Negative Pressure Wound Therapy Compared to Petrolatum Gauze and a Bogota Bag to Manage Postoperative Midline Abdominal Wound Dehiscence: A Pilot, Nonrandomized Controlled Trial

Treating postoperative abdominal wound dehiscence following abdominal surgery using negative pressure wound therapy (NPWT) has shown promising results. Purpose: A study was conducted to evaluate the efficacy of NPWT for fascial closure/cutaneous cover compared to non-NPWT treatment using petrolatum gauze and a Bogota bag in patients with postoperative laparotomy wound dehiscence. Methods: A single center, prospective, nonrandomized pilot study was conducted. Using convenience sampling methods, consecutive patients on 6 different surgical units who were at least 18 years of age and who developed postoperative abdominal wound dehiscence following elective and emergency laparotomy from January 2017 to December 2018 were recruited. NPWT dressing with polyvinyl white foam sponge or loosely packed, saline-soaked petrolatum gauze followed by Bogota bag application were used and compared. Baseline patient demographics and history were collected, and patients were followed for an average of 6 months after surgery. Number of days until first signs of granulation tissue appearance, time until complete granulation tissue cover/fascial surgical closure, and hospital discharge were compared. Categorical variables (gender, comorbidities, presence or absence of stoma, exposure to prior radiotherapy) were expressed as proportions and analyzed using chi-squared test or Fischer’s exact test. Continuous variables such as age, body mass index, albumin, postoperative hospital stay, and number of days required for decision for fascial closure were expressed as Mean ± standard deviation and analyzed using an independent t test or Mann Whitney U test based on whether the data followed normal distribution. Postoperative day of wound dehiscence, the number of days for the appearance of granulation tissue, and the number of NPWT placements required also were assessed using Mean ± standard deviation and analyzed using an independent t test. A P value <.05 was considered significant. Results: Sixty (60) patients were included, but 4 in NPWT group and 10 in the non-NPWT group could not complete the study, leaving 26 patients in NPWT group and 20 patients in non-NPWT group. Demographic and surgical variables were not significantly different. Patients in both groups achieved complete wound coverage by surgical closure or healing by secondary intention. Days until first signs of granulation tissue (2.92 vs. 6.65; P <.001), number of days until fascial closure (15.50 vs. 29.50; P <.001), and length of postoperative hospital stay (24.30 vs. 37.90; P <.001) were significantly less in NPWT group. Two (2) patients (7.6%) in the NPWT developed a fistula during the 6-month follow-up period. No fistulas developed in the control group, and no intra-abdominal abscesses, ventral hernias, or wound dehiscence were reported in either group. Conclusion: Time until first signs of granulation tissue appearance and complete granulation tissue coverage was significantly shorter in the NPWT group, but time until definitive closure was not evaluated. Randomized, controlled clinical studies to compare definitive time to wound closure and long-term follow up to evaluate long-term complication rates, including the risk of developing fistulas, are warranted. 

Postoperative abdominal wound dehiscence following abdominal surgery is a chronic debilitating condition that contributes to patient morbidity and mortality and prolongs the length of hospital stay. The overall incidence of abdominal wound dehiscence following laparotomy is reported to be 0.25% to 3.5% in various studies, including a systematic review.^1-3 The incidence of abdominal wound dehiscence has not changed significantly over the past decade despite advancements in perioperative care, sophisticated suture techniques, and advanced suture materials. Multiple factors contribute to the risk of wound dehiscence. The most common factors include local factors such as wound infection, hematoma, and seroma; in addition, a retrospective study⁴ of 3500 laparotomies has shown that poor surgical fascial closure technique eventually can lead to wound dehiscence.

The mortality rate of abdominal wound dehiscence is reported to be as high as 34% to 45%.^1,3 The most serious immediate complication is evisceration, which requires immediate re-exploration. Other complications include dehydration and electrolyte imbalance following bowel exposure, enterocutaneous fistula (ECF), and a high incidence of ventral herniation (69%).^3-5 

There has been considerable advancement in the management of abdominal wound dehiscence over the past few decades. Techniques to address abdominal wound dehiscence include dynamic retention sutures, the zipper technique using a mesh sheet with a zipper, artificial burr (Wittmann patch) with 2 opposite Velcro sheets and hooks and loops connected in the middle, vacuum pack, and vacuum-assisted closure (ie, negative pressure wound therapy [NPWT]). In their systematic review, Hensbroek et al² showed that among the various methods, fascial closure was best achieved with artificial burr (with a closure rate of 90%) and NPWT (with a closure rate of 60%). However, all the studies included in the systematic review were retrospective case series; controlled prospective studies evaluating the efficacy of NPWT are sparse. 

