Perioperative outcomes and anesthetic challenges of robotic-assisted gynecologic surgery: a systematic review

Abstract

Background

Robotic-assisted gynecologic (RAS) surgery offers enhanced precision, ergonomics, and visualization compared with open (OP) and laparoscopic (LAP) approaches. However, it introduces distinct anesthetic challenges related to patient positioning, airway access, and postoperative recovery and analgesia. This review aims to evaluate perioperative outcomes across four domains: intraoperative anesthetic management, postoperative recovery, pain control strategies, and surgical performance.

Methods

We performed a systematic review of the current literature on adult women undergoing robotic-assisted gynecologic surgery under general anesthesia. Data on perioperative outcomes across the aforementioned domains were extracted via standardized selection, extraction, and bias assessment methods.

Results

A total of 479 papers were included in the study following screening of over 10,000 citations. Intraoperatively, airway complications were infrequently reported; however, a few cases of increased pulmonary pressures and hypercapnia emerged. Postoperatively, RAS was associated with shorter hospital stay (mean 2.52 days RAS vs. 1.85 LAP vs. 5.11 OP) and faster return to daily activities (mean 11.4 days RAS vs. 13.6 LAP vs. 13.2 OP), but evidence for validated quality of recovery scores was lacking. Pain intensity and analgesic requirements were found to be lower after RAS than after OP and sometimes LAP, with inconsistent findings across studies. Surgical outcomes favored RAS, showing reduced blood loss (mean 154 mL RAS vs. 149 LAP vs. 358 OP), lower conversion and major complications rates than other approaches. Oncologic metrics—including lymph node yield and R0 resection—were comparable across approaches.

Overall study quality was limited by a moderate to serious risk of bias across most outcomes.

Conclusion

Robotic-assisted gynecologic surgery is a feasible and safe alternative to laparoscopic and open approaches, with advantages in selected surgical outcomes and potential benefits in postoperative pain and recovery. However, further high-quality randomized studies are needed to confirm these benefits.

Systematic review registration

PROSPERO CRD420251105240

Supplementary Information

The online version contains supplementary material available at 10.1186/s44158-026-00358-y.

Introduction

Robotic-assisted surgery has increasingly transformed the field of gynecologic surgery, offering theoretical advantages over traditional laparoscopic and open approaches in terms of dexterity, ergonomics, and three-dimensional visualization—features that may facilitate more precise dissection and reduced perioperative morbidity [1–6]. At the same time, the physiological demands inherent to robotic surgery—particularly steep Trendelenburg positioning and prolonged pneumoperitoneum—introduce significant cardiopulmonary drawbacks. These include impaired pulmonary compliance, reduced functional residual capacity, diaphragmatic cephalad displacement, increased venous pressures, and reduced cardiac index. Recognizing these competing factors underscores why robotic gynecologic surgery represents a unique intersection of surgical innovation and anesthetic complexity.

The rapid expansion of robotic-assisted surgery has been accompanied by a wealth of research comparing perioperative outcomes across various surgical approaches. For example, Lenfant et al. (2023) [1] demonstrated improvements in blood loss and length of stay following robotic surgery compared with laparoscopic, vaginal, and open hysterectomy techniques, though persistent heterogeneity in operative metrics was noted. More recently, a systematic review and meta-analysis by Ricciardi et al. (2025) [2] concluded that robotic-assisted surgery offers meaningful advantages over laparoscopic and open surgery in oncologic settings, with improvements in operative time, hospital stay, conversion rate, and postoperative sequelae. Nonetheless, the analysis is limited by emphasis on 30-day outcomes and aggregation of varying oncologic procedures, which limits its ability to capture procedure-specific perioperative effects. Broader narrative and scoping reviews have further investigated robotic platforms, their evolution, and expanding indications in obstetrics and gynecology [3–6].

Despite this growing literature, existing reviews primarily focus on surgical performance and complication profiles, with limited attention to anesthetic implications. Yet robotic-assisted gynecologic surgery poses distinct challenges in airway management, ventilation, hemodynamic stability, and overall perioperative safety. Beyond intraoperative concerns, postoperative recovery, analgesia, and patient-centered outcomes remain active research fields. Although robotic platforms have been hypothesized to facilitate enhanced recovery through reduced pain, faster ambulation, and lower complication rates, data are heterogeneous, and the overall magnitude of perioperative benefits remains unclear.

The current systematic review addresses this gap by comprehensively assessing perioperative outcomes and anesthetic implications of robotic-assisted gynecologic surgery. Specifically, we synthesized available evidence across four key domains: (1) intraoperative anesthetic challenges, (2) postoperative recovery metrics, (3) analgesic strategies and pain-related outcomes, and (4) surgical performance, including technical and oncologic surgical results. This review provides a focused integration of anesthetic considerations within the broader perioperative landscape of robotic-assisted gynecologic surgery, aiming to clarify current evidence and identify critical targets for future research.

Materials and methods

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [7] guidelines. The study protocol was developed prior to data extraction and registered in the PROSPERO database (ID: CRD420251105240) on July 24, 2025.

