Ventral Hernia Repair: A Meta-Analysis of Randomized Controlled Trials

Abstract

Background:

From the patient's perspective, a ventral hernia (VH) can cause pain, adversely affect function, increase size, cosmetically distort the abdomen, and incarcerate/strangulate abdominal contents. The only known “cure” for a VH is surgical repair. The aim of this study was to review systematically the published randomized controlled trials (RCTs) of the surgical care of VH.

Method:

A search of the Ovid, MEDLINE, EMBASE, and Cochrane databases was performed to obtain reports of RCTs on the use of mesh reinforcement in abdominal wall hernia repair. The outcomes assessed were hernia recurrence and surgical site infection (SSI). The overall quality of evidence was assessed using Grading of Recommendations Assessment, Development, and Evaluation (GRADE). When feasible, based on assessment of heterogeneity, data were pooled and analyzed in a meta-analysis.

Results:

Of the 10,349 titles screened, 25 articles (23 studies) met the search criteria. Evidence indicates that mesh reinforcement in clean cases can decrease hernia recurrence (number needed to treat = 7.9) but increase risk of SSI (number needed to harm = 27.8). Placing mesh in the sublay position (as opposed to the onlay or underlay position) may decrease the risk of hernia recurrence and SSI.

Conclusions:

Mesh reinforcement is recommended for all VH repairs in a clean case (high grade of evidence). Sublay mesh location may result in fewer recurrences and SSIs than onlay or inlay placement, but further study is needed to confirm this hypothesis (moderate grade of evidence).

Ventral hernias (VHs) are among the most common pathologies experienced by patients and evaluated by clinicians [1]. Ventral hernias can significantly affect an individual's quality of life by causing pain, impairing function such as activities of daily living, and creating a bulge that can be disfiguring [2,3]. In addition, patients with VH are at risk for emergency surgery because of incarceration and strangulation of the abdominal contents.

Elective VH repair poses substantial risks, such as infection, re-admission, recurrence, and re-operation [4 –7]. Mesh reinforcement is being used increasingly, although it may carry a higher risk of surgical site infection (SSI). In particular, mesh infections and mesh erosion (e.g., chronic wound or enterocutaneous fistula) can follow mesh placement. A number of factors may affect these complication rates, including the position and site in which the mesh is placed and the type of mesh utilized. It is widely accepted that mesh should be used for VH repair in a clean setting; however, the type of mesh that should be utilized and where the mesh should be placed are still debated (Fig. 1) [4 –10].

FIG. 1.

Current sources of practice heterogeneity.

The societal and healthcare ramifications of these decisions can be substantial. Each year, the United States spends more than $3.2 billion on the operative management of nearly 400,000 VHs in adult patients [11]. The cost of different meshes is highly variable and may eclipse the expected cost of a typical operation. Short-term and long-term complication rates after VH repair remain high, with estimated SSI and hernia recurrence rates each being reported in the range of 25%–50%. Whereas on average, an SSI costs the healthcare system $28,000, a prosthetic infection imposes a staggering $100,000 cost, and a mesh erosion with enterocutaneous fistula is reported to cost $220,000 [12,13]. Current management of VHs funnels a subgroup of patients into a vicious cycle of cost and complications with VH repair, infection or recurrence, and re-operation [14]. There is a need to identify the best practice guidelines to balance the risk of SSI and hernia recurrence in various clinical settings.

Our objective was to perform a systematic review to identify randomized controlled trials (RCTs) of VH repair to answer three key questions:

1. When is mesh reinforcement indicated during VH repair?

2. What type of mesh is recommended for VH repair?

3. Where should mesh be placed during open VH repair?

The first key question will serve as a control to show that our methods are valid, as it is widely accepted that mesh reinforcement results in fewer recurrences with a slight increase in SSI. We aimed to provide answers to the second and third questions with this review and meta-analysis.

Methods

Search strategy

A review of the Ovid, MEDLINE, EMBASE, and Cochrane databases was performed by the authors on December 1, 2015. The search terms were “ventral hernia” (MeSH term includes incisional hernias and primary VHs such as umbilical hernias), and the search was limited to “randomized controlled trials.” References from selected articles were reviewed for additional papers not found in the literature search. Inclusion criteria included all completed RCTs on VH repair among adult patients in which there was reporting of incision infection, SSI, or hernia recurrence. Non-English-language studies were translated and evaluated in the same fashion as English-language studies. For Key Question 1, all studies comparing suture-only and mesh repair were included [15 –48]. For Key Question 2, all studies comparing two or more mesh types were assessed for inclusion. Mesh type was described by material and density, the latter being defined as low- (35 g/m²), mid- (35–70 g/m²), high- (70–140 g/m²), or super high- (>140 g/m²). Studies with a conflict of interest, with the manufacturing company of the mesh being studied without blinding of the outcomes assessment, were excluded. For Key Question 3, all studies with at least one arm assessing open VH repair with mesh placed in the onlay, sublay, or underlay position were included. Comparisons with laparoscopic underlay (hereinafter referred to as laparoscopic) or suture repair were used as controls. Exclusion criteria were failure to define the location in which meshes were placed or the use of mixed techniques without reporting the stratified results.

Methods to appraise the literature

The methodologic qualities of the studies were evaluated using the guidelines from the Critical Appraisal Skills Programme (CASP), and the overall quality of the evidence was assessed using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) criteria [49,50]. Studies were evaluated by two authors and disagreements resolved by a third author. A CASP score of >5 was considered a low risk for bias.

Data extracted from studies included interventions, number of patients, age, body mass index (BMI), hernia type and size as defined by the European Hernia Society, mesh material and density (where applicable), follow-up duration, and elective/emergency nature of the operation. These factors were chosen because they represent the most important variables for classifying VHs, assessing patient risk, and treatment of patients. The outcomes assessed were hernia recurrence and SSI.