The purpose of the present study was to evaluate the efficacy of NPWT for fascial closure/cutaneous cover compared to non-NPWT treatment in postoperative laparotomy wound dehiscence. 

Study design. The study was a single-center, prospective, parallel-arm, nonrandomized pilot study conducted over a period of 2 years in a tertiary care center in south India. Institutional Human Ethics Committee approval and patient informed consent were obtained. All information recorded was kept confidential, and patients were permitted to leave the study at any point without repercussion. All ethical principles included in the Declaration of Helsinki were followed in this study. 

Study patients. All consecutive patients at least 18 years of age who developed postoperative abdominal wound dehiscence following elective and emergency laparotomy were included in the study. Patients with an ECF, patients who developed postoperative ascites, and patients who required reoperation, had metastatic carcinoma, or multiorgan dysfunction were excluded from the study. Patients who developed wound dehiscence while on ventilator care and were anticipated to require ventilator support for more than 7 days and patients with evisceration through laparostoma also were excluded following enrollment, because NPWT application in these patients is technically not feasible/suitable. In short, patients with fascial disruptions without any additional complications were included in this study. 

Patients were recruited from 6 surgical units at the hospital. Patients in 3 units were enrolled in the NPWT group, and patients in the other 3 units were enrolled in the non-NPWT group. 

Sample size. Because this was a pilot study, a feasible sample size of 60 patients with 30 in each group was sought, considering the annual number of such cases in the authors’ hospital. A nonrandomized trial design and purposive sampling method was adapted due to logistical constraints in randomizing the patients. 

Study groups. Group A (NPWT group). Resident physicians in surgical units 1, 3, and 5 received a 2-week pretrial training on NPWT application under the supervision of the trained consultants; the trained residents applied the NPWT dressings. Before NPWT application, the wound was examined for existing slough, necrotic tissue, and pus discharge. Once slough, necrotic tissue, and pus were removed and the wound was considered fit for NPWT, sterile petrolatum gauze pieces were applied over the exposed bowel. A locally available sterilized medical grade polyvinyl alcohol foam sponge was placed at least 1 to 2 cm around the wound perimeter; 2 layers of polyvinyl white foam sponges with a Ryle’s tube (18 French size) were placed in between over the wound. An adhesive sterile drape was placed over the wound and snugly fit around the sponges. Waterproof adhesive bandage or plaster was applied along the entire edge of the drape to make it airtight. The Ryle’s tube was brought out through the upper part of the wound through a small tear in the adhesive drape. The Ryle’s tube was connected to a wall-mounted vacuum canister with wall suction through a connecting tube. Continuous negative pressure was set at 125 mm of Hg.^2-5 Brisk contraction of the foam sponge after application of negative pressure indicated a successful airtight seal dressing (see Figure 1).

The NPWT dressing was changed every 48 hours or if excessive dressing soakage was noted, whichever occurred first. Once the wound was covered with granulation cover or the wound contracted enough to carry out delayed secondary closure, the treating surgeon evaluated the wound for final closure technique—delayed secondary closure, split-thickness skin grafting (SSG), or healing by secondary intention. 

Group B (non-NPWT group). The exposed bowel was covered with petrolatum gauze pieces and then loosely packed with saline-soaked gauze. A Bogota bag then was applied using an urobag (see Figure 2) attached at both sides of the wound margin and joined in the middle to prevent exposure of the bowel. The dressing was changed at 24-hour intervals or sooner if the wound dressing was excessively soaked. Improvements in terms of the appearance of granulation tissue and contraction of the wound were monitored serially every 48 hours and at every change of the dressing in both groups. 

Parameters assessed. A data collection proforma, developed by the investigators and approved by the institute scientific advisory committee, was used to record the parameters. The deidentified proforma of each patient was used to record the demographic parameters such as age, gender, body mass index (BMI), comorbidity, type of surgery, type of stoma, albumin level, and primary and secondary outcome parameters. The primary outcome parameters studied were number of days until granulation tissue appearance and the number of days needed to make a decision regarding closure (surgical method or healing by secondary intention with granulation tissue cover). Data regarding the outcome parameters were recorded in the specified proforma by the principle investigator every 48 hours and whenever the NPWT or conventional dressing was changed. Secondary outcome parameters, including duration of postoperative hospital stay, development of an incisional hernia, and wound dehiscence also were noted. 

