High-energy open fractures of the tibia are frequently associated with tissue loss, wound contamination, and compromised vascularity. The management of these severe injuries remains a challenge for orthopedic reconstructive surgeons. Gustilo type IIIB injuries have a high rate of nonunion and/or osteomyelitis that often results in secondary amputation.1 Meta-analyses show that the functional outcomes of lower-limb salvage are no more successful than those of primary amputation, further complicating management decisions.2,3 With the continual improvements in reconstructive techniques, it is more commonplace to attempt lower-limb salvage when faced with these severe injuries.1
To help guide treatment decisions, studies have compared the timing of definitive fixation and soft tissue coverage with associated poor outcomes or eventual secondary amputation.4–9 Historically, “early” soft tissue coverage has been defined as wound closure within 7 days of injury and “late” coverage as any time after 7 days; these studies concluded that increased time to soft tissue coverage was a significant predictor of complications.5–8 Flap coverage within 72 hours of the injury was associated with shorter length of hospitalization, lower deep infection rates, and a smaller number of surgical procedures, although there was no significant difference in time to bone union, flap failure, and secondary amputation.4
The literature is replete with small retrospective studies reporting variable rates of reconstructive success in Gustilo type IIIB limb salvage. Many studies do not include cases of delay in soft tissue management of the injury secondary to patient instability, transfer from an outside institution after initial management, or the clinical evolution of an open tibia wound from Gustilo type IIIA to type IIIB during the first few weeks after injury. All of these conditions resulting in delay of soft tissue coverage are commonly encountered at a tertiary hospital orthopedic trauma service. Larger studies with consistent surgical practice methods during longer periods are necessary to guide treatment.
A retrospective analysis was performed of a group of consecutive patients who presented to the authors' level I trauma center during 13 years. All patients who underwent reconstructive lower extremity salvage surgery and soft tissue reconstruction with the final diagnosis of Gustilo type IIIB open tibia fracture were eligible. Patients who were transferred from outside institutions and those who required delayed soft tissue coverage for a variety of reasons were also included.
Materials and Methods
A retrospective chart review included all patients in the soft tissue reconstruction database from 2001 to 2014. All available medical records, including radiographs and clinical reports, were reviewed. A total of 140 patients with 142 limbs with the final diagnosis of Gustilo type IIIB open fracture of the tibia were identified; all patients presented for soft tissue flap coverage within 6 weeks of the initial injury. The majority of patients presented to the authors' center for initial orthopedic treatment of an obvious IIIB injury. However, patients transferred to the authors' service after orthopedic stabilization of the bone but needing higher-level soft tissue care were also included. Two patients had bilateral fractures, while 3 had segmental tibial injuries requiring 2 flaps on the same limb. There were 145 injuries. Demographic variables, comorbidities, method of injury, all surgical procedures, and main outcome measures were collected into a standardized database.
Each patient was treated by an orthopedic trauma surgeon trained in micro-vascular reconstruction who managed the care from initial consultation to final follow-up. All patients were initially treated with surgical debridement, application of negative pressure wound therapy, and administration of intravenous antibiotics until definitive soft tissue closure. All patients who arrived at the authors' institution were given prophylactic antibiotics as per the standard of care for open fracture treatment. The authors could not account for antibiotic administration performed at outside institutions before transfer to their facility, but it was assumed to be consistent with standard open fracture prophylaxis. Flap surgeries were performed as soon as possible after definitive bone fixation or, in the case of transfers, as soon as wounds were clean and fixation was appropriate. Each injury and associated wound were evaluated by an experienced orthopedic trauma surgeon (B.J.H.) who regularly performs soft tissue coverage including free and local flaps. The amount of debridement and temporary coverage before definitive fixation and definitive soft tissue coverage was assessed and performed on an individualized basis and could not be standardized. Repeat irrigation and debridements were routinely performed 48 hours apart based on operating room availability and individual wound characteristics. All patients had at least one debridement performed. Free flaps were chosen when local flaps were not possible, primarily because of wound size and zone of injury. Bone morphogenetic protein gels were placed at the fracture sites under flaps at final soft tissue surgery in the setting of plate or nail treatment with direct bone apposition and/or small segmental defects. Bone morphogenetic protein was not used with initial Ilizarov frame application for segmental bone loss but was used at the docking site at the conclusion of bone transport. Bacterial cultures were not routinely obtained unless osteomyelitis was suspected based on symptoms and imaging. Osteomyelitis was defined as positive results on bacterial cultures.