Inclusion and exclusion criteria

Inclusion and exclusion criteria were defined according to the PICO Framework (Population, Intervention or Exposure, Comparison or Control, Outcome(s)). Eligible studies focused on (a) adult women (≥ 18 years); (b) undergoing robotic-assisted gynecological surgery under general anesthesia; (c) performed for benign or malignant indications; (d) with at least one of the following outcome measures: intraoperative anesthetic challenges, postoperative recovery outcomes, postoperative pain control, and surgical outcomes. Exclusion criteria comprised (a) studies of pediatric patients; (b) non-gynecologic robotic surgery; (c) abstracts, pre-clinical studies, commentaries, editorials, letters, opinion articles, and trial protocols; and (d) non-English publications. Only observational studies (prospective and retrospective cohort, case–control, and cross-sectional studies) and clinical trials were considered eligible for inclusion.

Outcome measures

Primary outcomes were categorized into four domains:

• PICO 1—intraoperative anesthetic management: airway management, ventilation settings, hemodynamic alterations, positioning-related complications.

• PICO 2—postoperative recovery: length of hospital stay (LOS), return to daily activities, time to ambulation, Quality of Recovery (QoR-15).

• PICO 3—analgesia and pain outcomes: postoperative pain scores (NRS or VAS, at rest and on movement), analgesic requirements (opioid and non-opioid), incidence of postoperative nausea and vomiting (PONV), time to mobilization.

• PICO 4—surgical performance: operative time, blood loss, conversion rate, complications (Clavien-Dindo Classification), lymph node yield, R0 resection, wound dehiscence, instrumental or technical failures, need for reintervention or re-hospitalization, patient cosmetic satisfaction, fever, or infectious complications.

Search strategy

Following approval on July 25, 2025, a comprehensive search was conducted for each PICO question in PubMed®, Embase®, the Cochrane Central Register of Controlled Trials®, Web of Science®, and Scopus®, covering the period from January 1, 2005, to July 25, 2025. Searches were restricted to human studies and English-language publications. Representative search terms included “robotic gynecologic surgery”, “airway management”, “hemodynamics”, “patient positioning”, “postoperative period”, “length of stay”, “time to ambulation”, “quality of recovery”, “postoperative pain”, “PONV”, “operative time”, “blood loss”, “conversion to open surgery”, “Clavien-Dindo”, “lymph node yield”, “R0 resection”, “reintervention”, “dehiscence”, and “hospital readmission”. The complete search strategy is provided in the Supplementary materials.

Quality control and reproducibility

Study selection was performed using the CADIMA® [8] platform. Following deduplication, two reviewers working in predefined pairs (MC/FB, EI/GT, DC/IS, SG/TP) independently screened the titles and abstracts of all retrieved records. Articles were included if both reviewers agreed that the eligibility criteria were met. Disagreements were resolved through discussion and, when consensus was not reached, arbitration by a third reviewer. In the second stage, full-text screening was conducted, with discrepancies again resolved through arbitration by a third party. Prior to screening, inter-reviewer agreement was assessed for both phases by performing a consistency check on five sample articles, yielding a Cohen’s κ coefficient of 0.61 and 0.65, respectively.

Data extraction and harmonization

Data extraction was independently conducted by two reviewers via a standardized extraction form. Extracted variables included study title, first author, year and country of publication, study design, type of surgery, comparator, clinical outcomes, study limitations, and risk of bias. Reported measures were converted to a common metric where feasible and justified (e.g., length of stay was standardized to days, opioid consumption was harmonized to morphine milligram equivalents [MME]); when conversion was not possible or risked misrepresentation, outcomes were synthesized narratively and stratified by metric. Data were expressed as mean ± standard deviation (SD), median (range), or median [interquartile range, IQR], unless otherwise specified. Where feasible, median (range) values were converted into estimated mean (SD) using standard Wan/Luo formulas.

Data synthesis

Owing to substantial clinical (e.g., surgical indication, procedure type, pathology, patient demographics, and perioperative management) and methodological heterogeneity (e.g., study design, outcome definitions, follow-up duration), a formal meta-analysis of comparative effect sizes was not performed. Instead, for outcomes reported on comparable continuous scales, we generated descriptive pooled estimates to summarize overall trends across surgical approaches. When at least three studies reported means and standard deviations (SDs) for a given outcome, or when conversion from medians and ranges or interquartile ranges (IQRs) to estimated means and SDs was feasible using the Wan–Luo formulas, pooled estimates were calculated. Sample size–weighted mean values were derived from aggregate study-level data using fixed-effect weighting. Fixed-effect weighting was used solely to summarize central tendency across study arms and does not imply an assumption of clinical or methodological homogeneity. Each study arm (robotic-assisted surgery [RAS], laparoscopy [LAP], or open procedure [OP]) was treated as an independent analytical unit for weighting purposes, without double counting participants across comparisons.

