Abstract
Background and Objective:
Minimally invasive endometriosis surgery can be done laparoscopically or with robotic assistance. The choice of modality of this procedure could impact postprocedure pain levels and the need for opioids to manage pain. The objective of this study is to evaluate differences in opioid prescription fill rates following endometriosis surgery between robotic-assisted and laparoscopic approaches.
Methods:
A retrospective cohort study using the Merative MarketScan (United States employer-based claims database) from 2016 to 2021. Opioid prescription fills were examined in three time periods: preoperative (12 months to 30 days before surgery), perioperative (29 days before to 14 days after), and postoperative (15 days to 24 months after). Logistic regression models estimated average marginal effects (AME) for perioperative and postoperative opioid use, adjusting for demographics, health conditions, concomitant procedures, and prior opioid use.
Results:
Of the 28,088 individuals who underwent endometriosis surgery, 6.24% (1,752 patients) had robotic-assisted procedures. In the postoperative period, laparoscopic surgery was associated with a higher probability of postoperative opioid prescription fills compared to robotic-assisted surgery (adjusted AME = 3.2; 95% CI: 0.7, 5.7; P ≤ .01). Among patients with baseline pain disorders, robotic-assisted surgery was associated with lower postoperative opioid prescription fills (adjusted AME = 3.2; 95% CI: 0.2, 6.2; P = .04).
Conclusions:
Robotic-assisted procedures were associated with a reduced likelihood of opioid prescription fills in the postoperative period following surgery compared to laparoscopic procedures, including for patients with baseline pain disorders.
Keywords: Endometriosis, Laparoscopy, Opioid use, Pain disorders, Robotics
INTRODUCTION
Endometriosis is a chronic condition that affects 10% of women globally, leading to significant symptoms such as dysmenorrhea, dyspareunia, chronic pain, and infertility.1 Current treatments include pain management with nonsteroidal anti-inflammatory drugs (NSAIDs), acetaminophen, and hormonal therapies. Surgical excision of endometriosis remains the mainstay treatment for achieving symptom relief and restoring organ function through the complete removal of endometriotic tissue. While conventional laparoscopy has long been the standard for endometriosis resection, robotic-assisted surgery has gained traction due to its potential for greater precision and reduced surgeon fatigue. However, its benefits over laparoscopy, particularly regarding postoperative pain and opioid use, remain unclear.2
Despite expectations that surgery reduces pain and opioid use, prior research has not clarified whether surgical modality influences postoperative opioid prescribing patterns. Most studies have examined opioid use in gynecologic surgery broadly rather than focusing on endometriosis excision. Furthermore, few have stratified opioid use by postoperative timeframes or accounted for baseline pain disorders, both of which may influence prescribing trends.
This study aims to fill these gaps by analyzing opioid prescription fill rates in the preoperative, perioperative, and postoperative periods following robotic-assisted and laparoscopic endometriosis surgery. Utilizing data from the Merative MarketScan Research Database, this study seeks to determine whether robotic-assisted surgery for endometriosis is associated with differences in postoperative opioid prescription fill rates compared to laparoscopic excision. The findings could inform surgical decision-making and postoperative pain management in endometriosis patients.
METHODS
Data Source
This study employed retrospective claims data analysis utilizing the Merative MarketScan Research Databases. This aggregated database encompasses paid claims and encounter data from over 273 million unique individuals, sourced from employers, health plans, as well as Medicare and Medicaid programs. The database documents enrollment, inpatient, outpatient, and prescription drug service usage in the United States.3
Due to the observational nature of this study and the use of de-identified patient data, the Institutional Review Board at the University of South Florida determined that the study was exempt from review.
Cohort
Individuals with an endometriosis procedure identified in the Merative data from January 1, 2016, to December 31, 2021, were included in the primary analysis. To be eligible, individuals were required to have an insurance plan with prescription coverage and continuous enrollment from one year before to two years after the procedure. Individuals were excluded if they had missing demographic data, lacked continuous prescription coverage, or had procedures coded without clear identification of surgical modality. A total of 28,088 individuals met these criteria (Figure 1).
Figure 1.
Cohort formation.
Endometriosis Procedure and Modality
Endometriosis procedures were identified through current procedural terminology (CPT) code 58662. Robotic-assisted procedures were identified when a secondary CPT code indicated the use of robotic assistance (see Table 1 for CPT and International Classification of Diseases [ICD]-10 procedure codes). Complexity-increasing concomitant procedures (e.g., bowel resections, bladder or ureteral repairs, and other advanced pelvic procedures) were identified using CPT codes and grouped to reflect multivisceral involvement in advanced endometriosis surgery; these are also listed in Table 1.
Table 1.