The initial analysis included calculation of frequencies for all variables, as well as means, standard deviations, and medians for continuous variables such as age and time to soft tissue coverage. Bivariable comparisons of possible associations of categorical variables were undertaken using Pearson's chi-square test or Fisher's exact test. Outcomes for these analyses were primary nonunion, osteomyelitis, and secondary amputation. The “exposures” in these analyses included sex, age, smoking, hypertension, and diabetes, as well as method of injury, surgical characteristics, procedures (eg, type of flap, initial flap failure, vein graft, definitive fixation method), and other “outcomes.” Differences in mean values of continuous variables (eg, age, time to soft tissue coverage) were examined using t tests.
The results of these preliminary analyses were used to construct logistic regression models to examine the role of time to soft tissue coverage in the development of primary outcomes with control for demographics, health behaviors, and clinical and surgical characteristics. All analyses were conducted using SPSS version 24 software (IBM, Armonk, New York). P≤.05 was considered statistically significant. Logistic regression results were reported with odds ratios (OR) and 95% confidence intervals (CI).
The patients included 108 males and 32 females with a mean±SD age of 39.4±16.6 years (Table 1). Motor vehicle accidents were the most common mechanism of injury, causing 78% of fractures. Definitive fixation methods and surgical procedures are summarized in Table 2. A free flap was used 78% of the time, and 55 (38.7%) of the patients had definitive fixation and definitive soft tissue coverage (“fix and flap”) performed on the same day. The mean time from injury to definitive fixation was 5.6 days (range, 0 to 47 days), and the mean time to soft tissue coverage was 14.0 days (range, 0 to 86 days). The time to soft tissue coverage was greater than 7 days in 57.9% (n=81) of the patients and 3 days or less in 13.6% (n=19).
Demographic and Injury Characteristics of the Patients (N=140)
Characteristics of Surgical Procedures Among the Patients (N=145 Initial Surgeries)
Full weight-bearing status and successful limb salvage were achieved by 83% of the patients. Secondary amputation was required for 20 patients (1 patient had both legs amputated), with 52% of these amputations attributed to refractory osteomyelitis (Table 3). Of the 113 free flaps performed, 8 failed in the acute postoperative period. All of these returned to the operating room for hematoma evacuation or exploration of anastomosis, with 3 resulting in salvage. Of the 5 that could not be initially salvaged, 1 patient decided not to have further reconstruction and chose to have an amputation. The other 4 underwent secondary soft tissue reconstruction. Of these 4, 1 secondary reconstruction was successful and 3 went on to have a secondary amputation due to soft tissue infection or osteomyelitis. Figure 1 illustrates the range of times of soft tissue coverage for patients with or without secondary amputation performed; several patients with the longest time to soft tissue coverage had successful limb salvage rather than a secondary amputation.
Time to the soft tissue coverage (days) for patients with and without secondary amputation performed.
Complete bone union was achieved in 126 of the fractures; 57% of these unions occurred “primarily” with no additional bony surgeries after the soft tissue coverage (Table 3). Osteomyelitis occurred in 30% of the patients, with 78% of these infections successfully eradicated with additional procedures. Individuals with diabetes and a free flap performed with a vein graft were more likely to experience osteomyelitis (P=.034 and .067, respectively). Time to soft tissue coverage was similar for those with and without primary nonunion (P=.61). Time to soft tissue coverage for those with osteomyelitis (mean, 17.1 days) was slightly longer than that for those without osteomyelitis (mean, 12.7 days); however, the difference was not statistically significant (P=.10).
Results of the logistic regression model are provided in Table 4. Time to soft tissue coverage was not a significant predictor of osteomyelitis (P=.310), while diabetes was associated with an adjusted OR of 3.67 (95% CI, 1.01–13.36) (P=.048). In an adjusted logistic regression model examining primary nonunion, those with osteomyelitis were 8 times more likely to experience a primary nonunion (P<.001), whereas time to successful soft tissue coverage had a nonsignificant association with nonunion (P=.170). A logistic regression model examining secondary amputation also found that time to soft tissue coverage was not a significant predictor of amputation (P=.730), while significant positive associations were noted for both osteomyelitis (P<.001) and hypertension (P=.016).
