Significant Reduction of Wound Infections With Daily Probing of Contaminated Wounds: A Prospective Randomized Clinical Trial

Hypothesis  Local wound management using a simple wound-probing protocol (WPP) reduces surgical site infection (SSI) in contaminated wounds, with less postoperative pain, shorter hospital stay, and improved patient satisfaction.

Design  Prospective randomized clinical trial.

Setting  Academic medical center.

Patients  Adult patients undergoing open appendectomy for perforated appendicitis were enrolled from January 1, 2007, through December 31, 2009.

Interventions  Study patients were randomized to the control arm (loose wound closure with staples every 2 cm) or the WPP arm (loosely stapled closure with daily probing between staples with a cotton-tipped applicator until the wound is impenetrable). Intravenous antibiotic therapy was initiated preoperatively and continued until resolution of fever and normalization of the white blood cell count. Follow-up was at 2 weeks and at 3 months.

Outcome Measures  Wound pain, SSI, length of hospital stay, other complications, and patient satisfaction.

Results  Seventy-six patients were enrolled (38 in the WPP arm and 38 in the control arm), and 49 (64%) completed the 3-month follow-up. The patients in the WPP arm had a significantly lower SSI rate (3% vs 19%; P = .03) and shorter hospital stays (5 vs 7 days; P = .049) with no increase in pain (P = .63). Other complications were similar (P = .63). On regression analysis, only WPP significantly affected SSI rates (P = .02). Age, wound length, body mass index, abdominal circumference, and diabetes mellitus had no effect on SSI. Patient satisfaction at 3 months was similar (P = .69).

Conclusions  Surgical site infection in contaminated wounds can be dramatically reduced by a simple daily WPP. This technique is not painful and can shorten the hospital stay. Its positive effect is independent of age, diabetes, body mass index, abdominal girth, and wound length. We recommend wound probing for management of contaminated abdominal wounds.

Postoperative surgical site infections (SSIs) remain a major source of illness and a cause of death in the surgical patient.¹ These infections number approximately 500 000 per year among an estimated 27 million surgical procedures and account for approximately one-quarter of the estimated 2 million nosocomial infections in the United States each year.^2,3 Infections result in pain, longer hospitalization, poor cosmesis, incisional hernias, and overall higher costs.⁴

Except for perioperative systemic antibiotics, no single local wound therapy has been proven to decrease the SSI rate in large randomized trials. Examples of local therapy considered include topical antibiotics, subcutaneous drain placement, and delayed primary closure.^5-8 There is no standard wound management for contaminated wounds, but the topic has not been aggressively pursued for decades.

The most experience with management of contaminated wounds is from appendectomy for perforated appendicitis. Appendectomy is the most commonly performed emergency surgical procedure in the United States (280 000 procedures per year).² Surgical site infection is the most common postoperative complication for this procedure. Before the use of antibiotics, postoperative SSI after appendectomy for a perforated appendicitis was as high as 70%.⁹ Since the introduction of routine prophylactic antibiotic use, infection rates in the most contaminated groups have reduced drastically.¹⁰ Infection rates in US National Nosocomial Infection Surveillance system hospitals are reported to be 7.1% for contaminated wounds, although this number can increase to 50% in certain subgroups.¹¹

Wound probing, a simple bedside technique that may reduce SSI in contaminated wounds, combines the benefits of primary (less labor for wound care and less pain) and secondary (lower rates of SSI) wound closure. The efficacy of wound probing was evaluated in only 1 retrospective study¹² that involved pediatric patients undergoing an endorectal intestinal pull-through procedure for ulcerative colitis. The results showed a dramatic decrease in SSIs among patients who underwent wound probing (24.4% vs 4.3%). Some surgeons use this method with anecdotally good results.

We sought to prospectively validate wound probing in a prospective randomized clinical trial. The objective was to demonstrate the short- and long-term superiority of wound probing over primary closure in the management of contaminated wounds while quantifying the pain that may be attributed to this procedure.

We conducted a prospective, randomized, single-blind, single-institution controlled trial. Eligible patients included all adults with perforated appendicitis admitted to the Los Angeles County University of Southern California Medical Center's nontrauma Emergency Surgery Service who were scheduled for or who had undergone open appendectomy within 24 hours before enrollment (Table 1). Approval to perform the trial was obtained from the Health Sciences Institutional Review Board at the University of Southern California. Written informed consent was obtained before surgery or within 24 hours after surgery. Perforated appendicitis was defined as the finding of perforation along the length of the appendix and/or free pus in the abdomen. Patients with gangrenous appendicitis alone were considered ineligible.

Study patients started perioperative intravenous antibiotic therapy per the standard appendectomy protocol at our hospital (ertapenem sodium, 1 g/d, or levofloxacin, 500 mg/d, plus metronidazole hydrochloride, 500 mg every 8 hours for patients with penicillin allergy). Antibiotic therapy was continued until patients were afebrile for longer than 24 hours and had a normal or decreasing white blood cell count.

