Bolster Versus Negative Pressure Wound Therapy for Securing Split-Thickness Skin Grafts in Trauma Patients

Introduction Trauma affects people of all ages with considerable economic impacts on society. Soft tissue wounds resulting from trauma often cannot be reapproximated due to undue tension on the skin. With significant soft tissue loss, wound coverage is needed to expedite healing. Successful coverage of a soft tissue defect with a split-thickness skin graft (STSG) requires immobilization of the graft to prevent infection, hematoma, and desiccation. The ideal securing dressing would achieve all of the above requirements. The traditional means of securing a STSG has been with a cotton bolster dressing for at least five days.[1] Success with STSG has been limited in areas where the contour is complex and on mobile surfaces, such as the perineum and inguinal region.[2] Over the past two years at the authors’ institution, the surgeons’ approach to securing STSG in trauma patients has changed. More recently, negative pressure wound therapy (NPWT) (Vacuum-Assisted Therapy®, V.A.C.®, Kinetic Concepts, Inc., San Antonio, Texas) has been used to treat the majority of patients. Since changing from cotton bolster dressings to NPWT, it appeared the number of repeat grafts decreased. This study evaluates these experiences with NPWT of STSG in comparison to STSG secured with cotton bolster dressings. Patients and Methods After securing Institutional Review Board approval, charts of 40 trauma patients requiring STSG were reviewed. These patients were admitted between January, 2001, and January, 2003, to Charleston Area Medical Center (CAMC), a level I trauma center in the capital city of West Virginia. These 40 patients represented all trauma patients that received STSG during the study time period with the exception of two burn patients who had heavily contaminated wounds. Due to the large degree of contamination and extremely large size of these two wounds, they were rendered incomparable and excluded from the review. Both wounds were treated with bolster dressings. Patients were assigned by surgeon preference to have their skin grafts secured by either NPWT or a cotton bolster. Patient medical records were examined for patient age, size of skin graft in square centimeters, length of dressing placement in days, total length of hospital stay in days, and need for repeat grafting during the same hospital admission. Percentage graft take was noted; however, the nature of the comments by the attending physician were subjective and ranged from “good take” to 75–100 percent. These partial-graft take percentages were considered unreliable, and the resulting lack of inter-rater reliability between the various charting physicians was a concern. The operational definition of “graft failure” may not have been consistent between physicians; however, a failure would have triggered, at the very least, a suggestion of a revision or repeat graft, which would have been noted in the chart. Therefore, all grafts were considered either completely successful or total failures. Any graft that failed or had to be revised was considered a total failure. Hence, all others were considered to be surviving grafts; “graft survival” was then used as the endpoint. Skin grafting was performed by general/trauma, plastic, or orthopedic surgeons. Skin was harvested with an electric dermatome (thickness of 0.012–0.016 inches). Donor skin was then meshed. In both groups (NPWT, cotton bolster) the STSG was secured by either staples or absorbable sutures circumferentially. In the NPWT group, a layer of nonadhering dressing was applied over the STSG, followed by placement of the vacuum sponge/tubing. Subsequently, an occlusive dressing was used to form a tight seal around the area. In the case of cotton bolster, the cotton was irrigated with saline and secured with nonabsorbable nylon sutures (a nonadhering dressing layer is not used, as the cotton bolster is generally soaked before removal, eliminating any graft-dressing adherence difficulties). The dressing was left in place for a mean of 4.8 days in the NPWT group and 5.2 days in the cotton bolster group. Differences in patient age, length of stay, wound type, wound site, graft size, duration of graft, and graft failure were compared between the two groups. The Fisher’s Exact Test (or c2 where appropriate) was performed on categorical variables. For analysis, one trichotomous variable, wound site, was collapsed into “extremity, other.” Continuous variables were examined using two-tailed, unmatched t-tests. For all reported analyses, a significance level of 2-tailed p0.05 was used. All analyses were performed using SAS for Windows, Version 8.1 (SAS Institute, Inc., Cary, North Carolina). Results From January, 2001, to January, 2003, 40 trauma patients underwent STSG at a level I trauma center. Seventeen patients with a total of 21 wounds had NPWT used to secure their grafts, whereas 23 patients with a total of 25 wounds had cotton bolster dressings. There was no significant difference in mean patient age or in the mean length of hospital stay between the groups, although the mean stay in NPWT patients tended to be slightly longer (Table 1). Overall indications for graft placement included soft-tissue loss (n=34) and fasciotomy wound coverage (n=12). There was no statistically significant difference in the site or type of wound grafted between the groups (by Fisher’s Exact Test). The average length of time the dressing was left in place was approximately five days for each group. There was also no significant difference in the mean graft size between the groups, although the bolster group had slightly larger grafts overall (Table 2). Only one graft failure occurred, which was in the cotton bolster group. The graft failure was 300cm2 on a soft tissue injury on a buttock. The failure differential between the two groups was not significant (by Fisher’s Exact Test) (Table 2). The average cost of the cotton bolster was approximately $18.50 per graft, which was significantly lower than the average cost of the NPWT, approximately $1,000.00 per graft (Table 3). Discussion Soft tissue defects pose a difficult problem for the trauma surgeon. Nonetheless, most wounds are amenable to STSG coverage after granulation tissue is present. The ideal STSG dressing has three main components: elimination of fluid collection, immobilization of the graft, and stabilization of the graft on an irregular surface. Secondarily, protecting the graft from desiccation is essential, as well as providing an environment that limits infection risks. Beginning in January 2001, the trauma service in Charleston, West Virginia, changed its technique of securing STSG. Previously, a cotton bolster was used to secure a graft as described by Rudolph, et al., in 1990.1 However, with the advent of negative-pressure dressings, wounds have been noted to develop granulation tissue faster, allowing for quicker reepithelization.