Statistical analysis

Studies were evaluated for clinical homogeneity and appropriateness for data synthesis. Meta-analysis was performed to compare all direct treatment using fixed-effects and random-effects models. Statistical heterogeneity was assessed using I², where any value >50% was considered heterogeneous, and the heterogeneity p value, where a value <0.1 indicated significant statistical heterogeneity. In cases of statistical heterogeneity, the random-effects model was applied; otherwise, the results of the fixed-effects model are reported [51,54]. Sensitivity analyses selected a priori included only studies at low risk for bias, and studies evaluating the most commonly used mesh type (typically polypropylene), primary VHs, and incisional hernias.

A multiple-treatment meta-analysis was adapted from the guidelines developed by the International Society for Pharmacoeconomics and Outcomes Research Task Force to compare the different VH repair methods [51,54]. Unlike a traditional meta-analysis, which considers only studies with direct comparisons, a multiple-treatment, or network, meta-analysis considers studies with multiple competing treatments and combines direct and indirect evidence to determine the most effective treatment [51]. A network of treatment was created to map the relations between available studies. A multiple-treatment meta-analysis was performed using onlay as the reference treatment to compare all repair types by estimating the fixed-effects and random-effects models in the Bayesian framework. The Bayesian approach estimates the odds ratio, the 95% credible interval (CI), and the probability that each repair is the best using the Markov-chain Monte Carlo (MCMC) technique [55,56]. The CI is the Bayesian probability interval that is similar in use to the frequentist confidence interval; however, the CI also incorporates information from the prior distribution, which can be thought of as existing data or information, into the estimate, whereas confidence intervals are based solely on the existing data. Model fit was assessed by the deviance information criterion (DIC). A heterogeneity p value was calculated to determine statistical heterogeneity. For a heterogeneity p value <0.05, a random-effects model was used. Otherwise, a fixed-effects model was used [56]. Analysis was performed with the statistical software R version 2.15.0 [58].

Results

Search results

A total of 10,349 unique titles were retrieved, of which 25 (23 studies; 25 articles) were included. The overall search strategy is shown in Fig. 2.

FIG. 2.

PRISMA flow diagram of search results.

All included articles and key articles excluded are reported in Table 1. Additional study details are reported in Supplementary Table 1.

Table 1.

