ORIGINAL RESEARCH


https://doi.org/10.5005/jp-journals-10030-1433
Panamerican Journal of Trauma, Critical Care & Emergency Surgery
Volume 12 | Issue 03 | Year 2023

Lower Limb Trauma: Salvage Through Vacuum-assisted Closure in an Ecuadorian Tertiary Hospital


Andrea Villarreal-Juris1https://orcid.org/0000-0002-0977-3179, Ramiro V illarreal-Juris2https://orcid.org/0000-0002-8533-1698, Jaime D yer-Rolando3

1,3Department of General Surgery, Luis Vernaza Hospital, Guayaquil, Guayas, Ecuador

2Fire Department of the Metropolitan District of Quito, Quito, Ecuador

Corresponding Author: Andrea Villarreal-Juris, Department of General Surgery, Luis Vernaza Hospital, Guayaquil, Guayas, Ecuador, Phone: +593979032563, e-mail: andrea.fima_95@hotmail.com

Received: 05 July 2023; Accepted: 18 November 2023; Published on: 30 December 2023

ABSTRACT

Introduction: Major trauma causes shear, compressive, and torsional forces to the lower limbs, resulting in complex and devastating soft tissue and bone injuries that challenge the most experienced surgeons. For such injuries, the standard of care is early debridement and coverage, which in recent years has been ubiquitously done using vacuum-assisted closure (VAC).

Objective: Describe the short-term outcomes of adult patients with devastating lower limb injuries from January to December 2022 at the Luis Vernaza Hospital in Ecuador.

Materials and methods: Observational, retrospective, and analytical study. Data was obtained from electronic medical records and analyzed using Statistical Package for the Social Sciences (SPSS) 27.0. Yan’s classification was used. Management was early wound excision and coverage using VAC. Main outcomes—wound healing; limb mobility at 180–190 days; infection; length of stay; days-to-skin graft; and mean total surgical time.

Results: A total of 13 patients were included, all male, mean age—31.4 years (20–45 years), mean length of stay—38.9 days (24–65 days). The most common mechanism of injury was motorcycle (58.85%); the left lower limb was the most affected (61.5%); 53.9% of injuries were pattern 2B and 46.1% pattern 3, which increased length of stay (p = 0.004); mean total surgical time (p = 0.349), days-to-skin graft (p = 0.002); and a 4x higher probability of secondary healing [odds ratio (OR)—4.031, p = 0.005]. No significant difference was observed in conventional vs silver-impregnated foam dressings and infection, length of stay, days-to-skin graft, and mean total surgical time. Minor infection significantly increased days-to-skin graft (p = 0.002). All patients had limb mobility by 180–190 days follow-up and none underwent amputation.

Conclusion: The VAC appears to be a useful tool in managing catastrophic lower limb injuries (including pattern 3). Early VAC coverage allows early large defect coverage until surgical reconstruction, may reduce infection, and preserve mobility thus aiding limb salvage.

RESUMEN

Introducción: Los traumatismos mayores traducen fuerzas de compresión, torsión y abrasión en las extremidades, provocando lesiones que plantean un desafío para los cirujanos debido a su complejidad, incluyendo fracturas y una pérdida devastadora de tejidos blandos. El desbridamiento temprano y la cobertura han sido durante mucho tiempo el estándar de atención para estas lesiones, incluido el cierre asistido por vacío (VAC) como una herramienta casi omnipresente en los últimos años.

Objetivo: Describir los resultados a corto plazo en pacientes ≥16 años que sufrieron traumatismos de miembros inferiores durante enero a diciembre de 2022 en el Hospital Luis Vernaza.

Materiales y métodos: Estudio observacional, retrospectivo, analítico. Los datos se obtuvieron a partir de la historia clínica electrónica y se analizaron con SPSS 27.0. Se utilizó la clasificación de Yan. El manejo fue la escisión radical de la herida y la cobertura inmediata con VAC. Resultados principales: cicatrización de heridas; movilidad de las extremidades; infección; estancia hospitalaria; días-hasta el injerto de piel; tiempo quirúrgico total medio.