Baseline demographic information was recorded, such as age, sex, ethnicity, height, weight, circumferential abdominal girth measurements, and comorbidities.

After obtaining informed consent, patients were randomized in a 1:1 fashion to loose primary skin closure with staples (control arm) or loose primary skin closure with staples and wound-probing protocol (WPP arm). Randomization was performed using a random draw from a group of 100 sealed envelopes, 50 of which contained designation for wound probing or non–wound probing.

Appendectomy was performed via standard right lower quadrant incision. Intraoperative cultures were taken. The wound was closed with staples placed at 2-cm intervals. Unblinded research staff performed the daily wound assessments and procedures. The patient's primary surgical team was blinded to the randomization arm and was encouraged to treat the patient and manage the wound as they deemed appropriate.

Patients randomized to the control arm underwent daily swab of their closed appendectomy incision with povidone-iodine. Patients randomized to the WPP arm underwent standardized wound probing. Similar to the control condition, patients in the WPP arm underwent povidone-iodine solution swab of their incision. Next, a dry, sterile cotton-tipped applicator was used to penetrate the skin and soft tissue between the skin staples, reaching down to the external oblique fascia. Any fluid that extruded as a result of the probing was blotted dry. The wound was then covered with a dry dressing. This WPP procedure was performed only once in the morning and repeated every day until the wound was dry and no longer penetrable. This procedure was not repeated later in the day if there was continued fluid drainage.

On postoperative days 1, 3, and 5, quantitative bacterial cultures were obtained from the cotton-tipped applicators used to probe the wound in the WPP arm. If the wounds of control- or WPP-arm patients were deemed to be infected, cultures of the drainage from these wounds were also taken.

During their hospitalization, patients underwent daily assessment of vital signs and white blood cell count. Wound assessment included daily monitoring based on the ASEPSIS criteria for SSI (additional treatment; presence of serous discharge, erythema, purulent exudate, and separation of the deep tissues; isolation of bacteria; and duration of inpatient stay) (Table 2).¹³ Patients with an ASEPSIS score greater than 20 were considered to have an SSI. In cases of purulent drainage and an ASEPSIS score less than 20, the wound was considered infected per the Centers for Disease Control and Prevention definition of nosocomial SSIs.¹⁴ If an SSI was diagnosed in patients in either arm of the study, the wound was opened and loosely packed with moist gauze. The gauze dressing was changed 3 times daily, allowing for healing by secondary intention.

Patients were queried daily for their pain score based on the Wong-Baker FACES visual analog scale.¹⁵ For those in the WPP arm, the pain score was recorded after the probing of their wound.

Length of hospital stay was recorded along with any unexpected adverse events related or unrelated to treatment, which were reported to the institutional review board. Study participants were followed up as outpatients at 2 weeks and at 3 months after discharge. Wounds were assessed for infection, and pain scores were recorded. Each patient's satisfaction with the operative result was recorded according to a 3-point scale (satisfied, somewhat satisfied, or not satisfied).

The primary outcome measure was the rate of postoperative SSIs, which was objectively based on the daily ASEPSIS score for the wound and the daily assessment by the blinded surgical team. Secondary outcomes included length of hospital stay, pain score, and patient satisfaction with the surgical results at 2 weeks and at 3 months.

The purpose of this study was to evaluate the WPP for management of contaminated wounds. To our knowledge, this is the first prospective randomized controlled trial evaluating wound probing. The objective was to show superiority of the WPP compared with primary skin closure on the basis of reduced SSI and reduced hospital stay, without a significant increase in pain score.

Our hospital has an approximately 25% SSI rate after open appendectomy for perforated appendicitis with primary wound closure. To detect a 50% reduction in SSI between the 2 procedures at the .05 level of significance and 80% power, we would need to randomize 336 patients. We perform 300 appendectomies annually, of which one-third are for perforated appendicitis. This is a pilot study randomizing 100 patients in a prospective controlled trial to evaluate the efficacy and feasibility of wound probing.

Statistical analysis included the 2-sample t test with equal variances, χ² test, and Fisher exact test for categorical variables, which are presented as mean (SD). For nonnormal numerical variables, the Wilcoxon rank sum test was performed, and data are presented as median and range.

During a 2-year period, 76 patients were enrolled in the trial and equally distributed between the control and WPP arms. One patient in the control arm withdrew on postoperative day 2 (Figure). Patients in the 2 arms had similar characteristics, including age, sex, and race (Table 3). They were also similar with respect to factors and comorbidities that may affect wound healing or SSI after open appendectomy, including diabetes mellitus, body mass index, and abdominal girth.

