The Benign COMPARE Study: Comparing Perioperative Outcomes of Benign Minimally Invasive Laparoscopic, da Vinci Robotic, and Open Procedures: A Systematic Review and Meta-Analysis of Evidence

Minimally invasive surgery has fundamentally transformed surgical disease management. Compared with open approaches, conventional laparoscopic surgery is associated with smaller incisions, reduced perioperative morbidity, decreased postoperative pain, shorter hospital stays, faster recovery, and improved cosmetic outcomes. 1 Despite these well-established benefits, traditional laparoscopy remains constrained by several technical limitations, including 2D visualization with reduced depth perception, camera instability due to reliance on handheld systems, limited instrument articulation with rigid tools, and ergonomic challenges that may contribute to surgeon fatigue and musculoskeletal strain. Collectively, these factors impose a steep learning curve, particularly for complex procedures performed in anatomically confined spaces. 2 , 3 The introduction of the da Vinci surgical system (Intuitive Surgical, Sunnyvale, CA) represented a major technological advancement in minimally invasive surgery by addressing many of these limitations. Robotic-assisted surgery using the da Vinci robotic–assisted surgery (dV-RAS) offers enhanced 3D visualization, improved instrument articulation with greater degrees of freedom, motion scaling, tremor filtration, and stable camera control, enabling greater precision and control during dissection and reconstruction. These advantages have been particularly impactful in deep, narrow, or anatomically complex operative fields such as the chest, abdomen, and pelvis. 4 – 8 Benign surgical populations account for a substantial proportion of global surgical volume across gynecologic, gastrointestinal, bariatric, urologic, and hernia-related procedures. Many remain technically challenging due to patient anatomy and disease complexity. Consequently, dV-RAS has been rapidly adopted across a wide range of benign surgical indications, and numerous studies have evaluated perioperative outcomes for individual procedures. 4 , 5 However, whether these technical advantages translate into consistent, clinically meaningful improvements across diverse benign procedures compared with established laparoscopic and open approaches remains uncertain. This study builds upon the previously published Comparing Perioperative Outcomes of Oncologic Minimally Invasive Laparoscopic, da Vinci Robotic, and Open Procedures (COMPARE) study on oncologic procedures, 9 extending the evaluation of dV-RAS to 13 high-volume benign procedures, and analyzes perioperative clinical and patient-centered outcomes.

This systematic review and meta-analysis was conducted and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Supplemental Tables 1 and 2, https://links.lww.com/AOSO/A655 and https://links.lww.com/AOSO/A656 ) and was prospectively registered in the PROSPERO international database of systematic reviews (CRD420251001402) on March 7, 2025. The study protocol is available upon request. PubMed, Embase, and Scopus databases were searched to identify studies published between January 1, 2010, and December 31, 2024, with the final search conducted on October 22, 2025. The search strategy incorporated combinations of robotic-related terms (“da Vinci,” “robot*,” and “minimally invasive”) and procedure-specific keywords, including “hysterectomy,” “myomectomy,” “endometriosis resection,” “inguinal hernia repair,” “ventral hernia repair,” “hiatal hernia repair,” “cholecystectomy,” “appendectomy,” “colorectal,” “adrenalectomy,” “pyeloplasty,” “Roux-en-Y gastric bypass,” and “sleeve gastrectomy,” along with benign indication terms, such as “benign,” “leiomyoma,” “colitis,” and “Cushing’s.” The complete search strategies for each database are detailed in Supplemental Table 3, https://links.lww.com/AOSO/A657 . Study screening was conducted using a dual-review process supported by EndNote 10 , 11 and the Robot Screener tool in AutoLit (Nested Knowledge), 12 with final inclusion decisions made by human reviewers. Full-text articles of potentially eligible studies were subsequently assessed based on predefined inclusion and exclusion criteria. Studies were eligible for inclusion if they met the following criteria: (1) reported outcomes of interest for patients undergoing surgery involving the pelvis (hysterectomy, myomectomy, and endometriosis resection), groin (inguinal hernia repair), anterior abdominal wall (ventral hernia repair), diaphragm/hiatus (hiatal hernia repair), abdominal cavity (cholecystectomy, appendectomy, and colorectal surgery), retroperitoneum (adrenalectomy and pyeloplasty), or upper abdomen (Roux-en-Y gastric bypass and sleeve gastrectomy) performed using the da Vinci robotic–assisted surgical system, and at least 80% of cases were performed for benign indications; (2) were peer-reviewed publications between January 1, 2010, and December 31, 2024; and (3) were a randomized controlled trial, database study, or prospective or retrospective cohort study comparing dV-RAS with laparoscopic or open surgery, with at least 20 patients in each cohort. An exception was made for benign hysterectomy, the most widely