Resultados: Se incluyeron trece pacientes, todos varones. Edad media: 31.39 años (20 – 45 años), estancia media: 38.85 días (24 – 65 días). La causa más frecuente fue accidente de motocicleta (58.85%); el miembro inferior izquierdo fue el más afectado (61.54%). El 53.85% de las lesiones fueron Patrón 2B y el 46.15% Patrón 3. Este último aumentó la estancia hospitalaria (W: 7.0, p = 0.004); tiempo quirúrgico total medio (t: 0.978, p = 0.349) y días-hasta el injerto de piel (t: 1.678, p = 0.002); y tuvo una probabilidad 4 veces mayor de curación secundaria (OR: 4.031, p = 0.005). No hubo diferencias significativas entre los apósitos de espuma convencionales e impregnados de plata en términos de infección, duración de la estancia, días-hasta el injerto de piel y tiempo quirúrgico. La infección menor aumentó significativamente de manera proporcional a los días-hasta el injerto de piel (t: -3.902, p = 0.002). En todos los casos, la movilidad de las extremidades se mantuvo a los 180 y 190 días de seguimiento; y el rescate de la extremidad fue exitoso.

Conclusión: VAC es una herramienta útil en las lesiones por traumatismos de las extremidades inferiores (incluidas aquellas Patrón-3), permitiendo una gran cobertura del defecto hasta la reconstrucción quirúrgica, la disminución de la infección y la preservación de la movilidad, logrando salvar la extremidad.

How to cite this article: Villarreal-Juris A, Villarreal-Juris R, Dyer-Rolando J. Lower Limb Trauma: Salvage Through Vacuum-assisted Closure in an Ecuadorian Tertiary Hospital. Panam J Trauma Crit Care Emerg Surg 2023;12(3):136–143.

Source of support: Nil

Conflict of interest: None

Keywords: Degloving injuries, Limb salvage, Lower limb, Vacuum-assisted closure

Palabras Clave: Extremidad inferior, Lesiones por desenguantamiento, Recuperación del miembro, Terapia de presión negativa para heridas

INTRODUCTION

For patients younger than 45 years, major trauma remains the leading cause of death. Due to its complexity, many traumatic lower extremity injuries are very challenging particularly when catastrophic soft tissue, bone, and vascular injuries occur alone or in combination. Often, unsatisfying results, including amputation, result from extensive limb damage, massive blood loss, and subsequent infection.

Decades ago, a classic study by Godina recommended definitive wound coverage within 72 hours of injury to complex extremity injuries, which reduced secondary complications and skin flap loss. In this study, the highest rate of flap failure and infection was free flap reconstruction performed between 72 hours and 3 months.1

Godina’s study did not differentiate according to the mechanism of injury, classification of the initial injury, failure to report concomitant injuries in polytrauma patients, or comorbidities. Overall, it analyzed a heterogeneous population and included reconstructions of both the upper and lower extremities, thus, adding significant statistical confounders. Furthermore, this landmark study was done at a time when vacuum-assisted closure (VAC), a contemporary perioperative tool for wound management, was not readily available or widely used.

While most evidence suggests that early wound coverage is undoubtedly advantageous, VAC application at present has evidently introduced a degree of flexibility regarding the specific timing of surgical reconstruction in these patients, sustained by multiple studies recommending a wider range of optimal timing varying from immediately after injury, to several months later.2,3 In practice, it is not always feasible to perform an early and complex microsurgical reconstruction (within the first 72 hours), even in academic centers. Often transferred patients are admitted late or the expert surgical team is not available. It may also be unwise to perform complex reconstruction on polytrauma patients with life-threatening injuries.4

Regardless, in recent years, the application of VAC has become the standard of care for most such injuries,5 and much less frequently cryopreserved split-thickness skin grafts procured from degloved flaps or artificial dermal replacements are used.6

Antibiotics are also standard of care for extensive wound management of trauma limb salvage pathway. The longer the delay to start antibiotics and wound coverage, the greater the risk of infection and subsequent secondary amputation (p = 0.001).7 Notably, many patients with major lower extremity trauma have fractures. Orthopedic fracture surgery may be complex and carry up to 27% infection risk. Postoperative wound care also includes the type of dressing, which is known to reduce the risk of infection.8

Objective

Describe short-term outcomes of adult patients with extensive lower limb injury between January and December of 2022 at the Luis Vernaza Hospital in Ecuador.

