Use of Vacuum-assisted Closure in Pediatric Open Fractures With a Focus on the Rate of Infection

Materials and Methods

After internal Institutional Review Board approval, International Classification of Diseases (ICD-9) codes for “open wound” and “complication of open wounds” in a patient age range up to 18 years was retrospectively compiled over a 4.5-year period at a pediatric level I trauma center. This yielded a dataset of 10,108 potential patients with the ICD-9 code. Chart review of these cases yielded 28 patients with a total of 37 open fractures treated with wound VAC. Inclusion criteria was any open fracture initially treated with VAC therapy during this time period. There were no exclusion criteria.

All patients received prophylactic antibiotics according to standard of care guidelines. In addition, all patients underwent emergent irrigation and debridement of their wounds followed by application of a VAC dressing unless medically contraindicated because of traumatic or neurologic instability. These patients, once stable, underwent definitive irrigation and debridement of their fracture. The fixation method (immobilization with casting, external fixation, or internal fixation) was not a priority/variable as the diversity of the fractures and their severity precluded obtaining useful comparisons between fracture types. Likewise, time to union, malalignment, and nonunion were difficult to compare given the diversity of fractures and fracture locations.

Eleven of the 28 patients had their open fracture classified at the time of presentation using the Gustillo-Anderson system. No attempts were made to retrospectively grade the open fractures. The primary outcome measure of this study was documented evidence of infection or surgical intervention for infection. Superficial infection was broadly defined as any documented signs or symptoms of infection from in-hospital notes, clinical follow-up notes, or operative reports that required but resolved with administration of antibiotics.

Deep infection was defined as any documented signs or symptoms of infection necessitating a return trip to the operating room for debridement resulting in positive intraoperative cultures. Osteomyelitis was defined as positive bone cultures taken at the time of operative debridement for a deep infection. Length of time the VAC dressing was used was at the discretion of the surgeon for wound management. The VAC dressing was typically changed an average of 48 to 72 hours regardless of location (within the operating room or at home). Follow-up of patients was conducted via chart review and included admission/hospital documentation, clinic notes, and operative notes.

Results

The age of the 28 patients with open fractures treated with a VAC ranged from 2 to 17 years (average age, 12 years). Information regarding hospitalization and treatment of these 28 patients is found in Table . The average number of days spent in the hospital was 12 days (range, 3–33 days). The average number of operations, while in the hospital, related to their open fracture (either fixation or for repeat irrigation and debridement) was 4 (range, 2–12), with an average of 1 operation after discharge (range, 0–6). Thirteen of the 28 patients were involved in a motor vehicle collision, motorcycle, or pedestrian accident (Table ). The majority of fractures involved the lower extremities (29/37); the exact location of all fractures is found in Table .

Table 1. Patient, Hospitalization, and Treatment Information

Table 2. Mechanism of Injury

Table 3. Fracture Location

Final wound coverage (closure, skin graft, flap, etc) is summarized in Table . Nineteen of the 37 fractures were able to be covered with delayed wound closure or were left to be closed via secondary intention. These three patients treated via secondary intention were able to be discharged home with home VAC dressing changes with eventual conversion to wet to dry dressings when the wound had healed sufficiently that the VAC sponge was no longer able to be applied (these included 2 patients with foot wounds and 1 with a femur wound). Thirty-one of the 37 patients were able to be closed via secondary intention, underwent delayed closure, or closure via skin graft or Integra (Integra LifesSciences Corporation, Plainsboro, New Jersey) with skin graft. Six patients required either local or free flap coverage. Only 2 amputations were necessary, with both of these toe phalanx amputations in patients who sustained metatarsal fractures. After amputation, no further complications were noted. One amputation occurred within the acute setting, resulting in delayed wound closure (hence a delayed closure as well as amputation within the statistics). The other amputation occurred later as a result of contractures that developed about the foot that failed conservative management.

Table 4. Final Soft Tissue Coverage of Fractures

There was no evidence of superficial infection in any of the 37 fractures; however, 2 patients were found to have deep infections throughout the course of treatment. Neither patient developed osteomyelitis. One of these fractures involved a lawnmower injury in which the patient sustained open femur, tibia, and patella fractures. The patient was to undergo coverage via Integra followed by skin grafting. However, this patient’s wound was found to be infected after Integra placement. Skin grafting was delayed, and the patient subsequently underwent 6 more operations including delayed Integra grafting with delayed skin grafting. Cultures at the time of debridement identified infection with Pseudomonas and Enterobacter cloacae. The patient underwent antibiotic therapy for 6 weeks with no further complications/infections after completion of antibiotic therapy. The second infection occurred in a patient who sustained a grade III calcaneus fracture after being involved in a motorcycle collision. Infected tissue was found on the third irrigation and debridement (of 5 total). Cultures taken at that time showed E. cloacae and coagulase negative Staphylococcus. The patient was treated with a 4-week course of antibiotics and had no further infection or other complication. The patient was discharged from the hospital with home wound VAC changes, and was later converted to wet-to-dry dressing changes with ultimate closure/healing of the wound without further complication.