adopted benign robotic procedure, for which retrospective cohort studies were excluded. This modification was justified by the abundant availability of high-level evidence on this procedure. Studies were excluded for the following reasons: (1) non-English-language publication; (2) pediatric study population; (3) non-peer-reviewed health technology assessment; (4) evaluation of alternative surgical techniques (eg, single-port, single-site, or hand-assisted surgery); (5) absence of stratified outcome reporting by treatment arm; (6) reporting of aggregated data across multiple procedures or indications outside the scope of this review; (7) lack of reported outcomes of interest; and (8) inclusion of overlapping patient populations with redundant data and similar conclusions. Primary outcomes were 30-day perioperative clinical endpoints, including conversions to open surgery, operative time (OT), blood transfusions, length of hospital stay, intraoperative and postoperative complications, surgical site infections (SSIs), readmissions, reoperations, and mortality. Secondary outcomes included pain and recovery-related measures, specifically Visual Analog Scale (VAS) pain scores and pain medication usage (including analgesics and opioids) within 30 days of surgery, and the patient-reported quality-of-life outcomes: activities of daily living and return to work. To enable consistent cross-procedure comparisons, the analysis was restricted to commonly shared perioperative outcomes; procedure-specific outcomes were not evaluated. Clinical outcomes were extracted by 2 researchers using a standardized Excel form and verified for accuracy. Disagreements were resolved by consensus or a third reviewer. Extracted data included the first author, publication year, study design, sample size, country, data source, and outcomes of interest. Continuous outcomes were standardized to mean and standard deviation and unit of measure, and binary outcomes to event counts and sample size. Studies reporting outcomes that could not be standardized for quantitative pooling were included in the systematic review but excluded from the meta-analysis, with reasons documented in the PRISMA flow diagram. Meta-analyses were performed using R software. 13 , 14 Forest plots were generated for each outcome and comparison (dV-RAS vs laparoscopic and open) and summarized in a results table. Analyses were conducted separately for each comparison. Effect measures for perioperative outcomes pooled across the 13 benign procedures were expressed as both an odds ratio (OR) and a risk difference with a 95% CI for binary outcomes or as a weighted mean difference (MD) with a 95% CI for continuous outcomes. When the direction and significance of binary outcome estimates were consistent, results were presented as ORs. In cases of discrepancy between OR and risk difference, both measures were reported. Heterogeneity was assessed using the Cochran Q (χ 2 ) test and the I 2 statistic, with Q evaluating the presence of heterogeneity and I 2 quantifying magnitude. Given the limited power of Q with few studies, emphasis was placed on I 2 as a robust, scale-independent measure. A fixed effect model was used when P ≥ 0.05 and I 2 ≤ 50%, and a random-effects model was applied otherwise. Study weights were derived by R software 13 , 14 using a combination of sample size and outcome variance and were used to generate pooled estimates, standard deviations, proportions, and 95% CIs. Statistical significance was defined as a 2-sided P < 0.05. Multiple sensitivity analyses were performed, including temporal comparisons between the early (2010–2017) and later (2018–2024) dV-RAS eras following the da Vinci Xi introduction in 2014 and VAS outcomes stratified by assessment timing. Additional analyses by study design, procedure type, and risk of bias explored heterogeneity, with subgroup differences assessed using χ 2 tests. Risk of bias was independently assessed by 2 reviewers (A.Y., N.M.P.) using the Cochrane Risk of Bias in Non-Randomized Studies of Interventions 15 and Cochrane risk-of-bias tool for randomized trials 16 tools for nonrandomized and randomized studies, respectively. Publication bias was evaluated through visual inspection of funnel plots and supplemented with the Egger test when >10 studies reported nonnull values for an outcome. Data collection forms, extracted datasets, individual study forest plots, and the R code are available upon request. Several authors are employees of Intuitive Surgical, the manufacturer of the da Vinci surgical system. To safeguard objectivity and reproducibility, the protocol, eligibility criteria, and outcomes were prospectively registered in PROSPERO before study initiation; predefined eligibility criteria were applied irrespective of the direction of study findings; and study screening, data extraction, and risk-of-bias assessment were each performed in duplicate, with disagreements resolved by consensus. A complete log of excluded studies with reasons for exclusion is provided. The academic authors without industry employment independently accessed and verified the data, reviewed study selection, analyses, and interpretation, and retained final authority over the manuscript content and the decision to submit.

