Incisional hernia (IH) is a frequent complication after abdominal surgery and remains a major source of post-operative morbidity [1–4]. The incidence ranges from 5%–20% after laparotomy to approximately 1.7%–24.3% after laparoscopic procedures [4, 5]. Incidence can reach 30% in high-risk settings [1, 4]. IH commonly occurs after vascular, colorectal, bariatric, and upper gastrointestinal surgery [6]. Several patient- and procedure-related risk factors have been identified, including chronic diseases, obesity, emergency surgery, midline incision, and infection [7–9]. IH may present with chronic symptoms such as an abdominal bulge or discomfort, but may also lead to small or large bowel obstruction [4]. Diagnosis is primarily clinical, with the addition of ultrasound and CT scan to confirm the defect [10]. Several classification systems have been described, with the European Hernia Society classification for IH published in 2009 being the most widely used [11]. Management can either consist of a watchful waiting therapy or a surgical repair with fascial closure and reinforcement with a mesh [12]. The cost of IH repair was estimated at $15,899 per case in 2006 [13]. Although surgery for IH is nowadays well-established, prevention has become a major focus in recent years [1]. Prophylactic mesh reinforcement instead of abdominal closure by conventional suture-only has been proposed as a strategy to reduce the risk of IH [3]. Even if the small-bite suture technique is necessary to decrease the rate of IH, it is not sufficient, as the long-term follow-up of the STITCH-it trial revealed that 34% of patients still developed an incisional hernia [14]. Initially, the use of prophylactic mesh reinforcement was studied in patients undergoing open abdominal aortic aneurysm repair, followed by bariatric surgery, and finally colorectal and upper gastrointestinal surgery, including randomized controlled trials (RCTs) [2]. Growing evidence suggests that prophylactic mesh reinforcement significantly reduces the incidence of IH [2]. Several anatomical planes can be used to place a mesh (onlay, retromuscular, intraperitoneal); however, the optimal anatomical plane to prevent IH without increasing surgical site infection (SSI) or occurrence (SSO) remains a topic of ongoing debate [1, 3]. Therefore, this network meta-analysis (NMA) compares anatomical positions for prophylactic mesh reinforcement in elective and emergency settings to compare their relative efficacy and morbidity profiles.

27 RCTs [17–43] were included in the analysis. Among them, 23 RCTs (85%) were primary index publications and 4 (15%) were follow-up reports. 24 studies (89%) had a two-arm design, and 3 (11%) were three-arm comparisons. According to the Cochrane Risk Of Bias 2 (ROB-2) tool for RCTs, 15% of studies were judged to be at low risk of bias, 74% raised some concerns, and 11% were considered at high risk of bias. Included studies involved colorectal surgery (4%), Upper-GI/bariatric surgery (15%), vascular surgery (19%), and mixed surgical populations (63%). Procedures were elective (74%), emergency (15%), and mixed (11%). A total of 4327 patients were analyzed, with a median follow-up of 24 months. Of these, 1891 patients were allocated to the suture-only group, and 2436 to the prophylactic mesh group. In the suture group, fascial closure was achieved with interrupted suture (7%), continuous small-bite (14%), and continuous large-bite (79%). Suture materials were non-absorbable (20%), quickly absorbable (8%), and slowly absorbable (72%). In the mesh group, mesh placements were onlay (38%), inlay-strip (7%), retromuscular/preperitoneal (37%), and intraperitoneal (18%). Inlay-strip consisted of the interposition of a mesh between the right and left fascia, which were all closed together. Meshes were absorbable in 23% and non-absorbable in 77%. A summary of the included studies is available in Supplementary Appendix 2.

This systematic review and network meta-analysis assesses the effectiveness of prophylactic mesh reinforcement after midline laparotomy compared with suture-only fascial closure in both elective and emergency surgeries. In addition, the NMA enabled indirect comparisons between mesh positions to assess which anatomical plane may provide the most favorable balance between efficacy and morbidity. To our knowledge, this is the largest NMA on this topic to date, including 27 RCTs and 4327 patients. This study extends the previous NMA by Tansawet et al. (2020), which included 20 RCTs until 2019 [4]. They found that onlay mesh and retromuscular prophylactic augmentation were effective for preventing IH. In contrast, the present study incorporates a larger cohort of patients, more recent publications, an emergency surgery population, and an additional analysis of wound morbidity comparing the risk of SSO and seroma. Our analyses provide a more comprehensive and clinically relevant assessment of prophylactic mesh reinforcement in the current surgical practice.

