Outcomes and Factors Influencing Time to Healing in Patients Undergoing Surgical Debridement With Primary Closure.

Alexandra Tilt, MD; Reuben A. Falola, MD, MPH; Anagha Kumar, MA; Tessa J. Campbell, MD; Jacob M. Marks, MD; Christopher E. Attinger, MD; Karen K. Evans, MD

Abstract and Introduction

Abstract

Introduction: Plastic surgeons are often consulted by other surgical teams for management of wound dehiscence following abdominopelvic surgery.

Objective. The purpose of this study is to determine whether operative debridement and primary closure of abdominopelvic wounds are safe and expeditious for patients.

Materials and Methods: A retrospective analysis was conducted on a database of patients who underwent operative debridement and closure at a single institution between January 2011 and December 2015 for dehisced abdominal or pelvic wounds acquired from prior obstetric, gynecologic, transplant, plastic, or general surgery procedures.

Results: Of the 163 patient records identified, 43 patients met inclusion criteria. The median time from final debridement and primary surgical closure to complete wound healing was 27 days. Time to healing differed significantly by index procedure type (P = .004), with obstetric procedures requiring the shortest median time (12.0 days) and general surgery procedures requiring the longest (39.5 days). Wound healing took 3.6 times longer for patients with diabetes (P = .046) and 11.4 times longer for patients who experienced delayed superficial wound healing or redehiscence (P = .003). Nevertheless, with the exception of 4 patients who died of other causes, all wounds (39/39; 100%) achieved complete wound closure.

Conclusions: Operative debridement and closure of abdominopelvic wound dehiscence through a multidisciplinary team approach with plastic surgery results in expeditious wound healing with minimal complications, and it may be safer and more cost effective than healing by secondary intention.

Introduction

Wound dehiscence following abdominopelvic surgery is an uncommon but serious complication of general, transplant, obstetric, gynecologic, and plastic surgeries. The incidence of surgical wound dehiscence of the abdomen reported in the literature is about 3%.^[1,2] This complication of abdominopelvic surgery can lead to prolonged hospital stays, increased cost to the health care system, and increased mortality.^[3] In addition, abdominal wound dehiscence has been associated with a high incidence of incisional hernia, low body image, and low quality of life.^[4]

Due to the severity of the sequelae of postoperative abdominal wound dehiscence, research has been devoted to identifying risk factors for this complication. Several studies^[2,5,6] have shown age, gender, diabetes, hypertension, obesity, steroid use, infection, hemodynamic instability, and malignancy correlate to abdominal wound dehiscence. Some research has focused on validating risk scores based on these variables in order to predict which patients are at a higher risk of wound dehiscence, guiding clinical management to help prevent this complication of abdominopelvic surgery.^[7]

Although identification of risk factors may lead to prevention of postoperative abdominal wound dehiscence, there is little consensus on the management of wound dehiscence in patients once it occurs. If the wound is superficial, many patients are treated nonoperatively through healing by secondary intention aided by dressing changes.^[8] The use of negative pressure wound therapy (NPWT) also has been shown to be effective in developing granulation tissue faster than dressing changes alone, albeit with a higher cost.^[9,10] In the obstetrics and gynecology literature, minor wound dehiscence following cesarean section (C-section) has been treated successfully with secondary closure under local anesthesia.^[8,11]

When conservative methods fail, plastic surgery consultation may be warranted. However, little research has been devoted to surgical management, specifically operative debridement with primary closure, of these wounds in the plastic surgery literature. One recent retrospective review^[12] analyzed donor site complications from abdominally based free flap breast reconstruction and found delayed primary closure of dehisced wounds led to significantly faster healing times than healing by secondary intention. Presented herein is the first retrospective analysis focused on outcomes of simple primary closure of wound dehiscence following various types of abdominopelvic surgery.

