Subatmospheric Wound Therapy With a Sealed Gauze Dressing
An intention to treat analysis was performed for all outcomes. There were 18 patients (40%) with initially infected wounds in the GSUC arm vs 13 patients (31%) in the VAC arm (P = 0.38). There was no statistically significant difference in the anatomical distribution of the wounds between the GSUC and VAC arms (P = 0.59). A majority of the patients in both the infected and noninfected groups had wounds on the trunk. A subgroup analysis was used to compare the efficacy of GSUC and VAC for the management of infected wounds. However, due to the limited number of observations after day 7—most patients were discharged from the hospital or underwent operative treatment for their wounds—these analyses included data only through the first week (Figure 2 and Figure 3).
(Enlarge Image)
Figure 2.
Percent change from baseline in wound surface area. Among the infected and noninfected groups, there was a statistically significant reduction in wound surface area in both VAC and GSUC groups. However, the rate of change between the treatment groups was not significantly different for the infected group (P = 0.73) and the noninfected group (P = 0.77). Mean values with error bars representing ± 1 SEM are presented.
(Enlarge Image)
Figure 3.
Percent change from baseline in wound volume. Among the infected and noninfected groups, there was a statistically significant reduction in wound volume in both VAC and GSUC groups. However, the rate of change between the treatment groups was not significantly different for the infected group (P = 0.22) and the noninfected group (P = 0.58). Mean values with error bars representing ± 1 SEM are presented.
None of the patients in either treatment group developed complications related to SAWT, evidence of systemic sepsis during the study period, or progressing or necrotizing wound infections under the dressings. At 48 hours, > 10 colony-forming units per gram of tissue were present in 8 of 18 (44%) patients in the GSUC arm and in 6 of 13 (46%) patients in the VAC arm. However, at 96 hours, there were < 10 colony-forming units per gram of tissue in all of the tissue biopsies.
Among the infected group, there was a statistically significant reduction in wound surface area in both treatment arms. The average rate of change was a decrease of 4.4% per day for GSUC and 4.8% per day for VAC (P < 0.001 for both). However, the rate of change between the arms was not significantly different (P = 0.73 for the treatment-by-time interaction), with the estimated difference (VAC - GSUC) being - 0.3%, 95% CI (-2.3, 1.6). There also was a statistically significant reduction in wound volume in both treatment arms. This represented a mean decrease of 7.8% per day for GSUC and 9.7% per day for VAC (P < 0.001 for both). As with surface area, the rate of volume reduction was not significantly different between the 2 arms based on the treatment-by-time interaction (P = 0.22). The estimated VAC – GSUC difference was -2.0%, 95% CI (-5.1, 1.2).
Among the noninfected group, there was also a statistically significant reduction in wound surface area in both treatment arms. The average rate of change was a decrease of 4.7% per day for GSUC and 5.0% per day for VAC (P < 0.001 for both). However, the rate of change between the arms was not significantly different based on the treatment-by-time interaction (P = 0.77), with the estimated difference VAC - GSUC being -0.3%, 95% CI (-2.1, 1.6). There also was a statistically significant reduction in wound volume in both treatment arms. This was a mean decrease of 8.9% per day for GSUC and 9.7% per day for VAC (P < 0.001 for both). As with surface area, the rate of volume reduction was not significantly different between the 2 treatment arms (P = 0.58 for the treatment-by-time interaction). The estimated VAC - GSUC difference was -0.8%, 95% CI (-3.6, 2.0) ( Table 2 ).
Overall, patients in the GSUC arm reported less pain (SPID = 0.50, 95% CI -0.11 to 1.11) compared to patients in the VAC arm (SPID = 1.73, 95% CI 0.91 to 2.54, P = 0.019). For patients with infected wounds, average SPID was 0.50 in the GSUC arm compared to 2.16 for patients in the VAC arm. This difference approached statistical significance (P = 0.056). However, the presence of infection did not significantly alter the effect of dressing type on SPID (P = 0.41 for the treatment by infection status interaction) (Table 3).