Articles Included and Key Articles Excluded

	Key questions
Series	1	2	3	Study summary and reason(s) for exclusion	CASP score
Abdel-Baki et al. (2007) [19]	I	NA	I	A small, single-center RCT of emergency repair of small/medium ventral hernias. Suture repair was compared with onlay high-density polypropylene mesh. Major concerns include limited description of how patients were allocated.	5
Abo-Ryia et al. (2015) [32]	NA	NA	I	A small single-center trial comparing onlay and sublay repair of primary ventral hernias. This study suffered from inadequate reporting of the randomization strategy, failure to blind, and inadequate power.	5
Afifi (2005) [33]	NA	NA	I	A small, single-center RCT of recurrent incisional hernias repaired with onlay or underlay of HDPP. This study suffered from a small sample (i.e., under-powered) and inadequate description of patients and their hernias (e.g., size).	3
Ammar (2010) [20]	I	NA	I	A small, single-center RCT of emergency umbilical hernia repairs in cirrhotic patients. Suture repair was compared with high-density polypropylene mesh. Major concerns include eight exclusions after randomization (patients with contamination were dropped from the analysis).	4
Arroyo et al. (2001) [21]	I	NA	E	A large, single-center RCT of low/moderate-risk patients undergoing umbilical hernia repair. Suture repair was compared with inlay high-density polypropylene mesh. Concerns include limited description of patient comorbidities, in particular obesity. Exclusion of KQ3: Assessed inlay mesh position was excluded because there is no equipoise in its use among surgeons	7
Barbaros et al. (2007) [34]	NA	NA	I	A small single-center study comparing laparoscopic with open ventral hernia repair with onlay repair using HDPP. The study was not blinded, the types of hernias (primary vs. incisional) were not balanced, the patients and outcomes were poorly described, and the study was underpowered.	4
Bensaadi et al. (2014) [30]	NA	E	NA	A single-center RCT of elective open repair of small ventral hernias comparing two mesh types (HDPP and PP/PTFE). Exclusion of KQ2: Concerns about this study include the high risk of bias and that the company producing the “superior” mesh funded it. The study was not blinded, the randomization was underpowered, and the outcomes were subjective.	7
Bessa et al. (2015) [35]	NA	NA	I	A small single-center study comparing laparoscopic with open ventral hernia repair with sublay repair using HDPP. This study suffered from inadequate reporting of randomization strategy, was not blinded, was inadequately powered.	5
Carbajo 1998	NA	NA	E	Exclusion of KQ3: Did not define mesh location with open ventral hernia repair group.	NA
Conze 2005	NA	I	NA	A large, multi-center RCT of elective open ventral incisional hernia repair comparing two mesh types (HD and LDPP). The study was underpowered because of poor recruitment; however, to date, this represents the best RCT comparing two mesh types. It is an effectiveness trial, as a variety of heavyweight meshes were used, and fixation was institution dependent.	10
De Vries Reilingh et al. (2007) [15]	I	NA	I	A small, single-center RCT of ventral incisional hernia repair. Suture repair was compared with polytetrafluoroethylene (PTFE) placed in an underlay position. The PTFE product is a super-high-density mesh. Concerns include limited description of the randomization, two exclusions after randomization (patients with contamination were dropped from the analysis), small sample, and relevance, as PTFE is being abandoned, in particular for open ventral hernia repair.	6
Eker et al. (2013) [36]	NA	NA	I	A large multi-center trial comparing laparoscopic with open sublay repair. The mesh type was not reported. This study suffered from inability to blind and heterogeneity in the mesh type used. Nevertheless, despite its deficiencies, this is one of the better studies.	7
Itani et al. (2010) [37]	NA	NA	I	A large, multi-center Veterans Affairs trial comparing laparoscopic and onlay repair with HDPP mesh. This study was stopped early because of safety concerns, given the high rate of incision complications with open repair. This study is one of the best trials assessing laparoscopic ventral hernia repair.	10
Grubnik 2014 [52]	NA	E	NA	A small trial comparing laparoscopic repair using two types of mesh and two mesh fixation techniques. Concerns include a small sample and lack of blinding. Exclusion of KQ2: This trial was excluded because of multiple interventions. It thus was difficult to determine if the effect was attributable to mesh type or to mesh fixation.	7
Kaafarani et al. (2009/20100 [44]	NA	NA	E	Exclusion of KQ3: Study patients have been reported previously.	NA
Korenkov et al. (2002) [22]	I	E	I	A multi-center RCT of repair of small/medium ventral incisional hernias. Suture repair was compared with two interventions: Onlay of high-density polypropylene mesh or autodermal reinforcement (biologic mesh). Despite the short follow-up (just over one y), a difference in outcomes was noted between suture and synthetic mesh. Exclusion of KQ2: Concerns about this study include its limited follow-up and inadequate power. However, the greatest weakness is its lack of generalizability; use of autodermal mesh is not routine and is irrelevant in today's practice.	10
Lal et al. (2012) [16]	I	NA	I	A small single-center RCT of elective ventral incisional hernias. Suture repair was compared with sublay repair using high-density polypropylene mesh. Concerns include limited description of randomization, small sample, and limited description of the study population.	4
Luijendijk et al. and Burger et al. (2000/2004) [23,24]	I	NA	I	A large, multi-center RCT of elective small/medium ventral incisional hernia repairs. Suture repair was compared with sublay repair using high-density polypropylene mesh. The authors reported the longest follow-up to date of a ventral hernia RCT demonstrating stunning rates of hernia recurrence. This study is considered the apex of ventral hernia outcomes research.	10
Malik (2015) [53]	NA	NA	E	Violated principles of randomization by analyzing failed laparoscopic repairs as “open repairs.”	NA
Misra et al. (2006) [38]	NA	NA	I	A small single-center trial comparing laparoscopic and sublay repair using HDPP mesh. This study had the limitation of inability to blind.	8
Moreno-Egea et al. (2013) [31]	NA	E	NA	A single-center, single-surgeon RCT that assessed laparoscopic repair of small incisional hernias using two mesh types (polyethylene and LDPP). Exclusion of KQ 2: Concerns about this study include its generalizability and effect on modern practice. Reporting an unheard of 0 infection and 0 hernia recurrence rates in more than 100 repairs. No differences were noted between the meshes.	7
Moreno-Egea et al. (2002) [39]	NA	NA	I	A small single-center trial comparing laparoscopic and open sublay repair of Spigelian hernias using HDPP mesh. This study had the limitations of inability to blind, inadequate power, and no reported SSIs or recurrences.	7
Navarra et al. (2007) [17]	NA	NA	I	A small single-center trial comparing laparoscopic and open sublay repair using LDPP mesh. This study had the limitations of inability to blind, inadequate power, and unusually low reported rates of complication.	5
Othman 2012	NA	NA	E	Exclusion of KQ3: Did not define mesh location in open ventral hernia repair group.	NA
Olmi et al. (2007) [48]	NA	NA	E	Exclusion of KQ3: Mixed mesh location with open ventral hernia repair group.	NA
Pawlak et al. (2015) [18]	NA	E	NA	A single-center RCT that assessed laparoscopic repair of ventral hernias using two meshes (MDPP and LDPP) and fixation devices. Exclusion of KQ2: Although this study was well designed and blinded, the investigators utilized two interventions (mesh type and fixation type) without having a factorial design, which makes it difficult to determine if the effect was attributable to the mesh type or to the fixation type.	10
Polat et al. (2005) [25]	I	NA	I	A small, single-center RCT of umbilical hernia repairs. Suture repair was compared with two controls: A three-dimensional mesh and onlay mesh, both of high-density polypropylene. Concerns include poorly described randomization schema and small, diluted sample.	7
Pring et al. (2008) [40]	NA	NA	I	A small single-center trial comparing laparoscopic and open underlay repair with PTFE. The study had limitations of inability to blind and inadequate power.	4
Rickert et al. (2012) [29]	NA	I	NA	A multi-center RCT that compared two meshes (MDPP and LDPP) for open ventral hernia repair. A concern with this study is that the company producing the “superior” mesh funded it. However, the study was well designed and double blinded, minimizing the risk of bias.	9
Rogmark et al. (2013) [41]	NA	NA	I	A large multi-center trial comparing laparoscopic and open sublay repair with HDPP. This study was limited by the inability to blind.	7
Venclauskas et al. (2010) [42]	I	NA	I	A large single-center RCT of ventral incisional hernia repairs. Suture repair was compared with two controls: onlay and sublay low-density polypropylene. Limitations of this study include poorly described randomization schema.	7
Wéber et al (2002/2010) [27,28]	I	NA	I	A non-English-language large, multi-center RCT of small/medium ventral hernias. Suture repair was compared with sublay mesh repair, and onlay was compared with sublay. Limitations included poorly described randomization schema. Differences in the hernia recurrence rate were substantial.	7

CASP Score: out of 11.

Key Questions: 1. When is mesh reinforcement indicated during ventral hernia repair? 2. What type of mesh is recommended for ventral hernia repair? 3. Where should mesh be placed during open ventral hernia repair?

CASP = Critical Appraisal Skills Programme; E = exclude; HD = high density (polypropylene or polyethylene); HDPP = high-density polypropylene; I = include; LDPP = low-density polypropylene; MDPP = mid-density polypropylene; NA = not applicable; PP/PTFE = polypropylene and polytetrafluoroethylene composite.

