“Comparative safety and efficacy of robotic TAPP and IPOM techniques in ventral hernia repair: a systematic review and meta-analysis of Short-term Outcomes”

This systematic review and meta-analysis was conducted following the Items for Systematic Reviews and Meta-Analyses Protocols (PRISMA) checklist [31] and the principles of Cochrane Handbook for Systematic Reviews of Interventions version 6.2, We used online databases: SCOPUS, Web of Science, Medline (PubMed), and Cochrane, till February 2025. The search strategy was based on Medical Subject Headings (MeSH) and included the following keywords and Boolean operators: “Robotic” AND “IPOM OR “intraperitoneal onlay mesh” OR “intraperitoneal onlay mesh plus” OR “r-IPOM” OR “IPOM+” AND “TAPP OR “Transabdominal preperitoneal repair” OR “preperitoneal” OR “sublay mesh” AND “Ventral hernia” OR “Hernia”, to ensure reaching out to all the potential included studies. A word cloud in Figure S summarizes the most frequent terms from all retrieved titles and abstracts (Table S).

The study was structured according to the PICO framework. P: Ventral hernia patients, I: Robotic mode of repair implementing rIPOM (Intraperitoneal Onlay Mesh repair)/rIPOM+, C: Robotic mode of repair implementing rTAPP (Trans-abdominal preperitoneal mesh repair), O: outcomes included a range of surgical endpoints, such as seroma, operative time, intraoperative bleeding, recurrence, SSI, closure of the hernia defect, and excision of the hernia sac, encompassing both short-term and long-term clinical results.

This comprehensive literature search yielded a total of 58 studies that were screened for eligibility. After duplicate removal using EndNote, 28 studies remained and were screened based on titles and abstracts. Full-text assessment was then performed to identify primary studies reporting outcomes of either r-TAPP, r-IPOM, or both in the context of ventral hernia repair.

Studies were excluded if they met any of the following criteria: (1) conducted on animal models, (2) focused on hernia types other than ventral hernia, (3) employed surgical techniques other than r-IPOM or r-TAPP, or (4) lacked sufficient outcome data for analysis.

Ultimately, eleven studies fulfilled all eligibility criteria and were included in the meta-analysis: Gockal et al. 2019 [14], Kudsi et al. 2020 [15], Wijerathne et al. [16] Bindal et al. 2024 [17], Kennedy et al. 2018 [18], Chelliah et al. 2024 [19], Ferraro et al. 2023 [20], Baur et al. 2021 [21], Bauer et al. 2024 [22], Kudsi et al. [23], and Gokcal et al. 2020 [24]. The study selection process is illustrated in the PRISMA flow diagram, as shown in Fig. 1 [31].

Titles, abstracts, AND full-texts were screened by Y.O.B, A.M.A, and N.M.A, and further reviewed by A.S.A and Z.M.W, to ensure meticulous study selection according to the criteria mentioned. They also extracted the data using Excel sheets to obtain outcomes and baseline characteristics.

A.M.A and N.E.M assessed the quality of included studies using the Newcastle-Ottawa Scale (NOS) used for the assessment of case-control studies and cohort studies [32]. NOS assesses each study by dividing it into 3 elements: selection, comparison, and exposure or outcome, whether the study is a case-control or cohort study, respectively. For each category, each fulfilled item scores one star, and for comparability, the maximum score is two stars. Study quality is determined by the number of stars. Studies of high-quality score (7–9), moderate quality studies score (4–6), whereas low-quality studies score (0–3), The assessment of nine studies was conducted revealing that, One study was rated as high-quality and eight of the studies were rated as moderate quality as shown in Table 1. For assessment of the risk of bias, we used the ROBINS-I V2 tool to evaluate our studies since most of them were non-randomized cohort studies. Each study was assessed across seven domains: confounding, selection of participants, classification of interventions, deviations from intended interventions, missing data, measurement of outcomes, and selection of the reported result. Risk levels are color-coded as low, moderate, serious, or critical. Most studies were rated as having a low to moderate risk of bias, with no studies judged to be at critical risk. The overall assessment of risk of bias is considered to be low.