A diverse group of organisms was cultured from the wounds, including Gram-positive and Gram-negative bacteria, combinations of multiple organisms, and fungus. In the GSUC arm, there were 3 patients (17%) with a Gram-positive infection, 4 patients (22%) with a Gram-negative infection, 9 patients (50%) with a polymicrobial infection, 1 patient (6%) with a fungal infection, and 1 patient (6%) from whom no organisms were recovered. All of the Gram-positive organisms were methicillin-resistant Staphylococcus aureus (MRSA), and vancomycin-resistant Enterococcus was present in 2 patients (11%). In the VAC arm, there were 4 patients (31%) with a Gram-positive infection (all MRSA), one patient (8%) with a Gram-negative infection, 6 patients (46%) with a polymicrobial infection, and 2 patients from whom no organisms were recovered (15%). Thirteen out of 18 patients (72%) in the GSUC arm, and 11 out of 13 patients (85%) in the VAC arm, received systemic antimicrobial therapy. Based on clinical indications, and the judgment of the primary treating physicians, broad-spectrum antimicrobial therapy was initiated before the culture results were available for 12 patients in the GSUC arm, and 11 patients in the VAC arm. Furthermore, systemic antimicrobial therapy was initiated prior to obtaining cultures in 3 patients (2 in the VAC arm and 1 in the GSUC arm). Ultimately, no organisms were isolated from these cultures. Decisions about the specific drugs used, dosages, and duration of treatment were made by the primary attending physicians.
Subatmospheric wound therapy failed in 2 of the 31 patients in the subgroup with infected wounds (6.5%), both of whom were in the VAC arm of the trial. One of the patients had an infected perineal wound where the VAC seal could not be maintained after 4 days; this patient successfully crossed over to GSUC treatment. Another patient had an infected abdominal and flank wound after evacuation of a retroperitoneal hematoma; the VAC seal could not be maintained because of excessive fluid drainage. This patient also successfully crossed over to GSUC therapy after 4 days. Data collected after patients crossed over to the other arm of the study was not included in the analysis of the results.
Results
Primary Outcomes
An intention to treat analysis was performed for all outcomes. There were 18 patients (40%) with initially infected wounds in the GSUC arm vs 13 patients (31%) in the VAC arm (P = 0.38). There was no statistically significant difference in the anatomical distribution of the wounds between the GSUC and VAC arms (P = 0.59). A majority of the patients in both the infected and noninfected groups had wounds on the trunk. A subgroup analysis was used to compare the efficacy of GSUC and VAC for the management of infected wounds. However, due to the limited number of observations after day 7—most patients were discharged from the hospital or underwent operative treatment for their wounds—these analyses included data only through the first week (Figure 2 and Figure 3).
(Enlarge Image)
Figure 2.
Percent change from baseline in wound surface area. Among the infected and noninfected groups, there was a statistically significant reduction in wound surface area in both VAC and GSUC groups. However, the rate of change between the treatment groups was not significantly different for the infected group (P = 0.73) and the noninfected group (P = 0.77). Mean values with error bars representing ± 1 SEM are presented.
(Enlarge Image)
Figure 3.
Percent change from baseline in wound volume. Among the infected and noninfected groups, there was a statistically significant reduction in wound volume in both VAC and GSUC groups. However, the rate of change between the treatment groups was not significantly different for the infected group (P = 0.22) and the noninfected group (P = 0.58). Mean values with error bars representing ± 1 SEM are presented.
Safety
None of the patients in either treatment group developed complications related to SAWT, evidence of systemic sepsis during the study period, or progressing or necrotizing wound infections under the dressings. At 48 hours, > 10 colony-forming units per gram of tissue were present in 8 of 18 (44%) patients in the GSUC arm and in 6 of 13 (46%) patients in the VAC arm. However, at 96 hours, there were < 10 colony-forming units per gram of tissue in all of the tissue biopsies.