Key question 1: When is mesh reinforcement indicated during VH repair?

There were ten RCTs and 12 papers that reported outcomes of suture-only vs. mesh-reinforced VH repairs (Table 2). The patients and techniques were similar in clinical characteristics, allowing us to combine and synthesize the data. Studies included both primary and incisional VHs. Patients were all adults, and morbidly obese patients were largely excluded. Funnel plots did not suggest significant publication bias, and Forest plots demonstrated a lower risk of recurrence but a higher risk of SSI using mesh Fig. 3.

FIG. 3.

Funnel and forrest plot of (A) recurrence rates and (B) surgical site infection among randomized controlled trials comparing suture-only to mesh reinforcement.

Table 2.

Suture-Only Versus Mesh Reinforcement Study Assessment and Outcomes

			SSI		Recurrence
Series	Hernia type	N	RR	95% CI	RR	95% CI	CASP score
Abdel-Baki et al. [19]	Mixed	42	0.67	0.12–3.59	0.11	0.01–1.94	5
Ammar [20]	Primary	80	1.60	0.52–4.93	0.09	0.01–1.50	4
Arroyo et al. [21]	Primary	200	0.67	0.11–3.90	0.09	0.01–0.69	7
de Vries Reilingh et al. [15]	Incisional	37	3.17	1.02–9.87	1.16	0.67–2.04	6
Korenkov et al.^a [22]	Incisional	100	7.65 5.86	0.43–137 0.29–117	0.63 1.18	0.15–0.69 0.32–4.29	10
Lal et al. [16]	Incisional	62	0.38	0.08–1.79	0.13	0.01–2.49	4
Luijendijk et al. and Burger et al.^b [23,24]	Incisional	181 3 y 7 y	8.07 14.9	0.42–154 0.86–262	0.50 0.68	0.31–0.92 0.49–0.95	10
Polat et al.^a [25]	Primary	50 3D onlay	0.94 1.13	0.06–13.9 0.08–16.6	0.19 0.23	0.01–3.68 0.01–4.35	7
Venclauskas et al.^a [26]	Incisional	161 onlay sub	3.79 0.54	0.84–17.1 0.05–5.77	0.95 0.09	0.47–1.92 0.12–0.67	7
Wéber et al. [27,28]	Mixed	364	–	–	0.30	0.17–0.52	7

Hernia type defined by the European Hernia Society: Primary, incisional, or mixed (both primary and incisional).

Two mesh groups and one suture group.

Initial study reported at 24 mos' median follow-up; subsequent study reported mesh infections at 70–80 mos.

CASP = Critical Appraisal Skills Programme; CI = confidence interval; RR = relative risk; SSI = surgical site infection.

SSI and Recurrence: Suture is the control, and mesh is the intervention.

Figures in boldface indicate statistically significant differences.

On meta-analysis, mesh is associated with a significantly lower risk of hernia recurrence (odds ratio [OR] 0.35; 95% CI 0.26 indicated 0.48, I² = 26.4%, heterogeneity p = 0.200) but a higher risk of SSI (OR 1.73; 95% CI 1.00–3.00), I² = 25.8%, heterogeneity p = 0.258). Sensitivity analyses demonstrate similar trends for all subgroups and are reported in Supplementary Table 2. The absolute risk difference of mesh vs. suture for recurrence is 12.6% (8.7% vs. 21.3%) and for SSI is 3.5% (8.6% vs. 5.1%). The number needed to treat (prevent a hernia recurrence) is 7.93, whereas the number needed to harm (cause an SSI) is 27.82.

Three studies reported cases with contamination caused by violation of the gastrointestinal tract, but only two studies reported the outcomes of these cases separately to allow analysis [16,20,21]. Fourteen of sixteen patients had adequate data to analyze the impact of contamination by intervention. Eight patients were repaired using suture, and one (12.5%) developed an SSI, whereas six patients were repaired using an onlay high-density polypropylene mesh, and three (50%) developed an SSI (p = 0.245; relative risk [RR] 4.00; 95% confidence interval 0.54–29.6).

Multiple RCTs demonstrated a slightly higher risk of infectious complications but a substantially lower risk of hernia recurrence with reinforcement using mesh. This finding has been well established, and the analysis of this question validates our methodology for the remaining two key questions. Mesh reinforcement is recommended during clean repairs of both primary and incisional VHs (the strength of this recommendation is high; the grade of recommendation also is high). No RCTs are availble to guide management of contaminated VHs.

Key question 2: What type of mesh is recommended for VH repair?

There were six RCTs that compared different types of mesh (Table 3); however four were excluded because of a significant risk of bias and lack of generalizability (see Table 1).

Table 3.

Mesh Type Study Assessment and Outcomes

				SSI		Recurrence
Series	Hernia type	N	Control vs. intervention	RR	95% CI	RR	95% CI	CASP score
Bensaadi et al. [30] Excluded	Mixed	83	HDPP vs. PP/PTFE	0.33	0.02–7.77	0.11	0.01–1.95	7
Conze et al. [14]	Incisional	165	HD mesh vs. LDPP	1.14	0.58–2.24	2.31	0.93–5.71	10
Korenkov et al.^a [22] Excluded	Incisional, S/M L	67 60	HDPP vs. Autodermal	0.7 2.67	0.14–3.54 0.56–12.71	1.86 1.07	0.45–7.65 0.23–4.87	9
Moreno-Egea et al. [31] Excluded	Incisional	102	Polyethylene vs. LDPP	UC	UC	UC	UC	7
Pawlak et al. [18] Excluded	Mixed	50	MDPP vs. LDPP	–	–	11.0	0.64–189	10
Rickert et al. [29]	Incisional	80	MDPP vs. LDPP	^b	^b	^b	^b	9

CASP Score out of 11.