MATERIALS AND METHODS

After the approval of the Institutional Ethics Research Board, an observational, retrospective, and analytical study was conducted. Data was obtained from the hospital’s electronic medical records and analyzed using Statistical Package for the Social Sciences (SPSS) 27.0. The main outcomes included—wound healing; limb mobility at 180–190 days; infection; length of stay; days-to-skin graft; and mean total surgical time.

A descriptive report of the clinical-demographic characteristics was made. The qualitative variables were described through frequencies and percentages and the quantitative variables with percentages and central tendency measures.

The student’s t-test was performed comparing the mechanism of injury and age, mean total surgical time, days-to-skin graft, and infection; type of dressing and length of stay, and days-to-skin graft and mean total surgical time. The Mann–Whitney U test was performed comparing length of stay and injury pattern. A Chi-squared analysis was performed comparing the type of dressing and infection and the type of healing and injury pattern. For the later, the odds ratio (OR) was calculated. A p-value < 0.05 was established for statistical significance.

Inclusion Criteria

All adult (≥16 years old) patients admitted with extensive degloving injury of the lower extremity, involving more than one-fourth of the circumference of the affected areas (ankle, calf, or thigh) and located proximal to the metatarsals.

Injury Classification

Injuries were classified according to Yan’s Classification9 in pattern 1 (purely degloving injury), noncircumferential (1A) or circumferential degloving (1B); pattern 2 (degloving injury involving deep soft tissues), noncircumferential (2A) or circumferential degloving (2B); and pattern 3 (degloving injury with long-bone fractures) (Table 1).

Table 1: Demographic characteristics
Injury causes (%, n)
Motorcycle accident 58.85 (7)
 Pedestrian hit by car 15.39 (2)
 Road traffic accident 30.77 (4)
Injury location
 Left lower limb 61.54 (8)
 Right lower limb 38.46 (5)
Comorbidities (%, n)
 Arterial hypertension 15.39 (2)

Infection

Minor infections were defined as antibiotic-treated superficial soft tissue infection; and, major infections as osteomyelitis or requiring reoperation for incision and drainage.

Surgical Techniques

Following the initial resuscitation, suitable patients were taken to surgery for the extremity. The degloved skin was excised according to the principles and standard surgical approach. Sharp scalpels were used to remove necrotic tissue. Immediate coverage of the wound was performed using the VAC system. The foams utilized were open pore reticulated polyurethane foams (GranuFoam® or GranuFoam Silver® on propensity-to-infection tissue) and/or polyvinyl alcohol foams (WhiteFoam® for protection of critical areas with vascular, tendinous or bone exposure). The VAC system was then covered with an adhesive drape and connected to a collector (Canister®) through the tubing. The software-controlled negative pressure device was set to 50 up to 150 mm Hg target pressure (high, moderate, or low intensity) and continuous or intermittent pressure control according to the condition of the soft tissues during surgery. Later split-thickness skin grafting was performed at the surgeons’ discretion.

Healing

A wound free of dressings was defined as a healed wound. Healing was classified as primary or secondary and as healing early (within 2 weeks of the reconstruction) or late (after 2 weeks).

Limb Function

Limb mobility was evaluated at 180–190 days after a reconstruction follow-up visit at the outpatient clinic of the department of general surgery.

RESULTS

A total of 13 patients with devastating lower limb injuries were included in the present study, all male. Mean age was 31.4 years (range—20–45 years). Mechanisms of injury included mostly motorcycle crashes (58.8%) or pedestrian hit by a vehicle (15.4%), and most were in the left lower limb (61.5%). The mean hospital length of stay was 38.8 days (range—24–65 days). According to Yan’s classification, 53.8% were pattern 2B injuries and 46.2% were pattern 3. For the later, all patients had Gustilo 3B tibial fractures. Concerning comorbidities, arterial hypertension was present in 15.4% of patients (Table 1).

