The tibia is the most commonly fractured long bone, with 24% of shaft fractures presenting as open injuries.1 The traumatic opening introduces an extra layer of complexity, increasing the risk of infection and local wound complications. Open tibia fractures are more prone to infection than nontibia open fractures (10.5% vs 5.3%, respectively), probably because of the subcutaneous location of the bone and challenges associated with soft tissue management and muscular coverage.2 To reduce these risks, management of open tibia shaft fractures requires timely antimicrobial therapy2 and thorough surgical debridement.3
Previous studies showed that definitive plate fixation in open tibia shaft fractures is associated with a high rate of infection, particularly with increased Gustilo-Anderson types.4,5 Plating has fallen out of favor because of the negative effects of the surgical dissection required for placement of appropriately sized surface implants. Intramedullary nailing has become the treatment of choice for open tibia fractures because it offers superior biomechanical properties and excellent clinical outcomes.6,7
To assist in maintaining fracture reduction during intramedullary nailing, several techniques have been used, including blocking screws, percutaneous clamping, manual reduction with an assistant, and adjunctive plate fixation.8–10 A provisional plate (mini-fragment or small fragment plate) may be used after debridement of the traumatic wound, and the plate is either retained or removed before definitive closure.
The current study assessed whether the use of a provisional plate increased the risk of complications in open tibial shaft fractures. In addition, the study analyzed the subgroup of patients who were treated with a provisional plate to determine whether removal of the plate was associated with complications. Despite multiple studies describing the provisional plating technique, few studies have looked at its association with complications.11
Materials and Methods
A retrospective study was conducted of patients who were treated with intramedullary nailing for an open tibia fracture at a level 1 trauma center over an 8.5-year period. A list was compiled from the hospital surgical billing database, using International Classification of Diseases, Ninth Revision (823.30, 823.32), and Current Procedural Terminology (27759) codes. Patients were included if they met the following criteria: (1) 18 years or older; (2) treated between January 1, 2005, and June 30, 2013; (3) diagnosed with open diaphyseal fracture of the tibia; and (4) treated operatively with intramedullary nailing, with or without provisional plate fixation.
All patients underwent thorough debridement of the traumatic wound. The fracture was reduced, and in certain cases, a mini-fragment or low-profile small fragment (one-third tubular) provisional plate was used (off-label use) to maintain reduction at the discretion of the attending orthopedic trauma surgeon. Intramedullary nailing was the definitive method of fixation. Implants were either retained, with the wounds closed before nailing, or the provisional plate was removed after nailing and before definitive wound closure. Indications for provisional plate fixation included transverse or short oblique fractures that were not amenable to a distal clamp and simple segmental fractures with short working lengths. The authors' protocol is to administer intravenous antibiotics on arrival to the emergency department for all patients with open fractures.
The primary end point was to evaluate whether the use of a provisional plate before intramedullary nailing increases the risk of complications. Electronic medical records were reviewed for patient age, sex, body mass index, diabetes, smoking history, mechanism of injury, side of injury, Gustilo-Anderson type, and AO/Orthopaedic Trauma Association (OTA) fracture classification.12,13 Operative reports and radiographs were reviewed to determine whether provisional plating was used, and all cases were categorized as either yes or no. Those who had a provisional plate were further stratified into 2 groups: plate removed vs plate retained at the time of definitive closure.
Because few patients had complications, outcome measures for postoperative complications were classified into the following groups: (1) infection (superficial or deep); (2) noninfectious complication (wound healing, delayed union, nonunion, or compartment syndrome); and (3) any complication (combination of infection and noninfectious complications). Patients with complications were independently compared with those who did not have complications. Potential confounders determined in the univariate analysis and added to the multivariate logistic regression model were age, Gustilo-Anderson type, and AO/OTA fracture classification. To minimize potential bias, data were collected and verified by 2 independent surgeons (M.L., R.A.H.). Data were also independently verified and checked for statistical validity (L.R.).
Statistical analysis included descriptive statistics and baseline comparisons of the provisional plate group. Chi-square statistics were used for categorical variables, and independent t tests were used for continuous variables. Multiple logistic regression analyses were conducted to answer the research questions, with adjustments for confounding variables. Results were considered significant at alpha=.05 (2-tailed). Statistical analysis was performed with SPSS version 18 software (IBM, Armonk, New York).