Hernia type defined by the European Hernia Society: Primary, incisional, or mixed (both primary and incisional).

Two mesh groups stratified by size.

Unable to calculate because there were no recurrences and only one SSI.

CASP = Critical Appraisal Skills Programme; CI = confidence interval; HD = high-density (polypropylene or polyethylene); HDPP = high-density polypropylene; L = large (>10 cm); LDPP = low-density polypropylene; MDPP = mid-density polypropylene; PP/PTFE = polypropylene and polytetrafluoroethylene composite; RR = relative risk; S/M = small/medium (<10 cm width); SSI = surgical site infection; UC = unable to calculate as neither group had a reported hernia recurrence or infection rate.

Based on the two included studies, low-density mesh (colloquially termed “light-weight mesh”) may be associated with higher odds of hernia recurrence compared with high-density mesh; however, the evidence is limited. No cases of contamination were reported in either of the studies. High-density mesh (compared with low-density mesh) is recommended for VH repair in a clean setting (the strength of the recommendation is weak; the grade of the recommendation is low).

Key question 3: Where should mesh be placed during open VH repair?

There were 19 RCTs and 20 papers that compared locations for mesh placement and met the criteria for assessment (Table 4). Nine manuscripts were excluded for various reasons, including undefined or mixed mesh locations, previously published data (overlap), inlay mesh location, or inadequate study information (see Table 1).

Table 4.

Mesh Location Study Assessment and Outcomes

				SSI		Recurrence
Series	Hernia type	Control/intervention	N	RR	95% CI	RR	95% CI	CASP score
Abdel-Baki et al. [19]	Mixed	Suture onlay	42	0.67	0.12–3.59	0.11	0.01–1.94	5
Abo-Ryia et al. [32]	Primary	Onlay sublay	60	0.67	0.12–3.71	1.00	0.68–2.47	5
Afifi [33]	Incisional	Onlay underlay	41	1.16	0.08–17.3	0.09	0.01–1.47	3
Ammar [20]	Primary	Suture onlay	80	1.60	0.52–4.93	0.09	0.01–1.50	4
Barbaros et al. [34]	Mixed	Onlay lap	46	0.5	0.10–2.46	0.35	0.01–8.11	4
Bessa et al. [35]	Mixed	Onlay sublay	80	0.33	0.01–7.94	1.00	0.15–6.76	5
de Vries Reilingh et al. [15]	Incisional	Suture underlay	37	3.17	1.02–9.87	1.16	0.67–2.04	6
Eker et al. [36]	Incisional	Sublay lap	194	0.85	0.24–3.07	1.29	0.68–2.47	7
Itani et al. [37]	Incisional	Onlay lap	146	0.13	0.03–0.52	1.5	0.56–4.00	10
Korenkov et al. [22]	Incisional	Suture onlay	72	7.65	0.43–137	0.63	0.15–0.69	10
Lal et al. [16]	Incisional	Suture onlay	62	0.38	0.08–1.79	0.13	0.01–2.49	4
Luijendijk et al. and Burger et al. [23,24]	Incisional	Suture sublay	181 3 y 7 y	8.07 14.9	0.42–154 0.86–262	0.50 0.68	0.31–0.92 0.49–0.95	10
Misra et al. [38]	Mixed	Sublay lap	66	0.21	0.05–0.90	1.94	0.18–20.4	8
Moreno-Egea et al. [39]	Primary	Sublay lap	22	1.00	0.02–46.4	1.00	0.02–46.4	7
Navarra et al. [17]	Incisional	Sublay lap	24	0.33	0.01–7.45	0.92	0.02–43.0	5
Polat et al. [25]	Primary	Suture onlay	32	0.94	0.06–13.9	0.19	0.01–3.68	7
Pring et al. [40]	Mixed	Underlay (24) lap (30)	54	0.19	0.02–1.62	0.77	0.05–11.8	4
Rogmark et al. [41]	Incisional	Sublay (67) lap (64)	131	0.07	0.01–0.48	1.05	0.02–52.0	7
Venclauskas et al. [42]	Incisional	Suture onlay sublay	161	3.79 0.54	0.84–17.1 0.05–5.77	0.95 0.09	0.47–1.92 0.12–0.67	7
Wéber [28,29]	Mixed	Sublay onlay	370	–	–	0.60	0.37–0.98	7

Figures in boldface indicate statistically significant differences.

CASP Score out of 11.

Hernia type defined by the European Hernia Society: Primary, incisional, or mixed (both primary and incisional).

CI = confidence interval; HD = high density (polypropylene or polyethylene); HDPP = high-density polypropylene; L = large (>10 cm); LDPP = low-density polypropylene; MDPP = mid-density polypropylene; PTFE = polytetrafluoroethylene composite; RR = relative risk; S/M = small/medium (<10 cm width); SSI = surgical site infection.

Although there are 20 RCTs comparing various locations for mesh placement, there are few direct comparisons of onlay, sublay, and underlay (n = 4). Instead, a network comparison is needed to assess the different locations (Fig. 4). The studies were similar enough to synthesize the data. In this setting, suture-only and laparoscopic VH repair represent the controls. Based on the findings from Key Question 1 and previous meta-analyses, it is anticipated that the suture-only group will have the highest risk of hernia recurrence whereas both groups (suture-only and laparoscopic) will have the lowest risks of SSI. Fourteen cases of contamination were reported in three studies; however, the outcomes associated with these cases were not reported separately.

FIG. 4.

Study network meta-analysis.

