Meta-analysis on Materials and Techniques for Laparotomy Closure: The MATCH Review

https://doi.org/10.1007/s00268-017-4393-9 SCIENTIFIC REVIEW Meta-analysis on Materials and Techniques for Laparotomy Closure: The MATCH Review N. A. Henriksen 1 E. B. Deerenberg 2 L. Venclauskas 3 R. H. Fortelny 4 M. Miserez 5 F. E. Muysoms 6 Ó Société Internationale de Chirurgie 2018 Abstract Background The aim this systematic review and meta-analysis was to evaluate closure materials and suture techniques for emergency and elective laparotomies. The primary outcome was incisional hernia after 12 months, and the secondary outcomes were burst abdomen and surgical site infection. Methods A systematic literature search was conducted until September 2017. The the RCTs was evaluated by at least 3 assessors using critical appraisal checklists. Meta-analyses were performed. Results A total 23 RCTs were included in the meta-analysis. There was no evidence from RCTs using the same suture technique in both study arms that any suture material (fast-absorbable/slowly absorbable/non-absorbable) is superior in reducing incisional hernias. There is no evidence that continuous suturing is superior in reducing incisional hernias compared to interrupted suturing. When using a slowly absorbable suture for continuous suturing in closure, the small bites technique results in significantly less incisional hernias than a large bites technique (OR 0.41; 95% CI 0.19, 0.86). Conclusions There is no high- evidence available concerning the best suture material or technique to reduce incisional hernia rate when closing a. When using a slowly absorbable suture and a continuous suturing technique with small tissue bites, the incisional hernia rate is significantly reduced compared with a large bites technique. Introduction & N. A. Henriksen nadiahenriksen@gmail.com 1 2 3 4 5 6 Department Surgery, Zealand University Hospital, Koege, Denmark Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands Lithuanian University Health Sciences, Kaunas, Lithuania Department General, Visceral and Oncological Surgery, Medical Faculty, Wilhelminenspital & Sigmund Freud University, Vienna, Austria University Hospitals, KU Leuven, Louvain, Belgium AZ Maria Middelares, Ghent, Belgium Incisional hernia is a frequent problem after abdominal surgery with an incidence varying from 10 to 69% depending on the type surgery, length and method follow-up and patient characteristics [1 5]. Incisional hernias develop due to insufficient healing the abdominal wall after surgery. The defect in the abdominal wall allows for protrusion intraperitoneal content causing a variety symptoms ranging from discomfort and impaired body image to incarceration and ischemia the contents the hernia sac. Besides significant morbidity and impaired life, incisional hernias are costly to treat [6, 7]. Well-known patient-related risk factors for incisional hernia formation are smoking, obesity, re and

postoperative wound complications [7, 8]. Additionally, the suture material and the surgical technique used to close a wound are important surgical determinants the risk developing an incisional hernia. To reduce the incidence incisional hernia, an international group experts developed the European Hernia Society guidelines on the closure abdominal wall incisions [9]. The recommendations in this guideline for the closure midline incisions included a continuous suture technique, performed with a small bites technique and a slowly absorbable suture material. These recommendations were mainly based on the evidence from systematic reviews on the subject [10 12]. However, the randomized controlled trials (RCTs) included in these reviews generally compared a continuous suture technique with a slowly or non-absorbable suture to an interrupted suture technique with a fast-absorbable suture. A comparison limited to those studies evaluating only one variable between study arms (same technique performed with different suture materials or different techniques performed with same suture material) was not performed. It was the hypothesis that in order to evaluate a certain suture material or technique, the same suture material should be used in both arms with various techniques and vice versa. Therefore, this systematic review and metaanalysis was done to assess the evidence from published RCTs comparing closure materials or techniques for laparotomies with a primary outcome