A total of 23,497 unique references were screened, resulting in 385 unique publications included in the review. Of these, 366 unique publications were included in the meta-analysis. This included 13 randomized controlled trials, 21 prospective cohorts, 101 database studies, and 231 retrospective cohort studies, published over 14 years from 32 countries, representing 1,469,565 dV-RAS, 10,699,092 laparoscopy, and 1,730,656 open patients. There were 254 publications that compared dV-RAS to laparoscopy, 37 publications that compared dV-RAS to open surgery, and 75 publications that compared all 3 surgical approaches. A detailed PRISMA flow diagram is provided in Figure 1 . The 366 publications included in the meta-analysis reported on 13 different benign procedures, with some studies reporting on >1 procedure (Table 1 ). Comparative evidence between dV-RAS and open surgery was available for 9 of the 13 benign procedures, while all procedures were represented for the dV-RAS and laparoscopy comparison. A bibliography by procedure and by whether the study was included in the meta-analysis is provided in Supplemental Appendix A https://links.lww.com/AOSO/A678 . Detailed characteristics of the studies by procedure are provided in Supplemental Table 4, https://links.lww.com/AOSO/A658 , while Supplemental Table 5, https://links.lww.com/AOSO/A659 , summarizes publications excluded from the meta-analysis. Number of Included Studies per Procedure (Meta-Analysis) The study counts reported by procedure are not mutually exclusive. DB indicates retrospective database; Lap, laparoscopic; MA, meta-analysis; PRO, prospective cohort; RETRO, retrospective cohort; SLR, systematic literature review. Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram of the systematic literature review (SLR) search. Flowchart showing the identification and screening process. 1 Approximately 500 records were excluded using the Robot Screener tool in AutoLit (Nested Knowledge) and its determination of low probability of inclusion based on the human-based screening decisions up to that point. All other records were screened by hand. 2 Alternative techniques included single-port or transanal. 3 Insufficient sample size was defined as N<20 per cohort. 4 Data for surgical approach, procedures, or indications were not reported separately. 5 The study reported on redundant patient populations, outcomes, and conclusions. 6 There were 23 unique reports (4 multispecialty, for a total of 27 nonunique studies when counted by procedure) that met Population, Intervention, Comparator/Control, Timepoints, Setting criteria, but that did not report data with sufficient details to be able to be pooled (eg, no variance or only odds ratio or mean difference or no sample size reported). MA indicates meta-analysis. The pooled meta-analysis results for dV-RAS versus laparoscopy and dV-RAS versus open surgery are presented in Tables 2 and 3 , respectively. Summary forest plots highlighting the most clinically relevant outcomes for each comparison are presented in Figures 2 and 3 . Meta-Analysis of Outcomes Pooled Across 13 Benign Procedures: dV-RAS Versus Lap Bold indicates statistical significance P 1 procedure. Fixed effect model (all others are random-effects models). Pain medications include opioids. RD indicates risk difference; SSI, surgical site infection. Meta-Analysis of Outcomes Pooled Across 13 Benign Procedures: dV-RAS Versus Open Bold indicates statistical significance P 1 procedure. Fixed effect model (all others are random-effects models). Pain medications include opioids. RD indicates risk difference. Summary forest plots for length of stay and blood transfusions. Forest plots illustrating length of stay for (A) dV-RAS versus laparoscopy and (B) dV-RAS versus open surgery and blood transfusions for (C) dV-RAS versus laparoscopy and (D) dV-RAS versus open surgery. Blue squares denote individual study effect estimates, while horizontal black lines indicate the corresponding 95% CIs. The black diamond represents the pooled effect estimate, with its width reflecting the 95% CI. COMPARE refers to a systematic review and meta-analysis evaluating perioperative