On network meta-analysis, sublay repair had the highest probability of being the best treatment associated with the lowest hernia recurrence rate (OR 0.74; 95% CI 0.01–0.39; probability of being the best = 56.5%), whereas underlay repair (OR 5.11; 95% CI 0.01–59.99; probability of being the best = 4.5%) and suture-only repair (OR 7.0; 95% CI 1.10–33.35; probability of being the best = 0.1%) had the lowest probability of being the best (Supplementary Table 3A). Whereas suture-only repair (OR 0.88; 95% CI 0.03–4.13; probability of being the best = 30.0%) and laparoscopic repair (OR 0.63; 95% CI 0.01–7.69; probability of being the best = 47.1%) had the highest probability of having the lowest rate of SSI. Among open mesh repair techniques, sublay repair (OR 1.39; 95% CI 0.01–6.43; probability of being the best = 23.8%) had the highest probability of being the best treatment (onlay 12.6%, underlay 0) (Supplementary Table 3B).

Sublay mesh placement may result in fewer recurrences and SSIs, but a study comparing only mesh location is needed prior to making any strong recommendations (the strength of the recommendation is moderate; the grade of the recommendation is moderate).

Discussion

During clean repairs of both primary and incisional VHs, mesh reinforcement is recommended (strong recommendation; high grade). High-density mesh should be utilized rather than low-density mesh (weak recommendation; low grade). Mesh should be placed as a sublay (in the retrorectus or preperitoneal space) when this is safe and feasible (moderate recommendation; moderate grade). These recommendations represent the most comprehensive synthesis of high-quality RCTs on the role of mesh reinforcement during VH repair.

Areas where controversy remains and the existing evidence fails to provide an adequate answer are: (1) The role of mesh reinforcement during contaminated cases; (2) the role of biologic and bioabsorbable meshes; (3) the role of mesh reinforcement for small (<0.5–1 cm width) primary VHs; and (4) The comparative effectiveness of different techniques and materials used in open and laparoscopic VH repair. In the few cases where contamination was noted, the data often were not reported in sufficient detail to draw conclusions. An RCT of commonly used treatments in contaminated VH repair, such as biologic mesh, bioabsorbable mesh, suture, or synthetic mesh, is needed.

Next, a wide variety of hernia types, including primary and incisional, and hernia sizes were assessed. However, mesh use for the smallest of hernia defects remains controversial. This is true particularly for very small primary VHs, where the risk of hernia recurrence presumably is the lowest. Many surgeons believe that the risk of recurrence is low enough for small defects that suture-only repair can provide long-term outcomes similar to those of mesh repair. In addition, surgeons are concerned that introducing mesh into a small defect will require making the hernia defect bigger. None of the studies performed a subgroup analysis of hernias <0.5–1 cm wide. However, in the meta-analysis, the odds of reducing recurrence by using mesh were greater with primary VHs than with ventral incisional hernias. An RCT assessing interventions of mesh vs. suture-only repair of small primary VHs may be needed, given the controversy among surgeons (NCT01635868). In addition, laparoscopic repair is a form of underlay repair, yet open underlay repair had a significantly higher recurrence rate although a lower SSI rate than laparoscopic underlay repair. Tradeoffs between these outcomes must be considered. Further studies may assess different techniques and materials for each of these repair types to optimize both strategies and to determine the best setting for each.

Although VHs are among the most common pathology seen by the clinician, few RCTs exist to guide the management of these patients. A number of challenges must be faced. First, there is limited standardization of terminology to describe hernia type and outcome measurements. Throughout the literature and in common use, the terms “umbilical hernia,” “ventral hernia,” and “incisional hernia” often are misapplied and misused. Only recently has a widely accepted classification system for VHs been developed. Second, the field of VH repair is intimately associated with industry, which often drives and supports low-quality studies at high risk for bias [57,58,59]. Increasing non-industry funding opportunities for hernia studies or careful planning and integrating blinding into RCTs (in particular, outcomes assessment) may reduce the risk of bias. Third, given the heavy industry ties, there is concern about publication bias. Studies supported by mesh companies with “negative” results may be less likely to be published. Although current practice is to report plans for an RCT on clinicaltrials.gov, no system exist to monitor reporting, particularly of negative studies. Fourth, surgeon skill may play a role in outcomes but is not easily measured in RCTs. However, effectiveness trials should assess a wide variety of surgeons, so this may make the results more generalizable.

Increased availability and access to local or large databases has resulted in an explosion of available information and reliance on database studies for decision-making. However, this type of data is meant only for hypothesis generation and quality improvement. Unfortunately, it often is used inappropriately to draw conclusions and infer causation. Without blinding and randomization, comparison between treatments is subject to known and unknown biases, which can lead to incorrect estimates of treatment effect. Instead of relying solely on database studies, RCTs should be used to guide answers to key clinical questions along with large databases for assessment of quality improvement. Its conclusions are limited by the quality of the studies analyzed. However, only high-quality RCTs were included, so this report may represent the best available evidence. In addition, differences between studies combined for a meta-analysis may invalidate the results; however, statistical and clinical heterogeneity were assessed prior to combining studies.

Conclusions

Among patients undergoing VH repair, mesh reinforcement should be utilized routinely when there is no contamination. Sublay mesh may result in fewer recurrences and SSIs, but further study is warranted comparing different mesh locations directly. The quality of evidence to support these recommendations is moderate to high. In cases where there is contamination, there is no high-quality evidence to guide treatment options. Future studies should assess contaminated VH repair, mesh type, the role of biologic and bioabsorbable mesh, and the repair of small defects.

Footnotes

Acknowledgments

We thank Blake E. Henchcliffe and Thomas O. Mitchell for assistance with data acquisition.

Author Disclosure Statement

The authors have no conflicts of interest with regard to this paper.

References

Liang

, Holihan

, Itani

, et al. Ventral hernia management: An expert consensus guided by systematic review and Delphi method. Appendix 1. Ann Surg, 2017; 265:80–89.

Bellows

, Robinson

, Fitzgibbons

, et al. Watchful waiting for ventral hernias: A longitudinal study. Am Surg, 2014; 80:245–252.