incisional hernia after 12 months and with secondary outcomes surgical site infection (SSI) and burst abdomen. Materials and methods A written study protocol was produced and registered on Prospero (CRD42015023689) before the initiation the systematic review and meta-analysis. The data are reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [13]. Inclusion and exclusion criteria The aim the systematic review and meta-analysis was to evaluate published RCTs comparing techniques and materials for fascial closure a. The primary endpoint was incisional hernia, and the secondary endpoints were SSI, burst abdomen/wound dehiscence and suture sinus formation. Incisional hernia was defined as a protrusion the scar with the use Valsalva s maneuver during clinical examination with a minimum 12-month follow-up after the primary. All types incisions (transverse/midline/oblique/para median) and all indications for surgery (both emergency and elective laparotomies) were included. Only human studies on adults C18 years age were included. Studies on mesh closure laparotomies and studies including historic suture materials such as catgut and stainless steel sutures for comparison were excluded. Search strategy A systematic search was done independently by two authors (NAH and FM) in the following databases: Medline, EMBASE, Cochrane, SCOPUS, CINAHL and Web-Science. The search was not restricted to certain languages or years publication. The last search was performed on September 3, 2017. In Medline and EMBASE, the search strategy was based on the Medical Subject Heading (MeSH) terms:, wound closure, sutures and abdominal wall hernias. The detailed search term for Medline was (( Wound Closure Techniques [Mesh] OR Sutures [Mesh] OR Surgical Procedures, Operative [Mesh] AND Laparotomy [Mesh] AND (Randomized Controlled Trial[ptyp] AND humans [MeSH Terms])) OR ( Wound Closure Techniques [Mesh] OR Sutures [Mesh] OR Surgical Procedures, Operative [Mesh] AND Laparotomy [Mesh] AND Hernia, Ventral [Mesh] AND (Randomized Controlled Trial[ptyp] AND humans [MeSH Terms]))). Evaluation papers and data extraction Firstly, the records were screened by title and abstract by two assessors independently (ED, LV). Secondly, the full texts were divided into two groups and each group was evaluated by two authors (NAH, ED, LV, RF) independently for eligibility with the use critical appraisal checklist for randomized controlled trials developed by the Scottish Intercollegiate Guidelines Network (SIGN) [14]. Only papers rated as acceptable or high by SIGN were included to limit the risk bias. Any disagreement between the two assessors was settled by discussion with a third evaluator (FM, MM). Data were extracted by two authors independently regarding the predefined outcomes (NAH, FM) and checked by co-authors (LV, RF). Non-English full texts were handled with same procedure by two individual assessors outside the author group. Selection outcomes to be included in the metaanalysis In the meta-analysis, sutures were divided into fast-absorbable/slowly absorbable and non-absorbable sutures

regardless whether the sutures were monilament or. Suture methods were divided into interrupted versus continuous suturing and small bites versus big bites technique, regardless whether a layered or mass closure technique was used. elective laparotomies were pooled in the same analysis, as was all types incisions (midline, transverse/oblique/paramedian). It was decided only to compare different suture types, when the same suture method was used in both arms. Likewise, suture methods were only compared, when the same type sutures was used for both methods. Statistical analysis The outcomes were pooled in conventional meta-analyses and reported as weighted odds ratios (OR) with 95% confidence intervals (95% CI) using the random effects model and illustrated with forest plots. Heterogeneity was explored using I 2 statistics. Funnel plots were used to assess possible publication bias. The Cochrane risk bias tool was used to assess the risk bias. Kappa statistics were used to assess the agreement between two assessors. Statistical analyses were performed with Review Manager Stware version 5.3 (The Nordic Cochrane Centre, Copenhagen, Denmark). Results Literature search Out 1818 citations, a total 23 RCTs with 10,130 patients were included in the meta-analyses (Fig. 1). The level agreement between the two assessors was moderate for screening abstracts and high for assessing eligibility the full texts. All studies were published in peer-reviewed journals from 1981 to 2015. One study was in French, and the remaining 22 studies were English. Study characteristics are listed in Table 1. Fig. 1 PRISMA flow diagram search strategy and study selection