outcomes of benign procedures across minimally invasive laparoscopic, da Vinci robotic–assisted, and open surgical approaches. df indicates degrees of freedom; IV, inverse variance; RYGB, Roux-en-Y gastric bypass. Summary forest plots for conversions, 30-day readmissions, and postoperative complications. Forest plots illustrating (A) conversion to open surgery dV-RAS versus laparoscopy, (B) 30-day readmissions dV-RAS versus open surgery, and 30-day postoperative complications for (C) dV-RAS versus laparoscopy and (D) dV-RAS versus open surgery. Blue squares denote individual study effect estimates, while horizontal black lines indicate the corresponding 95% CIs. The black diamond represents the pooled effect estimate, with its width reflecting the 95% CI. COMPARE refers to a systematic review and meta-analysis evaluating perioperative outcomes of benign procedures across minimally invasive laparoscopic, da Vinci robotic–assisted, and open surgical approaches. df indicates degrees of freedom; IV, inverse variance; RYGB, Roux-en-Y gastric bypass. Across 13 benign procedures, dV-RAS was associated with a 54% lower odds of conversion to open surgery [OR, 0.46 (0.37–0.57); P < 0.01; I 2 = 88%], had a lower likelihood of patients requiring blood transfusion [OR, 0.87 (0.77–0.98); P = 0.02; I 2 = 67%], and demonstrated shorter length of hospital stay [MD, −0.17 (−0.27 to −0.08); P < 0.01; I 2 = 97%]. For the secondary outcomes, dV-RAS was associated with lower postoperative pain scores [MD, −0.69 (−1.2 to −0.18); P < 0.01; I 2 = 95%]. The sensitivity analysis by timing of assessment showed significance at the 1-day timepoint ( P = 0.0317) but not at the 30-day timepoint ( P = 0.0811; Supplemental Fig. 28A, https://links.lww.com/AOSO/A665 ). In addition, fewer dV-RAS patients required postoperative analgesics within 30 days [OR, 0.88 (0.83–0.92); P < 0.01; I 2 = 48%] and showed a faster return to work by an average of 2 days [MD, −2.04 (−3.43 to −0.64); P < 0.01; I 2 = 81%]. dV-RAS and laparoscopic surgery patients did not differ in their ability to resume activities of daily living [MD, −0.58 (−1.65 to 0.49); P = 0.29; I 2 = 57%]. No statistically significant differences between dV-RAS and laparoscopic surgery were observed for intraoperative complications [OR, 0.83 (0.65–1.07); P = 0.16; I 2 = 67%], SSI [OR, 1.14 (0.88–1.46); P = 0.33; I 2 = 68%], reoperation [OR, 1.06 (0.91–1.23); P = 0.48; I 2 = 65%], readmission [OR, 1.06 (0.95–1.19); P = 0.27; I 2 = 92%], and mortality [OR, 1.06 (0.96–1.16); P = 0.26; I 2 = 0%] within 30 days. OT was significantly longer for dV-RAS by an average of 23 minutes [MD, 23.58 (18.92–28.24); P < 0.01; I 2 = 99%]. Compared with open surgery across the 13 procedures, dV-RAS was associated with significantly improved perioperative outcomes, including a lower likelihood of blood transfusion [OR, 0.31 (0.24–0.42); P < 0.01; I 2 = 82%], fewer intraoperative complications [OR, 0.64 (0.55–0.74); P < 0.01; I 2 = 36%], shorter length of hospital stay [MD, −1.97 (−2.36 to −1.58); P < 0.01; I 2 = 100%], fewer postoperative complications [OR, 0.54 (0.43–0.68); P < 0.01; I 2 = 98%], fewer SSIs [OR, 0.40 (0.28–0.56); P < 0.01; I 2 = 83%], fewer readmissions [OR, 0.66 (0.51–0.85); P < 0.01; I 2 = 94%], fewer reoperations [OR, 0.70 (0.51, 0.96); P = 0.03; I 2 = 85%], and decreased mortality [OR, 0.37 (0.29–0.46); P < 0.01; I 2 = 42%) within 30 days. Secondary outcomes reported in comparison to open surgery demonstrated a faster return to work by an average of 5 days for dV-RAS [MD, −5.19 (−9.27 to −1.12); P = 0.01; I 2 = 90%]. Patients undergoing dV-RAS had 37% lower odds of receiving postoperative pain medication within 30 days [OR, 0.63 (0.43–0.91); P = 0.01; I 2 = 74%]. There was no difference in pain scores or the resumption of activities of daily living. OT was, on average, 46 minutes longer for dV-RAS [MD, 46.34 (33.27–59.4); P < 0.01; I 2 = 100%]. For the temporal sensitivity analysis, a subset analysis was performed to assess the robustness of the primary outcomes comparing dV-RAS to laparoscopic and open surgical approaches across 2 temporal periods: an early era (2010–2017) and a