Codes Used to Identify Procedures and Diagnoses
| Procedure Modality | Code Type | Code |
|---|---|---|
| Robotic | CPT | S2900 |
| ICD-10 PR | 8E090CZ, 8E093CZ, 8E094CZ, 8E097CZ, 8E098CZ, 8E09XCZ, 8E0W0CZ, 8E0W3CZ, 8E0W4CZ, 8E0W7CZ, 8E0W8CZ, 8E0WXCZ, 8E0X0CZ, 8E0X3CZ, 8E0X4CZ, 8E0XXCZ, 8E0Y0CZ, 8E0Y3CZ, 8E0Y4CZ, 8E0YXCZ | |
| Procedure | ||
| Hysterectomy | ICD-10 PR | 0UT90ZL, 0UT90ZZ, 0UT94ZL, 0UT94ZZ, 0UT97ZL, 0UT97ZZ, 0UT98ZL, 0UT98ZZ, 0UT9FZL, 0UT9FZZ |
| CPT | 58150, 58152, 58180, 58200, 58210, 58260, 58262, 58263, 58267, 58270, 58275, 58280, 58285, 58290, 58291, 58292, 58293, 58294, 58541, 58542, 58543, 58544, 58548, 58550, 58552, 58553, 58554, 58570, 58571, 58572, 58573 | |
| Oophorectomy/Salpingectomy | ICD-10 PR | 0UT20ZZ, 0UT24ZZ, 0UT27ZZ, 0UT28ZZ, 0UT2FZZ, 0UT00ZZ, 0UT04ZZ, 0UT07ZZ, 0UT08ZZ, 0UT0FZZ, 0UT10ZZ, 0UT14ZZ, 0UT17ZZ, 0UT18ZZ, 0UT1FZZ, 0UT70ZZ, 0UT74ZZ, 0UT77ZZ, 0UT78ZZ, 0UT7FZZ, 0UT50ZZ, 0UT54ZZ, 0UT57ZZ, 0UT58ZZ, 0UT5FZZ, 0UT60ZZ, 0UT64ZZ, 0UT67ZZ, 0UT68ZZ, 0UT6FZZ |
| CPT | 58262, 58263, 58291, 58292, 58542, 58544, 58548, 58552, 58554, 58571, 58573, 58661, 45560, 57240, 57250, 57260, 57265, 57267, 57268, 57270, 57282, 57283, 57284, 57285, 57295, 57296, 57423, 57426, 58270, 58292, 58294 | |
| Cocomplexity Increasing Procedure | CPT | 44145, 44146, 44150, 44160, 44180, 44187, 44188, 44202, 44204, 44205, 44207, 44208, 44210, 44212, 44213, 44227, 44238, 45100, 45300, 45330, 45378, 45395, 45999, 44120, 50431, 50600, 50684, 50693, 50780, 50785, 50947, 50949, 51860, 51865, 51999, 52332, 52341, 52344, 52345, 50949, 58999 |
| Diagnosis | ||
| Pain Disorders—not commonly related to endometriosis | ICD-9 DX | 711.*, 712.*, 713.*, 714.*, 715.*, 716.*, 717.*, 718.*, 719.*, 725,726.*, 727.*, 728.*, 729.3*, 729.7*, 729.8*, 729.9*, 730.*, 731.*, 732.*, 733.*, 734, 735.*, 736.*, 737.2*, 737.4*, 738.1*, 710, 710.1, 710.3, 710.4, 710.5, 710.8, 710.9, 729, 729.2, 729.4, 729.5, 729.6, 737, 737.8, 737.9, 738, 738.2, 738.3, 738.6, 738.7, 738.8, 738.9, 739, 739.5, 739.6, 739.7, 739.8, 739.9, 722.30, 722.32, 722.33, 722.70, 722.72, 722.73,722.80, 722.82, 722.83, 722.90, 722.92, 722.93, 737.1, 737.3, 738.4, 738.5, 739.2, 739.3, 739.4, 756.10, 756.11, 756.12, 756.19, 805.4, 805.8, 839.2, 839.42, 846, 846.0, 847.1, 847.2, 847.3, 847.9, 721.3* − 721.9*, 722.2*, 724.*, 756.13, 721.0*, 721.1*, 722.0*, 722.31, 722.71, 722.81, 722.91, 839.0, 839.1, 847.0, 784.0, 375.15, 786.59, 530.5, 536.8, 617.*, 710.2, 388.3, 780.52 |
| ICD-10 DX | M00.*, M01.*, M02.*, M05.*, M06.*, M08.*, M11.*, M12.*, M13.*, M14.*, M15.*, M16.*, M17.*, M18.*, M19.*, M20.*, M21.*, M22.*, M23.*, M24.*, M25.*, M32.*, M33.*, M34.*, M36.*, M42.*, M61.*, M62.*, M65.*, M6.*, M67.*, M70.*, M71.*, M72.*, M75.*, M76.*, M77.*, M80.*, M81.*, M84.*, M85.*, M86.*, M87.*, M88.*, M89.*, M90.*, M91.*, M92.*, M93.*, M94.*, M95.*, M40.1*, M40.4*, M40.5*, M41.4*, M41.5*, M43.8*, M60.0*, M60.1*, M60.2*, M79.6*, M43.9*, M35.1*, M35.2*, M35.3*, M35.4*, M35.5*, M35.6*, M35.7*, M35.8*, M35.9*, M96.4*, R25.2*, R26.2*, R29.4*, R29.898, S12.000K, S12.001K, S12.100K, S12.101K, S12.200K, S12.201K, S12.300K, S12.301K, S12.400K, S12.401K, S12.500K, S12.501K, S12.600K, S12.601K, S42.009K, S42.009P, S42.209K, S42.209P, S42.91SK, S42.92SK, S82.009P, S82.009Q, S82.009R, S92.819K, S92.819P, S92.909K, S92.909P, S92.919K, S92.919P, S99.209K, S99.209P, S99.219K, S99.219P, S99.229K, S99.229P, S99.239K, S99.239P, S99.249K, S99.249P, S99.299K, S99.299P, M48.40*A, M48.41*A, M48.42*A, M48.43*A, M48.44*A, M48.45*A, M48.46*A, M48.47*A, M48.48*A, M48.50*A, M48.51*A, M48.52*A, M48.53*A, M48.54*A, M48.55*A, M48.56*A, M48.57*A, M48.58*A, S02.91*K, S02.92*K, S42.90*K, S52.90*K, S62.90*K, S72.90*K, S82.90*K, S52.90*M, S72.90*M, S82.90*M, S52.90*N, S72.90*N, S82.90*N, S52.90*Q, S72.90*Q, S82.90*Q, S52.90*R, S72.90*R, S82.90*R, S42.90*P, S42.91*P, S42.92*P, S52.90*P, S62.90*P, S72.90*P, S82.90*P, S22.9**K, S32.9**K, M40.0*, M40