According to Yan’s classification, 62.5% of the injuries of the left lower limb were pattern 2B, and 37.5% were pattern 3. For the injuries of the right lower limb, 40% were pattern 2B, and 60% were pattern 3. Half of the pattern 3 injuries were due to motorcycle crashes (50%). Hospital length of stay was longer in patients with pattern 3 injuries (45 days, range—31–61 days) than pattern 2B (33.57 days, range—24–65 days). The same was observed in days-to-skin graft in pattern 2B (20.8 days, range—16–27 days) vs pattern 3 (16.7 days, range—12–26) which also had longer surgical time (1012.83 minutes, range—745–1,292 minutes vs 852.86 minutes, range—510–1,585 minutes). Minor infection occurred more frequently in pattern 3 (66.67%), and mobility at 180–190 days was preserved in all patients, thus, no amputations. Patients with pattern 2B injuries required a mean of 5.3 surgical interventions (range—4–7) before skin grafting while pattern 3 required 6.2.57 For skin graft pattern 2B was done after required a mean of 2.1.15 operations while pattern 3 required 2.8.24 Most pattern 2B patients (85.7%) healed primarily while all pattern 3 patients (100%) healing was secondary (Table 2).

Table 2: Injuries according to Yan’s classification
Yan’s classification
Pattern 2B Pattern 3
Injury location (%, n)
 Lower left limb 62.5 (5) 37.5 (3)
 Lower right limb 40 (2) 60 (3)
Injury cause (%, n)
 Motorcycle accident 57.14 (4) 50 (3)
 Pedestrian hit by car 0 33.33 (2)
 Road traffic accident 42.86 (3) 16.67 (1)
Length of stay (days, range) 33.57 (24–65) 45 (31–61)
Mean total surgical time (minutes, range) 852.86 (510–1585) 1012.83 (745–1292)
 General surgery (minutes, range) 610 (395–985) 717.83 (565–867)
 Plastic surgery (minutes, range) 242.86 (115–600) 295 (180–425)
 Mean follow-up (days, range) 184.29 (180–190) 183.33 (180–190)
Days-to-skin graft (days, range) 16.71 (12–26) 20.83 (16–27)
Minor infection (%, n) 14.29 (1) 66.67 (4)
 Pseudomonas aeruginosa MBL (%, n) 14.29 (1) 33.33 (2)
 Klebsiella pneumoniae KPC (%, n) 0 33.33 (2)
Mobility (%, n) 100 (7) 100 (6)
Healing (%, n)
 Primary 85.71 (6) 0 (0)
 Secondary 14.29 (1) 100 (6)

MULTIVARIATE ANALYSIS

Age

Shapiro–Wilk test confirmed normal data distribution (W—0.916, p = 0.442), and Levene’s test showed variance homogeneity (F—1.721, p = 0.216) (Tables 3 and 4). Student’s t-test was performed (Table 5)—though not significant, older patients presented more pattern 3 injuries vs pattern 2B (t—1.502, p = 0.161, Cohen’s d—0.836) with a great effect (Fig. 1).

Table 3: Test of normality for age, length of stay, mean total surgical time, and days-to-skin graft
Test of normality (Shapiro–Wilk) W p
Age Pattern 2B 0.916 0.442
Pattern 3 0.864 0.203
Length of stay Pattern 2B 0.637 <0.001
Pattern 3 0.891 0.325
Mean total surgical time Pattern 2B 0.841 0.101
Pattern 3 0.966 0.864
Days-to-skin graft Pattern 2B 0.855 0.137
Pattern 3 0.949 0.730

