the first study investigating the role of abdominal wall blocks in a population comprising only patients who had robotic surgery for EC. We believe restricting the study population, as we have done, is clinically relevant given the rising incidence of the disease and the widespread use of robotic surgery. Due to disease and patient factors, as well as staging procedures in selected patients, endometrial cancer surgery may be of longer duration and increased complexity compared to hysterectomies performed for benign indications. We found that abdominal wall blocks were associated with a 15 % lower risk of intravenous opioid use. This difference was largely accounted for by a 17 % less frequent use of IV hydromorphone. We were unable to assess the standard resource units of opioids used between groups for statistical significance since each drug has a distinct dose defined as a resource unit in the data set. Non-opioid analgesics were used in more than 90 % of patients in both groups; slightly more frequently in the patients who had a block (RR 1.05 [1.02–1.08], p < 0.001). The selection of intravenous and oral opioids included in our analysis was guided by standard practice in management of post-operative pain as reflected in the information supplied by hospitals to Vizient. As mentioned above, consistent with previous studies ( Leitao et al., 2013 ), nearly all patients received fentanyl, as a component of general anesthesia, so we analyzed the data including and excluding Fentanyl. We observed that approximately 10 % more of the patients who had a block were given fentanyl (95 % vs. 85 %) perhaps suggesting a concern that the block alone might not be adequate to attain adequate pain relief. In the present series, there was no association of blocks with any (including either oral or IV) opioid use and only a modest decrease in intravenous opioids. Our results are consistent with the existing body of research suggesting that block use in the setting of robotic hysterectomies is not associated with large scale reductions in opioid use ( Hughes et al., 2021 , Shin et al., 2020 , El Hachem et al., 2015 , Ghisi et al., 2016 , Hansen et al., 2021 , Hotujec et al., 2015 , Hutchins et al., 2015 , Jadon et al., 2021 , Lopez-Ruiz et al., 2022 , Mathew et al., 2019 , McDonald et al., 2022 , Torup et al., 2015 ). The largest available systematic review and meta -analysis was performed by Shin et al. and investigated the effect of transverse abdominis plane (TAP) blocks on pain scores and opioid use in laparoscopic and robotic assisted hysterectomies. The authors did not identify a significant reduction in either opioid use or pain scores and concluded that TAP blocks should not be routinely used in this setting ( Hughes et al., 2021 ). Limited data exists regarding cost-effectiveness of abdominal wall blocks for endometrial cancer patients. A study by Seagle et al. found that TAP blocks performed for laparoscopic hysterectomies for EC are only cost-effective when coupled with same-day discharge ( Yap et al., 2018 ). In our study, the mean LOS and total direct cost both trended higher among patients who had a block, although the differences were not statistically significantly different. Our study has several limitations. Due to the structure and availability of data in the Vizient CDB, we were unable to differentiate between different types of abdominal wall block (transversus abdominis plane (TAP) or quadratus lumborum (QL) or analgesic use. This is potentially relevant given that previous evidence suggests improved pain control with QL blocks in obstetric and gynecologic surgeries (Cesarean sections and total abdominal hysterectomies) ( Seagle et al., 2017 , Bacal et al., 2019 , Nedeljkovic et al., 2020 ). However, these studies predominantly evaluated open procedures, in which the utility of regional anesthesia has already been established. QL block data in minimally invasive gynecologic surgery is reported in only one study (Hansen et al,) which showed no difference in opioid use in patients undergoing laparoscopic hysterectomy with versus without QL block ( Ghisi et al., 2016 ). Moreover, there are significant technique differences in administration of blocks (i.e., ultrasound guided vs laparoscopically guided) which could have an impact on analgesic effectiveness ( Hotujec et al., 2015 , Mathew et al., 2019 ). The CDB does not provide data on the type of local anesthetic used (liposomal formulation or not). However, only one study by Hutchins et al showed improved pain control with TAP blocks performed with liposomal bupivacaine compared to bupivacaine in patients undergoing laparoscopic hysterectomy for EC ( Hotujec et al., 2015 ), while the consensus in anesthesiology literature suggests no added benefit to using liposomal bupivacaine formulations ( Tan et al., 2020 , Otremba et al., 2022 , Zhu et al., 2020 ). Another limitation is our inability to distinguish between the phases of care in which the opioids were given (intra-op, PACU or floor). Future prospective studies will be designed to capture these variables. The mean LOS was greater than 2 days in both groups, which is longer than we would have expected for robotic hysterectomy. At our institution, patients undergoing robotic hysterectomy for EC are typically discharged same day or observed overnight and discharged home approximately 24 h after surgery. Our dataset excluded patients converted to laparotomy and the comorbidity rates observed in this cohort closely approximate the documented prevalence in the target population. The overall complication rate is similarly low between groups (4 % vs 2 %) and is comparable to published data ( Wright et al., 2013 ). It is unclear if this longer LOS reflects differences in regional / institutional practices or if there are other contributing factors. An additional limitation is that pain scores are not recorded in the dataset. In absence of this information, we cannot ascertain if the differences noted in analgesic use are correlated with severity of pain or of the adequacy of post-operative pain relief.

Endometrial cancer (EC) is the most common gynecologic malignancy in developed countries ( Koskas et al., 2021 ). Robotic surgery has quickly increased in frequency ( Muaddi et al., 2021 ) and has been established as a safe and effective approach to EC treatment. Studies suggest superior outcomes compared to traditional laparoscopy (shorter hospital stays, decreased post-operative pain, and quicker recovery) with the added benefit of increased feasibility in obese and morbidly obese patients ( Cohn et al., 2016 , Alshowaikh et al., 2021 , Betcher et al., 2014 , Fader, 2017 , Leitao et al., 2013 ). Since its introduction to gynecologic surgery in 2016, the Enhanced Recovery After Surgery (ERAS) protocol has been associated with improved outcomes; early literature documents these improvements primarily among patients having open operations ( Nelson et al., 2019 , de Groot et al., 2016 ). Recent evidence suggests that ERAS guidelines can also be safely implemented in minimally invasive cases, resulting in less pain and fatigue and improved ambulation ( de Groot et al., 2016 ). Optimizing peri -operative analgesia using a multi-modal, opioid sparing regimen is a key tenet of ERAS protocols ( Nelson et al., 2019 ). Therefore, regional anesthesia has been increasingly used in the management of peri -operative pain with the goal of minimizing opioid use. While evidence shows improved pain scores when regional anesthesia is used in open gynecologic and obstetric procedures, studies have shown equivocal results in minimally invasive procedures ( Nelson et al., 2019 , de Groot et al., 2016 , Modesitt et al., 2016 ). Our study aims to investigate the association between abdominal wall block use and in-hospital post-operative opioid use among patients undergoing a robotic hysterectomy for endometrial cancer. To our best knowledge, this is the first study investigating this question exclusively in an endometrial cancer population.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.