Methods
The original GSUC trial was prospective, randomized, and designed according to CONSORT guidelines (Figure 1). Participants were men and women age 18 years or older who were admitted to the University of Chicago Medical Center with acute wounds resulting from trauma, dehiscence, or surgery between October 2006 and May 2008. Table 1 illustrates exclusion criteria. All eligible patients were offered the opportunity to participate in the study. Upon enrollment into the study, each patient was assessed for the presence of a wound infection. Patients with grossly necrotic wounds or systemic sepsis caused by wound infection were initially excluded, but were considered eligible after the wound was adequately debrided, and the patient's septic state had resolved. Wound infection was defined as the presence of an abscess with purulent material, with or without associated cellulitis, and a positive rapid slide quantitative Gram stain showing > 10 colony forming units per gram of tissue. Tissue from the initial wound biopsy was retained for culture to guide antimicrobial therapy. Once patients were allocated to either the infected or the noninfected wound category, a stratified randomization scheme was used to assign VAC or GSUC therapy in a 1:1 ratio.
(Enlarge Image)
Figure 1.
CONSORT statement flow diagram. Forty-five patients were randomized to the GSUC intervention group (18 patients with infected wounds) and 42 patients were randomized to the VAC intervention group (13 patients with infected wounds).
In the VAC arm, GranuFoam (KCI, San Antonio, TX) black sponge was applied to wounds and sealed with an occlusive plastic cover. Continuous suction at 75 mm Hg to 125 mm Hg was initiated and the dressing was changed every 48 hours, as recommended by VAC therapy guidelines. In the GSUC arm, a gauze dressing (Kerlix 4.5 inch roll, Covidien, Mansfield, MA) moistened with 0.9% normal saline was applied to the wounds. A red rubber catheter (Bard Medical, Covington, GA) was placed in the center of the dressing, and the dressing was then sealed with an occlusive cover (Ioban 2 Antimicrobial Incise Drape, 3M, St. Paul, MN). Continuous wall suction at 75 mm Hg to 80 mm Hg was applied and the dressings were changed daily, as recommended for optimal wound healing in the original descriptions of this technique.
Patients with infected wounds were managed the same as the noninfected cohort described in the original GSUC trial, except that in addition, the VAC or GSUC dressing was soaked with quarter-strength Dakin's solution (Century Pharmaceuticals, Indianapolis, IN) 3 times a day for 30 minutes while suction was on hold. After 48 hours of treatment with Dakin's solution, a second wound biopsy and rapid slide quantitative gram stain were performed. If > 10 colony-forming units per gram of tissue were still present, a mixture of liquid sulfamylon and nystatin (mafenide acetate [5%] 50 gms and nystatin powder 15 gms [100,000 units per gram] in 1 liter water as prepared by the inpatient pharmacy) irrigation was substituted for the Dakin's solution. On the other hand, if the second wound biopsy < 10 colony forming units per gram of tissue, wound irrigation was discontinued. Another wound biopsy and quantitative rapid slide gram stain were performed at 96 hours from the start of treatment with GSUC or VAC. The study protocol specified that SAWT would be considered to have failed if > 10 colony-forming units per gram of tissue remained at 96 hours, and SAWT would be discontinued in that event. Patients received systemic antimicrobial drugs at the discretion of the primary physicians as clinically indicated.
The outcomes used to assess safety included development of an enlarging or necrotizing wound, evidence of systemic sepsis (fevers, leukocytosis, and/or hemodynamic changes), or any other clinical complication thought to be related to treatment with SAWT. The outcomes used to assess efficacy were changes in wound surface area and volume. The dimensions of each wound were documented at each dressing change. Wound size was calculated as described by Xakellis and Frantz:
Pain associated with each type of dressing change was assessed using self-reported pain levels. Patients were asked to rate their pain level according to the 0–10 linear analog scale immediately prior to, during, and after removal of the dressing. The Sum of Pain Intensity Differences (SPID) was used to facilitate comparison of pain levels between groups. The SPID score was calculated using the following formula: (pain during dressing change) – (pain before dressing change) + (pain after dressing change) – (pain during dressing change) = SPID score.