Table 1 Overview included studies Paper Quality assessment Number patients (group A/group B) Agrawal et al. [34] Acceptable 147 (40/ 47/45/ 42) Berretta et al. [16] Acceptable 191 (63/ 63/65) Bloemen et al. [2] 456 (233/ 233) Bresler et al. [17] 235 (70/ 71/62) Cameron et al. [21] 284 (143/ 141) Carlson et al. [22] Acceptable 225 (112/ 113) Colombo et al. [18] Acceptable 614 (308/ 306) Type Emergency midline Elective midline Elective midline Elective midline Intervention Comparison Length followup (months) Suture-to-wound length ratio, stitch size Outcome measures Non-absorbable continuous mass closure (polypropylene 1-0) Non-absorbable interrupted mass closure (polypropylene 1-0) Non-absorbable monilament continuous mass closure (Premilene 1-0) Nonabsorbable interrupted fascial closure (Ethibond 2-0) Non-absorbable monilament continuous single layer (Prolene 1-0) PDS continuous suture (I or II) Non-absorbable monilament interrupted mass closure (Prolene 1-0) Non-absorbable monilament nylon continuous mass closure (Ethilon 0-0 loop) monilament polyglyconate continuous (Maxon 1-0) continuous mass closure (polyglactin 1-0) interrupted mass closure (polyglactin 1-0) monilament continuous mass closure (PDS 1-0) monilament continuous single layer (PDS 1-0) Fastabsorbable continuous suture polyglactin monilament interrupted mass closure (PDS 1-0) monilament polyglyconate continuous mass closure (Maxon 0-0 loop) polyglycolic acid Interrupted Smead- Jones (Dexon 1-0) 48 mo 4:1 ratio, N/A suture sinus. 36 mo 4:1 ratio, N/A scar pain 35 mo 4:1 ratio, 1 cm from fascial edge SSI, suture sinus 12 mo N/A, N/A suture sinus. 15 mo N/A, N/A palpable knots, wound pain, suture sinus 24 mo N/A, N/A wound pain, suture sinus 33 mo N/A, 1.5-2 cm from fascial edge and 1.5-2 cm apart. Smead- Jones near 0.5 cm and far 2.0 cm. dehiscence

Table 1 continued Paper Quality assessment Number patients (group A/group B) Corman et al. [23] Acceptable 161 (49/ 53/59) Deerenberg et al. [4] 560 (284/ 276) Deitel et al. [19] 84 (42/ 42) Donaldson et al. [44] Acceptable 231 (80/ 74/77) Gislason et al. [29] Acceptable 599 (203/199/ 197) Gurjar et al. [24] Acceptable 200 (100/ 100) Gys et al. [30] Acceptable 132 (67/ 65) Type Elective midline Elective midline elective paramedian elective subcostal, transverse and midline elective subcostal and midline Intervention Comparison Length followup (months) Suture-to-wound length ratio, stitch size Outcome measures Monilament non-absorbable polypropylene interrupted single layer (Prolene) Multifilament non-absorbable uncoated nylon interrupted single layer (Nurolon) monilament, continuous single-layer small bites (PDS 2-0) monilament polyglyconate continuous single layer (Maxon 1-0) Continuous polypropylene (Prolene 1-0) Continuous mass polyglyconate (Maxon loop) Continuous mass polyglactin (Vicryl) interrupted single layer (Vicryl) monilament, continuous mass closure big bites (PDS 1-0 loop) polyglycolic acid continuous single layer (Dexon 1-0) Continuous chromic catgut 1-0 Continuous polyglycolic acid (Dexon 1-0) Interrupted polyglactin (Vicryl) 19 mo N/A, N/A suture sinus 12 mo 4:1 ratio, small bites 0.5 cm from fascial edge and apart, big s 1 cm from fascial edge and apart SSI, burst abdomen 24 mo N/A, N/A seroma, SSI 12 mo N/A, N/A suture sinus 12 mo N/A, 1.5 cm from fascial edge and 1.5 cm apart burst abdomen Continuous polypropylene, intermittent Aberdeen knot at every 4 stitch (Prolene 1-0) Non-absorbable polyamide continuous layered closure with (Ethilon 1-0) Interrupted simple stitch polypropylene (Prolene 1-0) monilament polyglyconate continuous layered closure (Maxon 1-0) 12 mo N/A, 2 cm from fascial edge and 2 cm apart 12 mo N/A, 1 cm from fascial edge and 1 cm apart SSI, dehiscence SSI, burst abdomen