Logistic Regression Models of Predictors of Osteomyelitis, Primary Nonunion, and Secondary Amputation
Logistic regression modeling in which time to soft tissue coverage was categorized as “on or before 7 days” or “after 7 days” found a 36% nonsignificant increase (P=.46) in the likelihood of osteomyelitis with after 7 days relative to before. However, somewhat surprisingly, a 54% decreased risk of nonunion (P=.08) was found when soft tissue coverage occurred after 7 days. Similar models using tissue coverage before or after 72 hours also revealed no effect of time to soft tissue coverage for both osteomyelitis and nonunion. Models to determine associations from time of definitive fixation to time of definitive soft tissue coverage found a 4% decreased likelihood of a nonunion (P=.035) with longer time differences between definitive fixation and soft tissue coverage. There were no significant associations with an increase in time difference and osteomyelitis (OR, 1.01; 95% CI, 0.98–1.04; P=.48) or amputation (OR, 1.02; 95% CI, 0.98–1.07; P=.26).
There is general agreement that timely and robust soft tissue coverage of type IIIB open tibial fractures once definitive fracture stabilization has been performed is key when attempting reconstruction.2,3 Many studies have concluded that delay in soft tissue coverage will lead to higher rates of deep infection and decreased rates of union; therefore, primary amputation is often considered more appropriate in the setting of delay.5–8 The results of the current study do not support these opinions, instead revealing that delays to soft tissue coverage were not significant predictors of osteomyelitis or nonunion or of secondary amputation. Early definitive soft tissue coverage is the ideal, and it was accomplished in a single fix and flap procedure for 55 of the patients in this study (38.7% of injuries). However, early or immediate coverage is often not the reality, as severely traumatized tissue often takes many days to evolve clear margins of viability. One study that strongly recommends early soft tissue reconstruction reported that the majority of the definitive soft tissue coverage was performed beyond 5 days, and even beyond 2 weeks in many cases.10 The patients in the current study with longer time to soft tissue coverage were delayed for a variety of reasons, with delayed consultation being the most common. Despite these delays in presentation to the authors' service, the time from when the authors assumed care to the time of definitive soft tissue coverage was typically within a 7-to 10-day interval and therefore similar to historical studies.5–8,10
This study was larger than most previous retrospective analyses, and the demographics of the patients were comparable to those of the overall population; consequently, it is unlikely that the results were based on bias created by preselection of a healthier population. The prevalence of method of injury was also similar to that of prior studies. This study's findings also highlight the concept that high-energy lower extremity open fracture wounds initially thought to be IIIA can evolve during the first few weeks into a more complex injury due to the initial trauma, surgical exposure, and fixation choices. This study is unique in that the authors did not initially categorize a determined time for “early” vs “late” soft tissue coverage. Use of continuous “time to” variables allowed for greater precision in the estimates obtained. The authors performed additional analyses using these categorical classifications to compare their findings with those of historical studies.4–8 These analyses showed that neither of the categorical coverage variables significantly predicted osteomyelitis or nonunion; however, the inverse association with nonunion could simply reflect the need for additional treatments outside of the typical 1-week window. The authors also analyzed the association between the time from definitive fixation (rather than injury) and the time of soft tissue coverage, with delays once again not being predictive of poor outcomes.