.2*, M41.00, M41.04-M41.08, M41.20, M41.24-M41.27, M41.3*, M41.80, M41.84-M41.87, M41.9, M43.00, M43.04-M43.10, M43.14-M43.19, M43.27, M43.28, M46.40, M46.44-M46.48, M47.10, M47.14-M47.16, M47.814-M47.819, M48.00, M48.04-M48.08, M48.10, M48.14-M48.20, M48.24-M48.27, M48.30, M48.34-M48.38, M48.9, M51.04-M51.06, M51.44-M51.47, M51.84-51.87, M51.9, M53.2*4-M53.2*9, M53.3, M53.9, M54.04-M54.08, M54.14-M54.18, M54.3*-M54.6, M54.89, M54.9, M96.1-M96.3, M96.5, M99.02-M99.04, M99.83, M99.84, Q76.2, Q76.4*, S22.009A, S23.101A, S23.111A, S23.121A, S23.123A, S23.131A, S23.133A, S23.141A, S23.143A, S23.151A, S23.153A, S23.161A, S23.163A, S23.171A, S23.3XXA, S23.8XXA, S23.9XXA, S32.009A, S32.10XA, S32.2XXA, S33.101A, S33.2XXA, S33.5XXA, S33.6XXA, S33.8XXA, S33.9XXA, M41.02, M41.03, M41.22, M41.23, M41.82, M41.83, M43.01, M43.02, M43.03, M43.11, M43.12, M43.13, M43.6, M47.11, M47.12, M47.13, M47.811, M47.812, M47.813, M48.01, M48.02, M48.03, M50.0*, M50.2*, M50.8*, M50.9*, M53.0, M53.1, M53.82, M54.01, M54.02, M54.03, M54.11, M54.12, M54.13, M54.2, S13.101A, S13.111A, S13.121A, S13.131A, S13.141A, S13.151A, S13.161A, S13.171A, S13.181A, S13.4XXA, S13.8XXA, R51.*, K59.*, N80.*, K30.*, M60.9*, M79.1*, M79.9*, M60.8*, K22.4*, H93.1*, M35.0*, M54.81*, R07.82*, R07.89*, G47.00*, H04.129, M54.20 | |
| Pain Disorders—potentially related to endometriosis | ICD-9 DX | 346.*, 595.1, 780.71, 307.81, 729.1, 338.0, 724.2, 625.7*, 524.6*, 564.*, 729.1, 723.*, 739.1 |
| ICD-10 DX | G43.*, K58.*, M26.6*, N30.10*, N30.11*, N94.81*, R53.82*, G44.201, G44.209, G89.0, M79.7, M79.15, M54.5, M99.01 | |
| Substance Use Disorder | ICD-9 DX | 291, 292, 303–305 |
| ICD-10 DX | F10–F19, F55 | |
| Smoking | ICD-9 DX | 305.1, 649.0, 649.00, 649.01, 649.02, 649.03, 649.04, V15.82, 989.84 |
| ICD-10 DX | F17.2, 099.33, P04.2, P96.81, T65.2, Z57.31, Z71.6, Z72, Z77.2, Z87.8 | |
Outcomes
We examined perioperative (29 days before to 14 days after surgery) and postoperative (15 days to 24 months after surgery) opioid prescription fill rates.
Opioid use was identified in the outpatient prescription claims using National Drug Classification (NDC) codes, which captures outpatient dispensation well.4 Claims that had a negative days’ supply or negative quality listed were excluded (this excluded 6 claims).
Opioid use was examined in two ways; first as a binary outcome examining whether a prescription had been filled and second as the total days supply of opioids that were prescribed in the postoperative period.
Demographic and Health Conditions
Demographic information was obtained from the enrollment file, and included age, and location (urban vs rural). We included location as access to laparoscopy may differ from access to robotic surgery in rural areas.5
Substance use, smoking, and pain disorders were analyzed as potential opioid use predictors in the year before surgery. Pain disorders were classified as potentially related to endometriosis (e.g., chronic pelvic pain, fibromyalgia, migraines) or unrelated (e.g., arthritis, neck pain). ICD-10 codes for these conditions are detailed in Table 1.
Concomitant Procedures
Individuals undergoing endometriosis surgery may also receive additional surgical interventions during the same procedure. These concomitant procedures, including hysterectomy, oophorectomy/salpingectomy, and other procedures aimed at addressing more extensive disease, could significantly affect postoperative outcomes, such as pain levels and opioid use. For example, surgeries involving gastrointestinal or urinary tract endometriosis are generally more complex and may present higher rates of perioperative and postoperative challenges. We used these complexities as surrogates for more advanced stages of endometriosis, as direct severity data was not available. These interventions are classified as “complexity-increasing procedures,” with specific CPT codes listed in Table 1.