During the study period, 143 patients were diagnosed with open diaphyseal tibia fracture. After exclusions for age (n=7), death (n=4), acute amputation (n=1), and insufficient follow-up (n=27), 104 patients remained in the analysis (4 bilateral tibia fractures, N=108 extremities). Of the 108 extremities that were treated with intramedullary nailing, 35 underwent provisional plating and 73 did not.
Average patient age was 37 years (±12.9) (SD), and most patients were male (75%) and nonsmokers (68.9%) (Table 1). Average body mass index was 26.9 kg/m2 (±5.3), and 3.9% of patients had a history of diabetes mellitus. In most cases, the mechanism of injury was related to road traffic accidents (approximately 67% motor vehicle collision, motorcycle collision, or pedestrian struck) vs falls, crush injuries, and other mechanisms. Average follow-up was approximately 10 months. Overall, 27.8% of patients had a complication (16.7% superficial or deep infection; 11.1% other). Of the infections, 5 were associated with wound breakdown. A provisional plate was used in 32.4% of extremities (n=35), and of these, 57.1% (n=20) were retained permanently and 42.9% (n=15) were removed before definitive wound closure. In a subgroup analysis of noninfectious complications, particularly delayed union (n=1) and nonunion (n=8), only 2 cases of nonunion had a provisional plate (n=1 removed; n=1 retained).
Demographic and Clinical Characteristics by Provisional Plate Use
Comparison of patients with and without a provisional plate showed no significant differences with respect to body mass index, smoking status, history of diabetes, mechanism of injury, or side of injury (Table 1). However, those with a provisional plate were significantly older (t=2.74, df=102, P=.007), and fewer were classified as AO/OTA 42-A (37.1% vs 67.1%, P=.01) compared with the non-plated group. Therefore, age and AO/OTA classification were added as confounding variables in the overall analysis. Additionally, because Gustilo-Anderson type approached statistical significance at baseline by provisional plate use (P=.066, greater percentage of provisional plate use with more complex fractures) and was associated with having a complication (odds ratio, 2.40; 95% confidence interval, 1.01–5.68; P=.046, Table 2), Gustilo-Anderson type was also controlled for in the overall multivariate analysis.
Logistic Regression Analysis: Risk of Infection and Any Complication
In the multiple logistic regression analysis (Table 2), after the authors controlled for age, AO/OTA classification, and Gustilo-Anderson type, provisional plate use did not significantly increase the risk of infection (adjusted odds ratio, 1.64; 95% confidence interval, 0.51–5.32; P=.41) or the risk of any type of complication (infection and noninfectious) (adjusted odds ratio, 1.24; 95% confidence interval, 0.46–3.35; P=.67) (Table 3).
Provisional Plate Use and Complications
In the provisional plate subgroup (n=35), after intramedullary nailing, the plate was retained in 20 extremities and removed in 15 extremities. Of those who had the plate retained, 60% did not have a complication (n=12) and 40% had a complication (n=8) (25% infection; 15% other complication). Compared with those who had the provisional plate removed, 73.3% did not have a complication (n=11) and 26.6% had a complication (n=4) (13.3% infection; 13.3% other complication). Plate removal did not significantly decrease the odds of infection (adjusted odds ratio, 0.43; 95% confidence interval, 0.07–2.69; P=.36) or any complication (adjusted odds ratio, 0.55; 95% confidence interval, 0.12–0.46; P=.44) compared with retention of the provisional implant (adjusted for AO/OTA classification and Gustilo-Anderson type) (Table 4).
Provisional Plate Retained Versus Removed and Complications
Intramedullary nailing of open tibial shaft fractures has been shown to be the most effective treatment option, with good clinical outcomes reported.6,7 However, because of the high-energy mechanism of injury, significant comminution of the fracture, and high risk of complications, intramedullary nailing can pose a challenge. Because obtaining and maintaining fracture reduction can be difficult, surgeons have sought adjunct procedures to facilitate intramedullary nailing.