Postoperatively, patients had similar incisional lengths (mean, 8 cm). Patients were afebrile, with a normal heart rate and normalization of their white blood cell count within a similar postoperative period (Table 4). Patients in the control arm reached a wound ASEPSIS score of 0 significantly earlier than those in the WPP arm (2.0 vs 4.0 days; P = .007); however, the range of scores was much wider among the control patients (0-20 vs 0-7 days). Those in the WPP arm had a significantly shorter hospital stay (5 vs 7 days; P = .049).

Follow-up visits at 2 weeks were attended by 34 of 37 control-arm patients (92%) and 37 of 38 WPP-arm patients (97%). At 3 postoperative months, follow-up visits were attended by 25 of 37 control-arm patients (68%) and 24 of 38 WPP-arm patients (63%). Postoperative wound complications were significantly lower in the WPP arm (19% vs 3%; P = .03) (Table 5). The patients were similar in all other postoperative complications, including intra-abdominal abscess, although there was a trend toward more infectious and noninfectious complications in the control arm. Long-term follow-up included the pain score, wound ASEPSIS score, and patient satisfaction. The patients had similar long-term outcomes at 2 weeks and 3 months postoperatively (Table 6).

Wound probing is a simple daily technique that can be applied to contaminated abdominal wounds to significantly reduce SSIs without increasing pain. In our study, we standardized all patients to open appendectomy via right lower quadrant wounds. The wounds were probed daily until the wound was closed and no longer draining, which took approximately 3 days and rarely as long as 5 days. The patients who underwent wound probing had a significantly lower SSI incidence (19% vs 3%) and did not report increases in pain from the probing technique. Patients who had SSIs required an opening of the wound and packing with saline-soaked gauze 3 times a day. By reducing SSI, there was a significantly shorter hospital stay (7 vs 5 days) and also a reduction in time spent by the health care providers attending to wound care.

Surgical practice varies with regard to wound management. Even for a procedure as common as open appendectomy, there is little consensus concerning optimal wound management for perforated appendicitis. Primary closure is a common choice and results in failure in 7% to 50% of patients.¹¹ When primary closure fails or is deemed inappropriate, the wound is opened and allowed to close by secondary intention (ie, packing an open wound multiple times a day). Closure by secondary intention can be painful to the patient, can be labor intensive, and has a poor cosmetic outcome. Nevertheless, it is the most widely used closure technique in adults with contaminated wounds. To avoid these problems, some surgeons prefer to delay primary closure (ie, packing an open wound for 4 days followed by primary closure).¹⁶ Compared with closure by secondary intention, delayed primary closure offers improved cosmetic outcome and reduction in morbidity and wound care costs once closed. However, the morbidity and costs of packing the wound for the first 4 days are similar to those of closure by secondary intention. In obese patients, all 3 methods can result in amplification of the morbidities.

Based on our results, we recommend that wound probing be practiced for contaminated wounds due to open appendectomy for perforated appendicitis. We believe that our results can also be extrapolated to other contaminated abdominal wounds regardless of their incisional length. Wound probing for all contaminated laparotomy wounds has been a common practice of one of us (S.T.) and of other surgeons, with anecdotal results until now. The favorable results from our prospective randomized study have strengthened our belief in this technique.

The method by which wound probing reduces SSIs is not clearly understood at this time. We surmise that this technique allows for drainage of contaminated fluid within the soft tissue, thus reducing the bacterial count while maintaining a moist wound for improved healing. We are currently analyzing the bacteriology data gathered from this study to gain insight into the process by which wound probing helps in wound healing in the face of contamination.

We recommend that wound probing be practiced in the first-line management of incisions after open surgery, such as bowel resection, in the face of contamination. Use of this technique will reduce SSI and decrease hospital stay, both of which may result in improved long-term outcomes such as fascial dehiscence and incisional hernias, reduced labor for wound care by nursing staff and physicians, improved cosmetic outcome with primary closure, and possibly reduced overall cost.

Correspondence: Shirin Towfigh, MD, Department of Surgery, Cedars-Sinai Medical Center, 8635 W Third St, 650W, Los Angeles, CA 90048 (shirin.towfigh@cshs.org).

Accepted for Publication: March 24, 2010.

Author Contributions:Study concept and design: Towfigh, Yacoub, and Berne. Acquisition of data: Towfigh, Clarke, Yacoub, Mason, and Katkhouda. Analysis and interpretation of data: Towfigh and Pooli. Drafting of the manuscript: Towfigh and Pooli. Critical revision of the manuscript for important intellectual content: Towfigh, Clarke, Yacoub, Pooli, Mason, Katkhouda, and Berne. Statistical analysis: Towfigh and Pooli. Obtained funding: Towfigh and Yacoub. Administrative, technical, and material support: Towfigh, Katkhouda, and Berne. Study supervision: Katkhouda.

Funding/Support: This study was supported in part by a James H. Zumberge Faculty Research and Innovation fund grant from the University of Southern California.

Previous Presentation: This study was presented at the 81st Annual Meeting of the Pacific Coast Surgical Association; February 14, 2010; Kapalua, Hawaii.