Statistical Analysis
We first examined whether there were differences in baseline demographic, health conditions, and concomitant procedures between those who had a laparoscopic endometriosis procedure and those that had a robotic-assisted endometriosis procedure; differences were compared using χ2 tests for binary/categorical variables and t-tests for continuous variables.
Next, we compared rates of opioid prescription fills between those who had a laparoscopic endometriosis procedure and those who had a robotic assisted endometriosis procedure. Differences were examined in 3 time periods: preoperative (365 to 30 days before the procedure), peri-operative (29 days before to 14 days after the procedure), and postoperative (15 days to 24 months after the procedure). Differences were compared using χ2 tests.
Binary logistic models were used to obtain average marginal effects (AME) and 95% confidence interval (CI) of perioperative and postoperative opioid use, comparing those whose endometriosis procedure was performed with robotic assistance to those whose procedure was performed laparoscopically. The adjusted model examining perioperative opioid use included preoperative opioid use, demographics, health conditions, and concomitant procedures; the adjusted postoperative opioid use model also included perioperative opioid use. We also provided descriptive information on the total days supply of opioids in the postoperative period among individuals in each group that had at least one opioid prescription.
To determine whether opioid use differences were due to surgical modality rather than pre-existing pain conditions, four subanalyses were conducted. The first, excluded all individuals who had baseline pain disorders not associated with endometriosis (e.g., arthritis, neck pain). The second, only included individuals who had a complex procedure. The third, stratified results by presence of baseline pain disorders. The fourth, focused on patients with surgery for endometriosis alone in patients with and without baseline pain disorders, excluding those who had a concomitant hysterectomy or oophorectomy/salpingectomy to control for the potential impact of these procedures on long-term pain outcomes. These subanalyses also used logistic models to obtain AMEs.
RESULTS
Of the 28,088 individuals who had an endometriosis procedure, 1,752 (6.24%) were performed with robotic assistance. Patients undergoing robotic-assisted procedures were more likely to be older, live in urban areas, smoke and undergo concomitant surgical procedures compared to those undergoing conventional laparoscopy (Table 2).
Table 2.
Individual Demographic and Health Characteristics and Concomitant Procedures, by Modality of Endometriosis Procedure
| Characteristics | Laparoscopic; n (%) | Robotic Assisted; n (%) | Difference P-Valueb |
|---|---|---|---|
| Age; mean (SD) | 34.7 (10.0) | 38.4 (9.0) | <.01 |
| Rurality | |||
| Rural | 3,824 (14.5) | 136 (7.80) | <.01 |
| Urban | 19,651 (74.6) | 1504 (85.8) | |
| Missing | 2,859 (10.9) | 114 (6.5) | |
| Baselinea Health Characteristics | |||
| Substance Use Disorders | 1,644 (6.2) | 104 (5.9) | .60 |
| Smoking | 7,071 (26.9) | 588 (33.5) | <.01 |
| Pain Disorders | 18,011 (68.4) | 1262 (72.0) | <.01 |
| Overlap with Endometriosis | 13,356 (50.7) | 960 (54.7) | <.01 |
| Do not Overlap with Endometriosis | 15,482 (58.8) | 1031 (58.8) | .99 |
| Concomitant Procedures | |||
| Hysterectomy | 2,703 (13.2) | 495 (38.7) | <.01 |
| Oophorectomy/Salpingectomy | 8,291 (31.5) | 899 (51.3) | <.01 |
| Complexity Increasing Procedure | 1,032 (3.9) | 144 (8.2) | <.01 |
In the year before the procedures.
Test is t test for continuous variable (age) and χ2 test for binary/categorical variables (all others).
There was no significant difference in the proportion of individuals filling opioid prescriptions in the two preoperative periods between the two groups (Figure 2; Table 3). However, in the perioperative period, opioid fill rates were higher among those who had a robotic-assisted procedure (86.5%) compared to those who had a laparoscopic procedure (83.7%). This difference did not remain significant after adjusting for demographics, health conditions, and concomitant procedures (adjusted AME = –2.1; 95% CI: –3.9, 0.4; P = .02) (Table 4).
Figure 2.
Percent of population filling an opioid prescription before and after Endometriosis procedure, by procedure modality.
Table 3.
Opioid Use before and After Endometriosis Procedure
| Time to Procedure | Filled an Opioid Prescription | Difference P-Value | |
|---|---|---|---|
| Laparoscopic; n (%) | Robotic Assisted; n (%) | ||
| Before (−365 to −30 days) | |||
| 2 (12–6 months) | 4,742 (18.0) | 300 (17.1) | .34 |
| 1 (6 months to 3 days) | 5,131 (19.5) | 368 (21.0) | .13 |
| Perioperative (−29 to 14 days) | 22,041 (83.7) | 1517 (86.5) | <.01 |
| After (15 days to 2 years) | |||
| 1 (14 days to 6 months) | 4,530 (17.2) | 295 (16.8) | .68 |
| 2 (6–12 months) | 4,890 (18.6) | 286 (16.3) | .02 |
| 3 (12–18 months) | 4,766 (18.1) | 287 (16.4) | .07 |
| 4 (18–24 months) | 4,464 (17.0) | 247 (14.1) | <.01 |
Table 4.