Provisional plating is an accepted technique to maintain alignment and facilitate internal fixation.11,14,15 Archdeacon and Wyrick11 described a technique known as as reduction plating, in which one-third tubular plates or mini-fragment plates are used. Plates with unicortical fixation were used to help to maintain reduction before definitive fixation, and the authors' preference was to retain the plate to assist with nailing.11 Dunbar et al14 described the technique with a thicker, 3.5-mm small fragment dynamic compression plate (DCP) or limited contact dynamic compression plate (LC-DCP) (Synthes, Paoli, Pennsylvania) to maintain fracture reduction during placement of intramedullary nails in 32 Gustilo-Anderson type III open tibial shaft fractures. They concluded that provisional plating provided stability and maintained accurate reduction during reaming and insertion of the nail. All plates were removed after insertion of the tibial nail before final wound closure. The current authors consider this necessary because of concerns about soft tissue with thicker plates. Grimsrud and Siebler15 described provisional plate fixation in both bone forearm fractures as a temporary internal external fixator. Although this technique is beyond the scope of the current study, it is another novel concept in provisional fixation.
In the current study, the overall complication rate for open tibia fractures was similar to that cited in the literature. The superficial and deep infection rate was 16.7% compared with a deep infection rate of 13.3% reported by Dunbar et al.14 After the authors controlled for age, AO/OTA classification, and Gustilo-Anderson classification, use of a provisional plate did not significantly increase the odds of infection or any other measured complication (nonunion, delayed union, or compartment syndrome) compared with patients who did not have a provisional plate.
This study was conducted to determine whether there was an increased risk of complications with retention vs removal of a provisional plate. Some surgeons consider plate removal important to decrease the mass effect of hardware in the region of a potentially tenuous wound. Other surgeons propose closing the wound over the provisional implant before intramedullary nailing to preserve the reamings. In the current study, removing the provisional plate was not associated with a decrease in infection or any complication. However, this study had a small subgroup sample size and the investigation would need to be replicated in a larger cohort. Nonetheless, local soft tissue may respond better with plate removal. The authors cannot comment on whether retention of the implant positively affected fracture union because that was not the focus of this study.
The benefits, risks, and costs of using a provisional plate must be weighed. Although the current study did not directly measure the fracture gap, provisional plating can be a valuable technique to decrease the fracture gap. In contrast, because provisional plating of tibia fractures is typically performed through the traumatic wound, there is concern that this technique causes additional trauma to the already compromised soft tissue envelope. However, an earlier report argued that a provisional plate actually required less soft tissue stripping than bone reduction clamps placed directly through a traumatic wound.13 The current authors' technique is to be meticulous with soft tissue handling and work through the debridement wound. Further, a 2.0-mm or 2.7-mm, 4-hole plate plus cortical screws can cost approximately $550 to $840 retail (DePuy-Synthes, Paoli, Pennsylvania), but the cost is variable by facility type and contracted volume discounts. Because provisional plating can expedite intramedullary nailing, the cost of the provisional plate can be offset by shortened operating room time.14,16,17 With operating room costs ranging from $22 per minute to $133 per minute, the time saved during nailing with a provisional plate to assist with reduction should be considered.18
This study had several limitations. First, this was a retrospective chart review and was restricted to the available data collected from electronic medical records. Second, the decision to use provisional plate fixation and the decision to retain or remove the implant were based on the surgeon's preference, which can introduce a selection bias. Third, the study did not account for the time interval between injury and initial dosage of antibiotics in the emergency department or the time interval between injury and initial operative debridement as confounding variables. Current data suggest that early administration of appropriate antibiotics is the most important variable for preventing wound infections in open fractures.19 However, the authors' protocol for all open fractures is to administer intravenous antibiotics on arrival to the emergency department, with irrigation and debridement performed the same day or scheduled as the first case the next day. This study did not measure the biomechanical effect of using a provisional plate to aid intramedullary nailing. In a separate biomechanical study with synthetic long bone models, a provisional mini-fragment plate enhanced compressive forces for fixation with dynamic compression plates.20 Further studies are needed to explore this effect with intramedullary nailing. However, based on the current findings, the effect is likely to be negligible, given the difference in size between the nail and the plates. Finally, the results of this study should be interpreted with caution. Because this study is underpowered (post hoc power, 0.15), with a sample size of 108, nonsignificant differences in results should not dictate a change in practice until the findings are replicated in studies with larger sample sizes.