Confidence intervals around pooled means were calculated to reflect the precision of the summary estimate. Dispersion around pooled values reflects combined within-study variability and between-study heterogeneity rather than true patient-level variance. Accordingly, pooled estimates were interpreted as descriptive summaries rather than precise effect measures, and no formal inferential statistical testing or comparative meta-analytic modeling was performed. p-values reported in the manuscript derive exclusively from individual primary studies. When numerical harmonization was not possible or judged likely to misrepresent the original findings, results were synthesized narratively using structured tables in accordance with PRISMA recommendations.

Risk of bias assessment and GRADE

Risk of bias was assessed using the Cochrane RoB-2 tool for randomized controlled trials (RCTs) [9] and ROBINS-I V2 [10] for non-randomized studies. Before assessment, reviewers completed a calibration exercise to standardize the interpretation of the tools. Two reviewers independently evaluated each study and recorded judgments as “low risk,” “some concerns/moderate risk,” or “high risk/critical risk,” according to tool-specific guidance. Disagreements were resolved through discussion or, when needed, arbitration by a third reviewer. Quality of evidence for each outcome was assessed using the GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach [11].

Results

Figures 1, 2, 3, and 4 illustrate the study flow diagrams for the four PICO questions. The database search yielded a total of 1368 citations for PICO 1, 5408 for PICO 2, 857 for PICO 3, and 3547 for PICO 4. After removing duplicates, 1146, 3053, 619, and 2460 records, respectively, were included in the initial screening procedure at the title and abstract level. Following full-text assessment of 1091 potentially eligible studies across all PICOs, a total of 479 publications met the inclusion criteria and were included in the final systematic review. An overview of the included studies and their demographics is provided in the Supplementary materials.

Fig. 1.

Flowchart of included studies for PICO 1

Fig. 2.

Flowchart of included studies for PICO 2

Fig. 3.

Flowchart of included studies for PICO 3

Fig. 4.

Flowchart of included studies for PICO 4

PICO 1—intraoperative anesthetic challenges

Airway complications (i.e., accidental tube misplacement) were not reported in any of the included studies, possibly due to underreporting. Similarly, no major hemodynamic complications were documented, aside from sporadic episodes of intraoperative hypotension requiring vasopressors, with no consistent differences across surgical approaches.

Ventilation changes were commonly observed. Both robotic (RAS) and laparoscopic (LAP) approaches were associated with elevations in end-tidal CO₂ (EtCO₂). For example, Singla et al. described higher EtCO₂ levels following pneumoperitoneum in RAS compared with open surgery (OP) (35.80 ± 4.87 vs. 34.11 ± 2.05; p = 0.038), possibly reflecting reduced lung compliance [12]. Additionally, significantly higher peak inspiratory pressures (Ppeak) in RAS compared to OP (30.9 ± 7.1 vs. 23.3 ± 6.8; p < 0.001) emerged in Gali et al. [13].

Pulmonary complications—including atelectasis and pneumonia—occurred at similar rates in RAS and LAP (p = 0.673) [13], and some cases of conversion to laparotomy due to intolerance of the Trendelenburg position were registered in the robotic group.

Positioning-related challenges were more pronounced in RAS, where extended use of steep Trendelenburg angles exceeding 30° was associated with increased optic nerve sheath diameter (ONSD) compared to OP (5.61 ± 0.57 vs. 5.04 ± 0.51 mm; p < 0.001) [12], suggesting an increase in intracranial pressure.

Across descriptive pooled data from 1823 patients, robotic-assisted surgery showed the longest operative time (166.8 ± 68.0 min; 95% CI 162.7–171.0), followed by LAP (162.5 ± 54.2 min; 95% CI 157.1–167.9) and OP (125.1 ± 47.8 min; 95% CI 120.5–129.7).

Trendelenburg angles ranged from 30° to 40° but were rarely reported, thus limiting adequate generalizability.

Overall, intraoperative anesthetic challenges were heterogeneously reported and frequently limited to surrogate physiological parameters, restricting the ability to perform formal quantitative synthesis beyond operative time (Table 1).

Table 1.

Summary of intraoperative anesthetic outcomes across included studies (PICO 1)

Domain	N. citations	Comparator	Overall result	Interpretation
Airway management issues	0	N/A	No study reported accidental tube misplacement or major airway-related adverse events	Absence of reported events; cannot exclude underreporting
Ventilation changes	11	RAS vs. LAP vs. OP	Two articles reported increased EtCO2 levels in RAS compared to LAP and OP. One study evidenced significantly higher Ppeak in RAS compared to OP Several cases of RAS conversion to OP due to intolerance to the Trendelenburg position and/or pneumoperitoneum were observed Sporadic pulmonary complications (e.g., ARF, atelectasis, pneumonia) were registered	Possible confounding due to elevated BMI and COPD
Hemodynamic alterations (hypotension, vasopressors)	0	N/A	No study systematically reported major hemodynamic instability or vasopressor requirements	Absence of reported events; cannot exclude underreporting
Positioning complications	1	RAS vs. LAP vs. OP	Significantly higher average ONSD in RAS (5.61 ± 0.57) vs. OP (5.04 ± 0.51); p < 0.001. Expressed as change in ONSD (mm)	Single-study evidence. Possibly linked to steep Trendelenburg in RAS
Operative time (mean(SD)) (min)	9	RAS vs. LAP vs. OP	Pooled mean ± SD (95% CI): 166.8 ± 68.0 (162.7–171.0) RAS vs. 162.5 ± 54.2 (157.1–167.9) LAP vs. 125.1 ± 47.8 (120.5–129.7) OP	Operative time consistently longer in minimally invasive approaches; pooling limited by heterogeneity
Trendelenburg positioning (°)	3	No direct comparator	> 30° RAS	Limited reporting prevents quantitative synthesis
Pneumoperitoneum pressure (mmHg)	0	N/A	No study systematically reported intraoperative pneumoperitoneum pressures	Data unavailable