We used the AMSTAR 2 tool to appraise the methodological quality of our Meta-analysis [33]. Two reviewers independently evaluated the study across 16 domains. Any discrepancies were first discussed between the reviewers, and if no consensus was reached, a third reviewer (AA) was consulted to resolve the disagreement. Overall confidence ratings were determined according to AMSTAR 2 guidelines [34], as per Table 2.

Primary outcomes of interest were postoperative seroma formation, hematoma incidence, SSI, and successful hernia defect closure. These were selected based on their clinical relevance in early postoperative recovery and potential implications for long-term outcomes. Secondary outcomes included operative time, hospital stay, bowel-related complications, recurrence, and reoperation.

We conducted meta-analyses for both binary and continuous outcomes using random-effects models to account for both within-study and between-study heterogeneity. All analyses were performed using the DerSimonian-Laird method for estimating between-study variance (tau²) as implemented in the ‘meta’ and ‘metafor’ packages in R version 4.0.0 [35, 36].

For binary outcomes (hematoma, seroma, surgical site infection, recurrence, and hernia defect closure), proportions were used as the summary measure with 95% confidence intervals calculated using the Clopper-Pearson exact method, which provides more reliable estimates for rare events [37]. Studies reporting zero events were incorporated by applying a continuity correction of 0.5 to both the numerator and denominator [38]. Between-group comparison was performed using the test for subgroup differences, with chi-squared statistics and corresponding p-values to determine whether differences between r-IPOM and r-TAPP approaches were statistically significant [39].

For continuous outcomes (length of hospital stay, operative time, and consule time), the mean difference (MD) was used as the effect measure with 95% confidence intervals. Weighted means were calculated for each surgical approach by weighting individual study means by their respective sample sizes. The variance of the mean difference was computed as the sum of the variances of each group’s mean [40].

Heterogeneity across studies was quantified using I² statistics for both binary and continuous outcomes, with values of 25%, 50%, and 75% considered to represent low, moderate, and high heterogeneity, respectively [41]. To further explore potential sources of heterogeneity for key outcomes, meta-regression analyses were performed. For instance, the influence of defect closure rate (%) as a continuous moderator variable on the overall complication rate (%) was examined using random-effects meta-regression models implemented in the ‘metafor’ package. The statistical significance of moderator variables was determined by the p-value of their respective regression coefficients. The statistical significance of the overall effect was assessed using Z-tests, with p < 0.05 considered statistically significant.

Comprehensive sensitivity analyses were conducted, including leave-one-out analyses to identify influential studies and alternative methods for heterogeneity estimation (maximum likelihood and restricted maximum likelihood) [42]. Publication bias was assessed using contour-enhanced funnel plots. All results were visualized using forest plots showing individual and pooled effect estimates with corresponding confidence intervals.

Fifteen studies (923 patients) from eight publications reported on hernial defect closure rates for r-IPOM+ (6 studies, 296 patients) and r-TAPP (9 studies, 627 patients). High heterogeneity (I² = 73.9%) warranted the use of the random-effects model. The pooled closure rate was 0.942 [95% CI: 0.880–1.000] for r-IPOM (I² = 81%), and 0.979 [95% CI: 0.956–1.000] for r-TAPP (I² = 67.5%). The combined pooled rate was 0.970 [95% CI: 0.949–0.991]. No significant subgroup difference was found between techniques (χ² = 1.19, df = 1, p = 0.2752). (Fig. 2, Supplementary Figs. 1)

Subgroup analyses comparing non-coated (mostly macroporous in our Meta-analysis) meshes with coated mesh types did not yield any statistically significant differences across the evaluated complications, including recurrence, hematoma, seroma, surgical site infection (SSI), or reoperation rates. Due to inconsistent definitions and considerable heterogeneity in mesh classification, along with the absence of meaningful statistical differences, the results of this subgroup analysis have been omitted from the main results and supplementary materials, as per Supplementary Fig. 3.

Meta-regression investiNgated the relationship between closure and complication rates. The unweighted model did not reach significance (r = −0.494, p = 0.1226), as shown in (Fig. 11, Supp. Figure 4). However, a sample-size weighted model showed a significant inverse correlation:

Complication Rate = 1.0290–1.0121 × Closure Rate (R² = 0.747, p = 0.0006), suggesting that higher closure rates predict lower complication rates.