Granulation tissue appearance was defined as the day on which the first signs of granulation tissue (appearance of pink granular buds that were soft to the touch) were observed by the treating surgeon after starting either NPWT or non-NPWT therapy. Surgical closure of the wound was defined as either delayed secondary suturing of the rectus sheath and skin or covering of the wound with SSG without fascial closure once the wound was considered ready for closure. Healing by secondary intention was defined as complete coverage of the contracted wound surface with flat healthy granulation tissue at the end of either NPWT or non-NPWT therapy. The final result of healing by secondary intention was formation of weak scar tissue (see Figure 3).

Data collection and follow-up. Baseline patient characteristics including demographic data, comorbidities, BMI, most recent albumin levels before the start of the study, presence or absence of a stoma, and exposure to radiotherapy were recorded in the data collection forms at the time of patient recruitment. Discharged patients were followed-up monthly for up to 6 months for development of complications, including incisional hernia.

Statistical analysis. The collected patient data were entered and tabulated in Microsoft Excel 2016 and analyzed using SPSS software, version 19.0 for Windows (IBM Corp). 

Categorical variables including gender, comorbidities, presence or absence of a stoma, and exposure to prior radiotherapy were expressed as proportions and analyzed using chi-squared test or Fischer’s exact test. Continuous variables such as age, BMI, albumin, duration of postoperative hospital stay, and number of days until fascial closure were expressed as Mean ± standard deviation and analyzed using an independent t test or Mann Whitney U test, based on whether the data followed normal distribution. Postoperative day of wound dehiscence, the number of days for the appearance of granulation tissue, and the number of NPWT placements required also were assessed using Mean ± standard deviation and analyzed using an independent t test. A P value <.05 was considered significant. 

A total of 60 patients were enrolled in the study. Four (4) patients in the NPWT group could not complete the treatment (2 patients subsequently required prolonged ventilator support, and 2 discontinued treatment, 1 due to pain and 1 due to development of ECF) and were excluded from analysis. Ten (10) patients in the non-NPWT group could not complete the study (4 patients later required prolonged ventilator support, 3 died before completion of therapy, and 3 patients were discharged before completion of therapy) and were excluded from the analysis. As such, a total of 46 patients were analyzed, 26 in the NPWT group and 20 in the non-NPWT group. Men (33; 71.7%) outnumbered women (13; 28.3%) in the study population (a male to female ratio of 2.5:1). Mean albumin level, BMI, age, and gender distributions were comparable in both the groups. The distribution of comorbidities, especially diabetes mellitus (8 vs. 4; P = .509) were comparable in both the groups. Both groups were similar in terms of a history of radiotherapy (2 vs. 4; P = .38) (see Table 1).

The NPWT groups included 6 ileostomy and 7 colostomy patients, and the non-NPWT group included 3 ileostomy and 5 colostomy patients; the number of patients with stoma was comparable between groups (65.4% vs. 40%; P = .087). Emergency surgeries (31; 67.4%) outnumbered elective surgeries (15; 32.6) in both groups (P = .35). The majority of index surgeries after which the abdominal wound dehiscence occurred involved perforation peritonitis (21 cases; 45.6%); this distribution was similar in both the groups (see Table 2).

No significant difference was noted in the postoperative day of wound dehiscence among the 2 groups (6.62 vs. 6.40 days for NPWT vs. non-NPWT, respectively; P = .604). A statistically significant difference was noted between the 2 groups in terms of the number of days required for granulation tissue formation (2.92 vs. 6.65, for NPWT vs. non-NPWT, respectively; P = .001). The mean number of NPWT dressing placements required to achieve fascial closure/complete granulation cover in the NPWT group was 7.46 ± 2.02. In the non-NPWT group, the mean number of dressing placements was 12.12 ± 3.20, a statistically significant difference (P = .001). The number of days required for fascial closure/complete granulation tissue cover was significantly different between groups (15.5 days vs. 9.5 days, for non-NPWT vs. NPWT; P = .001). The mean length of postoperative hospital stay was significantly lower in the NPWT group (24.3 days vs. 37.9 days; P = .001] (see Table 3).

At the end of therapy, patients in both the groups achieved wound cover either by surgical closure or healing by secondary intention. One (1) patient underwent SSG and 1 patient underwent delayed primary closure after surgical mobilization of skin flaps. The remainder of the patients’ wounds healed by secondary intention (see Table 4). 

Among NPWT patients, 1 developed an ECF during the course of treatment and 1 developed a vesicocutaneous fistula (VCF) post NPWT therapy during the follow-up period of 4 months. No ECF or VCF occurred in the non-NPWT group. No intra-abdominal abscesses, ventral hernias, or wound dehiscence was reported in either group during the follow-up period.