Discussion

Pediatric open fractures present the orthopedist with unique treatment considerations. Pediatric biology, soft tissue healing capacity, time to fracture healing, potential for bone grafting, open growth plates, and method of fixation all vary in comparison to comparable adult injuries. The introduction of VAC therapy as a treatment modality for soft tissue defects related to these wounds has given the orthopedist a valuable asset in aiding both wound coverage and closure. Vacuum assisted closure is gaining acceptance and use in a variety of situations within the pediatric population from general surgery use in closure of abdominal wounds and pilodonal cysts, soft tissue defects, scoliosis wounds, and burns and has demonstrated excellent results. ^2–10 However, despite its widespread use, little data is available to support its use within the pediatric trauma population.

The reasons behind the success of VAC in treating wounds varies. The use of the VAC has been postulated to decrease edema and purulent drainage (via removal of bacteria), increase blood flow, and thereby promote granulation tissue, cell/protein synthesis, and healing. ^6,10,19,20 DeFranzo et al ²¹ have documented a 4-fold increase in perfusion following VAC therapy with an increase of granulation tissue of 80% when compared to wounds treated with traditional wet to dry dressing changes. The use of the VAC within the pediatric population offers another advantage because VAC as dressings are typically changed every 48 to 72 hours as opposed to the twice (or more) daily dressing changes required for traditional wound care (ie, wet to dry). If necessary, the VAC may be changed on the hospital floor or at home. With use of interposing material to help decrease in-growth of granulation tissue into the VAC dressing and the addition of a little saline, most VAC dressing changes are tolerated with minimal pain. In addition, because it is an adherent dressing, mobility is increased allowing children both in and outside the hospital increased activity levels and ease of care.

As with adults, the majority of the literature focuses on the pediatric tibia as a model for open fracture treatment. However, many question whether open fractures in the pediatric population behave similarly to their equivalent adult counterparts. Reported rates of infection for grade III open tibia fractures in children range from 0% to 33%. ^{10,12–14,18,22} A retrospective review of 20 grade III tibia fractures followed for an average of 34 months demonstrated no deep infections. ¹³ Cullen et al ¹⁵ retrospectively reviewed their data from 83 open tibia fractures including 19 grade III fractures. Two superficial infections in the 19 grade III fractures occurred. Both were treated with oral antibiotics for 1 week with resolution of infections and no long-term consequences. ¹⁵ Likewise, Jones and Duncan ²² reviewed their experience with 83 open tibia fractures. As with Cullen, they reported a 2% superficial infection rate with no reports of deep infection. ²²

The above studies are in contrast to Buckley et al ¹⁴ in which a retrospective review of 20 grade III tibia fractures demonstrated an osteomyelitis rate of 15%. Hope and Cole ¹⁸ retrospectively reviewed 92 open tibia fractures, 21% of which were grade III. The overall infection rate across grades was reported as 11% (superficial wound infection in 7 patients, deep wound infection in 3 patients), with this number increasing to 21% for grade III injuries. ¹⁸ Gougoulias et al ¹⁷ in their systematic review of 714 open pediatric fractures, across all grades, found an overall infection rate of 6%. In our experience, use of the wound VAC for all open pediatric fractures yielded a 5.4% (2/37) deep infection rate with no incidence of osteomyelitis, chronic infection, or superficial infection.

Some recent studies have examined VAC therapy efficacy in open pediatric fractures. Dedmond et al ¹² examined the use of VAC in grade III tibia fractures. In their retrospective review of 15 patients, 5 (33%) eventually developed infections (3 requiring surgical intervention). The higher rate of infection associated with their study compared with ours is difficult to explain. First, both studies included small patient populations making any definitive conclusions difficult. Second, Dedmond et al ¹² focused on open tibia fractures whereas our study included all open fractures regardless of location. Further study is required to ascertain whether the use of the VAC in decreasing open tibia fracture infection rate differs from the generalized open pediatric fracture.