Average Marginal Effects (AME) of Perioperative and Postoperative Opioid Use, by Endometriosis Procedure Modality
| Outcome | Laparoscopic | Robotic Assisted | Unadjusted | Adjusted | ||
|---|---|---|---|---|---|---|
| n (%) | n (%) | AME (95% CI) | P-Value | AME (95% CI | P-Value | |
| Perioperative Opioid Use | 22,041 (83.7) | 1,517 (86.5) | −2.8 (−4.5, −1.1) | <.01 | −2.1 (−3.9, 0.4)a | .02 |
| Postoperative Opioid Use | 11,408 (43.3) | 685 (39.1) | 4.3 (1.9, 6.6) | <.01 | 3.2 (0.7, 5.7)b | .01 |
Adjusted for age, rurality, substance use at baseline, smoking at baseline, pain disorders at baseline, concomitant procedures, preoperative opioid use.
Adjusted for age, rurality, substance use at baseline, smoking at baseline, pain disorders at baseline, concomitant procedures, preoperative opioid use, perioperative opioid use.
In the postoperative period, 39.1% of individuals who underwent a robotic-assisted procedure and 43.3% of those who had a laparoscopic procedure filled an opioid prescription. The probability of postoperative opioid prescription fills was significantly higher among those who had a laparoscopic procedure, even after adjusting for individual characteristics, concomitant surgical procedures, and preoperative and perioperative opioid use, compared to those who had a robotic-assisted procedure (adjusted AME = 3.2; 95% CI: 0.7, 5.7; P ≤ . 01) (Table 4).
Opioid Use among Individuals without Baseline Pain Disorders Not Commonly Associated with Endometriosis
Among 11,575 individuals without pain disorders unrelated to endometriosis, 723 (6.3%) underwent robotic-assisted surgery. Initially, robotic-assisted surgery was linked to increased perioperative opioid use, while laparoscopic surgery was associated with greater postoperative opioid use. However, after adjusting for confounders, these differences were no longer statistically significant (Table 5).
Table 5.
Average Marginal Effects (AME) of Perioperative and Postoperative Opioid Use among Individuals That Did Not Have Pain Disorders Not Commonly Related to Endometriosis, by Endometriosis Procedure Modality
| Outcome | Laparoscopic | Robotic Assisted | Unadjusted | Adjusted | ||
|---|---|---|---|---|---|---|
| n (%) | n (%) | AME (95% CI) | P-Value | AME (95% CI) | P-Value | |
| Perioperative Opioid Use | 8,904 (82.1) | 614 (84.9) | −2.9 (−5.6, −0.02) | .04 | −2.1 (−5.0, 0.7)a | .14 |
| Postoperative Opioid Use | 3,786 (34.9) | 223 (30.8) | 4.0 (0.6, 7.5) | .02 | 3.0 (−0.7, 6.6)b | .11 |
Adjusted for age, rurality, substance use at baseline, smoking at baseline, endometriosis-related pain disorders at baseline, concomitant procedures, baseline opioid use.
Adjusted for age, rurality, substance use at baseline, smoking at baseline, endometriosis-related pain disorders at baseline, concomitant procedures, baseline opioid use, perioperative opioid use.
Opioid Use among Individuals Who Had Complex Procedures
Among the 1176 procedures that were considered complex, 144 (12.2%) were done with robotic assistance and 816 (88.2%) were done laparoscopically. In this population, there was no difference in odds of perioperative or postoperative opioid prescription fills between the two groups (Table 6).
Table 6.
Average Marginal Effects (AME) of Perioperative and Postoperative Opioid Use among Individuals Who Had Complex Endometriosis, by Endometriosis Procedure Modality
| Outcome | Laparoscopic; n (%) | Robotic Assisted; n (%) | Unadjusted | Adjusted | ||
|---|---|---|---|---|---|---|
| AME (95% CI) | P-Value | AME (95% CI) | P-Value | |||
| Perioperative Opioid Use | 878 (85.1) | 123 (85.4) | −0.3 (−6.5, 5.8) | .91 | 0.3 (−5.9, 6.4)a | .93 |
| Postoperative Opioid Use | 450 (43.6) | 57 (39.6) | 4.0 (−4.5, 12.6 | .36 | 0.3 (−5.6, 12.5)b | .45 |
Adjusted for age, rurality, substance use at baseline, smoking at baseline, pain disorders at baseline, concomitant procedures, baseline opioid use.
Adjusted for age, rurality, substance use at baseline, smoking at baseline, pain disorders at baseline, concomitant procedures, baseline opioid use, perioperative opioid use.
Opioid Use Among Individuals, by Baseline Pain Disorders Diagnosis
In a stratified analysis of 19,273 patients (68.62%) with baseline pain disorders, robotic-assisted surgery was associated with significantly higher perioperative opioid use (adjusted AME = –2.5; 95% CI: –4.5, –0.6; P < .01), while laparoscopic surgery was linked to significantly higher postoperative opioid use (adjusted AME = 3.2; 95% CI: 0.2, 6.2; P = .04) (Table 7). In contrast, no significant differences in perioperative or postoperative opioid use were observed between surgical modalities among patients without baseline pain disorders.
Table 7.
