Fascial closure techniques at stoma reversal: a systematic review and network meta-analysis of randomized controlled trials Authors: E Meyer, J Douissard, E Liot, G Meurette, F Ris, J Meyer Published: 2026 ABSTRACT AbstractBackgroundIncisional hernia after stoma reversal is a frequent complication, yet the optimal method for fascial closure remains uncertain. This network meta-analysis (NMA) synthesized randomized evidence comparing closure techniques and prophylactic mesh reinforcement.MethodsA systematic search of MEDLINE (PubMed) and CENTRAL (Cochrane) was performed through October 2025. Randomized controlled trials in adults comparing fascial closure strategies during stoma reversal reporting on the incidence of incisional hernia and/or surgical site infection were included. Random-effects frequentist and Bayesian NMAs were conducted using MetaInsight powered by R packages netmeta, gemtc, BUGSnet, and bnma. Treatment effects were expressed as odds ratios (OR) and risk differences (RD) with 95% confidence or credible intervals.ResultsSix RCTs compared standard closure, resorbable mesh, and non-resorbable mesh. Both mesh types reduced incisional hernia compared with no mesh (non-resorbable mesh: OR 0.13 [0.03–0.49]; resorbable mesh: OR 0.54 [0.35–0.84]) without increasing surgical-site infection. Bayesian results and SUCRA rankings were consistent.ConclusionFascial reinforcement, particularly with non-resorbable mesh, significantly reduces the risk of stoma-site incisional hernia without increasing infection rates. These findings support prophylactic reinforcement in selected patients.Supplementary InformationThe online version contains supplementary material available at 10.1007/s10151-026-03315-7.Keywords: Stoma, Ileostomy, Colostomy, Closure, Surgical site infection, Mesh Introduction Incisional hernia (IH) following stoma reversal is a frequent and clinically relevant complication, occurring in up to one-third of patients. It often leads to chronic discomfort, aesthetic concerns, and even reoperation. Although various closure techniques have been explored, the optimal method for fascial closure during stoma reversal remains unclear. In contrast to skin closure, which has been extensively studied and standardized, fascial closure strategies have only recently been subjected to rigorous randomized evaluation. Over the past decade, several randomized controlled trials (RCTs) have assessed whether prophylactic mesh reinforcement at the stoma site can reduce the incidence of IH without increasing the risk of surgical site infection (SSI). These studies used different suture methods (e.g., Duramesh), mesh materials (synthetic, biological, or biosynthetic), and placements (onlay, sublay, or retromuscular), producing heterogeneous results. Previous systematic reviews have addressed aspects of stoma-site closure, but randomized evidence comparing mesh types and closure strategies remains fragmented, underscoring the need for a unified synthesis. We therefore performed a systematic review and network meta-analysis (NMA) to synthesize all available RCT evidence comparing fascial closure techniques after stoma reversal, with a particular focus on the incidence of IH and SSI. Methods A systematic search was conducted in MEDLINE (via PubMed) and CENTRAL (Cochrane Library). Given that the present network meta-analysis was restricted to randomized controlled trials, the search focused on MEDLINE and CENTRAL, which together capture the vast majority of indexed surgical RCTs. The search covered all available records from database inception through October 2025. The search combined controlled-vocabulary and free-text terms related to stoma, ileostomy, colostomy, closure, fascia, suture, and mesh, and was restricted to RCTs in adults (Table S1). Grey literature and conference abstracts were also screened and considered for eligibility. We included trials enrolling adult patients undergoing ileostomy or colostomy closure and reporting on IH and/or SSI at the stoma site. Non-randomized studies, reviews, studies without extractable outcome data, and reports with insufficient follow-up were excluded. Two reviewers independently screened all titles, abstracts, and full texts using the Rayyan web platform (Qatar Computing Research Institute, Doha, Qatar), with disagreements resolved by consensus. The methodological quality of all included randomized controlled trials was evaluated using the Cochrane Risk of Bias 2.0 (RoB2) tool. Results of the assessment are summarized in Table S2. The quantitative synthesis was carried out using MetaInsight (Owen et al., Res Syn Meth 2019;10:569–581). Analyses were performed under both frequentist and Bayesian random-effects frameworks. All frequentist calculations were executed using the R package netmeta, which implements random-effects network models based on graph theory. Bayesian computations were performed through R packages gemtc, BUGSnet, and bnma. Results were expressed as odds ratios (OR) with 95% confidence intervals (CI) in the frequentist framework and as posterior ORs with 95% credible intervals (CrI) in the Bayesian framework. Risk differences (RD) were also calculated to provide absolute effects. Ranking probabilities were derived using the surface under the cumulative ranking curve (SUCRA). The no-mesh strategy served as the reference treatment in all analyses. Results Of 208 studies screened, six RCTs were