Significant results suggest a deviation from normality

Table 4: Equality of variances for age, length of stay, mean total surgical time, and days-to-skin graft
Test of equality of variances (Levene’s) F Degrees of freedom1 Degrees of freedom2 p
Age 1.721 1 11 0.216
Length of stay 0.043 1 11 0.840
Mean total surgical time 1.037 1 11 0.330
Days-to-skin graft 0.029 1 11 0.996
Table 5: Student’s t-test and Mann–Whitney U test for age, length of stay, mean total surgical time, and days-to-skin graft
95% confidence intervals for location parameter
Test Statistic Degrees of freedom p Location parameter Standard error difference Lower Upper Effect size Standard error effect size
Age Student −1.502 11 0.161 −7.333 4.883 −18.081 3.414 −0.836 0.599
Length of stay Mann–Whitney 7.000 0.004 −12.855 −30.000 6.346 × 10–6 −0.667 0.322
Mean total surgical time Student −0.978 11 0.349 −159.976 163.652 −520.172 200.220 −0.544 0.575
Days-to-skin graft Student −1.678 11 0.002 −4.119 2.455 −9.523 1.285 −0.933 0.610

For the student t-test, effect size is given by Cohen’s d, and the location parameter is given by mean difference. For the Mann–Whitney test, the effect size is given by the rank biserial correlation, and the location parameter is given by the Hodges–Lehmann estimate

Fig. 1: Age

Length of Stay

Shapiro–Wilk test did not prove a normal data distribution (W—0.637, p ≤ 0.001) (Tables 3 and 4), so the Mann–Whitney U test was applied (Table 5), finding pattern 3 injuries significantly increased length of stay with a great effect (W—7.0, p = 0.004, rank-biserial correlation—0.667) (Fig. 2).

Fig. 2: Length of stay

Mean Total Surgical Time

Normal data distribution was proven with Shapiro–Wilk’s test (W—0.841, p = 0.101), and variance homogeneity with Levene’s test (F—1.037, p = 0.330) (Tables 3 and 4), which allowed student’s t-test to be applied (Table 5), showing pattern 3 injuries not significantly increased mean total surgical time with a great effect (t—0.978, p = 0.349, and Cohen’s d—0.575) (Fig. 3).

Fig. 3: Mean total surgical time

Days-to-skin Graft

Normal data distribution was proven with Shapiro–Wilk’s test (W—0.855, p = 0.137), with Levene’s test proved variance homogeneity (F—0.029, p = 0.996) (Tables 3 and 4), so student’s t-test was performed (Table 5), showing that pattern 3 injuries significantly increased days-to-skin graft with a great effect (t—1.678, p = 0.002, Cohen’s d—0.933) (Fig. 4).

Fig. 4: Days-to-skin graft

Type of Dressing

Normal data distribution was shown by the Shapiro–Wilk’s test (Table 6), and variance homogeneity with Levene’s test (length of stay—F = 0.231, p = 0.640; mean total surgical time—F = 0.147, p = 0.709; and days-to-skin graft—F = 0.037, p = 0.851), allowing student’s t-test showing no significant difference between conventional and silver-impregnated foam dressings in terms of length of stay, days-to-skin graft, and mean total surgical time (Table 7). A Chi-squared test was performed between the type of dressing and infection, also finding a nonsignificant correlation between these outcomes (Χ2 = 2.236, p = 0.135), with a moderate effect (phi coefficient and Cramer’s V—0.415).

Table 6: Normality test for type of dressing
Test of normality (Shapiro–Wilk) W p
Length of stay Conventional dressing 0.850 0.124
Silver-impregnated dressing 0.805 0.066
Total surgical time Conventional dressing 0.975 0.930
Silver-impregnated dressing 0.883 0.281
Days-to-skin graft Conventional dressing 0.972 0.914
Silver-impregnated dressing 0.924 0.537

Significant results suggest a deviation from normality

Table 7: Student’s t-test for type of dressing
Independent samples t-test
95% confidence intervals for Cohen’s d
t Degrees of freedom p Cohen’s d Standard error Cohen’s d Lower Upper
Length of stay 0.340 11 0.740 0.189 0.559 −0.908 1.278
Total surgical time −0.063 11 0.951 −0.035 0.556 −1.125 1.056
Days-to-skin graft 0.535 11 0.604 0.297 0.562 −0.806 1.388

Type of Healing

Chi-squared test was performed between injury pattern and type of healing, revealing a significant correlation between the outcomes (Χ2 = 12.203, p < 0.001) (phi coefficient and Cramer’s V—0.857) (Tables 8 and 9). Subsequently, it was found that pattern 3 injuries had a 4-time higher probability of undergoing secondary healing (OR—4.031, p = 0.005) (Table 10).