Average Marginal Effects (AME) of Perioperative and Postoperative Opioid Use among Individuals with and without Baseline Pain Disorders, by Endometriosis Procedure Modality
| Baseline Concurrent Pain Disorder Status; Outcome | Laparoscopic | Robotic Assisted | Unadjusted | Adjusted | ||
|---|---|---|---|---|---|---|
| n (%) | n (%) | AME (95% CI) | P-Value | AME (95% CI) | P-Value | |
| Did not have Concurrent Pain Disorders | ||||||
| Perioperative Opioid Use | 6,804 (81.8) | 411 (83.5) | −1.8 (−5.2, 1.6) | .30 | −1.1 (−4.6, 2.5)a | .56 |
| Postoperative Opioid Use | 2,812 (33.8) | 148 (30.1) | 3.7 (−0.5, 7.9) | .08 | 2.8 (−1.5, 7.2)b | .21 |
| Had Concurrent Pain Disorders | ||||||
| Perioperative Opioid Use | 15,237 (84.6) | 1,106 (87.6) | −3.0 (−4.9, −1.1) | <.01 | −2.5 (−4.5, −0.6)a | .01 |
| Postoperative Opioid Use | 8,596 (47.7) | 537 (42.6) | 5.2 (2.4, 8.0) | <.01 | 3.2 (0.2, 6.2)b | .04 |
Adjusted for age, rurality, substance use at baseline, smoking at baseline, concomitant procedures, baseline opioid use.
Adjusted for age, rurality, substance use at baseline, smoking at baseline, concomitant procedures, baseline opioid use, perioperative opioid use.
Opioid Use Among Individuals, by Baseline Pain Disorders Diagnosis in Individuals Who Underwent Endometriosis Surgery without a Hysterectomy or Oophorectomy/Salpingectomy
In a stratified analysis of 18,570 patients who underwent endometriosis excision without a concomitant hysterectomy or oophorectomy/salpingectomy, 66.9% had a baseline pain disorder. Among patients without baseline pain disorders, there was no significant difference in perioperative opioid use between robotic-assisted and laparoscopic surgery. However, robotic-assisted surgery was associated with significantly lower postoperative opioid use (adjusted AME = 6.1; 95% CI: 0.3, 11.8; P = .04). Among patients with baseline pain disorders, robotic-assisted surgery was linked to higher perioperative opioid use (adjusted AME = –4.9; 95% CI: –7.6, –2.3; P < .01) but lower postoperative opioid use compared to laparoscopy, a difference that remained significant after adjusting for confounders (adjusted AME = 5.5; 95% CI: 1.1, 9.9; P < .01) (Table 8).
Table 8.
Average Marginal Effects (AME) of Perioperative and Postoperative Opioid Use among Individuals by Baseline Pain Disorders Diagnosis in Individuals Who Did Not Have a Concomitant Hysterectomy, Oophorectomy or Salpingectomy, by Endometriosis Procedure Modality
| Baseline Concurrent Pain Disorder Status; Outcome | Laparoscopic | Robotic Assisted | Unadjusted | Adjusted | ||
|---|---|---|---|---|---|---|
| n (%) | n (%) | AME (95% CI) | P-Value | AME (95% CI) | P-Value | |
| Did not have Concurrent Pain Disorders | ||||||
| Perioperative Opioid Use | 4,778 (81.0) | 206 (80.2) | 0.8 (−4.1, 5.8) | .74 | 0.7 (−4.2, 5.7)a | .78 |
| Postoperative Opioid Use | 2,086 (35.4) | 76 (29.6) | 5.8 (.01, 11.5) | .047 | 6.1 (0.3, 11.8)b | .04 |
| Had Concurrent Pain Disorders | ||||||
| Perioperative Opioid Use | 15,237 (84.6) | 1,106 (87.6) | −5.1 (−7.8, −2.5) | <.01 | −4.9 (−7.6, −2.3)a | <.01 |
| Postoperative Opioid Use | 8,596 (47.7) | 537 (42.6) | 5.1 (0.9, 9.3) | .02 | 5.5 (1.1, 9.9)b | .01 |
Adjusted for age, rurality, substance use at baseline, smoking at baseline, concomitant procedures, baseline opioid use.
Adjusted for age, rurality, substance use at baseline, smoking at baseline, concomitant procedures, baseline opioid use, perioperative opioid use.
DISCUSSION
In this large retrospective cohort study, we found that robotic-assisted endometriosis surgery was associated with a lower likelihood of opioid prescription fills in the two years postsurgery compared to laparoscopic surgery. This association persisted among patients with baseline pain disorders, although laparoscopic cases showed lower perioperative opioid prescription fills compared with robotic cases, this difference was no longer significant after adjustment for important confounders. These findings suggest that surgical modality may influence postoperative opioid use patterns over time, particularly in populations at higher risk of opioid use.