Verhelst

, Timmermans

, van de Velde

, et al. Watchful waiting in incisional hernia: Is it safe?. Surgery, 2015; 157:297–303.

Berger

, Li

, Hicks

, et al. Development and validation of a risk-stratification score for surgical site occurrence and surgical site infection after open ventral hernia repair. J Am Coll Surg, 2013; 217:974–982.

Liang

, Goodenough

, Martindale

, et al. External validation of the ventral hernia risk score for prediction of surgical site infections. Surg Infect, 2015; 16:36–40.

Simon

, Frelich

, Gould

, et al. Inpatient outcomes after elective versus nonelective ventral hernia repair. J Surg Res, 2015; 198:305–310.

Helgstrand

, Rosenberg

, Kehlet

, et al. Nationwide prospective study of outcomes after elective incisional hernia repair. J Am Coll Surg, 2013; 216:217–228.

Holihan

, Nguyen

, et al. Mesh location in open ventral hernia repair: A systematic review and network meta-analysis. World J Surg, 2015; 2016:40:89–99.

Gray

, Vick

, Graham

, et al. Variation in mesh placement for ventral hernia repair: An opportunity for process improvement?. Am J Surg, 2008; 196:201–206.

10.

Funk

, Perry

, Narula

, et al. Current national practice patterns for inpatient management of ventral abdominal wall hernia in the United States. Surg Endosc, 2013; 27:4104–4112.

11.

Poulose

, Shelton

, Phillips

, et al. Epidemiology and cost of ventral hernia repair: Making the case for hernia research. Hernia, 2012; 16:179–183.

12.

Slade

, Carlson

. Takedown of enterocutaneous fistula and complex abdominal wall reconstruction. Surg Clin North Am, 2013; 93:1163–1183. Available at: https.www.facs.org/quality-programs/acs-nsqip/about/businesscase/roicalc

13.

Holihan

, Alawadi

, Martindale

, et al. Adverse events after ventral hernia repair: The vicious cycle of complications. J Am Coll Surg, 2015; 221:478–485.

14.

Conze

, Kingsnorth

, Flament

, et al. Randomized clinical trial comparing lightweight composite mesh with polyester or polypropylene mesh for incisional hernia repair. Br J Surg, 2005; 92:1488–1493.

15.

de Vries Reilingh

, van Goor

, Charbon

, et al. Repair of giant midline abdominal wall hernias: “Components separation technique” versus prosthetic repair: Interim analysis of a randomized controlled trial. World J Surg, 2007; 31:756–763.

16.

Lal

, Laghari

, Soomro

. A comparative study of anatomic repair versus mesh repair in paraumbilical hernia. Med Channel, 2012; 19:110–113.

17.

Navarra

, Musolino

, De Marco

, et al. Retromuscular sutured incisional hernia repair: A randomized controlled trial to compare open and laparoscopic approach. Surg Laparosc Endosc Percutan Tech A, 2007; 17:86–90.

18.

Pawlak

, Hilgers

, Bury

, et al. Comparison of two different concepts of mesh and fixation technique in laparoscopic ventral hernia repair: A randomized controlled trial. Surg Endosc, 2015; 2016; 30:1188–1191.

19.

Abdel-Baki

, Bessa

, Abdel-Razek

. Comparison of prosthetic mesh repair and tissue repair in the emergency management of incarcerated para-umbilical hernia: A prospective randomized study. Hernia, 2007; 11:163–167.

20.

Ammar

. Management of complicated umbilical hernias in cirrhotic patients using permanent mesh: Randomized clinical trial. Hernia, 2010; 14:35–38.

21.

Arroyo

, García

, Pérez

, et al. Randomized clinical trial comparing suture and mesh repair of umbilical hernia in adults. Br J Surg, 2001; 88:1321–1323.

22.

Korenkov

, Sauerland

, Arndt

, et al. Randomized clinical trial of suture repair, polypropylene mesh or autodermal hernioplasty for incisional hernia. Br J Surg, 2002; 89:50–56.

23.

Luijendijk

, Hop

, van den Tol

, et al. A comparison of suture repair with mesh repair for incisional hernia. N Engl J Med, 2000; 343:392–398.

24.

Burger

, Luijendijk

, Hop

, et al. Long-term follow-up of a randomized controlled trial of suture versus mesh repair of incisional hernia. Ann Surg, 2004; 240:578–583.

25.

Polat

, Dervisoglu

, Senyurek

, et al. Umbilical hernia repair with the Prolene hernia system. Am J Surg, 2005; 190:61–64.

26.

Venclauskas

, Maleckas

, Kiudelis

. One-year follow-up after incisional hernia treatment: Results of a prospective randomized study. Hernia, 2010; 14:575–582.

27.

Wéber

, Baracs

, Horváth

. [“Onlay” mesh provides significantly better results than “sublay” reconstruction: Prospective randomized multicenter study of abdominal wall reconstruction with sutures only, or with surgical mesh—Results of a five-years follow-up] (Hun). Magy Seb, 2010; 63:302–311.

28.

Wéber

, Horváth

. [Results of ventral hernia repair: comparison of suture repair with mesh implantation (onlay vs sublay) using open and laparoscopic approach–prospective, randomized, multicenter study]. Magy Seb, 2002; 55:285–289.

29.

Rickert

, Kienle

, Kuthe

, et al. A randomised, multi-centre, prospective, observer and patient blind study to evaluate a non-absorbable polypropylene mesh vs. a partly absorbable mesh in incisional hernia repair. Langenbecks Arch Surg, 2012; 397:1225–1234.

30.

Bensaadi

, Paolino

, Valenti

, et al. Intraperitoneal tension-free repair of a small midline ventral abdominal wall hernia: Randomized study with a mean follow-up of 3 years. Am Surg, 2014; 80:57–65.

31.