[3,4] Increasing success in local wound management, coupled with the limited nursing care associated with NPWT, has led to the adoption of NPWT for securing STSG. Early studies of the NPWT device by Argenta[4] and Blackburn[2] appeared promising, with it emerging as a more effective way to secure skin grafts than traditional bolsters. Recently, Sherer, et al.,[5] retrospectively compared bolster versus NPWT in trauma patients and found a statistically significant decrease in repeat grafting during the same hospitalization when NPWT was used in securing the STSG.[5] However, some criticisms of their work include a large difference in graft size between the two groups and the lack of a cost-effectiveness comparison. In this present study, there was no statistically significant difference in mean graft size, compared to a 300 percent difference in the Scherer, et al., study. As described previously, zero patients in this NPWT group required repeat in-hospital grafting. In the bolster group, one graft failed and required repeat grafting. The difference in failure rate between modalities is extremely small because of a very low failure rate in both modalities. It should be noted that the one graft failure was on a large wound (300 cm2, more than twice either mean) that also involved substantial soft tissue loss. Graft failures are most likely related to the relative size and type of wound treated rather than the treatment mechanism. In heavily contaminated or edematous wounds, graft failure tends to be more common, so perhaps the choice of treatment mechanism would have a greater influence in that subset of patients. The difference in the length of stay (LOS) between the two groups was not found to be statistically significant; however, it does approach significance. This finding may be incidental in that the LOS ranges are quite sizeable for both groups. The average length of dressing placement was very similar for both groups (4.8 days, NPWT; 5.2 days, bolster). The bolster length of treatment was more variable than that of NPWT, most likely because the manufacturer recommends duration of five days for NPWT placement. The average length of dressing placement in the bolster was also skewed by two unusually long bolster placements (10 and 14 days); with those two outliers removed, the bolster mean drops to 4.7 days. Importantly, there is a significantly increased cost of approximately $1,000 more per graft treated with NPWT than a cotton bolster—NPWT costs are over 50 times that of a cotton bolster dressing (Table 3, p0.0001). The total costs incurred by these patients during their entire stay were quite substantial, given an overall mean length of stay of 17.7 days across both groups (means: 15.3 days, cotton bolster; 20.9 days, NPWT). It is perhaps obvious that the percentage of each patient’s total cost attributable to the dressing is rather small. However, this does not necessarily negate the importance of the cost difference, particularly given the continued acceleration in healthcare spending growth, which has not been matched by economic growth or growth in reimbursements.[6] Opportunities for cost savings that do not mar clinical outcomes deserve consideration. This study has several limitations. As a retrospective study, group assignment (NPWT vs. bolster) was not controlled, randomized, or blinded. While the groups were found to be statistically similar in most ways, assignment could have been biased. The major limitation of this study is the small sample size of only 46 grafts. This sample size is too small to be able to resolve a difference in failure rates. Based on the failure rates, it would take nearly 250 patients entering a prospective randomized trial to decisively answer the question whether or not NPWT is better (assuming a desired power of 0.80, a=0.05). However, even this small sample calls into question the widespread clinical use of NPWT—as the extremely high costs incurred with use of NPWT may outweigh any slight treatment gains over that of a traditional cotton bolster for most grafts. The excessive cost of routinely using NPWT on all grafts seems to exceed any potential medical benefits on the rare few that fail. A prospective, randomized trial could perhaps definitively answer whether or not NPWT is clearly a superior treatment well above and beyond the traditional cotton bolster, but as stated previously, this would require what may be considered a larger than practical sample size. It may be more prudent to conduct a trial of these two modalities in grafts that are known to have high associated failure rates, such as heavily contaminated or edematous wounds. Perhaps in these high-risk cases, the high costs associated with NPWT may be rationalized. Conclusion NPWT is a promising device to secure STSG, with zero split-thickness skin grafts failing in the NPWT group. However, a significantly higher cost is incurred with the use of this device. For routine wounds that are small and have a contour that is not complex, use of a cotton bolster dressing is clinically as effective as NPWT and substantially more cost effective. A randomized, prospective trial treating grafts with high associated failure rates and these two treatment modalities would help answer questions about the appropriate use of NPWT.