To our knowledge, no prior studies have directly compared laparoscopic and robotic-assisted endometriosis surgery in terms of pain outcomes and opioid use. However, research on opioid use across different surgical specialties has yielded conflicting results. Bastawrous et al reported lower opioid requirements in patients undergoing robotic-assisted sigmoidectomy compared to laparoscopy, likely due to reduced tissue trauma.6 Conversely, studies on robotic-assisted sleeve gastrectomy found no significant difference in opioid use.7
In the context of gynecology, a study found that hysterectomies via minimally invasive approaches (laparoscopy and robotics) were associated with lower opioid use compared to open hysterectomy, though no significant differences were noted between the two minimally invasive modalities.8 Similarly, Hachem et al prospectively evaluated 91 gynecologic patients undergoing surgery for benign and malignant indications and found that robotic surgery was associated with similar pain scores and narcotic requirements compared with conventional laparoscopy. Our large, multi-institutional analysis suggests that robotic and laparoscopic surgery are equivalent in patients without pain disorders and complex procedures. But robotic assistance may be associated with lower rates of postoperative opioid prescription fills, among patients with baseline pain disorders. These differences may reflect variations in study design, subgroup analysis, and duration of follow-up.9 Delgado et al found that patients undergoing robotic-assisted endometriosis surgery used twice as many opioids postoperatively compared to those undergoing robotic surgery for other benign conditions. This suggests that opioid use may be driven more by the pain burden of endometriosis rather than the surgical modality itself. Our study builds on these findings by stratifying patients based on baseline pain conditions and adjusting for key confounders, providing a more targeted analysis of opioid use patterns in endometriosis patients specifically.10
Although previous studies have grouped laparoscopic and robotic approaches together when analyzing opioid use in gynecologic surgery,11–13 our study is the first to compare these modalities in a high-risk patient population. We found that robotic-assisted surgery was associated with reduced opioid use in patients undergoing excision alone, as well as in those undergoing concomitant hysterectomy/oophorectomy. This distinction is important, as hysterectomy and oophorectomy may influence long-term pain outcomes.
The differences between robotic and laparoscopic surgery may be attributed to technical advantages of the robotic system. Robotic-assisted surgery offers enhanced articulation, a greater range of motion, and superior tremor control, which allow for finer dissection and reduced tissue trauma.14 The increased precision may reduce postoperative pain by minimizing collateral tissue damage.15,16 Furthermore, improved visualization with magnification may facilitate a more complete resection of endometriotic lesions, potentially reducing long-term pain and opioid reliance.
Given the ongoing opioid crisis, strategies to minimize opioid exposure are increasingly relevant. Substantial evidence suggests that opioid exposure during surgery and the postoperative period can lead to prolonged use, highlighting the need for surgical techniques that minimize opioid requirements.17,18
Strengths of this study include its large sample size and the use of comprehensive statistical adjustments. Also, we only included patients who had continuous enrollment from one year before to two years after the endometriosis procedure. However, limitations must be considered.
Several limitations must be considered. We assessed opioid prescription fills rather than actual consumption, which may not fully represent opioid use. While prescription fills cannot substitute for direct measures of consumption or pain scores, they provide valuable insight into prescribing practices and potential patient need. Other studies, including that by Hachem et al, have measured actual narcotic consumption in gynecologic surgeries and found no difference in consumption rates between laparoscopic and robotic approaches.9 Future prospective studies that incorporate both prescription data and direct consumption measures are warranted to validate and extend these findings in endometriosis patients. There is a possibility of underreporting of comorbidities such as pain disorders, smoking, and substance use could have influenced prescribing patterns. The identification of endometriosis and robotic cases relied on CPT and ICD coding, which can be inconsistently reported, leading to potential misclassification and limiting our ability to fully capture surgical complexity or disease stage.
Another limitation is that important variables such as body mass index, use of nonopioid pain medications, and access to complementary treatments (e.g., acupuncture, pelvic floor physical therapy) were not available in the dataset. These factors, which may differ between urban and rural populations, could confound results. Additionally, the dataset did not include endometriosis stage, prior surgical history, operative duration. To address this limitation, we identified concomitant surgical procedures—such as gastrointestinal and urinary tract operations—performed at the time of surgery and used these as surrogates for advanced stage disease and increased surgical complexity. We also acknowledge that data on surgeon expertise or other characteristics, complication rate was not available which could limit our results since complications can increase opioid consumption. While we accounted for overlapping pain conditions as detailed in Table 1, we did not capture mental health conditions such as anxiety and depression, which may influence pain perception and opioid requirements. Incorporating these variables into endometriosis and chronic pain research is challenging, as misclassification is common and there is a risk of attributing pain to a mental health disorder rather than to endometriosis, which may lead to misinterpretation. Socioeconomic variables, which were also unavailable in this dataset, may further contribute to long-term opioid use and warrant inclusion in future studies.
Our study revealed that 6.24% of endometriosis surgeries utilized robotic assistance, below the national rate of 24.2% for various surgical specialties.19 This lower rate is a limitation of our dataset and may reflect disparities in access to robotic systems. While this mirrors real-world practice patterns, it may also introduce confounding related to institutional or regional availability of robotic platforms. Potential billing inaccuracies may also contribute to the underreporting of robotic surgeries. Additionally, there could be incomplete billing of concurrent endometriosis procedures, especially in cases where higher reimbursing codes like those for hysterectomies are prioritized, potentially leading to underreporting of surgeries that included concomitant endometriosis procedures.
Our findings highlight the need to investigate the mechanisms by which robotic surgery may reduce opioid use and whether these benefits vary by surgery type and procedural complexity. Additionally, implementing standardized opioid-sparing approaches, such as multimodal analgesia and enhanced recovery after surgery (ERAS) principles, should be emphasized regardless of surgical approach.