Table 1 continued Paper Quality assessment Number patients (group A/group B) Hsiao et al. [45] 340 (184/ 156) Krukowski et al. [25] Acceptable 757 (374/ 383) Lewis et al. [26] Acceptable 200 (105/ 95) Millbourn et al. [27] Acceptable 737 (381/ 356) Osther et al. [31] Acceptable 204 (100/ 104) Richards et al. [32] Acceptable 571 (286/ 285) Sahlin et al. [33] 988 (345/ 339) Type Elective midline, subcostal, paramedian and transverse, paramedian, transverse and oblique, paramedian, transverse and oblique, transverse and paramedian Intervention Comparison Length followup (months) Suture-to-wound length ratio, stitch size Outcome measures monilament polydioxanone continuous layered closure with (PDS loop 0-0) Absorbable polyglactin continuous layered closure (Vicryl 0-0) 24 mo N/A, 1.5 cm from fascial edge and 1.5 cm apart SSI Non-absorbable monilament continuous mass closure (polypropylene) Non-absorbable polypropylene, continuous layered closure (Prolene 1-0) continuous single-layer small bites (PDS 2-0) polyglyconate, single-layer interrupted sutures (Maxon 0-0) monilament continuous mass closure (PDS) polyglycolic acid, interrupted Smead- Jones (Dexon 1-0) continuous single-layer big bites (PDS 1-0) polyglycolic acid single-layer interrupted sutures (Dexon 0-0) 12 mo N/A, stitch size 1 cm apart and 1 cm from edge 60 mo N/A, 1 cm from edge, 1-2 cm apart 12 mo 4:1 ratio, small bites 0.5 cm from fascial edge and apart, big bites 1 cm from fascial edge and apart 12 mo N/A, 1,5 cm from fascial edge and 1 cm apart wound pain SSI, dehiscence SSI, dehiscence SSI, dehiscence Non-absorbable continuous layered polypropylene (Prolene 0-0) polyglycolic acid layered interrupted Smead- Jones (Dexon 0-0) 12 mo N/A, 1,5 cm from fascial edge and 1 cm apart hematoma monilament polyglyconate, continuous (Maxon 0-0) polyglactin, interrupted (Vicryl 0-0) 12 mo N/A, N/A SSI, dehiscence

Outcome measures Suture-to-wound length ratio, stitch size Intervention Comparison Length followup (months) Interrupted Vicryl 12 mo 4:1 ratio, N/A SSI, burst abdomen Continuous slowly absorbable (MonoPlus) Continuous slowly absorbable (PDS) 12 mo N/A, N/A suture sinus, wound pain polyglactin, singlelayer interrupted (Vicryl 1-0) polyglactin, single layer continuous (Vicryl 1-0) polydioxanone single layer continuous (PDS 0-0) Non-absorbable single layer continuous (nylon 1-0) Risk bias Seven the studies were considered high with the SIGN critical appraisal checklist, and the overall risk bias was low. The most frequent source bias was performance bias, as 12 the studies did not report whether participants or personnel were blinded to the allocation. The funnel plots revealed no signs publication bias. Type A total 6 studies included only laparotomies [4, 15 19]. Nine the studies included both emergency and laparotomies [2, 20 27], but none them reported outcome data separately for the emergency or elective procedures. A total 5 studies included emergency and elective procedures through all kinds incisions [28 32], but again outcome data were not reported separately for emergency and elective procedures. Only two the studies [31, 32] sub-grouped the outcome data regarding incision type and Sahlin et al. reported more incisional hernias and wound dehiscences in midline laparotomies, whereas Richards et al. found no significant differences regarding type incision. Reporting suture technique The suture technique differed widely between the studies, making it difficult to extract data on varying suture methods apart from continuous versus interrupted and small bites versus big bites technique. In 9 the studies, the stitch size was not reported. Six studies [2, 4, 15, 19, 26, 33] reported that a 4:1 suture technique was used, but only half them reported the specific measurements [4, 19, 26]. Table 1 continued Type Number patients (group A/group B) Paper Quality assessment Elective midline Acceptable 625 (210/ 205/ 210) Seiler et al. [20] Acceptable 1539 (365/ 379/ 370/ 377) Wissing et al. [28] PDS, polydioxanone Primary outcome incisional hernia The vast majority the studies defined an incisional hernia with clinical examination as a protrusion the scar with the use Valsalva s maneuver. Only five studies used radiological examination in the identification incisional hernias [2, 4, 15, 19, 33]. There was no reporting on differences incisional hernia rates between elective and emergency laparotomies [2, 20 32]. There were no significant differences on incisional hernia rates when comparing non-absorbable and slowly absorbable sutures or slowly absorbable and fast-absorbable sutures (Fig. 2). Further, there were no significant differences on incisional hernia rate between non-absorbable and fast-absorbable sutures compared (Fig. 2).