Data reflect the number of citations reporting each domain and the comparators assessed. Operative time was summarized using sample size–weighted pooled mean estimates derived from study-level data, restricted to studies providing or allowing estimation of mean (SD). Other outcomes were synthesized narratively due to sparse reporting and heterogeneity. Absence of reported complications does not imply absence of risk. No eligible studies reported airway complications, hemodynamic instability, or pneumoperitoneum pressures

Abbreviations: RAS robotic-assisted surgery, LAP laparoscopy, OP open procedure, ONSD optic nerve sheath diameter, ARF acute respiratory failure, EtCO₂ end-tidal carbon dioxide, BMI body mass index, COPD chronic obstructive pulmonary disease, Ppeak peak inspiratory pressure, CI confidence interval, N/A not available

PICO 2—postoperative recovery metrics

Despite heterogeneity in reporting, minimally invasive approaches were consistently associated with shorter postoperative hospital stay compared with open surgery. Laparoscopy showed the shortest length of stay (LOS), followed closely by robotic-assisted surgery, while open procedures required substantially longer hospitalization. Across the included studies, the pooled, sample size–weighted mean LOS was 2.52 ± 3.22 (95% CI 2.46–2.57, N = 12,980) for robotic surgery, 1.85 ± 2.18 (95% CI 1.82–1.87, N = 40,264) for laparoscopic surgery, and 5.11 ± 5.08 (95% CI 5.02–5.20, N = 11,952) for open surgery. These findings suggest a clear gradient in postoperative recovery, with laparoscopic surgery requiring the shortest hospital stay, followed by robotic-assisted procedures, and open surgery remaining the most resource-intensive approach (Fig. 5).

Fig. 5.

Length of hospital stay (LOS) according to surgical approach. Pooled, sample-size-weighted mean LOS for robotic-assisted, laparoscopic, and open procedures. Error bars represent 95% confidence intervals around the pooled estimates and reflect uncertainty of the summary measure

Data on functional recovery were more limited. Five studies reported outcomes related to return to activities of daily living (ADL). Across all measures, RAS consistently revealed faster return to ADL compared with OP, with reported p-values ranging from 0.04 to < 0.0001. On the contrary, RAS and LAP did not differ significantly in included studies. Across studies reporting time to return to daily activities, the weighted pooled mean recovery time was shorter after robotic surgery (11.4 ± 14.9 days, 95% CI 9.4–13.4, N = 212) compared with laparoscopic (13.6 ± 15.7 days, 95% CI 11.0–16.1, N = 146) and open surgery (13.2 ± 12.1 days, 95% CI 11.9–14.5, N = 339). However, these findings are driven by small samples and cannot be generalized.

One study [14] specifically assessed time to ambulation and found substantially faster recovery in patients undergoing robotic-assisted surgery (2.9 ± 1.1 days) compared with laparoscopic (4.4 ± 2.7 days) and open techniques (13.4 ± 9.4 days; p < 0.001).

None of the selected studies measured the Quality of Recovery (QoR-15) score as an outcome, thus leaving uncertainty regarding this aspect (Table 2).

Table 2.

Summary of postoperative recovery metrics across included studies (PICO 2)

Domain	N. citations	Comparator	Overall result	Interpretation
Length of stay (days)	191	RAS vs. LAP vs. OP	Pooled mean ± SD (95% CI): 2.52 ± 3.22 (2.46–2.57) RAS vs. 1.85 ± 2.18 (1.82–1.87) LAP vs. 5.11 ± 5.08 (5.02–5.20) OP	Minimally invasive approaches are generally associated with shorter LOS
Return to activity (days)	5	RAS vs. LAP vs. OP	Pooled mean ± SD (95% CI): 11.4 days ± 14.9 (9.4–13.4) RAS vs. 13.6 days ± 15.7 (11.0–16.1) LAP vs. 13.2 days ± 12.1 (11.9–14.5) OP	Limited data, trend toward earlier return to activity with RAS
QoR-15 Score	0	N/A	Not reported	Patient-centered recovery not assessed
Return to ambulation (days)	1	RAS vs. LAP vs. OP	Shorter return to ambulation in patients undergoing RAS (2.9 ± 1.1) compared with LAP (4.4 ± 2.7) and OP (13.4 ± 9.4); p < 0.001	Single-study evidence; heterogeneous definitions