Efficacy
Among the infected group, there was a statistically significant reduction in wound surface area in both treatment arms. The average rate of change was a decrease of 4.4% per day for GSUC and 4.8% per day for VAC (P < 0.001 for both). However, the rate of change between the arms was not significantly different (P = 0.73 for the treatment-by-time interaction), with the estimated difference (VAC - GSUC) being - 0.3%, 95% CI (-2.3, 1.6). There also was a statistically significant reduction in wound volume in both treatment arms. This represented a mean decrease of 7.8% per day for GSUC and 9.7% per day for VAC (P < 0.001 for both). As with surface area, the rate of volume reduction was not significantly different between the 2 arms based on the treatment-by-time interaction (P = 0.22). The estimated VAC – GSUC difference was -2.0%, 95% CI (-5.1, 1.2).
Among the noninfected group, there was also a statistically significant reduction in wound surface area in both treatment arms. The average rate of change was a decrease of 4.7% per day for GSUC and 5.0% per day for VAC (P < 0.001 for both). However, the rate of change between the arms was not significantly different based on the treatment-by-time interaction (P = 0.77), with the estimated difference VAC - GSUC being -0.3%, 95% CI (-2.1, 1.6). There also was a statistically significant reduction in wound volume in both treatment arms. This was a mean decrease of 8.9% per day for GSUC and 9.7% per day for VAC (P < 0.001 for both). As with surface area, the rate of volume reduction was not significantly different between the 2 treatment arms (P = 0.58 for the treatment-by-time interaction). The estimated VAC - GSUC difference was -0.8%, 95% CI (-3.6, 2.0) ( Table 2 ).
Secondary Outcome: Pain Associated With Dressing Change
Overall, patients in the GSUC arm reported less pain (SPID = 0.50, 95% CI -0.11 to 1.11) compared to patients in the VAC arm (SPID = 1.73, 95% CI 0.91 to 2.54, P = 0.019). For patients with infected wounds, average SPID was 0.50 in the GSUC arm compared to 2.16 for patients in the VAC arm. This difference approached statistical significance (P = 0.056). However, the presence of infection did not significantly alter the effect of dressing type on SPID (P = 0.41 for the treatment by infection status interaction) (Table 3).
Microbiological Analysis of Wounds
A diverse group of organisms was cultured from the wounds, including Gram-positive and Gram-negative bacteria, combinations of multiple organisms, and fungus. In the GSUC arm, there were 3 patients (17%) with a Gram-positive infection, 4 patients (22%) with a Gram-negative infection, 9 patients (50%) with a polymicrobial infection, 1 patient (6%) with a fungal infection, and 1 patient (6%) from whom no organisms were recovered. All of the Gram-positive organisms were methicillin-resistant Staphylococcus aureus (MRSA), and vancomycin-resistant Enterococcus was present in 2 patients (11%). In the VAC arm, there were 4 patients (31%) with a Gram-positive infection (all MRSA), one patient (8%) with a Gram-negative infection, 6 patients (46%) with a polymicrobial infection, and 2 patients from whom no organisms were recovered (15%). Thirteen out of 18 patients (72%) in the GSUC arm, and 11 out of 13 patients (85%) in the VAC arm, received systemic antimicrobial therapy. Based on clinical indications, and the judgment of the primary treating physicians, broad-spectrum antimicrobial therapy was initiated before the culture results were available for 12 patients in the GSUC arm, and 11 patients in the VAC arm. Furthermore, systemic antimicrobial therapy was initiated prior to obtaining cultures in 3 patients (2 in the VAC arm and 1 in the GSUC arm). Ultimately, no organisms were isolated from these cultures. Decisions about the specific drugs used, dosages, and duration of treatment were made by the primary attending physicians.
Failure of Intervention
Subatmospheric wound therapy failed in 2 of the 31 patients in the subgroup with infected wounds (6.5%), both of whom were in the VAC arm of the trial. One of the patients had an infected perineal wound where the VAC seal could not be maintained after 4 days; this patient successfully crossed over to GSUC treatment. Another patient had an infected abdominal and flank wound after evacuation of a retroperitoneal hematoma; the VAC seal could not be maintained because of excessive fluid drainage. This patient also successfully crossed over to GSUC therapy after 4 days. Data collected after patients crossed over to the other arm of the study was not included in the analysis of the results.
SHARE