Regarding IH prevention, this study supports prophylactic mesh reinforcement after a midline laparotomy compared with suture-only. The benefit is verified through multiple RCTs and MAs [1–4]. Guidelines from the European Hernia Society in 2022 already suggested that prophylactic mesh reinforcement should be considered for elective laparotomy [44]. However, uncertainty has persisted regarding the optimal anatomical plane for mesh placement. A recent meta-analysis by Lagger et al. found a significant decrease in IH with the use of retromuscular prophylactic mesh reinforcement in elective settings [45]. The I-PREVENT-AAA trial also described that retromuscular and onlay prophylactic mesh were effective for incisional hernia prevention [46]. In our study, both onlay and retromuscular/preperitoneal reinforcement were associated with a significant reduction of IH. Although onlay mesh achieved the highest P-score for IH prevention, this ranking should not be over-interpreted as definitive evidence of superiority over other mesh positions. Considering the heterogeneity of the included studies and the indirect nature of several comparisons, onlay mesh may represent an appropriate option for prophylactic mesh reinforcement in both elective and emergency settings, even if emergency NMA should be interpreted with caution, given the star-shaped network model and the inability to assess local inconsistency. These results are encouraging from a clinical perspective, as onlay mesh reinforcement offers several practical advantages. Technically the onlay mesh placement is a relatively easy and rapid intervention and it has the advantage to leave the retromuscular plane intact for further incisional repair should one occur. While the retromuscular mesh placement offers an excellent anatomical position used by many abdominal wall surgeons to repair incisional hernia, it requires greater expertise, involves a learning curve and more extensive dissection, with potential risks of peritoneal breach, bleeding, or injury to the neurovascular structures [44, 47, 48]. This complexity cannot be performed without prolonging the ongoing surgical procedure, without a formal teaching or a dedicated surgical team to close the abdominal wall. Although the present NMA did not assess operative time, cost, or resource utilization, these technical differences may influence the reproducibility and feasibility of prophylactic mesh reinforcement in routine surgical practice.

The occurrence of SSI remains one of the principal concerns limiting wider adoption of mesh prophylaxis among surgeons. However, several previous meta-analyses demonstrated no significant increase in SSI after mesh reinforcement [1, 3, 4]. Our study confirms these findings, as no statistically significant increase in SSI was noted after onlay mesh or retromuscular/preperitoneal mesh reinforcement. Altogether, these data suggest that mesh reinforcement is a safe and effective technique for IH prevention without significantly increasing SSI.

SSO remains another important consideration after prophylactic mesh implantation, although not all anatomical planes appear to have the same risk profile. In our study, pairwise MA showed a significant increase in SSO following mesh reinforcement compared with suture-only. NMA further identified onlay mesh as the principal risk factor for SSO, particularly in elective settings, whereas this risk was not confirmed in emergency surgeries. These results, nevertheless, should be interpreted with caution as the network model showed global inconsistency, showing disagreement between direct and indirect evidence. This could be explained in part by the heterogeneous definition of the SSO across the different studies, leading to the global inconsistency observed in the SSO network. Regarding seroma, the association between seroma and onlay mesh appeared more robust, as no global or local inconsistencies were detected. To date, the literature has reported conflicting results, with studies reporting higher seroma rates after onlay mesh reinforcement, while others have found no significant differences [1, 3, 4].

Our data suggest that onlay prophylactic mesh reinforcement provides an effective strategy for IH prevention without a significant increase in the SSI rate, but at the cost of increased wound morbidity, primarily due to seroma formation. This is probably linked to an extensive dissection of the subcutaneous space, leaving space for the development of a seroma, as mentioned in some studies [49]. Interestingly, the use of absorbable mesh did not statistically increase the risk of SSO. Further studies are needed to clarify the impact of the type and the size of onlay mesh reinforcement on SSO incidence.

A major strength of the present study is the use of RCTs only to reduce selection bias and inclusion of emergency surgeries. The use of NMA analysis further enabled statistical comparison and ranking of the different anatomical planes of mesh placement, providing insights beyond those achievable with conventional pairwise MA alone. Finally, for the vast majority of primary and secondary outcomes, no significant global or local inconsistency was detected, supporting the overall robustness of the network estimates.

Mortality was listed as a secondary outcome. However, reporting was inconsistent across RCTs and was not meaningful for a statistical analysis.