Data reflect the number of citations reporting each domain and the comparators assessed. Pooled mean values were calculated for length of stay and return to activities of daily living using sample size–weighted estimates derived from study-level summary data, restricted to studies reporting or allowing estimation of mean and standard deviation. Dispersion reflects combined within- and between-study variability. Time to ambulation was reported in a single study and is presented descriptively. No included study reported Quality of Recovery-15 scores

Abbreviations: RAS robotic-assisted surgery, LAP laparoscopy, OP open procedure, LOS length of stay, QoR-15 Quality of Recovery-15 questionnaire, N/A not available

PICO 3—postoperative pain outcomes and analgesic requirements

Postoperative pain was the most frequently assessed recovery outcome and was reported in 29 studies.

Although substantial heterogeneity was present in pain assessment tools, timing, and analgesic protocols, a consistent trend emerged indicating lower postoperative pain intensity following robotic-assisted surgery compared with open procedures [15], while comparisons between robotic-assisted and laparoscopic approaches were more variable and often nonsignificant. Seventeen studies reported acute postoperative pain at approximately 24 h using a 0–10 NRS or VAS scale (with higher-range VAS values rescaled accordingly). Across these studies, the pooled, sample size–weighted mean pain score was 3.1 ± 2.1 (95% CI 3.0–3.2, N = 793) for robotic-assisted surgery, 3.4 ± 2.3 (95% CI 3.2–3.6, N = 810) for laparoscopic surgery, and 4.6 ± 2.2 (95% CI 4.3–4.9, N = 280) for open surgery.

Analgesic requirements were evaluated in 23 studies. Overall, opioid consumption—expressed as morphine milligram equivalents (MME)—tended to be similar between RAS and LAP [16, 17]. In contrast, RAS showed a clear opioid-sparing effect compared with OP [18, 19]. Use of non-opioid analgesics was inconsistently reported. In six studies reporting opioid consumption as MME within the first 24–48 h, pooled mean opioid use was 36.2 ± 41.5 mg (95% CI 31.4–41.0, N = 292) for RAS, 25.0 ± 32.3 mg (95% CI 21.3–28.7, N = 294) for LAP, and 132.9 ± 105.0 mg (95% CI 91.8–174.0, N = 25) for OP. However, these pooled estimates should be interpreted cautiously due to heterogeneity in time windows and analgesia protocols, and open surgery small sample size.

Postoperative nausea and vomiting (PONV) was infrequently assessed, with only five studies reporting this outcome. Limited evidence suggested lower PONV rates following RAS compared with OP (18), particularly in the immediate postoperative period; however, findings were inconsistent and derived from small cohorts (Table 3).

Table 3.

Summary of postoperative pain outcomes and analgesic requirements across included studies (PICO 3)

Domain	N. Citations	Comparator	Overall Result	Interpretation
Postoperative pain scores	29	RAS vs. LAP vs. OP	Pooled mean ± SD (95% CI): 3.1 ± 2.1 (3.0–3.2) RAS vs. 3.4 ± 2.3 (3.2–3.6) LAP vs. 4.6 ± 2.2 (4.3–4.9) OP	Lower pain after RAS vs OP; RAS vs LAP differences modest and inconsistent
Analgesic requirements (MME)	23	RAS vs. LAP vs. OP	Pooled mean ± SD (95% CI): 36.2 ± 41.5 (31.4–41.0) RAS vs. 25.0 ± 32.3 (21.3–28.7) LAP vs. 132.9 ± 105.0 (91.8–174.0) OP	Open surgery associated with substantially higher opioid use; heterogeneity in protocols
PONV incidence	5	RAS vs. LAP vs. OP	PONV incidence was markedly higher in OP compared with RAS, both in the PACU and on the ward. Despite these differences, antiemetic use was similar across groups. Interestingly, one analysis suggested OP conferred a better reported outcome for nausea and vomiting	Very limited data; small cohorts; non-standardized reporting
Time to mobilization	0	N/A	Not reported	Outcome not assessed

Data indicate the number of studies reporting each outcome domain and the surgical comparators assessed. Pooled mean pain scores and opioid consumption were calculated using sample size–weighted estimates derived from study-level summary data, restricted to studies reporting or allowing estimation of mean and standard deviation. Dispersion reflects combined within- and between-study variability. Pain scales reported as VAS 0–100 mm or 0–150 mm were rescaled to a 0–10 metric. Postoperative nausea and vomiting was inconsistently reported, and no study provided data on time to mobilization

Abbreviations: RAS robotic-assisted surgery, LAP laparoscopy, OP open procedure, PONV postoperative nausea and vomiting, PACU post-anesthesia care unit, SD standard deviation, N/A not available

No data on postoperative mobilization was reported in the included studies.