Materials and Methods

A retrospective chart review of electronic medical records was conducted at a single academic medical center (MedStar Georgetown University Hospital, Washington, DC) of all patients undergoing operative debridement and primary closure of abdominal wounds from January 2011 to December 2015. Georgetown University Institutional Review Board (Washington, DC) approval was granted for this study (IRB 2018–0008). Records were identified using International Classification of Diseases, Ninth Revision, codes 879.2, 879.3, and 879.5, which code for open wounds of the anterior or lateral abdominal wall, uncomplicated, or complicated, respectively.

Inclusion criteria consisted of patients who had documented dehiscence of an abdominal incision following primary closure during a prior surgery involving the abdomen or pelvis by any surgeon. These patients also must have had at least 1 operative dictation by a plastic surgeon at the study institution documenting subsequent surgical debridement with primary closure of the wound.

All patients with abdominal wall wounds in this study were managed with operative debridement, intraoperative cultures, and an antibiotic regimen determined by infectious disease consultation. The decision to proceed with immediate or delayed primary closure was based on a combination of culture data and appearance of the wound immediately after debridement in the operating room. Wounds deemed appropriate for closure displayed a granular base with adequate bleeding and no evidence of biofilm or necrotic tissue. The closure technique focused on meticulous soft tissue handling and wound approximation with vertical mattress sutures using Prolene (Ethicon Inc, Somerville NJ), with minimal deep suture placement. Staples also were used for epithelial approximation. See case example in Figure 1.

Patients were excluded based on the following criteria: incomplete records or missing data on date of wound onset; lack of follow-up; use of mesh during any surgical procedure following wound dehiscence; use of skin graft, pedicled flap, or free flap to close the wound; and combination of operative debridement and primary closure of the wound with other surgical procedures, including hernia repair or component separation.

Data for each patient were recorded, including patient demographics, comorbidities (obesity, hypertension, diabetes, chronic obstructive pulmonary disease, coronary artery or peripheral vascular disease, chronic renal failure, solid organ transplant recipient), date of initial surgery, type of surgery, date of wound onset, presence of infection, hospital admission date, date(s) of operative debridement and primary closure, and hospital discharge date. Each procedure for operative management of the wound was counted as 1 debridement, including the procedure during which primary closure was achieved. Wound healing was defined as complete reepithelialization of the incision, and the date of wound healing was approximated based on documentation from a postoperative clinic visit at the MedStar Georgetown Center for Wound Healing (Washington, DC). In addition, complications for each patient were recorded, including any development of subsequent wound infection or redehiscence, hernia formation, and death.

Statistical analysis was performed using the statistical software R (R Foundation for Statistical Computing, Vienna, Austria). Data analysis consisted of fitting Cox proportional hazards models to estimate the relative risk of healing. Kaplan-Meier curves and descriptive statistics in the form of means (standard deviations [SD]) and counts (percentages) are presented in this paper. The Peto-Peto modification of the Gehan-Wilcoxon test was performed to test for statistically significant differences in the survival curves.

Results

Over the 4-year period, a total of 163 patient records were identified for review, of which 43 patients met inclusion criteria. The average age of these patients was 54 years and 65.1% of patients were female (Table 1). Caucasians accounted for 46.5% of patients, while 27.9% of were African-American. The average body mass index (BMI) of patients was 31.93. Patients with diabetes accounted for 23.3% (10/43) of the study group; 41.9% (18/43) of patients had hypertension, 18.6% (8/43) had hyperlipidemia, and 7.0% (3/43) had chronic renal failure. Of all 43 patients, 14.0% (6/43) were current smokers and 41.9% (18/43) were former smokers. Initial abdominopelvic surgeries for each patient were indexed as follows: 46.5% (20/43) general surgery, 16.3% (7/43) transplant, 14.0% (6/43) gynecologic, 11.6% (5/43) obstetrics, and 11.6% (5/43) plastic surgery.