The prevalence of osteomyelitis in the current study (30%) is within the range seen in similar literature, but higher than some recent studies,3,4,8,11 primarily due to the heterogeneity of the patients. Studies with lower prevalence often excluded IIIB injuries that were not amenable to fixation with intramedullary nails, and most did not include those injuries with bone loss requiring bone transport. These studies also did not include patients transferred for definitive soft tissue coverage after outside initial treatment, instead studying only patients presenting to their center immediately after injury for definitive care. This study's heterogeneous inclusion criteria may have negatively impacted the 17% rate of secondary amputation, given the trend observed in the results. A previously performed meta-analysis of 28 observational studies cited secondary amputation rates of 0% to 27%, which is consistent with the current study.3 Although other studies noted secondary amputation rates of 11%, 9%, 40%, 18%, and 28%, several of these did not include patients with delayed presentations.2,8
This study had several potential limitations. First, it was a retrospective chart review, including radiographic evidence and detailed operative, clinical, and hospital course records. This should not have jeopardized the conclusions, however, because the main outcomes of nonunion, osteomyelitis, and amputation were readily identifiable within the charts. Second, although the authors diligently performed soft tissue coverage and limb salvage as expeditiously as possible once consulted, they were unable to control any of the treatment that occurred before the patients presented. Many patients were diagnosed with Gustilo type IIIA injuries by another surgeon at the onset of care, being transferred to the authors' service once the injury was definitively diagnosed as Gustilo type IIIB. The delayed presentation allows for the natural variability in timing to soft tissue coverage. Third, this study occurred during a 13-year period in which a wide variability of bone and soft tissue injury was identified, and the care of the fracture was subject to a variety of fixation techniques. Although it would be ideal to evaluate a homogeneous patient population with midshaft fractures all treated with a single technique, this was not the patient population encountered in this study. It is therefore impossible to assess the role that any one fixation technique may have played in the outcomes. For example, although external fixation is generally not a preferred definitive fixation method, several of the patients in this study were treated with Ilizarov frames because of highly comminuted segmental injury and/or initial acute bone loss exceeding 2 cm. Likewise, plates were used for definitive fixation of proximal and distal metaphyseal fractures, especially when the flaps could be placed for coverage either at the same surgical setting or within 48 to 72 hours. This variability allows for increased generalizability of the study results to the real-world situations where patients with Gustilo type IIIB open tibia fractures often present in delayed and/or less than ideal situations.
Appropriate soft tissue coverage performed even up to 6 weeks from the time of initial injury can still yield positive outcomes in the salvage of severe open tibia fractures. This large retrospective study was heterogeneous. Therefore, it is generally applicable to currently practicing orthopedic and reconstructive surgeons at most level I trauma/tertiary care centers because these challenging presentations and situations are common. This study's results indicate that although there is a trend toward increased likelihood of adverse outcomes with increasing delay, the inability to obtain early soft tissue coverage should not be the primary deterrent to attempting limb reconstructive surgery.
- Hallock GG. Evidence-based medicine: lower extremity acute trauma. Plast Reconstr Surg. 2013;132(6):1733–1741. doi:10.1097/PRS.0b013e3182a80925 [CrossRef]
- Busse JW, Jacobs CL, Swiontkowski MF, Bosse MJ, Bhandari MEvidence-Based Orthopaedic Trauma Working Group. Complex limb salvage or early amputation for severe lower-limb injury: a meta-analysis of observational studies. J Orthop Trauma. 2007;21(1):70–76. doi:10.1097/BOT.0b013e31802cbc43 [CrossRef]
- Saddawi-Konefke D, Kim HM, Chung KC. A systematic review of outcomes and complications of reconstruction and amputation for type IIIB and IIIC fractures of the tibia. Plast Reconstr Surg. 2008;122(6):1796–1805. doi:10.1097/PRS.0b013e31818d69c3 [CrossRef]
- Chua W, De SD, Lin WK, Kagda F, Murphy D. Early versus late flap coverage for open tibial fractures. J Orthop Surg (Hong Kong). 2014;22(3):294–298. doi:10.1177/230949901402200305 [CrossRef]
- Cierny G III, Byrd HS, Jones RE. Primary versus delayed soft tissue coverage for severe open tibial fractures: a comparison of results. Clin Orthop Relat Res. 1983;(178):54–63.
- D'Alleyrand JC, Manson TT, Dancy L, et al. Is time to flap coverage of open tibial fractures an independent predictor of flap-related complications?J Orthop Trauma. 2014;28(5):288–293. doi:10.1097/BOT.0000000000000001 [CrossRef]
- Hwang KT, Kim SW, Sung IH, Kim JT, Kim YH. Is delayed reconstruction using the latissimus dorsi free flap a worthy option in the management of open IIIB tibial fractures?Microsurgery. 2016;36(6):453–459. doi:10.1002/micr.22428 [CrossRef]
- 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.