late era (2018–2024). Comparative analysis of secondary outcomes was limited, as many studies reporting quality-of-life measures were concentrated in only the late era. All evaluated outcomes for the dV-RAS versus laparoscopic comparison demonstrated consistency across the 2 timeframes, except for intraoperative complications ( P = 0.01; Supplemental Table 6, https://links.lww.com/AOSO/A660 ). During the earlier era (2010–2017), there was no statistical difference between the 2 approaches in intraoperative complications [OR, 1.29 (0.83, 2.01); P = 0.25]. In contrast, in the later era (2018–2024), dV-RAS was associated with a significantly lower likelihood of intraoperative complications relative to laparoscopy [OR, 0.67 (0.51–0.88); P = 0.005]. All other primary outcomes were comparable across the 2 time periods. For the dV-RAS versus open surgery comparison, pooled results are shown in Supplemental Table 7, https://links.lww.com/AOSO/A661 . Despite the temporal differences, all outcomes favored dV-RAS over open surgery consistently within the early and late time periods. Effect sizes differed for blood transfusions [early: OR, 0.24 (0.22–0.26); P < 0.0001] versus [late: OR, 0.39 (0.30–0.51); P < 0.0001; P < 0.01], SSIs [early: OR, 0.23 (0.18–0.29); P < 0.0001] versus [late: OR, 0.41 (0.28–0.60); P < 0.0001; P = 0.01], and mortality [early: OR, 0.16 (0.08–0.33); P < 0.0001] versus [late: OR, 0.44 (0.35–0.56); P < 0.0001; P < 0.01] within 30 days. The remaining outcomes were comparable across periods, including OT, hospital stay, intraoperative and 30-day postoperative complications, readmissions, and reoperations. To assess other possible sources of heterogeneity (I 2 > 50%), sensitivity meta-analyses by procedure, study type, and risk of bias were performed for each outcome and presented in Table 4 . Procedure type emerged as the primary driver of heterogeneity, influencing multiple outcomes across both comparisons. The forest plots in Supplemental Figs. 1–27, https://links.lww.com/AOSO/A664 , show the analysis broken down by procedure type. Study type explained heterogeneity for select outcomes, including blood transfusion in the laparoscopy comparison ( P = 0.0119) and length of stay ( P = 0.0234), 30-day postoperative complications ( P = 0.0005), and SSI ( P = 0.0133) in the open comparison (Supplemental Tables 8 and 9, https://links.lww.com/AOSO/A662 and https://links.lww.com/AOSO/A663 ). In contrast, risk of bias (high/serious vs low/moderate) had minimal impact, with significant differences limited to length of stay ( P = 0.04) in the laparoscopic comparison and 30-day readmission ( P = 0.0015) in the open comparison. Potential Sources of Heterogeneity Across Perioperative Outcomes Bold indicates statistical significance P < 0.05. Unique studies. Fixed effect model (all others are random effects). Risk of bias assessment results can be found in Supplemental Table 4, https://links.lww.com/AOSO/A658 . Bias due to confounding (domain 1) was high or critical in 16 studies and serious in 138 studies. In addition, 15 studies had critical bias in participant selection (domain 2), and 114 studies had serious bias in the selection of participants. Bias in classification of intervention (domain 3) was serious in 1 study. All other domains scored low or moderate risk of bias. For bias assessment by study type, 3 of the 13 randomized controlled trials and 11 of the 21 prospective studies were rated as having critical or serious risk of bias. Publication bias for dV-RAS and laparoscopic comparisons suggested generally symmetric distributions for most outcomes (Egger test ≥ 0.05; Supplemental Figs. 29A–N, https://links.lww.com/AOSO/A666 ). Only 2 outcomes demonstrated significant asymmetry: length of hospital stay ( P = 0.0069) and intraoperative complications ( P = 0.0087). The Egger test was not performed due to an insufficient number of studies for activities of daily living and return to work (<10). Publication bias for dV-RAS versus open comparisons was minimal (Supplemental Figs. 30A–M, https://links.lww.com/AOSO/A667 ), except for 30-day reoperations ( P = 0.0010) and 30-day readmissions ( P = 0.0030).