Fig. 2 Forest plots incisional hernia rates comparing different suture types (non-absorbable/slowly absorbable/fast-absorbable) using the same suturing technique There were no significant differences between continuous and interrupted suturing using the same suture material in both arms (Fig. 3). Two studies [4, 26] including a total 1067 patients compared continuous suture with a slowly absorbable suture using either the small bites or big bites technique for closure, and there were significantly fewer incisional hernias using the small bites suture technique including only the aponeurotic layer (Fig. 3). Secondary outcome surgical site infection, wound dehiscence/burst abdomen, suture sinus The included RCTs randomized patients to different fascial closure techniques, but the skin closure technique was the same in both arms. Various criteria were used for definition SSI ranging from the definition by Center Disease Control and Prevention with the classification into superficial, deep and organ/space infection to the surgeon s clinical assessment the wound. Wound dehiscence and burst abdomen were poorly defined by most the studies. Wound dehiscence seemed to include both skin dehiscence with intact fascia and fascial disruption. In the meta-analysis, we defined burst abdomen as cases where a fascial dehiscence was described. Only one study reported that SSI was more common after emergency compared with elective [28]. Several studies found that the development an incisional hernia was preceded by a SSI in up to 40% the cases [26 30]. There were no significant differences on SSI and burst abdomen, when comparing the use non-absorbable with fast-absorbable sutures for fascial closure (OR 1.01; 95% CI 0.72 1.42 and OR 1.16; 95% CI 0.65 2.06), or non-

Fig. 3 Forest plot incisional hernia rates using a continuous suturing technique versus an interrupted suturing technique and forest plot incisional hernia using small bites versus big bites suturing with the same suture material in both arms absorbable with slowly absorbable sutures (OR 1.05; 95% CI 0.68 1.64 and OR 1.04; 95% CI 0.62 1.75), respectively. Further, there were no significant differences on SSI rate between slowly absorbable and fast-absorbable sutures (Fig. 4). Further, there were no differences on SSI and burst abdomen using a continuous or interrupted suture technique (OR 0.83; 95% CI 0.41 1.71 and OR 0.91; 95% CI 0.54 1.52). Millbourn et al. [26]. reported significantly fewer SSIs when using a small bites suture technique; however, Deerenberg et al. [4]. found no significant differences on SSI between small and big bites techniques (Fig. 4). There were no significant differences regarding burst abdomen between small and big bites techniques (OR 1.34; 95% CI 0.33 5.46) [4, 26]. Suture sinus formation, palpable knots and wound pain were reported to be a problem with the use non-absorbable sutures as compared to fast- and slowly absorbable sutures in several studies [20, 22, 24, 27, 29], whereas only one study found no significant differences on wound pain or the presence suture sinus between non-absorbable and slowly absorbable sutures [21]. Discussion The results this systematic review and meta-analysis show no significant differences on incisional hernia rate when comparing non-absorbable, slowly absorbable and fast-absorbable sutures using the same suture technique. Furthermore, using an interrupted or continuous suture technique with the same kind suture material did not affect incisional hernia rate. Lastly, a small bites continuous suture technique with a slowly absorbable polydioxanone (PDS) small-sized suture decreases the incisional hernia rate compared to a large bites technique with a larger suture. Non-absorbable or slowly absorbable sutures were not superior to fast-absorbable sutures in decreasing incisional hernia rate in this meta-analysis. In accordance, a recent systematic review by Bosanquet et al. [3]. did also conclude that there is no evidence that suture type has an intrinsic effect on incisional hernia rate. However, this finding is controversial compared with previous metaanalyses concluding that slowly absorbable sutures are superior to fast-absorbable sutures [11, 12]. These previous meta-analyses differ from the current meta-analysis, which may explain the different conclusions. Firstly, in the previous meta-analyses, only RCTs on midline laparotomies were included. Furthermore, RCTs comparing a continuous suture technique using slowly absorbable sutures with an interrupted suture technique using rapid absorbable suture were compared in the forest plots [12]. Lastly, both metaanalyses included papers that were rejected in the current meta-analysis because inadequate judged by the critical appraisal checklists.

Fig. 4 Forest plot surgical site infection using a slowly absorbable suture versus a fast-absorbable suture and forest plot surgical site infection using small bites versus big bites suturing with the same suturing material in both arms Although no significant differences between sutures could be found on incisional hernia and SSI rate, we agree with the recommendation the European Hernia Society to use a slowly absorbable suture when closing the fascia [9]. When considering the biology wound healing, using a slowly or non-absorbable suture for fascial closure seems appropriate. Fascial healing starts by recruiting inflammatory cells. Two to five days after, fibroblasts enter the wound and start producing collagen. During the proliferation phase the first 3 weeks, mainly type III collagen is produced and an extracellular matrix is created. Type III collagen consists thin, weak fibers and is replaced by strong and thick type I collagen during the following maturation phase [34, 35]. The last part the maturation phase is remodeling or realignment collagen fibers along tension lines and can take up to years [36]. The half-life tensile strength absorbable sutures like polyglactin 910 (Vicryl Ò) and polyglycolic acid (Dexon Ò )is around 2 3 weeks [37], suggesting an insufficient support the healing fascial tissue. The half-life tensile strength slowly absorbable PDS suture is 6 weeks [37]. Since fascia needs at least 14 days to regain its strength [34, 38], using a fast-absorbable suture might not provide long enough support to the healing fascia, although this is not supported by our data. Since suture sinus formation, palpable knots and wound pain were reported to be a problem with the use non-absorbable sutures [20 22, 24, 27, 29], a slowly absorbable suture is preferred over a non-absorbable suture. Although no significant differences could be found when comparing interrupted and continuous suture techniques using the same suture