This study is the first to report on the use of provisional plate fixation (mini-fragment/small fragment plate) and the associated complications. Use of a provisional plate did not result in a significantly increased overall complication rate. Although this was not the focus of the study, the authors believe that provisional plate fixation facilitates the surgical procedure and improves fracture alignment, potentially improving outcomes. This technique is a viable option for the treatment of open tibia fractures. The question of removal vs retention of provisional plates remains unanswered, but the current study opens the door for higher-powered studies in the future.
- Court-Brown CM, McBirnie J. The epidemiology of tibial fractures. J Bone Joint Surg Br. 1995; 77(3):417–421.
- Patzakis MJ, Wilkins J. Factors influencing infection rate in open fracture wounds. Clin Orthop Relat Res. 1989; 243:36–40.
- Sanders R, Swiontkowski M, Nunley J, Spiegel P. The management of fractures with soft-tissue disruptions. J Bone Joint Surg Am. 1993; 75(5):778–789.
- Bach AW, Hansen ST Jr, . Plates versus external fixation in severe open tibial shaft fractures: a randomized trial. Clin Orthop Relat Res. 1989; 241:89–94.
- Clifford RP, Beauchamp CG, Kellam JF, Webb JK, Tile M. Plate fixation of open fractures of the tibia. J Bone Joint Surg Br. 1988; 70(4):644–648.
- Melvin JS, Dombrosk DG, Torbert JT, Kovach SJ, Esterhai JL, Mehta S. Open tibial shaft fractures: II. Definitive management and limb salvage. J Am Acad Orthop Surg. 2010; 18(2):108–117. doi:10.5435/00124635-201002000-00005 [CrossRef]
- Foote CJ, Guyatt GH, Vignesh KN, et al. Which surgical treatment for open tibial shaft fractures results in the fewest reoperations? A network meta-analysis. Clin Orthop Relat Res. 2015; 473(7):2179–2192. doi:10.1007/s11999-015-4224-y [CrossRef]
- Collinge CA, Beltran MJ, Dollahite HA, Huber FG. Percutaneous clamping of spiral and oblique fractures of the tibial shaft: a safe and effective reduction aid during intramedullary nailing. J Orthop Trauma. 2015; 29(6):e208–e212. doi:10.1097/BOT.0000000000000256 [CrossRef]
- Ricci WM, O'Boyle M, Borrelli J, Bellabarba C, Sanders R. Fractures of the proximal third of the tibial shaft treated with intramedullary nails and blocking screws. J Orthop Trauma. 2001; 15(4):264–270. doi:10.1097/00005131-200105000-00005 [CrossRef]
- Robertson A, Giannoudis PV, Matthews SJ. Maintaining reduction during unreamed nailing of a segmental tibial fracture: the use of a Farabeuf clamp. Injury. 2003; 34(5):389–391. doi:10.1016/S0020-1383(01)00195-4 [CrossRef]
- Archdeacon MT, Wyrick JD. Reduction plating for provisional fracture fixation. J Orthop Trauma. 2006; 20(3):206–211. doi:10.1097/00005131-200603000-00007 [CrossRef]
- Gustilo RB, Anderson JT. Prevention of infection in the treatment of one thousand and twenty-five open fractures of long bones: retrospective and prospective analyses. J Bone Joint Surg Am. 1976; 58(4):453–458.