PICO 4—surgical outcomes

A total of 17,720 robotic (RAS), 23,907 laparoscopic (LAP), and 14,984 open (OP) cases with extractable mean ± SD operative times were pooled across all eligible study arms. Sample-size weighted aggregation of study-level data showed a mean operative time of 168.9 ± 84.7 min (95% CI 167.7–170.2) for RAS, 164.1 ± 78.6 min (95% CI 163.0–165.2) for LAP, and 156.9 ± 122.8 min (95% CI 155.3–158.5) for OP.

Intraoperative blood loss showed a consistent pattern across studies. Pooled, sample size–weighted mean blood loss was comparable between RAS (154 mL; 95% CI 149.9–158.1) and LAP (149 mL; 95% CI 145.4–151.6), whereas substantially higher blood loss was observed in open surgery (358 mL; 95% CI 350.3–365.9).

Across the included literature, conversion to open surgery consistently favored RAS over laparoscopy. Most robotic series reported 0% conversion, whereas laparoscopic conversion rates ranged from 2 to 15%, with some reports as high as 29% [20]. Disparities among approaches widened in higher-complexity procedures, e.g., oncologic staging, radical hysterectomy, para-aortic lymphadenectomy. The most common indications for conversion included dense adhesions, disseminated malignancy, and intraoperative hemorrhage [21].

Robotic-assisted surgery generally displayed equal or lower complication rates—defined as Clavien–Dindo ≥ II—compared with laparoscopy. Most RAS series reported complication rates of 0–10%, whereas LAP rates ranged from 2 to 20%, though variability was substantial and procedure-dependent. In contrast, open surgery consistently showed the greatest morbidity burden, with complication rates typically 20–40%—often two- to four-fold higher than minimally invasive approaches.

Lymph node yield was broadly comparable across surgical modalities, typically ranging from 12 to 30 nodes for RAS, 13–28 for LAP, and 15–30 for OP. R0 resection rates exceeded 95% across most cohorts, with no statistically significant differences observed among surgical approaches. Wound dehiscence, reintervention, and rehospitalization were infrequently reported and generally uncommon following minimally invasive surgery. Rates were typically < 1% for wound dehiscence and 1–5% for reintervention and rehospitalization after RAS and LAP, and consistently higher when reported for open surgery. Cosmetic satisfaction, assessed in six studies, was high for minimally invasive approaches, with robotic-assisted surgery yielding scores around 9/10 and comparable results to laparoscopy. Infectious complications—including fever, urinary tract infection, and wound or pelvic infections—were reported inconsistently. When available, rates were similar between robotic-assisted and laparoscopic surgery (approximately 1–10%) and substantially higher following open procedures (Table 4).

Table 4.

Summary of surgical outcomes across included studies (PICO 4)

Outcome	N. citations	Comparator	Overall result	Interpretation
Operative time (min)	211	RAS vs. LAP vs. OP	Pooled mean ± SD (95% CI): 168.9 ± 84.7 min (95% CI 167.7–170.2) RAS vs. 164.1 ± 78.6 min (95% CI 163.0–165.2) LAP vs. 156.9 ± 122.8 min (95% CI 155.3–158.5) OP	Minimally invasive approaches associated with modestly longer operative time
Blood loss (ml)	189	RAS vs. LAP vs. OP	Pooled mean ± SD (95% CI): 154 mL (95% CI 149.9–158.1) RAS vs. 149 mL (95% CI 145.4–151.6) LAP vs. 358 mL (95% CI 350.3–365.9) OP	Substantially lower blood loss with minimally invasive surgery
Conversion rate (%)	150	RAS vs. LAP	RAS conversion rates typically 0–2%; LAP 2–15% (up to 29% in complex cases)	Consistently lower conversion rate for RAS compared with LAP
Complications (Clavien-Dindo ≥ II) (%)	100	RAS vs. LAP vs. OP	1–10% RAS vs. 2–20% LAP vs. 20–40% OP	Favorable morbidity profile for RAS compared with LAP and OP
Lymph node yield (N)	88	RAS vs. LAP vs. OP	Lymph-node yield was similar across approaches (~ 12–30 nodes RAS vs. ~ 13–28 LAP vs. ~ 15–30 OP)	No consistent differences
R0 resection (%)	10	RAS vs. LAP vs. OP	Uniformly high (> 95%)	Comparable oncologic adequacy
Wound dehiscence (%)	40	RAS vs. LAP vs. OP	< 1% for RAS/LAP; higher for OP	Minimally invasive approaches favorable
Reintervention (%)	48	RAS vs. LAP vs. OP	Generally ~ 1–5% for RAS, ~ 2–7% for LAP, and slightly higher for OP	Low across all cohorts
Re-hospitalization (%)	38	RAS vs. LAP vs. OP	Typically ~ 1–5% for RAS and LAP, and slightly higher for OP	Low overall
Cosmetic satisfaction (%)	6	RAS vs. LAP vs. OP	High scores (~ 9/10) for minimally invasive surgery	Limited evidence
Fever/infectious complications (%)	40	RAS vs. LAP vs. OP	Similar for RAS and LAP (1–10%); higher for OP	Heterogeneous reporting

Data indicate the number of studies reporting each outcome domain and the surgical comparators assessed. Quantitative pooled estimates were calculated only for outcomes with sufficient studies reporting or allowing estimation of mean and standard deviation and are presented as sample size–weighted descriptive summaries. Other outcomes are synthesized narratively due to heterogeneity in definitions and reporting. Dispersion reflects combined within- and between-study variability rather than patient-level variance.