This study to evaluate the efficacy of NPWT in achieving early fascial closure/cutaneous cover compared to a conventional therapy in postoperative laparotomy wound dehiscence included more men than women, but the number of men in each group was not significantly different (P =1 .06). Similar studies^6,7 reported that 64% of the NPWT group and 64.8% of the conventional therapy group were men; the reason for male preponderance of abdominal wound dehiscence is not clear. One possibility is that men are more likely to develop higher intra-abdominal pressure and wall tension, which can strain the tissues and wound edges causing sutures to cut through the muscles.⁸  

In terms of age, approximately 58.6 % of the entire study population was older than 50 years. Similarly, the retrospective analysis by Ko and Jung⁶ reported that 58.2% in conventional therapy and 48% in the NPWT therapy group were between 50 and 70 years of age. Advanced age is associated with defective tissue repair due to the decreased functional integrity of the immune system.⁶ A large retrospective review conducted by Walming et al⁹ found that BMI >25 was a risk factor for wound dehiscence (P = .001). Considering the limited number of patients in this study, a larger sample size with a greater number of obese patients may provide valuable information on this association.  

Diabetes is traditionally considered a risk factor for wound dehiscence, but a review of 27 retrospective studies by van Ramhorst et al⁷ showed diabetes had no significant effect on wound dehiscence occurrence. In the present study, comorbidities such as diabetes and hypertension were comparable in both the groups. 

The proportion of patients with a history of emergency surgery was higher than those who had elective surgery. This is not an unexpected finding. van Ramshorst et al⁷ reported that patients undergoing emergency surgery had an 1.8 times greater risk of developing wound dehiscence (odds ratio = 1.8) The number of study participants who had a colostomy or ileostomy was high (21), which is also in concordance with the literature; a multivariate analysis of 265 patients by Yilmaz et al¹⁰ reported that having a colostomy or ileostomy had a statistically significant effect on the rate of wound dehiscence (P = .002).

In the current study, the most common indication for index surgeries associated with abdominal wound dehiscence was bowel perforation, most likely due to local factors such as wound infection following contamination from bowel contents. An extensive literature review on wound dehiscence by Cliby¹¹ showed that the Centers for Disease Control and Prevention (CDC) class IV wounds—dirty surgical wounds, including bowel perforation—were associated with a greater than 30% chance of wound infection. Several previous studies^1,7 have shown that malignancy and sepsis are risk factors for postoperative wound dehiscence. 

The current study found that NPWT therapy facilitates granulation tissue formation, but the rate of granulation tissue formation was not evaluated as a separate factor in this study. In their cohort study among 21 patients receiving NPWT, Heller et al¹² proposed that removal of exudates harboring bacteria by continuous suction helps granulation tissue formation. 

In the current study, the number of days required for fascial closure consideration or complete granulation tissue cover was significantly less in the NPWT group. In a cohort study to evaluate the effect of NPWT conducted among 29 trauma patients with an open abdomen, Suliburk et al¹³ showed that a mean of 7 days (range 3–18 days) was required until fascial closure following NPWT therapy. This may be because only 2 patients in the current study were trauma patients or because Suliburk et al¹³ applied NPWT therapy over a planned laparostoma but not on wounds due to dehiscence following primary fascial closure, while the present study examined post-laparotomy wound dehiscence. In their prospective evaluation of negative pressure fascial closure in 45 patients with laparostoma following trauma, Miller et al¹⁴ reported a delayed fascial closure rate of 88%, which was significantly higher than their previous similar studies. Similar to the present study, few studies have evaluated the ability to achieve wound cover as the end point. In their cohort study involving 21 patients, Heller et al¹² showed 9 achieved fascial closure and stable cutaneous cover by skin grafting, and 6 achieved healing by secondary intension.

In the current study, apart from the difference in time to fascial closure/complete granulation tissue coverage between groups, no difference was noted in the rate of reaching the final end point (ie, the number of days needed to make a decision regarding closure either by surgical method or healing by secondary intention with granulation tissue cover). After excluding the patients in whom the final outcome could not be studied due to loss to follow-up, mortality, or discontinuation of treatment due to a complication during the study period, most patients in both groups achieved surgical fascial closure or complete granulation cover of the wound. However, the time required to reach complete wound closure is critically important, in part because it affects health costs. 

