A vast amount of literature advocates the use of prosthetic mesh in clean operative fields; in contrast utilization of prosthetic grafts in contaminated/obstructed settings has been seldom described [3]. Most surgeons believe that permanent prosthetic materials for incisional hernia repair are contraindicated in the setting of gross contamination, which includes emergency presentation of incarcerated incisional hernia with associated bowel obstruction as a subset, due to risk of infection as high as 10% to 35% [4]. However, primary repair of incisional hernia has high recurrence rates, ranging between 10% and 50%; while adoption of prosthesis lowers recurrence rates to between 3% and 18% [5-7]but also increases the risk of local complications such as wound infections. Inspite of possible wound complications, the use of mesh in emergency repair of incarcerated incisional hernia with associated bowel obstruction is sometimes necessitated by the size or nature of defects and often the choice is between absorbable versus non-absorbable mesh in obstructed settings. There is little published literature evaluating various surgical approaches to these incredibly challenging patients. Therefore this is a subject of debate whether to use non-absorbable prostheses with obstructed or gangrenous bowel in potentially or truly infected operating fields [3].

We divided our patients into two groups on basis of wound closure. In one group, a 90 gram prolene mesh was fixed over the anterior rectus sheath after primary repair of hernia (standard onlay technique)and the wound was closed in layers. In other group, mesh was placed over the primarily repaired defect by onlay technique but the wound was left open. The wound was treated by daily dressings until complete neoepithilization of the wound i.e healing by secondary intention. All patients received prophylactic antibiotics, a triple regime consisting of kefzole, augmentin and ceftriaxone. In patients with strangulation who required bowel resection, antiobiotics were continued for five days.

The mean operative time was 170 minutes (SD ± 37). 17 (28.3%) patients had small size hernias (less than 5 cm). Of these 7 patients had single defect and 10 had multiple defects. 12 (20%) patients had medium size hernias (5-10 cm). Of these 7 patients had single defect and 5 had multiple defects. 5 (8.3%) patients had large hernias (more than 10 cm). 3 patients had single and 2 had multiple defects whereas missing data was encountered in 26 cases.

The mean length of hospital stay for all patients was 4.5 days (SD ± 2.4 days). The group of patients who had their wounds left open stayed longer (mean 5.6 ± 2.4 days) compared to the group of patients in whom the wound was closed primarily (mean 4.4 ± 2.3 days). However, the difference in the length of hospital stay between the two groups was insignificant (p value 0.868). During the subsequent clinic visits after discharge from the hospital, wound care was provided. The average duration of follow-up was 3.8 months (range 1-48 months).

Wound infection was defined as any wound that required the prescription of antibiotics and/or skin opening with or without debridement. There were no infections in the group of patients who had their surgical wounds left open. In the other group of patients with primary skin closure, seventeen out of 55 patients (31%) developed surgical wound complications. This group was further subdivided on the basis of need for concurrent bowel resection. In this group, 13 patients required concurrent bowel resection while 42 did not. In the subgroup of patients without concurrent bowel resection 12 developed wound complications (28.5%). Of these, 8 patients developed superficial surgical site infection which required wound debridement. Three of the four patients who developed deep infection were treated with local wound-care measures but one patient required complete removal of mesh. In the other subgroup of patients with concurrent bowel resection, two of the three patients developing superficial surgical site infection required wound debridement. One patient had deep infection and cellulitis each but none required surgical treatment.

Colonic operations are classified as contaminated and infected (class 3-4) procedures according to Altemeier classification. For this reason the use of mesh in potentially contaminated procedures has been strongly discouraged [7]. Morris et al. [8] suggest abandonment of the use of mesh for repairs in which open bowel is encountered. A trend of increased pain and more severe wound infections after mesh repair were the basis for discontinuation of randomized control trial by Korenkov et al., [9] highlighting the risk of using a foreign body in a hernia repair. Temudom et al. [10] in a series of 50 complex prosthetic giant ventral hernia repairs reported that the two patients with simultaneous bowel surgery subsequently required mesh removal. Others [9,10]recommend that intestinal resection be done first and hernia repair should be postponed for a second time.