Table 8: Contingency table for the type of healing and injury pattern
Yan’s classification
Primary/secondary healing Pattern 2B Pattern 3 Total
Primary healing Count 6.000 0.000 6.000
Expected count 3.231 2.769 6.000
% of total 46.154% 0.000% 46.154%
Secondary healing Count 1.000 6.000 7.000
Expected count 3.769 3.231 7.000
% of total 7.692% 46.154% 53.846%
Total Count 7.000 6.000 13.000
Expected count 7.000 6.000 13.000
% of total 53.846% 46.154% 100.000%
Table 9: Chi-squared test and effect for type of healing and injury pattern
Chi-squared tests
Value Degrees of freedom p
Likelihood ratio (X2) 12.203 1 <0.001
N 13
Table 10: Odds ratio for type of healing and injury pattern
95% confidence intervals
Odds ratio Lower Upper p
4.031 0.651 7.412 0.005

Infection

Normal data distribution was proven with Shapiro–Wilk’s test (W—0.904, p = 0.313), with Levene’s test showing variance homogeneity (F—1.705, p = 0.218) (Table 11). Student’s t-test was applied (Table 12), showing that as days-to-skin graft increased, minor infection also significantly increased(t = −3.902, p = 0.002, and Cohen’s d—2.224) (Fig. 5).

Table 11: Test of normality and equality of variances for infection
Test of Normality (Shapiro–Wilk) W p
Days-to-skin graft NO 0.904 0.313
Minor infection 0.920 0.530
Test of Equality of Variances (Levene’s) F Degrees of freedom1 Degrees of freedom2 p
Days-to-skin graft 1.705 1 11 0.218

Significant results suggest a deviation from normality

Table 12: Student’s t-test for infection
Independent samples t-test t Degrees of freedom p Cohen’s d Standard error Cohen’s d
Days-to-skin graft −3.902 11 0.002 −2.224 0.796

Student’s t-test

Fig. 5: Infection

DISCUSSION

As early as 1986, Godina reported 532 patients undergoing microsurgical reconstruction for extremity trauma. Patients were classified into three groups—(I) free-flap transfer within 72 hours of the injury (early), (II) between 72 hours and 3 months of the injury (delayed), and (III) 3 months to 12.6 years (late), with a mean of 3.4 years. The flap failure rate was 0.75% in the first group, 12% in the second, and 9.5% in the third group (p < 0.0005 early vs late; p < 0.0025 early vs late). Postoperative infection occurred in 1.5, 17.5, and 6%, respectively; and mean time to bone-healing was 6.8, 12.3, and 29 months, for each group, respectively. Average hospital length of stay was 27, 130, and 256 days; and the number of surgeries averaged 1.3, 4.1, and 7.8 for each group, respectively. Godina’s study concluded that an early approach had better outcomes than a late one.1 The study recommendations are controversial, even after decades,10 and support from many earlier publications.

Since then, a systematic review of patients undergoing free flap, muscle, or fascial flaps with VAC, including 21 studies from 1986 to 2015, showed a total flap failure rate of 1.4%, partial failure of 0.4%, and infection rate of 7.7%, respectively, when coverage was performed <72 hours after injury. The same rates rose up to 8.8, 1.8, and 11.6%, respectively, when coverage was performed >72 hours after injury. The systematic review concluded that flap failure and infection rates were lower when wound coverage was done early, within 72 hours of injury. These findings further support Godina’s recommendation of early soft tissue coverage and recommending VAC when reconstruction is not feasible within 72 hours of injury.10