kept for inclusion [1–6]. Study selection is detailed in the PRISMA flow diagram (Figure S1). Together, these studies enrolled more than 900 patients and compared three main fascial closure strategies: standard suture closure without mesh, reinforcement with a non-resorbable (synthetic) mesh, and reinforcement with a resorbable (biologic or biosynthetic) mesh (Table 1). The network connecting these interventions was complete, allowing both direct and indirect comparisons (Fig. 1). Notably, the trial by Katyar et al. did not report on the incidence of IH [3]. Risk-of-bias assessments for all trials are provided in Table S2. Table 1Characteristics of the randomized controlled trials included in the network meta-analysisStudy (year)Country/settingPatients, nAge, meanGender, maleBMI, meanInterventionComparatorPrimary outcome(s)Follow-up (months)ROCSS Trial (2020)International, multicentre65058.765%26.7Biological mesh (inlay)Standard closureIncisional hernia24Elzeftawy et al. [2]Egypt, single centre8852–53.654–56%34.7–35.5Synthetic mesh (retromuscular)Standard closureIncisional hernia9Katyar et al. [3]Pakistan, multicentre9562–6662–64%25.5–25.6Biological mesh (retromuscular)Synthetic mesh (retromuscular)SSI1Mäkäräinen et al. [4]Finland, multicentre9063–6662.8–66%25.9Biological mesh (retromuscular)Synthetic mesh (retromuscular)Incisional hernia10Olona et al. [6]Spain, multicentre (abstract)58–––Synthetic mesh (onlay)Standard closureIncisional hernia6Burghgraef et al. [5]Netherlands, multicentre7961–6659–65%26Synthetic mesh (retromuscular)Standard closureIncisional hernia12Summary of the six randomized controlled trials included in the network meta-analysis evaluating fascial closure techniques during stoma reversal. Interventions were grouped as standard suture closure, resorbable (biologic/biosynthetic) mesh reinforcement, and non-resorbable (synthetic) mesh reinforcementSSI surgical site infection Characteristics of the randomized controlled trials included in the network meta-analysis Summary of the six randomized controlled trials included in the network meta-analysis evaluating fascial closure techniques during stoma reversal. Interventions were grouped as standard suture closure, resorbable (biologic/biosynthetic) mesh reinforcement, and non-resorbable (synthetic) mesh reinforcement SSI surgical site infection Network geometry and pooled effect estimates for incisional hernia. Network plot showing the structure of available evidence across all included randomized controlled trials evaluating fascial closure techniques at stoma reversal. Each node represents an intervention (no mesh, resorbable mesh, non-resorbable mesh), and the thickness of connecting lines reflects the number of direct comparisons between treatments. The lower panel presents the random-effects frequentist network meta-analysis for the outcome of incisional hernia (5 randomized controlled trials), with odds ratios (OR) and 95% confidence intervals (CI) displayed relative to standard fascial closure without mesh (OR = 1) In the frequentist random-effects model, both mesh types significantly reduced the risk of IH compared with standard closure. The odds ratio for non-resorbable mesh versus no mesh was 0.13 (95% CI 0.03–0.49), while for resorbable mesh versus no mesh it was 0.54 (95% CI 0.35–0.84) (Fig. 1). When expressed as absolute effects, the risk difference was − 12% (95% CI − 17% to − 7%) for non-resorbable mesh and − 10% (95% CI − 14% to − 5%) for resorbable mesh compared with standard closure (Table 2). Table 2Summary of network meta-analysis results comparing fascial closure strategies at stoma reversalOutcomeComparisonModelOdds ratio (95% CI/CrI)Risk difference (95% CI)Incisional herniaNon-resorbable mesh vs no meshFrequentist0.13 (0.03–0.49) − 0.12 (− 0.17 to − 0.07)Resorbable mesh vs no meshFrequentist0.54 (0.35–0.84) − 0.10 (− 0.14 to − 0.05)Non-resorbable mesh vs no meshBayesian0.04 (0.001–0.37)–Resorbable mesh vs no meshBayesian0.34 (0.01–3.39)–Surgical-site infectionNon-resorbable mesh vs no meshFrequentist1.01 (0.50–2.02) + 0.02 (− 0.03 to + 0.08)Resorbable mesh vs no meshFrequentist1.31 (0.87–1.97) + 0.03 (− 0.02 to + 0.08)Non-resorbable mesh vs no meshBayesian1.04 (0.49–2.22)–Resorbable mesh vs no meshBayesian1.30 (0.70–2.50)–Random-effects network meta-analysis comparing non-resorbable (synthetic) and resorbable (biologic/biosynthetic) mesh reinforcement versus standard fascial closure without mesh at stoma reversal. Negative risk differences indicate reduction in event risk relative to no-mesh closureCI confidence interval; CrI credible interval Summary of network meta-analysis results comparing fascial closure strategies at stoma reversal Random-effects network meta-analysis comparing non-resorbable (synthetic) and resorbable (biologic/biosynthetic) mesh reinforcement versus standard fascial closure without mesh at stoma reversal. Negative risk differences indicate reduction in event risk relative to no-mesh closure CI confidence interval; CrI credible interval The Bayesian random-effects model yielded consistent directionality, with posterior OR = 0.04 (95% CrI 0.001–0.37) for non-resorbable mesh and OR = 0.34 (95% CrI 0.01–3.39) for resorbable mesh. Although credible intervals were wider because of the limited number of trials, both models ranked non-resorbable mesh as the most effective intervention. SUCRA probabilities were approximately 100% for