This large-scale systematic review and meta-analysis across 13 benign procedures suggests that the perioperative benefits associated with dV-RAS vary by outcome. The most consistent findings were lower conversion rates and shorter hospital stays relative to laparoscopic and open approaches, consistent with prior multispecialty analyses and observational studies. 9 , 17 – 19 Differences in other outcomes were modest and procedure-dependent, and should be interpreted in the context of substantial clinical and methodological heterogeneity. Collectively, these findings provide a macroscopic, evidence-based perspective on the evolving role of robotic surgery across commonly performed benign procedures, integrating evidence from diverse specialties and stages of adoption while recognizing important procedure-specific variability. The reduction in conversion to open surgery represents the most clinically important and consistent finding of this analysis. Overall, dV-RAS was associated with approximately 54% lower odds of conversion compared with laparoscopy, with significant reductions observed across several procedures, including benign hysterectomy, ventral and inguinal hernia repair, myomectomy, cholecystectomy, and benign colon surgery. These findings align with prior analyses, with Shah et al 18 reporting a 58.5% reduction in conversion and Gangemi et al 19 showing lower conversion odds across abdominal surgeries, including mixed benign and malignant procedures. Conversion to open surgery is an important intraoperative event associated with increased morbidity and prolonged hospitalization. 17 , 18 Patel et al 17 reported that conversion during cholecystectomy was associated with higher rates of postoperative complications, including increased SSIs and longer hospital stays. Similarly, Shah et al 18 reported a 1.77-day increase in length of stay among converted cases. Although the present analysis was not designed to directly evaluate the downstream consequences of conversion, the observed reduction in conversion risk with dV-RAS may represent a clinically meaningful advantage, particularly in technically demanding procedures where enhanced visualization, instrument articulation, and surgeon ergonomics have been proposed, though not established, as factors facilitating completion of minimally invasive surgery. 20 However, the magnitude of benefit varied across procedures, suggesting that patient selection, procedural complexity, and surgeon experience influence the extent to which robotic technology may reduce conversion risk. In contrast, many differences between dV-RAS and laparoscopy were statistically significant yet modest in magnitude. Beyond conversion, the most consistent benefits were shorter hospital stays and lower transfusion rates. Overall, dV-RAS was associated with a 0.17-day (4.1-hour) shorter hospital stay although the benefit varied by procedure. The largest reductions were observed in benign colon surgery (0.47 days), adrenalectomy (0.42 days), and benign hysterectomy (0.18 days), suggesting that improvements in postoperative recovery may be greatest in procedures with higher operative complexity or recovery burden. Although the overall reduction in length of stay was smaller than that reported in the oncologic COMPARE study (≈0.5 days), 9 it may still have clinical relevance in selected procedures by facilitating earlier recovery and reducing postoperative burden. Many benign procedures are increasingly performed in ambulatory or same-day discharge settings, particularly following robotic surgery, 21 , 22 limiting the ability of inpatient length-of-stay alone to capture meaningful differences in postoperative recovery and potentially underestimating overall patient benefit. Importantly, the shorter length of stay observed with dV-RAS may also be partially mediated by the substantial reduction in conversion rates, given the well-established association between conversion to open surgery and prolonged hospitalization. While the benefit at the individual patient level may be limited for procedures increasingly performed in ambulatory or same-day discharge settings, modest reductions in hospital stays across high-volume benign procedures could contribute to improved recovery trajectories and more efficient utilization of inpatient resources. Nevertheless, the clinical significance of the observed 0.17-day reduction should be interpreted within the context of procedure-specific differences and substantial heterogeneity. Similarly, dV-RAS was associated with 13% lower odds of blood transfusion than laparoscopy although this effect was not consistent across procedures. Procedure-specific analyses suggested that the overall finding was driven primarily by benign hysterectomy and ventral hernia repair, where blood loss may be more influenced by operative complexity and technical precision. Given the established association between perioperative transfusion, postoperative infection, and prolonged recovery, 23 these findings may be clinically relevant in procedures with higher baseline transfusion risk. By contrast, no significant differences were observed in intraoperative complications, 30-day postoperative complications, SSIs, readmissions, reoperations, or mortality compared with laparoscopy. Collectively, these findings suggest that the advantages of dV-RAS over laparoscopy are procedure-dependent and concentrated in selected perioperative outcomes. While individual benefits may be modest, their cumulative impact across high-volume benign procedures could meaningfully reduce healthcare resource utilization. However, substantial heterogeneity suggests that these pooled estimates reflect average effects rather than uniform benefits across all surgical settings. Compared with open surgery, dV-RAS was associated with consistently favorable perioperative outcomes, including lower odds of intraoperative complications (36%), blood transfusions (69%), 30-day postoperative complications (46%), SSIs (60%), readmissions (34%), reoperations (30%), and mortality (63%), along with a nearly 2-day shorter hospital stay. These findings are directionally consistent with the oncologic COMPARE study and broadly reflect the established benefits of minimally invasive surgery relative to open approaches. 