Many factors may determine the timing of soft tissue coverage and reconstruction in injured patients with extensive extremity traumatic injuries. Despite Godina’s findings suggesting that early coverage is associated with better outcomes, more recently developed advanced wound care strategies such as VAC are deemphasizing the 72-hour interval. Recent reports indicate that management should focus on more effective coordination of peritraumatic and perioperative care, thus individualizing the care to each patient, generally a very complicated polytraumatized patient.4

Other series on lower extremity reconstructions examined the perioperative factors that affected success. Of the 88 patients analyzed, eight had flap loss, eight had flap infections, and 23 had other adverse outcomes. Timing of the reconstruction, VAC system use, injury classification, prior wound status, or presence of polytrauma had no statistically significant impact on the primary outcome. Significance was seen in injury classification and severity on-site complications (p = 0.051) and flap-specific complications (p = 0.073), anyways, the subsequent subgroup analysis did not show significance. Also, no significance was proven with a logistic regression of any recipient site complication including all independent variables.4

In patients with lower limb fractures (including open fractures), VAC use has been shown to reduce flap infection and failure.11 A retrospective study of patients with Gustilo type 3B tibial fractures compared wounds covered with occlusive dressing and VAC. The study found a 33% infection rate and 11% flap failure rate in the first, and 10 and 6% in the second, respectively. Infection rates were significantly different in the two cohorts (p = 0.029); leading to the conclusion that VAC is associated with better outcomes, and may have contributed to the findings.11

Notably, reconstruction of complex extremity trauma, including VAC, may be performed in geriatric or patients with multiple comorbidities with similar functional results.12

Yan’s classification was used in the present study, but others could be used such as the Arnez classification.13 In the later, degloving injuries are classified in four patterns based on the extent of the tissue injury—abrasions/avulsions, noncircumferential degloving, circumferential single-plane injuries, and circumferential multiplane injuries. The Arnez classification may be used for the assessment of the outcome as well.

The present study demonstrated that a disproportionately higher number of degloving extremity injuries occur more commonly in male patients.14 And even though degloving may occur in any area of the body, the most common sites are lower extremities, trunk, scalp, and face.15,16

The surgical approach to degloving injuries may be demanding and varies according to the area to be treated. Mismanagement of such injuries may result in necrosis of the avulsed flap, infection, and even sepsis. Therefore, the first step in the management of such injuries is the reestablishment of skin coverage of the denuded area. Nonetheless, poor cosmetic results and inadequate elasticity are seen after immediate coverage with full- or split-thickness skin grafts from the avulsed skin. VAC system allows evacuation of wound secretions and blood, reducing the risk of seroma, hematoma, and infection while shortening the required time for grafting.5

However, VAC use in other scenarios such as closed surgical wounds on the lower limbs has been shown not to be cost-effective, according to an economic report from the United Kingdom using data from the Wound Healing in Surgery for Trauma multicenter randomized clinical trial.17 Furthermore, no difference was seen in deep surgical site infection at 90 days, disability, quality of life, or scar appearance at 3 or 6 months.18

CONCLUSION

The present study suggests that the VAC system may be a useful tool in managing lower limb trauma injuries (including pattern 3). VAC allows large defect coverage until surgical reconstruction can be done, may reduce infection rate, and help in preserving mobility and thus, aid in limb salvage.

ORCID

Andrea Villarreal-Juris https://orcid.org/0000-0002-0977-3179

Ramiro Villarreal-Juris https://orcid.org/0000-0002-8533-1698

REFERENCES

1. Godina M. Early microsurgical reconstruction of complex trauma of the extremities. Plast Reconstr Surg 1986;78(3):285–292. DOI: 10.1097/00006534-198609000-00001

2. Hou Z, Irgit K, Strohecker KA, et al. Delayed flap reconstruction with vacuum-assisted closure management of the open IIIB tibial fracture. J Trauma 2011;71(6):1705–1708. DOI: 10.1097/TA.0b013e31822e2823

3. Liu DSH, Sofiadellis F, Ashton M, et al. Early soft tissue coverage and negative pressure wound therapy optimises patient outcomes in lower limb trauma. Injury 2012;43(6):772–778. DOI: 10.1016/j.injury.2011.09.003