CONCLUSIONS
Our findings suggest that robotic-assisted endometriosis surgery may be associated with a lower likelihood of postoperative opioid prescription fills compared with laparoscopic surgery. This association was most apparent among patients with baseline pain disorders, whereas no difference was observed in subgroups without baseline pain disorders or undergoing complex surgical procedures. Although laparoscopic surgery was initially associated with lower perioperative opioid prescription fills, this difference was no longer significant after adjustment for key confounders. These results should be interpreted as hypothesis-generating, given the limitations of claims-based analyses. The potential role of robotic techniques in optimizing postoperative pain management may warrant further exploration, particularly in high-risk populations with chronic pain. Future prospective studies are needed to clarify the mechanisms underlying these associations, including the potential contributions of surgical precision, nerve preservation, and enhanced visualization to postoperative pain outcomes.
Footnotes
Conflict of interests: none.
Funding sources: Intuitive Surgical sponsored access to the Merative MarketScan Research Databases.
Disclosure: E.W., Y.L., and F.Z. reported being employed by Intuitive Surgical during the conduct of the study. The remaining authors report no conflict of interest.
References:
- 1.Organization WH. Endometriosis: overview. Accessed April 28, 2024. https://www.who.int/endometriosis/overview.
- 2.Köckerling F. Robotic vs. standard laparoscopic technique - what is better? Front Surg. 2014;1:15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Merative. MarketScan Research Databases for life sciences researchers. 2023. [Google Scholar]
- 4.Baser O, Samayoa G, Yapar N, Baser E, Mete F. Use of open claims vs closed claims in health outcomes research. J Health Econ Outcomes Res. 2023;10(2):44–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Graves J, Mackelprang J, Moore M, et al. Rural- urban disparities in health care costs and health service utilization following pediatric mild traumatic brain injury. Health Serv Res. 2019;54(2):337–345. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Bastawrous AL, Shih IF, Li Y, Cleary RK. Minimally invasive sigmoidectomy for diverticular disease decreases inpatient opioid use: results of a propensity score-matched study. Am J Surg. 2020;220(2):421–427. [DOI] [PubMed] [Google Scholar]
- 7.Saleh T, Ford J, Kindel T, et al. Inpatient opioid use and pain control after robotic versus laparoscopic sleeve gastrectomy. Surgery. 2024;175(3):599–604. [DOI] [PubMed] [Google Scholar]
- 8.Garzon S, Mariani A, Weaver AL, et al. Robotic-assisted hysterectomy for benign gynecologic disease in the United States: in-hospital use of opioid and non-opioid analgesics. J Robot Surg. 2024;18(1):182. [DOI] [PubMed] [Google Scholar]
- 9.Hachem LE, Acholonu UCJ, Nezhat FR. Postoperative pain and recovery after conventional laparoscopy compared with robotically assisted laparoscopy. Obstet Gynecol. 2013;121(3):547–553. [DOI] [PubMed] [Google Scholar]
- 10.Delgado SI, Koythong T, Turrentine MA, Sangi-Haghpeykar H, Guan X. Postoperative opioid use for patients with chronic pelvic pain undergoing robotic surgery for resection of endometriosis. J Robot Surg. 2022;16(2):421–427. [DOI] [PubMed] [Google Scholar]
- 11.Frost AS, Kohn J, Wang K, Simpson K, Patzkowsky KE, Wu H. Risk factors for postoperative narcotic use in benign, minimally-invasive gynecologic surgery. JSLS. 2022;26(3):e2022.00041. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Wen X, Kogut S, Aroke H, Taylor L, Matteson KA. Chronic opioid use in women following hysterectomy: patterns and predictors. Pharmacoepidemiol Drug Saf. 2020;29(4):493–503. [DOI] [PubMed] [Google Scholar]
- 13.Ackenbom MF, Dong S, Romanova A, et al. Postoperative opioid utilization in older women undergoing pelvic organ prolapse surgery. Female Pelvic Med Reconstr Surg. 2021;27(5):304–309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Moorthy K, Munz Y, Dosis A, et al. Dexterity enhancement with robotic surgery. Surg Endosc. 2004;18(5):790–795. [DOI] [PubMed] [Google Scholar]
- 15.Prasad SM, Prasad SM, Maniar HS, Chu C, Schuessler RB, Damiano RJ., Jr. Surgical robotics: impact of motion scaling on task performance. J Am Coll Surg. 2004;199(6):863–868. [DOI] [PubMed] [Google Scholar]
- 16.Reddy K, Gharde P, Tayade H, Patil M, Reddy LS, Surya D. Advancements in robotic surgery: a comprehensive overview of current utilizations and upcoming frontiers. Cureus. 2023;15(12):e50415. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Chan WV, Le B, Lam M, et al. Opioid prescribing practices for women undergoing elective gynecologic surgery. J Minim Invasive Gynecol. 2021;28(7):1325–1333.e3. [DOI] [PubMed] [Google Scholar]
- 18.Bougie O, Blom J, Zhou G, Murji A, Thurston J. Use and misuse of opioid after gynecologic surgery. Best Pract Res Clin Obstet Gynaecol. 2022;85(Pt B):23–34. [DOI] [PubMed] [Google Scholar]
- 19.Ng AP, Sanaiha Y, Bakhtiyar SS, Ebrahimian S, Branche C, Benharash P. National analysis of cost disparities in robotic-assisted versus laparoscopic abdominal operations. Surgery. 2023;173(6):1340–1345. [DOI] [PubMed] [Google Scholar]