Moreno-Egea

, Carrillo-Alcaraz

, Soria-Aledo

. Randomized clinical trial of laparoscopic hernia repair comparing titanium-coated lightweight mesh and medium-weight composite mesh. Surg Endosc, 2013; 27:231–239.

32.

Abo-Ryia

, El-Khadrawy

, Moussa

, et al. Prospective randomized evaluation of open preperitoneal versus preaponeurotic primary elective mesh repair for paraumbilical hernias. Surg Today, 2015; 45:429–433.

33.

Afifi

. A prospective study between two different techniques for the repair of a large recurrent ventral hernia: A double mesh intraperitoneal repair versus onlay mesh repair. Hernia, 2005; 9:310–315.

34.

Barbaros

, Asoglu

, Seven

, et al. The comparison of laparoscopic and open ventral hernia repairs: A prospective randomized study. Hernia, 2007; 11:51–56.

35.

Bessa

, El-Gendi

, Ghazal

, et al. Comparison between the short-term results of onlay and sublay mesh placement in the management of uncomplicated para-umbilical hernia: A prospective randomized study. Hernia, 2015; 19:141–146.

36.

Eker

, Hansson

, Buunen

, et al. Laparoscopic vs. open incisional hernia repair: A randomized clinical trial. JAMA Surg, 2013; 148:259–263.

37.

Itani

, Hur

, Kim

, et al; Veterans Affairs Ventral Incisional Hernia Investigators. Comparison of laparoscopic and open repair with mesh for the treatment of ventral incisional hernia: A randomized trial. Arch Surg, 2010; 145:322–328.

38.

Misra

, Bansal

, Kulkarni

, et al. Comparison of laparoscopic and open repair of incisional and primary ventral hernia: Results of a prospective randomized study. Surg Endosc, 2006; 20:1839–1845.

39.

Moreno-Egea

, Carrasco

, Girela

, et al. Open vs laparoscopic repair of spigelian hernia: A prospective randomized trial. Arch Surg, 2002; 137:1266–1268.

40.

Pring

, Tran

, O'Rourke

, Martin

. Laparoscopic versus open ventral hernia repair: A randomized controlled trial. Aust NZ J Surg, 2008; 78:903–906.

41.

Rogmark

, Petersson

, Bringman

, et al. Short-term outcomes for open and laparoscopic midline incisional hernia repair: A randomized multicenter controlled trial: The ProLOVE (Prospective Randomized Trial on Open versus Laparoscopic Operation of Ventral Eventrations) trial. Ann Surg, 2013; 258:37–45.

42.

Venclauskas

, Maleckas

, Kiudelis

. One-year follow-up after incisional hernia treatment: Results of a prospective randomized study. Hernia, 2010; 14:575–582.

43.

Carbajo

, Martín del Olmo

, Blanco

, et al. Laparoscopic treatment vs open surgery in the solution of major incisional and abdominal wall hernias with mesh. Surg Endosc, 1999; 13:250–252.

44.

Kaafarani

, Kaufman

, Reda

, et al. Predictors of surgical site infection in laparoscopic and open ventral incisional herniorrhaphy. J Surg Res, 2010; 163:229–234.

45.

Kaafarani

, Hur

, Campasano

, et al. Classification and valuation of postoperative complications in a randomized trial of open versus laparoscopic ventral herniorrhaphy. Hernia, 2010; 14:231–235.

46.

Kaafarani

, Hur

, Hirter

, et al. Seroma in ventral incisional herniorrhaphy: Incidence, predictors and outcome. Am J Surg, 2009; 198:639–644.

47.

Othman

, Metwally

, Bakr

, et al. Comparative study between laparoscopic and open repair of paraumbilical hernia. J Egypt Soc Parasitol, 2012; 42:175–182.

48.

Olmi

, Scaini

, Cesana

, et al. Laparoscopic versus open incisional hernia repair: An open randomized controlled study. Surg Endosc. 2007; 21:555–559.

49.

www.casp-uk.net/ Accessed December 2, 2015.

50.

www.gradeworkinggroup.org/ Accessed December 2, 2015.

51.

Jansen

, Fleurence

, Devine

, et al. Interpreting indirect treatment comparisons and network meta-analysis for health-care decision making: Report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 1. Value in health. J Int Soc Pharmacoecon Outcomes Res, 2011; 14:417–428.

52.

Grubnik

, Grubnik

, Vorotyntseva

. Laparascopic repair of incisional and ventral hernias with the new type of meshes: Randomized control trial. Wideochir Inne Tech Maloinwazujne, 2014; 9:145–151.

53.

Malik

. Laparascopic versus open repair of paraumbilical hernia. Is it a good alternative?. J Pak Med Assn, 2015; 8:865–868.

54.

Hoaglin

, Hawkins

, Jansen

, et al. Conducting indirect-treatment-comparison and network-meta-analysis studies: Report of the ISPOR Task Force on Indirect Treatment Comparisons Good Research Practices: part 2. Value in health. J Int Soc Pharmacoecon Outcomes Res, 2011; 14:429–437.

55.

Gelman

, Rubin

. Markov chain Monte Carlo methods in biostatistics. Stat Methods Med Res, 1996; 5:339–355.

56.

Salanti

, Marinho

, Higgins

. A case study of multiple-treatments meta-analysis demonstrates that covariates should be considered. J Clin Epidemiol, 2009; 62:857–864.

57.

Every-Palmer

, Howick

. How evidence-based medicine is failing due to biased trials and selective publication. J Eval Clin Pract, 2014; 20:908–914.

58.

Amiri

, Kanesalingam

, Cro

, Casey

. Does source of funding and conflict of interest influence the outcome and quality of spinal research?. Spine J, 2014; 14:308–314.

59.

Lundh

, Sismondo

, Lexchin

, et al. Industry sponsorship and research outcome. Cochrane Database Syst Rev, 2012; 12:MR000033.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.03 MB

0.02 MB

0.00 MB