Several limitations should also be acknowledged. Clinical heterogeneity was present among the included studies, with a mixed population undergoing vascular, bariatric, and colorectal surgeries. Heterogeneity also existed in surgical indications, mesh materials, follow-up duration, and baseline IH risk. The substantial heterogeneity observed in the pairwise meta-analysis of IH (I² = 69.5%) could be explained by differences in surgical population, baseline IH risk, closure technique, mesh characteristics, and follow-up duration. Meta-regression to further explore these potential sources of heterogeneity was not performed because of the limited number of studies within individual covariate categories. Variability in closure techniques was also identified, as most suture-only closures were performed with continuous large-bite, in contradiction with the recent guidelines advocating small-bite closures. This may have resulted in higher baseline IH rates in the control group and could have amplified the apparent benefit of prophylactic mesh reinforcement. As a substantial proportion of the evidence supporting prophylactic mesh reinforcement derives from trials using large-bite closure, which no longer represents the current standard of care, the magnitude of the observed benefit may not be directly transferable to contemporary clinical practice using small-bite closure. Differences in mesh materials, size, weight, pore size and fixation techniques may have influenced both efficacy and morbidity outcomes, thereby limiting comparisons between studies. Variation in follow-up duration may also have influenced IH detection, as longer follow-up provided more opportunity to identify late-occurring hernias. Clinical heterogeneity may also challenge the transitivity assumption underlying the NMA. Potential effect modifiers differed across trials, including surgical setting and indication, contamination status, baseline IH risk and obesity prevalence, mesh characteristics and fascial closure technique. These differences could influence relative treatment effects and, therefore, the validity of indirect comparisons. Additional heterogeneity was related to outcome definitions, particularly for endpoints such as SSO or seroma, which were not uniformly reported across trials. Most included studies were judged to raise some concerns according to the Cochrane ROB-2 tool, while only a minority were considered at low risk of bias. Furthermore, some network comparisons demonstrated significant global inconsistency, notably for SSO, and local inconsistency could not always be assessed because of a star-shaped network configuration with few direct head-to-head trials between mesh planes. Therefore, several estimates should be interpreted with appropriate caution. Treatment rankings based on P-score should also be interpreted cautiously, as statistical rankings do not equal clinical superiority. The certainty of evidence was not formally assessed using CINeMA, which represents another limitation. Confidence in the network estimates likely varies across outcomes. For IH, the absence of global or local inconsistency supports the consistency of the network estimates. However, substantial heterogeneity, limited direct evidence, and small sample sizes in some treatment nodes warrant caution. Confidence is lower for outcomes such as SSO, for which significant global inconsistency was detected, and for subgroup analyses based on sparse or star-shaped networks. These limitations should be considered when interpreting treatment effects and rankings.

Interpretation of the inlay node also warrants caution. Only one RCT contributed to this analysis, and the surgical technique did not represent a conventional inlay mesh reinforcement. Instead, a narrow mesh strip was incorporated within the suture, acting as a reinforced suture rather than a true bridge.

Another potential limitation is the decision to combine retromuscular and preperitoneal mesh placements into a single anatomical group. Retromuscular and preperitoneal planes are anatomically and biologically different. These differences imply distinct tissue interfaces and technical requirements for dissection and mesh placement. Therefore, these two approaches should not necessarily be considered biologically or technically interchangeable, and combining them into a single node may have obscured potential differences in efficacy or morbidity. A sensitivity analysis separating retromuscular (8 studies) and preperitoneal (4 studies) was performed. Both planes remained associated with a significant reduction in IH risk, especially preperitoneal, with results comparable to onlay mesh. However, this node included only 112 patients, compared with 745 for the retromuscular node and 926 for the onlay mesh node. Moreover, estimates for secondary outcomes became less precise, with wider confidence intervals, and no emergency surgery study evaluated preperitoneal placement. Therefore, we grouped the retromuscular and preperitoneal planes as a common posterior plane for primary analysis, as separating these planes into two groups would have substantially fragmented the available evidence. Moreover, given the limited number of RCTs, constructing a meaningful network model, especially for subgroup analysis (emergency vs. elective surgery), would have been compromised.

In conclusion, prophylactic mesh reinforcement effectively prevents incisional hernia after midline laparotomy in both elective and emergency settings, without increasing surgical site infection. Both onlay and retromuscular/preperitoneal placement reduced IH risk. Onlay achieved the highest treatment ranking but should not be regarded as definitively superior, given the heterogeneity and the limited direct evidence between planes. Its practical advantages—technical simplicity and preservation of the retromuscular plane for future repair—make it an appropriate option, at the cost of higher seroma rates. Further high-quality randomized trials are needed to define the optimal mesh plane for prophylactic reinforcement.