non-resorbable mesh, 65% for resorbable mesh, and near 0% for no mesh. For the outcome SSI, neither the frequentist nor the Bayesian analyses demonstrated a statistically significant difference between groups. The odds ratio for non-resorbable mesh compared with no mesh was 1.01 (95% CI 0.50–2.02) and for resorbable mesh 1.31 (95% CI 0.87–1.97) (Table 2). Corresponding risk differences were small and non-significant (+ 0.02 and + 0.03, respectively) (Table 2). Bayesian posterior estimates were similar, with OR = 1.04 (95% CrI 0.49–2.22) for non-resorbable and OR = 1.30 (95% CrI 0.70–2.50) for resorbable mesh. Discussion This network meta-analysis provides an integrated overview of all randomized evidence comparing fascial closure strategies at stoma reversal. Both frequentist and Bayesian random-effects models demonstrated that prophylactic mesh reinforcement—particularly with non-resorbable synthetic material—markedly reduces the risk of stoma-site IH without increasing the risk of SSI. The relative risk reduction exceeded 80% for synthetic mesh and about 50% for biologic or biosynthetic mesh, translating into an absolute reduction of roughly 10–12%. From a practical perspective, the number needed to treat to prevent one hernia is approximately ten patients. These findings confirm and extend the conclusions of the pivotal ROCSS trial [1], which first established the feasibility and safety of biologic mesh reinforcement. More recent studies, including the PRELOOP trial by Mäkäräinen et al. [4], have shown that synthetic meshes placed in a retrorectus or sublay position achieve even greater protection with similarly low SSI rates. Our results therefore support the view that infection risk is not materially increased by mesh placement, provided that a meticulous closure technique and an appropriate anatomical plane are used. The congruence between frequentist and Bayesian frameworks reinforces the robustness of these conclusions despite the moderate sample size of the network. The wider credible intervals observed in the Bayesian model reflect the limited number of available studies and the inherently conservative nature of Bayesian estimation under sparse networks. Previous conventional meta-analyses have generally pooled different mesh types and study designs into a single comparison, limiting the ability to distinguish between synthetic and biologic materials. The present network meta-analysis, restricted to randomized trials, provides a more granular comparison by evaluating resorbable and non-resorbable meshes as separate interventions, thereby offering comparative rankings that were not available in earlier reviews. Ongoing randomized trials are expected to clarify the optimal mesh type and placement and will help strengthen the evidence. Studies using synthetic non-resorbable meshes, such as the PHaLIR [7] (NCT03720262), P.E.L.I.O.N. [8] (DRKS00027921), ILEOCLOSE [9], PROMISS (NCT03750942), PROMESH [10] (NCT04510558), and NCT06511414 trials, are evaluating retromuscular or sublay polypropylene reinforcement compared with standard closure. In parallel, smaller trials are exploring biologic and biosynthetic meshes, including the SCAR trial [11] and the ChiCTR2200064995 trial [12]. Practical barriers to implementation include surgeon experience with retromuscular dissection, concerns about mesh infection in contaminated fields, and institutional variability in mesh availability. Emerging robotic approaches may further standardize the dissection and retrorectus mesh placement during stoma reversal but require an intra-abdominal access. Limitations of this network meta-analysis include the modest number of available randomized trials, heterogeneity in mesh materials and anatomical planes, and the presence of a single-study treatment node, which reduces the strength of indirect comparisons. The sparse network structure increases statistical uncertainty, although consistency analyses did not reveal major conflicts between direct and indirect evidence. In this synthesis of six RCTs, fascial mesh reinforcement—particularly with non-resorbable synthetic material—significantly reduces the risk of stoma-site IH without increasing SSI. The benefit is both statistically and clinically relevant. These findings support the use of prophylactic fascial reinforcement at stoma closure in selected patients, pending confirmation from ongoing multicentre trials that will further delineate optimal mesh type and technique. Supplementary Information Below is the link to the electronic supplementary material. Supplementary file1 Figure S1. PRISMA inclusion flowchart (PNG 98 KB) Supplementary file2 (DOCX 16 KB) Supplementary file3 (DOCX 17 KB) Author contributions EM and JM conceived and designed the study. EM and JM collected the data. All authors interpreted the data, contributed to the writing of the draft manuscript, and approved the final version of the manuscript. Funding None. Data availability No datasets were generated or analysed during the current study. Declarations Conflict of interest The authors declare no competing interests. Ethical and Informed Consent Ethical clearing was not required. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary file1 Figure S1. PRISMA inclusion flowchart (PNG 98 KB) Supplementary file2 (DOCX 16 KB) Supplementary file3 (DOCX 17 KB) Data Availability Statement No datasets were generated or analysed during the current study.