The literature is also varied regarding the need and incidence of soft tissue coverage in pediatric open fractures. Again, the primary study model for this comes from open tibia fractures. In their series of 20 patients with open tibia fractures, Bartlett et al ¹³ were able to close 1 wound with tissue adapters, 12 patients via secondary intention, 2 via delayed closure, 2 with flaps, and 3 with skin grafts. Buckley et al ¹⁴ noted 6/20 fractures (30%) required flap coverage.

The experience of Cullen et al ¹⁵ advocates primary closure of wounds if possible based on their series of 83 open tibia fractures. In total, 57 open fractures were closed primarily across grade I, II, and III fractures. For grade III injuries, this number decreased to 7/19 primarily closed. Three grade III injuries were treated by delayed closure and 1 via secondary intention. Six grade III injuries required flap coverage (32%). Likewise, Hope and Cole ¹⁸ reported primary closure in 55% of fractures across all grades; however, 23 of the 91 wounds required plastic surgery coverage/assistance, 11 required skin grafting, and 12 required local or free flap coverage (13%). Jones and Duncan ²² reported the use of 10 flaps in their series of 83 patients.

A systematic review of the literature by Gougoulias et al ¹⁷ reviewed 14 studies of open pediatric tibia fractures, including 714 patients. Flap coverage or skin graft was required for approximately 20% of patients (no breakdown by grade or between skin graft/flap). In a systematic review of 54 grade IIIB open tibia fractures, Glass et al ¹⁶ demonstrated flap coverage in 45/54 (83%). There are, however, those who support the use of the VAC for decreasing the number of flaps required for soft tissue coverage in these injuries.

Dedmond et al ¹² reported a 50% decrease in flap coverage specific for initial injury grade with use of the VAC in open tibia fractures. Shilt et al ¹⁰ used the VAC in pediatric lawnmower injuries, studying 31 patients, 16 of which received VAC therapy while 15 received traditional wet to dry dressing changes. Fractures were associated with 12/16 patients in the VAC group and 8/15 patients in the traditional dressing group. A statistically significant difference between groups regarding the necessity of free flap coverage was found with VAC therapy.

It is difficult to draw conclusions between our study and the literature about the use of the VAC in extrapolating its role to decrease the need for flap coverage. First, the range of patients requiring flap coverage reported in the literature is varied and sample sizes are small. Secondly, the literature primarily reports on the use of the VAC for open tibia fractures whereas our study included all open fractures, making comparisons possible, but, conclusions difficult. Therefore, no conclusions can be drawn on the role of VAC therapy for decreasing the need for secondary soft tissue coverage procedures.

As with all studies, there are limitations to this study. First, its retrospective design automatically induces bias as these patient’s injuries were thought sufficiently traumatic to warrant VAC usage by the treating surgeon. In addition, the wide range of fractures treated with the wound VAC for a variety of wound types and locations makes comparisons within the study population and historical controls difficult. Open fractures at different locations in the body differ in their natural history, have different soft tissue coverage issues and considerations, and different treatment goals. This is likely a primary reason for the differences in duration of VAC therapy and final disposition as to soft tissue coverage. Additionally, only 11 of 28 patients had their open fracture classified using the Gustillo-Anderson system at time of initial debridement making comparisons to the literature difficult. Retrospective grading of open fractures was not performed to avoid the potential for introducing further bias. As with many of the other examinations within the literature, our study sample size was small. However, it is difficult to obtain large numbers of open pediatric fractures severe enough to warrant the use of the VAC. Historical controls demonstrate roughly equal numbers of patients involved in their studies. Our study also did not include a control group. Therefore, we can only evaluate our data based on historically matched controls. While the VAC appears favorable to historical controls regarding reported rates of infection following open pediatric fractures, we have no direct comparison of the wound VAC to traditional dressing changes for this specific patient population. Future studies should focus on increasing patient numbers while prospectively studying outcomes with control groups. Future studies may also help determine if VAC decreases the need for flap coverage within the pediatric population.

Conclusion

The use of the VAC appears to be equally as safe and efficacious in reducing the incidence of infection in pediatric open fractures as historical controls. It should be considered a valuable tool for the orthopedist treating these injuries. Research has shown that the VAC decreases edema and bacterial load, increases perfusion, and promotes the formation of granulation tissue. In addition, given the relative ease of dressing changes, increased patient mobility, and containment of the wound within a closed, sealed environment, the use of the wound VAC is not only efficacious in reducing infection, but, advantageous within the pediatric population for its ease of usage.