Abbreviations: RAS robotic-assisted surgery, LAP laparoscopy, OP open procedure, SD standard deviation; CI confidence interval, N/A not available

Data interpretation, risk of bias, and GRADE assessment

Interpretation of pooled intraoperative and postoperative outcomes warrants caution. Quantitative pooled estimates were derived from aggregate study-level data and weighted by sample size. In the presence of substantial clinical and methodological heterogeneity across studies, these estimates are intended as descriptive summaries rather than precise patient-level effect measures. Reported dispersion reflects combined within-study variability and between-study heterogeneity. Accordingly, formal comparative statistical testing across pooled estimates was not undertaken. Outcomes that were inconsistently reported or measured using heterogeneous definitions were synthesized narratively.

Across all four PICO domains, ROBINS-I assessment revealed a consistent pattern of moderate to serious risk of bias. In PICO 1, over half of the studies were rated as having a serious overall risk, primarily due to confounding (D1) and participant selection (D2). PICO 2, encompassing a broader range of studies, showed predominantly moderate risk across domains, with low risk frequently observed in outcome measurement and reporting domains (D3–D5). PICO 3 and PICO 4 further reinforce these trends, with most studies revealing serious concerns in confounding and missing data (D6), contributing to serious overall bias ratings. Notably, only a small subset of studies across all PICOs achieved low overall risk.

Consistent with these findings, the overall certainty of evidence assessed using the GRADE framework was low or very low across all outcomes. Downgrading was primarily driven by serious risk of bias related to retrospective study designs, non-randomized comparisons, single-center cohorts, and residual confounding. Additional downgrading for inconsistency and indirectness was applied where outcome definitions and measurement timing varied substantially between studies. Overall, evidence across all outcomes was of low to very low certainty. Detailed ROBINS-I assessments and GRADE judgments for individual outcomes are provided in the Supplementary materials.

Discussion

This systematic review highlights key aspects of anesthetic and perioperative management in robotic-assisted gynecological surgery. Across all PICO domains, the evidence consistently favors minimally invasive techniques over open surgery for several clinically relevant outcomes. However, distinctions between robotic-assisted and conventional laparoscopy are more nuanced and frequently attenuated by heterogeneity and risk of bias.

Intraoperative anesthetic outcomes were sparsely and heterogeneously reported. Despite the well-established physiological effects of pneumoperitoneum and steep Trendelenburg positioning on respiratory mechanics and gas exchange [22, 23], only isolated studies reported increased end-tidal CO₂ or reduced lung compliance. Similarly, hemodynamic instability was inconsistently reported, with no clear differences between robotic-assisted (RAS), laparoscopic (LAP), and open surgery (OP). Importantly, the apparently low incidence of airway and hemodynamic complications likely reflects limited and heterogeneous reporting rather than a true absence of such events. While these findings may suggest that airway complications are uncommon when appropriate ventilatory strategies are applied, the lack of systematic monitoring and standardized outcome definitions substantially weakens this conclusion. Evidence regarding positioning-related physiological changes was similarly sparse. Prolonged steep Trendelenburg angles—often exceeding 30° and characteristic of robotic-assisted procedures—have been linked to raised optic nerve sheath diameter in one study [12] and, in rare cases, conversion to open surgery. Although postoperative pulmonary complications (PPCs) were not the focus of this review, recent evidence provides important contextual insights. In a large multicenter observational study published in 2025 [24], Serafini et al. reported nearly double the incidence of PPCs in RAS compared with laparoscopic surgery (19% vs. 9%, p < 0.001), a difference largely explained by prolonged mechanical ventilation rather than the surgical platform itself. While this study was not restricted to gynecologic procedures, its findings remain relevant. Altogether, these data suggest that the relative intraoperative respiratory stability does not necessarily translate into a reduced risk of PPCs. In this context, longer operative and ventilation times—particularly during the learning curve of robotic-assisted procedures—may reasonably increase PPC risk until greater procedural efficiency is achieved. Accordingly, the findings of the present meta-analysis should be interpreted within this broader clinical framework, recognizing that postoperative respiratory risk may not be fully captured by intraoperative anesthetic parameters alone.

Postoperative recovery outcomes generally favored minimally invasive surgery over open procedures, but distinctions between RAS and LAP were less clear. Length of stay was consistently shorter after laparoscopy, with robotic surgery showing slightly longer stays that may reflect patient selection, procedural complexity, institutional discharge protocols, and variability in surgeon experience rather than intrinsic differences. Evidence on functional recovery, such as return to daily activities, was limited and variable, though several small studies suggested earlier recovery after RAS compared with open surgery. Earlier resumption of daily functions represents a meaningful patient-centered benefit, aligning with the increasing emphasis on enhanced recovery pathways in perioperative medicine. Data on ambulation and validated patient-reported outcomes such as QoR-15 were largely absent, underscoring an important gap within the current literature. Future randomized trials should prioritize the systematic inclusion of such instruments to better characterize recovery trajectories beyond traditional hospital-based metrics.