4. Patterson CW, Stalder MW, Richardson W, et al. Timing of free flaps for traumatic wounds of the lower extremity: have advances in perioperative care changed the treatment algorithm? J Reconstr Microsurg 2019;35(8):616–621. DOI: 10.1055/s-0039-1688712

5. Dini M, Quercioli F, Mori A, et al. Vacuum-assisted closure, dermal regeneration template and degloved cryopreserved skin as useful tools in subtotal degloving of the lower limb. Injury 2012;43(6):957–959. DOI: 10.1016/j.injury.2011.03.020

6. Josty IC, Ramaswamy R, Laing JHE. Vaccum assisted closure: an alternative strategy in the management of degloving injuries of the foot. Br J Plast Surg 2001;54(4):363–365. DOI: 10.1054/bjps.2000.3565

7. Burtt KE, Badash I, Leland HA, et al. The efficacy of negative pressure wound therapy and antibiotic beads in lower extremity salvage. J Surg Res 2020;247:499–507. DOI: 10.1016/j.jss.2019.09.055

8. Costao ML, Achteno J, Knighto R, et al. Negative-pressure wound therapy compared with standard dressings following surgical treatment of major trauma to the lower limb: the WHiST RCT. Health Technol Assess 2020;24(38):1–86. DOI: 10.3310/hta24380

9. Yan H, Gao W, Li Z, et al. The management of degloving injury of lower extremities: technical refinement and classification. J Trauma Acute Care Surg 2013;74(2):604–610. DOI: 10.1097/TA.0b013e31827d5e00

10. Qiu E, Kurlander DE, Ghaznavi AM. Godina revisited: a systematic review of traumatic lower extremity wound reconstruction timing. J Plast Surg Hand Surg 2018;52(5):259–264. DOI: 10.1080/2000656X.2018.1470979

11. Joethy J, Sebastin SJ, Chong AKS, et al. Effect of negative-pressure wound therapy on open fractures of the lower limb. Singapore Med J 2013;54(11):620–623. DOI: 10.11622/smedj.2013221

12. Reiter G, Thomas B, Kühner C, et al. Reconstruction of lower limbs in old age-an interdisciplinary approach: strategies for trauma surgery, vascular surgery and plastic surgery. Chirurg 2019;90(10):806–815. DOI: 10.1007/s00104-019-01023-7

13. Arnez ZM, Khan U, Tyler MPH. Classification of soft-tissue degloving in limb trauma. J Plast Reconstruct Aesthet Surg 2010;63(11):1865–1869. DOI: 10.1016/j.bjps.2009.11.029

14. Lekuya HM, Alenyo R, Kajja I, et al. Degloving injuries with versus without underlying fracture in a sub-Saharan African tertiary hospital: a prospective observational study. J Orthop Surg Res 2018;13(1):2. DOI: 10.1186/s13018-017-0706-9

15. Hakim S, Ahmed K, El-Menyar A, et al. Patterns and management of degloving injuries: a single national level 1 trauma center experience. World J Emerg Surg 2016;11:35. DOI: 10.1186/s13017-016-0093-2

16. Latifi R, El-Hennawy H, El-Menyar A, et al. The therapeutic challenges of degloving soft-tissue injuries. J Emerg Trauma Shock 2014;7(3):228. DOI: 10.4103/0974-2700.136870

17. Png ME, Madan JJ, Dritsaki M, et al. Cost-utility analysis of standard dressing compared with incisional negative-pressure wound therapy among patients with closed surgical wounds following major trauma to the lower limb. Bone Joint J 2020;102-B(8):1072–1081. DOI: 10.1302/0301-620X.102B8.BJJ-2020-0126.R1

18. Costa ML, Achten J, Knight R, et al. Effect of incisional negative pressure wound therapy vs standard wound dressing on deep surgical site infection after surgery for lower limb fractures associated with major trauma: the WHIST randomized clinical trial. JAMA 2020;323(6):519–526. DOI: 10.1001/jama.2020.0059

________________________
© The Author(s). 2023 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by-nc/4.0/), which permits unrestricted use, distribution, and non-commercial reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.