In this study, patients in the NPWT group were discharged from the hospital earlier than patients in the non-NPWT group. Ko and Jung⁶ reported a considerable (although not significant) reduction in the length of hospital stay in their NPWT group (n = 25) compared to control (n = 185) (P = .06), but the NPWT sample size could have limited the statistical power. In the current study, the difference in time to discharge was expected, because time until complete granulation coverage/fascial closure was significantly shorter in the NPWT group. 

In the present study, 1 patient in the NPWT group developed an ECF (3.8 %), and 1 patient developed a VCF (3.8%) after discharge at follow-up of 4 months; no patients in the control group developed an ECF. Although an ECF has been reported as a possible complication of NPWT, the exact risk of ECF formation following NPWT therapy is not well established. In their prospective cohort study, Perez et al¹⁵ proposed that the most common risk factors for development of an ECF are underlying illness with a poor general condition rather than due to NPWT.  In their multicenter observational study, Bobkiewicz et al¹⁶ established that NPWT can be used safely in the treatment of an ECF. Other factors thought to play a role are intra-abdominal sepsis, sutured enterotomy, presence of intestinal anastomosis in the wound, mechanical trauma either due to injury or excessive intraoperative bowel handling, sutures in wound area causing cut through injuries to the bowel, and presence of exposed bowel in the wound.^12-16 NPWT-associated factors that can contribute are suction tubing, which may produce bowel erosions, and the use of more bowel adherent black polyurethane foam.^12-17 

A systematic review by Misky et al¹⁷ concluded that NPWT can be considered safe treatment for an ECF, but multiple treatment strategies were proposed  to reduce the risk of ECF occurrence. Strategies related to NPWT include covering the exposed bowel with a nonadherent sheet before applying the NPWT dressing, using fewer adherent white polyvinyl foam sponges, applying continuous suction instead of intermittent suction (thereby minimizing mechanical shear and deformation of the bowel), and maintaining suction pressure at 75 mm Hg rather than at 125 mm of Hg.^15-18 

One (1) VCF developed in a patient in the NPWT group who had received post-pelvic radiotherapy for carcinoma of the cervix and was operated for intestinal obstruction. The patient developed a delayed VCF at the 3-month follow-up after discharge; the fistula was likely due to the prior radiotherapy and extensive surgery. The incidence of ECF following NPWT in this study is similar to other studies. In their prospective evaluation, Miller et al¹⁴ reported an incidence of 2.3% in the study population, and in their prospective observational study of 37 patients, Perez et al¹⁵ reported an incidence of 3%. 

In the current study, no intra-abdominal abscess formation or ventral hernias occurred. The follow-up period in the present study was up to 6 months, which may not be sufficient to detect late complications such as an incisional hernia.^16,18 Miller et al¹⁴ reported a 2.3% incidence of incisional hernia at the mean follow up period of 185 days. 

Because this was a pilot study, only 60 patients were included, which might be a limiting factor for evaluating certain outcomes. Larger studies comparing more patients would provide more precise information. A shorter duration of follow-up (6 months) limited the ability to evaluate the late complications such as incisional hernia.^16,17 Hence, more prospective, randomized controlled trials involving large sample sizes and prolonged follow-up for ventral hernia formation is recommended. Measuring wound size is technically difficult due to change in the size of the wound during inspiration and expiration and with postural changes. The percentage reduction in the size of the wound was not quantified in the present study, similar to other studies. 

A single center, prospective, parallel-arm, nonrandomized pilot study was conducted to compare a NPWT and non-NPWT method of managing wound dehiscence among 46 patients. The number of days to complete granulation tissue coverage/fascial closure procedure was significantly lower in the NPWT group. The mean length of postoperative hospital stay was significantly higher in conventional group compared to the NPWT group. No incidence of intra-abdominal abscess, wound dehiscence, or incisional hernias was reported in either group at the end of the follow-up period, but 2 patients in the NPWT group (7.6%) developed a fistula. Prospective, randomized controlled trials involving larger samples and prolonged follow-up for ventral hernia formation are recommended. 

Dr. Chandrasekhar is a former junior resident; Dr. Sureshkumar is a professor; Dr. Manwar is a professor; Dr. Elamurugan is an associate professor; Dr. Nelson is a former senior resident; Dr. Anandhi is an associate professor, Department of General Surgery; and Dr. Palanivel is an associate professor, Department of Preventive and Social Medicine, Jawaharlal Institute of Postgraduate Medical Education and Research (JIPMER), Puducherry, India. Please address correspondence to: Sathasivam Sureshkumar, MS, DNB, MNAMS, FMAS, FICS, Professor of Surgery, JIPMER, Pondicherry- 605006, India; drsureshkumar08@gmail.com.