- Russell GG, Henderson R, Arnett G. Primary or delayed closure for open tibial fractures. J Bone Joint Surg Br. 1990;72(1):125–128. doi:10.1302/0301-620X.72B1.2298770 [CrossRef]
- Lack WD, Karunakar MA, Angerame MR, et al. Type III open tibia fractures: immediate antibiotic prophylaxis minimizes infection. J Orthop Trauma. 2015;29(1):1–6.
- Bosse MJ, MacKenzie EJ, Kellam JF, et al. An analysis of outcomes of reconstruction or amputation after leg-threatening injuries. N Engl J Med. 2002;347(24):1924–1931. doi:10.1056/NEJMoa012604 [CrossRef]
Demographic and Injury Characteristics of the Patients (N=140a)
|Sex, No. (%)|
| Male||108 (77)|
| Female||32 (23)|
| Mean (SD)||39.4 (16.6)|
|Diabetes, No. (%)|
| No||127 (91)|
| Yes||11 (9)|
|Smoker, No. (%)|
| No||70 (52)|
| Yes||66 (48)|
|Hypertension,b No. (%)|
| Yes||33 (24)|
|Other vascular issues,b No. (%)|
| Yes||29 (21)|
| Any vascular issue||46 (33)|
|Injury method, No. (%)|
| Motor vehicle accidentc||109 (78)|
| Fall||19 (13)|
| Gun shot||8 (6)|
| Mower||2 (1)|
| Auger||1 (1)|
| Other||1 (1)|
|Employment after injury, No. (%)|
| No||12 (9)|
| Yes||47 (33)|
| Disabled||46 (33)|
| Student||17 (12)|
| Retired||8 (6)|
| Homemaker||3 (2)|
| Deceased||1 (<1)|
| Unknown||6 (4)|
Characteristics of Surgical Procedures Among the Patients (N=145 Initial Surgeriesa)
|Surgical Characteristic||No. (%)|
| Free||102 (70)|
| Local||32 (22)|
| Both||11 (8)|
| Intramedullary nail||36 (25)|
| Open reduction and internal fixation||84 (57)|
| Ilizarov||14 (10)|
| External fixator||11 (8)|
|Type of flap|
| Latissimus dorsi||46 (32)|
| Rectus abdominis||28 (19)|
| Radial forearm||29 (20)|
| Gastrocnemius||21 (15)|
| Soleus||8 (5)|
| Sural||6 (4)|
| Other||7 (5)|
|Recipient site of flap|
| Distal tibia||85 (59)|
| Mid tibia||26 (18)|
| Proximal tibia||25 (17)|
| Distal and mid||3 (2)|
| Mid and proximal||4 (3)|
| Distal–mid–proximal||2 (1)|
| Posterior tibial||91 (60)|
| Anterior tibial||13 (9)|
| Popliteal||4 (2)|
| Femoral artery branch||2 (1)|
| Peroneal||1 (<1)|
| Rotation||41 (27)|
| No||131 (89)|
| Yes||16 (11)|
| No||98 (69)|
| Yes||43 (30)|
| Lost to follow-upb||1 (<1)|
| Osteomyelitis eradicated|
| No||9 (22)|
| Yes||31 (78)|
| No||81 (57)|
| Yes||60 (42)|
| Lost to follow-up||1 (<1)|
| Union after additional surgery|
| No||5 (10)|
| Yes||45 (90)|
| No||119 (83)|
| Yes, secondaryc||21 (16)|
| Yes, primary||2 (<1)|
| Reason for amputation|
| Osteomyelitis||11 (52)|
| Distal gangrene||2 (9.5)|
| Infected nonunion||2 (9.5)|
| Nonunion, no infection||3 (14)|
| Decubitus ulcers||1 (5)|
| Flap failure||1 (5)|
| Noncompliance, distraction osteogenesis||1 (5)|
Logistic Regression Models of Predictors of Osteomyelitis, Primary Nonunion, and Secondary Amputationa
|Predictor||Osteomyelitis||Primary Nonunion||Secondary Amputation|
|AOR||95 % CI||P||AOR||95% CI||P||AOR||95% CI||P|
|Time to soft tissue coverage, d||1.01||0.99–1.04||.310||0.98||0.95–1.01||.170||0.99||0.94–1.04||.730|
|First flap survived|
|Age at injury, y||1.01||0.98–1.03||.680||1.01||0.98–1.03||.520||0.95||0.91–0.99||.041|