material, we agree with the recommendation the European Hernia Society to use a continuous suture technique to close the fascia [9]. The superiority the combination a continuous technique with a slowly absorbable suture on the incidence incisional hernia has been suggested in other systematic reviews and meta-analyses [11, 12]. Furthermore, a continuous technique is faster than an interrupted technique, thereby reducing the length surgery [11]. Two RCTs proved that a small bites suture technique using 2-0 monilamented slowly absorbable suture material significantly reduces incisional hernia rate compared to a large bites 1-0 suture technique [4, 26]. This suture method was firstly described in the seventies and is also referred to as the 4:1 suture method, as the suture length should be at least four times as long as the incision [39]. Using twice the amount stitches including the aponeurosis only provides close to ideal conditions for fascial healing due to avoidance necrosis the rectus abdominis muscles and to optimal distribution forces leading to a lower incidence incisional hernia [4, 26, 40]. Both RCTs comparing small bites with a large bites technique have used a different suture size and needle size in the two study groups [4, 26]. In the small bites group, a 2-0

PDS suture was used with 20-mm and 31-mm needles, respectively. In the large bites group, a 1-0 PDS suture was used with 41-mm and 48-mm needles, respectively [4, 26]. Whether these differences in suture size or needle size have an impact on the outcome incisional hernias or SSI is not clear and thus may be a confounding factor. The incidences SSI reported in the included RCTs emphasize that wound infection remains a frequent complication after and should be scored carefully. Furthermore, it was reported in the RCTs that incisional hernias were preceded by SSI in up to 40% the cases [26 30], stressing that SSI is an important risk factor for incisional hernia formation. However, in this meta-analysis the suture material or suture method for fascial closure did not seem to influence the rate SSI and burst abdomen. Previous meta-analyses have evaluated whether subcutaneous skin closure or skin closure with sutures or staples had an impact on SSI, and there is no convincing evidence any method is superior to another [41, 42]. Optimizing all surgical-technical factors in closing a midline and the increasing use minimally invasive surgery unfortunately does not reduce incisional hernia rate to zero. Patients undergoing open surgery for abdominal aortic aneurysm and obese patients have a higher risk incisional hernia formation [8, 43]. In these high-risk patients, other interventions might be needed to further reduce the incidence incisional hernia, such as prophylactic mesh augmentation, which has proven to be effective in the prevention incisional hernias (RR 0.17; 95% CI 0.08 0.37) [9]. Furthermore, many patients with high-risk priles such as many previous laparotomies, emergency surgery and/or major abdominal surgical procedures will still have to undergo open surgical procedures through midline incisions. In these high-risk groups, the avoidance incisional hernia remains a challenge. Further studies on the optimal closure technique in emergency laparotomies are still needed and should include the small bites technique in one study arm. To the best our knowledge, there is only one ongoing RCT on midline emergency laparotomies comparing continuous all-layer suturing with slowly absorbable suture to an interrupted technique with fast-absorbable sutures [44]. The meta-analysis design is limited by the fact that the results depend on the included studies. For instance, the majority RCTs only used clinical examination for the identification incisional hernia, which could have underestimated the hernia rate. Besides, the stitch size was not reported in many studies. Furthermore, one could argue that only studies on laparotomies should be included to minimize heterogeneity. On the other hand, the reality rarely represents with only laparotomies. Further, it was not possible to analyze emergency and elective laparotomies separately as data in the RCTs were presented together. Patient-related risk factors for wound healing such as smoking, body mass index and nutrition status were not included in the metaanalysis; however, these factors were accounted for in the RCTs and did not differ between randomized groups. This meta-analysis is limited by the fact that it does not evaluate skin closure technique independently for the evaluation SSI; however, the included RCTs randomized patients to different fascial closure methods and the skin closure technique was the same in the randomized groups. Yet this is the first systematic review and meta-analysis to compare studies on the suture material using the same technique in both arms, which is essential to conclude anything on type sutures. Likewise, to evaluate the suture technique, the same suture type must be used in both arms. Further, to minimize risk bias, the SIGN checklists were used by two independent assessors for the evaluation the RCTs. The SIGN checklist includes two questions on comparability on the randomized groups except for the intervention, which means that only RCTs with comparable groups were included in this meta-analysis. Conclusions No suture material proved superior in this meta-analysis. However, a slowly absorbable suture material seems wise to use as it keeps its strength until the fascial tissue is healed. Further, a slowly absorbable suture does not increase the SSI rate compared to fast-absorbable sutures and decreases the risk wound pain and suture sinus, which is a risk when using non-absorbable material. Moreover, no difference between an interrupted suture technique and a faster continuous technique was found. On the other hand, this meta-analysis concludes that the bestevidenced technique for closure a incision is a small bites suture technique with a 2-0 slowly absorbable suture including the aponeurosis only in a suture-to-wound length ratio at least 4:1. Acknowledgements The authors would like to thank Dr. Frederik Helgstrand, Dr. Manuel López Cano and Dr. José Antonio Pereira for the evaluation non-english papers. Compliance with ethical standards Conflict interest NH, ED, LV and RF declare no conflict interest. MM declares conflict interest not directly related to the submitted work: grants and personal fees (speaker) from Aspide, Bard, Braun, Ethicon, Gore and (editor-in-chief) Springer. FM declares conflict interest not directly related to the submitted work: grants and personal fees (consultancy and speaker) from Medtronic, Sradim, Johnson & Johnson, Advance Medical, Bard Davol and Dynamesh.