Median times to wound healing were calculated based on various time points in surgical management (Table 2). The median time from date of initial abdominopelvic surgery to date of first debridement by plastic surgery was 40 days (mean 154.0, SD 272.2). The median time between first debridement and final debridement with closure was 3 days (mean 4.6, SD 7.6). The average total number of debridement procedures per patient, including final debridement and closure, was 2.05 (SD 0.87). The overall median time from date of primary closure to date the wound was documented as healed was 27 days (mean 106.8, SD 266.0). The median for total time of plastic surgery involvement, defined as date of first debridement to date wound healed, was 33 days (mean 111.4, SD 266.8).

A Kaplan-Meier curve was used to illustrate time to healing based on initial surgical procedure (Figure 2). Time to healing differed significantly by procedure type (P = .00437), with obstetric procedures requiring the least amount of time to heal (median 12.0 days). Median healing time was greater for plastic surgery patients (23.0 days), followed by transplant (28.0 days), gynecologic (38.0 days), and general surgery procedures (39.5 days).

Complications following debridement and primary closure included 3 patients who developed hernia (7.0%), 4 who had documented evidence of wound infection (9.3%), and 4 who died before their wounds healed (9.3%). Two patients (4.7%) experienced delayed superficial wound healing, requiring dressing changes until the wound had completely healed. Redehiscence of wounds occurred in 3 patients (7.0%), 2 of which were managed with NPWT and 1 was taken back to the operating room for repeat debridement and primary closure. Aside from the 4 patients who died, all wounds eventually healed (39/39, 100%).

Univariate Cox proportional hazards regression analysis was conducted to determine how each demographic variable and complication affected overall time to healing. Gender, BMI, diabetes, hypertension, hyperlipidemia, chronic renal failure, and smoking status did not significantly affect time to healing (Table 3). However, the type of procedure from which the wound was derived did significantly impact healing time. Compared with obstetric procedures (C-section), wounds from all other procedures were significantly more likely to require more time for healing (Table 4). Wounds resulting from general surgery procedures took 6.1 times longer to heal (RR, 0.164; P = .001), gynecologic procedures took 7.9 times longer to heal (RR, 0.127; P = .003), plastic surgery procedures took 3.6 times longer (RR, 0.280; P = .056), and transplant procedures took 5.3 times longer to heal (RR, 0.188; P = .010).

In a multivariate Cox proportional hazards model, diabetes, redehiscence, and procedure type were found to be significant (Table 5). Holding all other factors (gender, race, BMI, smoking status, hypertension, hyperlipidemia, chronic obstructive pulmonary disease, chronic renal failure, procedure type, total number of debridements, and redehiscence) constant, patients with diabetes took 3.8 times longer to heal than those without diabetes (RR, 0.265; P = .046). Again holding all other factors (including diabetes) constant, patients who experienced delayed superficial healing or wound redehiscence took 11.4 times longer to heal than those who did not (RR, 0.088; P = .003).

Compared with patients in the obstetrics procedure group (5/43), patients in the gynecologic surgery group (6/43) took 11.2 times as long to heal when all other factors were held constant (RR, 0.089; P = .018).

Discussion
This is the first retrospective analysis of outcomes following debridement and primary closure of abdominal wall wounds from many types of abdominopelvic surgery. Overall, the results are favorable with a median healing time of 27 days following wound closure. In a subgroup analysis within a retrospective review of donor site complications from abdominally based free flap breast reconstruction, Mirzabeigi et al[12] examined 11 patients who underwent delayed primary closure of abdominal wound dehiscence, either in the operating room or as an outpatient office procedure. They found that 90% of these wounds were healed within 1 month of delayed primary closure, compared with 24.6% of wounds treated with wet-to-dry dressing changes or NPWT.[12] Of note, that patient population consisted only of women who were healthy enough to undergo elective free flap breast reconstruction, while the patients herein healed abdominal wounds from many other types of surgery, including transplant and general surgery procedures.

The comorbidity profile of the current patients was fairly typical for surgical patients in an inner-city academic medical center: 23.0% with diabetes, 55.8% former or current smokers, and an average BMI of 31.93. In congruence with the literature,[13,14] patients with diabetes in the current analysis took significantly longer to heal their wounds than those without diabetes. Smoking status and higher BMI did not significantly correlate to longer healing times, as would be expected from the literature;[15–17] this was likely a manifestation of the relatively small number of patient records included. A larger analysis may have yielded more statistically significant results.