Leave-one-out analyses demonstrated robustness for most outcomes. In seroma analysis, excluding Kudsi C 2020 (r-TAPP) or Baur 2021 (r-IPOM) reduced heterogeneity notably. For SSI, omitting Kudsi C 2020 or Gockal 2019 (both r-IPOM) eliminated heterogeneity. Console time heterogeneity decreased from 50.6 to 18.1% after removing Kennedy et al. 2018. Other outcomes (hematoma, recurrence, reoperation, hospital stay, operative time) were not notably affected by single-study exclusion.

The risk of bias for non-randomized studies was evaluated using the ROBINS-I tool, covering seven domains. As illustrated in Fig. 12, most studies showed a low to moderate risk of bias. No study was assessed to have a serious or critical risk across any domain. The most frequent sources of moderate risk were related to participant selection and intervention classification, underscoring the inherent limitations of observational data despite overall acceptable quality.

Using the AMSTAR 2 tool, the methodological quality of our meta-analysis was evaluated across 16 domains. The overall score achieved was 13 out of 16, corresponding to 81%, which reflects a high methodological quality. According to established AMSTAR 2 interpretation criteria, a score exceeding 70% is considered indicative of robust methodological rigor. This result supports the reliability of our systematic approach and strengthens the validity of the findings presented in this meta-analysis.

The quality of evidence for all primary and secondary outcomes was assessed using the GRADE approach. Overall, the certainty of evidence was rated as high for most clinical outcomes, including hernia defect closure rates, hematoma formation, recurrence, reoperation, seroma formation, surgical site infection, and operative time across both r-IPOM and r-TAPP groups. Console time and hospital stay outcomes were graded as high certainty, despite moderate inconsistency (I² >50%) noted across some analyses. No outcomes were downgraded due to serious risk of bias, indirectness, or imprecision. These results suggest that the findings of this meta-analysis are supported by robust and reliable evidence, as shown in Table 5.

Our meta-analysis is among the first to systematically synthesize a comparison between r-IPOM/rIPOM + and r-TAPP in hernia repair using a comprehensive approach, integrating subgroup and meta-regression analyses. By applying rigorous sensitivity analyses, we ensured the robustness of our conclusions, addressing potential sources of heterogeneity with appropriate statistical adjustments.

Several limitations must be acknowledged. Firstly, high heterogeneity was evident in some outcomes, especially operative time and hospital stay, reflecting variations in surgical technique, patient populations, and institutional practices, including mesh types. Although random-effects models and subgroup analyses were employed, this statistical heterogeneity cannot be fully eliminated. Secondly, the reliance on predominantly retrospective studies introduces an inherent risk of bias, and the short follow-up periods limit assessment of long-term outcomes such as chronic pain and late recurrence. Additionally, the lack of uniform surgical definitions and protocols across studies poses another challenge to direct comparisons.

Moreover, although postoperative pain and adhesion formation are highly relevant clinical outcomes, especially when comparing r-IPOM/+ and r-TAPP approaches, these endpoints could not be reliably assessed in our meta-analysis due to inconsistent and incomplete reporting across the included studies. Pain was only explicitly measured in one study (Bindal et al. 2024) using VAS scores, where the r-IPOM/+ group reported significantly more pain. Two additional studies (Gockal et al., 2024; Baur et al., 2021) referenced postoperative pain qualitatively but lacked objective measurement scales. Adhesion formation was even less consistently reported, with only Bindal et al. noting the absence of adhesions during a one-year follow-up. This underreporting, likely due to the variability in follow-up durations and the subjective nature of pain assessments, limited the inclusion of these parameters in our primary outcomes.

One important limitation of our analysis was the inconsistency and heterogeneity in mesh reporting across studies. Therefore, we omitted mesh-based subgroup analyses to avoid misleading interpretations. This reflects a broader issue in robotic hernia surgery literature: despite technical advances, outcome reporting remains inconsistent. Establishing standardized mesh classification and reporting guidelines is crucial to improving scientific rigor and comparability.

Finally, none of the included studies reported on patient satisfaction, highlighting a critical gap in the current literature and the need for future studies to incorporate standardized patient-reported outcome measures.