The use of mesh is sometime necessitated by the size and the nature of defect. More than half of the patients in our study had either large or multi-loculated hernias. The defects were too large to be repaired primarily. McLanahan et al. [11] reported no increased infectious risk with the prosthetic placement in a series of mostly clean-contaminated wounds having mesh incisional herniorrhapy. Vix et al. [4], Birolini et al. [12], Geisler et al. [13] and more recently Machiaras et al. [7] report 10.6%, 20%, 7% and 15.7% wound related morbidity respectively with the use of mesh in clean-contaminated and contaminated procedures. Campanelli et al. [14] performed ten prosthetic hernia repairs in potentially contaminated areas and report that there were no major or minor complications after a 21 months follow-up period. These authors advocated the use of non-absorbable mesh in potentially contaminated and contaminated operations including colonic resections, with results as good as those observed in clean procedures.

The overall 28.3% infection rate in this study is significantly higher compared to the previous studies. Kelly et al. [3] reported 21% infection rate in a series of emergency and elective incisional hernia repairs. Infection rates were 21% and 4% as reported by Alaedeen et al [1] and Ahmed et al. [15] in a similar patient casemix. The higher than usual infection rate in this study is attributable to the unique set of patients. We focused exclusively on a subset of incisional hernia repair cases which presented with an obstructed bowel and required emergency surgery, which carry higher risk of post-operative complications and have less favorable outcome [2]. Davies et al [2] found 10% infection rates in patients requiring emergency repair for all abdominal hernias. None of above quoted studies has focused on the subset of patients addressed in our study. Most of these studies include elective contaminated cases as a bulk of their population.

In view of the high infection rates, various techniques for mesh placement, including onlay, sublay (retromuscular or extrafascial), or underlay (intraperitoneal or subfascial), have been investigated. Rives-Stoppa technique has been advocated to have low infection rates, ranging between 2% and 17% [16]. However, most of these studies did not focus exclusively on emergency repair of incarcerated hernias. In addition, this procedure is time consuming, as shown by longer mean length of operation time (131 minutes primary, 141 minutes mesh, 231 minutes Stoppa) in a study by Veillette et al [17]. Patients with incarcerated incisional hernia with bowel obstruction are usually hemodynamically unstable. Therefore utility of time-consuming Stoppa technique in such patients is yet to be established.

The infection rate also depends on the need for concurrent bowel resection. It is 38% for patients needing concurrent bowel resection and 28% for those without bowel resection. The high wound infection in patients without bowel resection is postulated to be secondary to bacterial translocation thus an obstructed bowel is a significant risk factor for wound infection.

However, in view of the high infection rates, the option in operating room with a large defect and potentially contaminated field is to use absorbable mesh for temporary closure and do a definitive repair as a second planned operation.

In this study, the wound was left open in five patients to heal secondarily with granulation tissue. The decision to do so was made intra-operatively in light of grossly contaminated surgical field which made placement of prosthetic mesh a risky option. In addition absorbable (biological) mesh is not available in Pakistan. Post-operatively, the wounds took 4-5 months to heal depending upon size of the defect. These patients were advised to continue routine activities and once or twice daily self-wound care using a guaze once granulation had occurred. Patients were followed in the clinic on monthly basis till the wound healed completely. We did not have chronic draining sinuses in our patients.

Figure ​Figure11

In our experience, use of prolene mesh in contaminated fields is associated with high rates of wound infection, however mesh removal is rarely needed. All except one patient were treated with wound debridement or antibiotics. The evidence that prolene mesh is resistant to infection is also borne out by studies on Lichenstein hernia repair [18]. The authors in view of the physical characteristics of prolene mesh, that is the mononfilament structure that allows the neutrophils and macrophages to eradicate bacteria, placed the mesh in an onlay position and allowed the wound to granulate and incorporate the mesh. The growth of granulation tissue through the intricese of the mesh is a unique phenomenon. This requires daily change of wound dressing and complete wound healing can take up to a year. The mesh eventually gets incorporated and epithelilized. In one patient mismatch in the wound contraction and mesh contraction resulted in partial auto explanation of the mesh as shown in the picture; this was excised.

Figure ​Figure22 and Figure ​Figure33