The overall complications were minimal, with only 1 patient requiring reoperation (2.3%). Not surprisingly, patients with delayed superficial healing or redehiscence of wounds took more than 11 times longer to heal than those without this complication. Aside from the 4 patients who died from causes unrelated to their wounds, all patients (39/39; 100%) eventually healed in a median of 27 days.

The data pooled all wounds that were managed with simple primary closure, including C-section incision dehiscence. These wounds occurred in healthy, young female patients and predictably healed faster, generally within 2 weeks, than other types of procedures. Interestingly, debridement and primary closure in this population has been encouraged within the obstetrics and gynecology literature for more than 25 years. In a study by Walters et al[18] from 1990, patients who underwent operative reclosure of surgical wounds healed in an average of 15.8 days, which is comparable to the median healing time of 12 days for the current subset of obstetric patients. Likewise, a study by Dodson et al[11] in 1992 showed that secondary closure of surgical wounds took an average of 17.6 days to heal, as compared with 61.2 days for wounds healed through secondary intention. It is important to note the discrepancy in healing time between obstetric and gynecologic patients, with gynecologic patients in the current analysis taking nearly 8 times longer to heal. This likely can be attributed to differences in comorbidities between these 2 groups, with gynecologic patients often being older and having advanced stage cancer.

As the expense of health care continues to be a prominent issue, the cost effectiveness of operative debridement and closure of abdominal wounds is an important concern. Mirzabeigi et al[12] showed that patients who underwent delayed primary closure had a significantly lower total cost of reconstruction compared with those whose wounds were treated with NPWT. In addition, these patients required fewer procedures for scar revision and hernia repair, fewer visits to the emergency room, and fewer hospital admissions.[12] Future studies should examine the setting of operative reclosure and associated costs in the present patient population. While the patients herein underwent usually 1 or 2 procedures in the operating room during an inpatient hospital stay, some research has suggested that reclosure of superficial wounds can be done under local anesthesia in the outpatient clinic.[8,12] Obviously, the complexity of the wound and comorbidities of the patient should dictate an appropriate setting for the procedure, and cost savings should not overshadow a high standard of care.

Limitations

There are a number of limitations to the current study. First, it is a retrospective chart review with no control group; thus, the primary endpoint of time to wound healing for patients undergoing operative debridement and closure cannot be directly compared with a group of similar patients who were not managed surgically. Furthermore, the comorbidity profile and surgical history of patients was quite heterogeneous and undoubtedly played a role in time to wound healing. As such, the median time to healing in some analyses did not match the mean time to healing, and wide standard deviations were reported.

Nevertheless, the current results do highlight the need for a multidisciplinary team approach to manage dehiscence after abdominopelvic surgery, and this research demonstrates that a simple protocol of debridement and primary closure can result in high healing rates with minimal complications. Secondary wound healing has an inherent risk of infection, especially in high-risk patients such as those with diabetes or immunosuppression. Primary closure may be more favorable for patients than long-term dressing changes, frequent postoperative clinic visits for wound management, and delayed return to work or daily activities. Future studies will include a control group to demonstrate that debridement and delayed primary closure is not only safe but superior to secondary wound healing.

Conclusions

Abdominal wall dehiscence is a relatively common occurrence following abdominopelvic surgery in busy tertiary care hospitals with high operative volume and high case mix index. Using a team approach with plastic surgery, timely closure of these wounds may be a simpler, more efficient, and more cost-effective modality to manage abdominal wall dehiscence than prolonged dressing changes or NPWT, which are associated with high costs for coordinated outpatient care and supplies. Patient satisfaction also may be improved with operative debridement and closure, as many studies have shown the significant psychosocial implications of having a chronic wound.^[4,19] Further prospective studies comparing primary closure of dehisced surgical wounds to other modalities of wound healing are underway.