Postoperative pain and analgesic needs were the most frequently studied pain endpoints, yet results remained heterogeneous. Several studies reported lower pain scores and reduced opioid requirements with RAS compared with open surgery; this may be related to smaller incisions, improved surgical precision, and reduced tissue trauma. However, many studies lacked standardized analgesic protocols or adequate adjustment for confounders. Evidence on postoperative nausea and vomiting was sparse and contradictory. While some studies suggested a lower incidence with RAS, this may reflect confounders such as anesthetic technique, perioperative antiemetic protocols, and patient risk profiles rather than a true surgical effect. Similarly, return to mobilization was rarely and inconsistently reported, preventing robust conclusions. Overall, evidence suggests a potential benefit of RAS in reducing postoperative pain and analgesic consumption, but further high-quality trials are needed to confirm these findings and clarify its impact on PONV and functional recovery.

Regarding surgical outcomes, RAS was consistently associated with lower estimated blood loss and lower conversion rates than laparoscopy. These findings were among the most robust across PICOs, although interpretation remains limited by observational study designs and potential selection bias, especially in complex oncologic cases. Although operative time was frequently longer in RAS compared with laparoscopy, this difference has narrowed over time, likely reflecting improvement in surgical experience and technological advancements [23, 25, 26]. In many settings, longer duration may be outweighed by ergonomic benefits, enhanced vision, and greater precision of dissection. The complication profile of RAS was comparable or slightly favorable to that of LAP and OP. The reduction in transfusion requirements, wound complications, and re-hospitalizations may contribute to faster recovery and greater patient satisfaction, reinforcing the safety profile of robotic-assisted procedures. Oncologic surrogate markers—lymph node yield and R0 resection—did not differ meaningfully across approaches, but these outcomes may be strongly confounded by tumor characteristics and surgeon practice patterns.

These analyses should be interpreted with caution and regarded as exploratory and hypothesis-generating only. Outcomes were synthesized using sample size–weighted mean estimates based on aggregate study-level data rather than individual patient data. While this approach allows descriptive comparison across surgical modalities, it does not fully account for substantial between-study heterogeneity. Consequently, the dispersion observed around pooled estimates reflects a combination of within-study variability and between-study differences, rather than a true patient-level variance. These considerations precluded formal comparative statistical testing of pooled outcome estimates across surgical approaches. Therefore, the results should be interpreted as indicating consistent directional trends rather than exact quantitative effect sizes.

Across all domains, the quality of evidence was limited by moderate to serious risk of bias, lack of standardized outcome definitions, and predominance of retrospective designs. Additional limitations include potential publication bias, institutional practice variability, and residual confounding that could not be adequately controlled using aggregate data. Economic considerations also warrant attention: although robotic platforms involve higher upfront costs, these may be partially offset by reductions in complications and length of hospital stay, an aspect not formally evaluated in this review. Finally, the environmental impact of different surgical approaches was beyond the scope of the present analysis and remains an important area for future investigation.

Conclusion

In conclusion, this systematic review suggests that robotic-assisted surgery represents a safe and feasible alternative to laparoscopic and open approaches in gynecologic surgery, with consistent advantages in selected surgical outcomes, such as reduced blood loss and lower conversion rates. However, evidence regarding anesthetic implications, postoperative recovery, and pain outcomes remains limited and heterogeneous, particularly when comparing robotic-assisted and conventional laparoscopy. From an anesthesiology perspective, these findings underscore the need for prospective, adequately powered randomized trials with standardized anesthetic protocols, comprehensive reporting of intraoperative physiological parameters, and inclusion of validated patient-centered recovery outcomes. Such efforts are essential to clarify the true clinical value of robotic-assisted surgery within modern perioperative care pathways.

Abbreviations

ADL

Activities of daily living

EBL

Estimated blood loss

LAP

Laparoscopy

LOS

Length of hospital stay

MME

Morphine milligram equivalents

NRS

Numerical Rating Scale (Pain)

OP

Open procedure

QoR

Quality of recovery

PACU

Post-anesthesia care unit

PONV

Postoperative nausea and vomiting

RAS

Robotic-assisted surgery

R0

Complete tumor resection with negative margins

VA

Vaginal approach

VAS

Visual Analog Scale (Pain)

Authors’ contributions

E.I., G.T.: Conceptualization, Methodology, Writing of original draft; V.T.: Conceptualization, Supervision, Investigation, Writing - review & editing; D.C., S.G., M.C. I.S., F.B., T.P.: Investigation, Data curation; L.N.S., F.S.: review & editing. All authors read and approved the final manuscript.

Funding

The study did not receive any institutional departmental funds, nor external, internal, or individual financial support.

Data availability

No datasets were generated or analyzed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.