9 Notably, a shorter length of stay was observed across all procedures evaluated, suggesting a consistent association between dV-RAS and faster postoperative recovery. At a population level, modest reductions in hospital stays across high-volume procedures may translate into meaningful efficiencies in aggregate resource utilization. Similarly, lower transfusion rates were observed across multiple procedures, particularly benign colon surgery, benign hysterectomy, cholecystectomy, myomectomy, and pyeloplasty, where improved visualization and instrument articulation have been proposed as possible, though unproven, contributing mechanisms. 4 However, these comparisons remain vulnerable to residual confounding. Patients undergoing open surgery frequently have greater disease severity, higher procedural complexity, or clinical characteristics that preclude a minimally invasive approach, factors that are incompletely captured in retrospective datasets. Consequently, although the direction of effect consistently favored dV-RAS, the magnitude of observed differences should not be attributed solely to robotic technology. Rather, these findings likely reflect both the benefits of minimally invasive surgery and differences in patient selection, which remain difficult to fully adjust for in observational studies. OT represented the principal outcome for which dV-RAS was associated with less favorable results. Consistent with prior meta-analyses and the oncologic COMPARE study, 9 dV-RAS was associated with a longer operative duration than both laparoscopic and open surgery. In this analysis, OT was 23.6 minutes longer than laparoscopy and 46.3 minutes longer than open surgery, consistent with previously reported differences of 17.7 and 40.9 minutes, respectively. 9 The magnitude varied substantially by procedure, ranging from a relatively small increase in cholecystectomy versus laparoscopy (11 minutes) to a substantially larger increase in benign colon surgery versus open surgery (83 minutes). In contrast, adrenalectomy and pyeloplasty demonstrated comparable OTs across surgical approaches, potentially reflecting similar patient positioning requirements and the offsetting effect of robotic docking relative to flank-incision exposure. These findings likely reflect a combination of robotic docking and setup requirements, operating-room workflow considerations, learning-curve effects, and the preferential use of robotic platforms in more technically complex cases. Importantly, despite longer operative durations, dV-RAS was generally associated with comparable or improved perioperative outcomes, including lower complications, transfusions, and shorter hospital stays versus open surgery. These findings suggest that modest increases in OT may not necessarily translate into worse postoperative outcomes and may, in selected procedures, be offset by downstream clinical benefits. As experience and technology advance, operative efficiency is expected to improve, with contemporary data suggesting narrowing time differences as teams progress along the learning curve. 24 Patient-centered outcomes remain an important evidence gap. Perioperative patient-reported pain, assessed using the VAS, was significantly lower with dV-RAS than laparoscopy but did not differ significantly from open surgery. Interpretation of these findings is limited by the small number of available studies, substantial heterogeneity, and inconclusive sensitivity analyses by the timing of assessment due to outliers and insufficient statistical power (Supplemental Figure 28, https://links.lww.com/AOSO/A665 ). Similarly, dV-RAS was associated with lower postoperative analgesic use and earlier return to work; however, these outcomes were infrequently and inconsistently reported. Reduced analgesic and opioid utilization aligns with prior evidence of improved pain control with minimally invasive approaches, particularly versus open surgery. 25 , 26 Lower opioid exposure may also have broader clinical implications given the risks of prolonged use and dependency. 27 In addition, increasing surgeon and team experience with robotic platforms may further optimize pain outcomes through improved operative efficiency and reduced tissue manipulation. 24 Return to activities of daily living was comparable across approaches, limiting conclusions regarding functional recovery. Although reduced analgesic use and earlier return to work suggest improved postoperative recovery, these findings should be interpreted cautiously given the limited and heterogeneous evidence base. An earlier return to work may also have broader societal implications, as reductions in recovery time can lessen productivity losses and facilitate earlier reintegration into normal activities. 28 , 29 Other recent studies have reported long-term results such as recurrence rates following hernia repair, 30 – 32 pyeloplasty, 33 , 34 and adrenalectomy. 35 , 36 In particular, population-based analyses have reported higher rates of bile duct injury with robotic compared with laparoscopic cholecystectomy 37 , 38 and higher long-term operative recurrence following robotic ventral hernia repair. 