References 1. Alnassar S, Bawahab M, Abdoh A et al (2012) Incisional hernia postrepair abdominal aortic occlusive and aneurysmal disease: five-year incidence. Vascular 20:273 277 2. Bloemen A, van Dooren P, Huizinga BF et al (2011) Randomized clinical trial comparing polypropylene or polydioxanone for midline abdominal wall closure. Br J Surg 98:633 639 3. Bosanquet DC, Ansell J, Abdelrahman T et al (2015) Systematic review and meta-regression factors affecting midline incisional hernia rates: analysis 14,618 patients. PLoS ONE 10:e0138745 4. Deerenberg EB, Harlaar JJ, Steyerberg EW et al (2015) Small bites versus large bites for closure abdominal midline incisions (STITCH): a double-blind, multicentre, randomised controlled trial. Lancet 386:1254 1260 5. Hoer J, Lawong G, Klinge U et al (2002) Factors influencing the development incisional hernia. A retrospective study 2,983 patients over a period 10 years. Chirurg 73:474 480 6. Gillion JF, Sanders D, Miserez M et al (2016) The economic burden incisional ventral hernia repair: a multicentric cost analysis. Hernia 20:819 830 7. van Ramshorst GH, Eker HH, Hop WC et al (2012) Impact incisional hernia on health-related life and body image: a prospective cohort study. Am J Surg 204:144 150 8. Sorensen L, Hemmingsen U, Kirkeby L et al (2005) Smoking is a risk factor for incisional hernia. Arch Surg 140:119 9. Henriksen NA, Helgstrand F, Vogt KC et al (2013) Risk factors for incisional hernia repair after aortic reconstructive surgery in a nationwide study. J Vasc Surg 57:1524 1530 10. Muysoms FE, Antoniou SA, Bury K et al (2015) European Hernia Society guidelines on the closure abdominal wall incisions. Hernia 19:1 24 11. Sajid MS, Parampalli U, Baig MK et al (2011) A systematic review on the effectiveness slowly-absorbable versus non-absorbable sutures for abdominal fascial closure following. Int J Surg 9:615 625 12. Van t Riet M, Steyerberg EW, Nellensteyn J et al (2002) Metaanalysis techniques for closure midline abdominal incisions. Br J Surg 89:1350 1356 13. Diener MK, Voss S, Jensen K et al (2010) Elective midline closure: the INLINE systematic review and metaanalysis. Ann Surg 251:843 856 14. Liberati A, Altman DG, Tetzlaff J et al (2009) The PRISMA statement for reporting systematic reviews and meta-analyses studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol 62:e1 e34 15. Scottish Intercollegiate Guidelines Network (SIGN). http://www. signacuk/methodology/indexhtml 16. Berretta R, Rolla M, Patrelli TS et al (2010) Randomised prospective study abdominal wall closure in patients with gynaecological cancer. Aust N Z J Obstet Gynaecol 50:391 396 17. Bresler L, Courbey PJ, Feldman L et al (1995) Results a controlled trial comparing 3 suture threads at slow resorption for the closure supra-umbilical midline laparotomies. Ann Chir 49:544 548 18. Colombo M, Maggioni A, Parma G et al (1997) A randomized comparison continuous versus interrupted mass closure midline incisions in patients with gynecologic cancer. Obstet Gynecol 89:684 689 19. Deitel M, Alhindawi R, Yamen M et al (1990) Dexon plus versus Maxon fascial closure in morbid obesity: a prospective randomized comparison. Can J Surg 33:302 304 20. Seiler CM, Bruckner T, Diener MK et al (2009) Interrupted or continuous slowly absorbable sutures for closure primary abdominal incisions: a multicenter randomized trial (INSECT: ISRCTN24023541). Ann Surg 249:576 582 21. Cameron AE, Parker CJ, Field ES et al (1987) A randomised comparison polydioxanone (PDS) and polypropylene (Prolene) for abdominal wound closure. Ann R Coll Surg Engl 69:113 115 22. Carlson MA, Condon RE (1995) Polyglyconate (Maxon) versus nylon suture in midline