39 , 40 These procedure-specific endpoints were outside the prespecified perioperative outcome framework of this review, and the present findings should not be interpreted as addressing them; favorable or comparable perioperative outcomes do not preclude clinically important differences in procedure-specific or long-term results. Future studies should incorporate validated instruments to assess functional recovery across surgical modalities and extend evaluation to other long-term outcomes. The temporal subgroup analysis revealed that the magnitude of benefit for dV-RAS over open surgery attenuated between the earlier (2010–2017) and later (2018–2024) eras for transfusions, SSI, and mortality. Several factors may contribute to this trend, including progressive improvements in open surgical techniques and perioperative care, evolving patient selection, broader adoption of enhanced recovery protocols, and a ceiling effect as baseline complication rates decline. Conversely, the emergence of significant intraoperative complications for dV-RAS versus laparoscopy in the later era (OR, 0.67 vs 1.29) may reflect greater robotic experience and increasing use in more complex cases beyond the learning curve. These temporal trends underscore the dynamic nature of comparative evidence and the importance of evaluating robotic outcomes within contemporary surgical practice. Key strengths of this study include a comprehensive synthesis across 13 benign procedures, focusing on standardized, clinically relevant perioperative outcomes to enable cross-procedure comparisons and evaluation of total practice-level impact. These findings are strengthened by a rigorous, Cochrane-guided methodological framework, ensuring robustness, transparency, and reproducibility. Nonetheless, several limitations must be acknowledged. Most included studies were retrospective and therefore susceptible to selection bias, residual confounding, and inconsistent data reporting. These studies were included to ensure representation of procedures with limited evidence, but their inclusion may have reduced the precision and validity of pooled estimates. Interprocedure variability was incorporated into the pooled analysis, combining procedures at different stages of surgical adoption and technical complexity. This includes relatively standardized procedures (eg, cholecystectomy and inguinal hernia repair), often performed earlier in a surgeon’s robotic experience, alongside more complex operations that require a longer learning curve. Differences in procedural complexity, surgeon experience, case selection, technological improvements, and advanced instrumentation may contribute to outcome variability and influence pooled estimates. Inclusion of 13 benign procedures across multiple specialties introduced considerable heterogeneity, reflecting differences in indications, patient populations, surgeon experience, and institutional practices. This heterogeneity was anticipated given the study design, which intentionally synthesized evidence across procedures with distinct clinical complexity, baseline risk profiles, perioperative pathways, and recovery trajectories. Procedure type was identified as an important source of variability, and the inclusion of procedures spanning a broad range of clinical settings likely contributed to the observed heterogeneity. In addition, this analysis could not eliminate possible confounding bias on pooled estimates without patient-level adjustment for case complexity, disease factors, and comorbidity, especially across multiple procedures/diseases. Subgroup analyses confirmed procedure type as the dominant source of this variability; however, the consistent direction of effect across procedures supports the robustness of the findings. Importantly, all primary analyses used random-effects models to account for between-study variability and provide more conservative summary estimates in the presence of heterogeneity. This analysis focused on FDA-relevant perioperative endpoints commonly used in medical device regulatory submissions, consistent with the outcome framework of the oncologic COMPARE study 9 to enable comparisons across diverse benign procedures. As a result, procedure-specific outcomes such as recurrence, functional outcomes, disease-specific success, or long-term durability were not evaluated. Although this approach facilitates cross-procedure comparisons, excluding procedure-specific endpoints may limit clinical interpretation for individual procedures and not fully capture the effectiveness of each surgical approach. Patient-centered outcomes were inconsistently reported, with most comparative studies focusing on traditional perioperative endpoints while often excluding measures such as pain, recovery, and return to work or normal activities. Consequently, this study highlights important evidence gaps, underscoring the need for standardized patient-reported outcomes alongside traditional clinical endpoints to evaluate the impact of robotic surgery on patient recovery and quality of life. As robotic technology continues to evolve and adoption expands across specialties, platform-agnostic studies with standardized outcome reporting will be essential to inform evidence-based clinical and technology decisions. Limited availability of dV-RAS versus open comparisons reflecting the widespread shift toward minimally invasive approaches for procedures such as sleeve gastrectomy, hiatal hernia repair, appendectomy, and endometriosis resection restricts generalizability to procedures where open surgery remains a routine alternative. This analysis was limited to studies evaluating the da Vinci robotic platform; therefore, findings may not be generalizable to emerging robotic systems with differing technologies, performance characteristics, and clinical outcomes.

Our findings suggest that dV-RAS is associated with favorable perioperative outcomes versus laparoscopic and open approaches across benign procedures, particularly lower conversion rates and shorter hospitalization. The evidence base is limited by a paucity of high-quality studies, with reliance on retrospective data constraining inference to associations rather than causal effects. Residual confounding, selection bias, surgeon experience, case mix, and institutional factors may substantially influence observed differences. Clinical relevance also varies by procedure and comparator. Prospective studies with stronger control of bias, patient-reported outcomes, and cost-effectiveness data are needed to better delineate the role of dV-RAS in benign surgical care.

The authors thank Heather Atchison for her assistance with coordinating meetings and facilitating communication with key opinion leaders throughout this endeavor.