abdominal incision closure: a prospective randomized trial. Am Surg 61:980 983 23. Corman ML, Veidenheimer MC, Coller JA (1981) Controlled clinical trial three suture materials for abdominal wall closure after bowl operations. Am J Surg 141:510 513 24. Gurjar V, Halvadia BM, Bharaney RP et al (2014) Study two techniques for midline fascial wound closure. Indian J Surg 76:91 94 25. Krukowski ZH, Cusick EL, Engeset J et al (1987) Polydioxanone or polypropylene for closure midline abdominal incisions: a prospective comparative clinical trial. Br J Surg 74:828 830 26. Lewis RT, Wiegand FM (1989) Natural history vertical abdominal parietal closure: prolene versus Dexon. Can J Surg 32:196 200 27. Millbourn D, Cengiz Y, Israelsson LA (2009) Effect stitch length on wound complications after closure midline incisions: a randomized controlled trial. Arch Surg 144:1056 1059 28. Wissing J, van Vroonhoven TJ, Schattenkerk ME et al (1987) Fascia closure after midline : results a randomized trial. Br J Surg 74:738 741 29. Gislason H, Gronbech JE, Soreide O (1995) Burst abdomen and incisional hernia after major gastrointestinal operations comparison three closure techniques. Eur J Surg 161:349 354 30. Gys T, Hubens A (1989) A prospective comparative clinical study between monilament absorbable and non-absorbable sutures for abdominal wall closure. Acta Chir Belg 89:265 270 31. Osther PJ, Gjode P, Mortensen BB et al (1995) Randomized comparison polyglycolic acid and polyglyconate sutures for abdominal fascial closure after in patients with suspected impaired wound healing. Br J Surg 82:1080 1082 32. Richards PC, Balch CM, Aldrete JS (1983) Abdominal wound closure. A randomized prospective study 571 patients comparing continuous vs. interrupted suture techniques. Ann Surg 197:238 243 33. Sahlin S, Ahlberg J, Granstrom L et al (1993) Monilament versus absorbable sutures for abdominal closure. Br J Surg 80:322 324 34. Agrawal V, Sharma N, Joshi MK et al (2009) Role suture material and technique closure in wound outcome following for peritonitis. Trop Gastroenterol 30:237 240 35. Lau FH, Pomahac B (2014) Wound healing in acutely injured fascia. Wound Repair Regen 22(Suppl 1):14 17 36. Midwood KS, Williams LV, Schwarzbauer JE (2004) Tissue repair and the dynamics the extracellular matrix. Int J Biochem Cell Biol 36:1031 1037 37. Velnar T, Bailey T, Smrkolj V (2009) The wound healing process: an overview the cellular and molecular mechanisms. J Int Med Res 37:1528 1542 38. Bourne RB, Bitar H, Andreae PR et al (1988) In-vivo comparison four absorbable sutures: Vicryl, Dexon Plus, Maxon and PDS. Can J Surg 31:43 45 39. Howes EL, Sooy JW, Harvey SC (1929) The healing wounds as determined by their tensile strength. JAMA 92:42 45 40. Jenkins TP (1976) The burst abdominal wound: a mechanical approach. Br J Surg 63:873 876 41. Harlaar JJ, van Ramshorst GH, Nieuwenhuizen J et al (2009) Small stitches with small suture distances increase closure strength. Am J Surg 198:392 395

42. Gurusamy KS, Toon CD, Davidson BR (2014) Subcutaneous closure versus no subcutaneous closure after non-caesarean surgical procedures. Cochrane Database Syst Rev 1:CD010425 43. Mackeen AD, Berghella V, Larsen ML (2012) Techniques and materials for skin closure in caesarean section. Cochrane Database Syst Rev 9:3577 44. Lau B, Kim H, Haigh PI et al (2012) Obesity increases the odds acquiring and incarcerating noninguinal abdominal wall hernias. Am Surg 78:1118 1121 45. Rahbari NN, Knebel P, Kieser M et al (2012) Design and current status CONTINT: continuous versus interrupted abdominal wall closure after emergency midline a randomized controlled multicenter trial [NCT00544583]. Trials 13:72