- Marsh JL, Slongo TF, Agel J, et al. Fracture and dislocation classification compendium: 2007. Orthopaedic Trauma Association classification, database and outcomes committee. J Orthop Trauma. 2007; 21(suppl 10):S1–S133. doi:10.1097/00005131-200711101-00001 [CrossRef]
- Dunbar RP, Nork SE, Barei DP, Mills WJ. Provisional plating of type III open tibia fractures prior to intramedullary nailing. J Orthop Trauma. 2005; 19(6):412–414. doi:10.1097/01.bot.0000153446.34484.70 [CrossRef]
- Grimsrud C, Siebler J. Small fragment locking plate as provisional internal fixation in severe open both-bone forearm fractures. J Hand Surg Am. 2014; 39(8):1615–1619. doi:10.1016/j.jhsa.2014.05.015 [CrossRef]
- Bassi JL, Mahindra P, Singh N, Selhi HS, Garg R. Provisional plating of type III open tibia fractures prior to intramedullary nailing [letter to the editor]. J Orthop Trauma. 2006; 20(65):372–373. doi:10.1097/00005131-200605000-00016 [CrossRef]
- Dunbar RB, Nork SE, Barei DP, Mills WJ. Provisional plating of type III open tibia fractures prior to intramedullary nailing [author reply]. J Orthop Trauma. 2006; 20(5): 373. doi:10.1097/00005131-200605000-00017 [CrossRef]
- Macario A. What does one minute of operating room time cost?J Clin Anesth. 2010; 22(4):233–236. doi:10.1016/j.jclinane.2010.02.003 [CrossRef]
- Patzakis MJ, Harvey JP Jr, Ivler D. The role of antibiotics in the management of open fractures. J Bone Joint Surg Am. 1974; 56(3):532–541.
- Li J, Schwartzbach C, Iliev I, Westbrook R, Theiss M. Mini plate can influence the primary healing of a long bone fracture fixed with a compression plate. Poster presented at: Orthopaedic Research Society 2015 Annual Meeting. ; March 28–31, 2015. ; Las Vegas, Nevada. .
Demographic and Clinical Characteristics by Provisional Plate Usea
|Characteristic||Provisional Plate Use||Total||P|
|Age, mean (SD), y||41.5 (±13.0)||34.3 (±12.3)||36.6 (±12.9)||.007b|
|Body mass index, mean (SD), kg/m2||26.2 (±5.3)||27.2 (±5.3)||26.9 (±5.3)||.38|
|Sex, No. (%)|
| Female||8 (23.5)||18 (25.7)||26 (25)||.81|
| Male||26 (76.5)||52 (74.3)||78 (75)|
|Diabetes mellitus, No. (%)|
| Yes||2 (5.9)||2 (2.9)||4 (3.9)||.47|
| No||32 (94.1)||66 (97.1)||98 (96.1)|
|Smoking status, No. (%)|
| Yes||12 (35.3)||20 (29)||32 (31.1)||.52|
| No||22 (64.7)||49 (71)||71 (68.9)|
|Mechanism of injury, No. (%)|
| Motor vehicle collision||7 (20.6)||11 (15.7)||18 (17.3)||.12|
| Motorcycle collision||11 (32.4)||23 (32.9)||34 (32.7)|
| Pedestrian struck||4 (11.8)||14 (20)||18 (17.3)|
| Fall (standing)||1 (2.9)||3 (4.3)||4 (3.8)|
| Fall (height)||3 (8.8)||9 (12.9)||12 (11.5)|
| Crush||7 (20.6)||2 (2.9)||9 (8.7)|
| Other||1 (2.9)||8 (11.3)||9 (8.7)|
|Side of injury, No. (%)|
| Left||18 (51.4)||29 (39.7)||47 (43.5)||.25|
| Right||17 (48.6)||44 (60.3)||61 (56.5)|
|Gustilo-Anderson type, No. (%)|
| I and II||15 (42.9)||45 (61.6)||60 (55.6)||.066|
| III (A and B)||20 (57.1)||28 (38.4)||48 (44.4)|
|AO/OTA classification, No. (%)|
| 42-A||13 (37.1)||49 (67.1)||62 (57.4)||.01b|
| 42-B||12 (34.3)||14 (19.2)||26 (24.1)|
| 42-C||10 (28.6)||10 (13.7)||20 (18.5)|
Logistic Regression Analysis: Risk of Infection and Any Complicationa
|Demographic||Univariate Odds Ratio (95% Confidence Interval)||Multivariate Odds Ratio (95% Confidence Interval)|
| Age||0.98 (0.94–1.02)||-|
| AO/OTA type 42-C||1.10 (0.28–4.39)||-|
| Gustilo type III||1.28 (0.45–3.61)||-|
| Provisional plate used||1.52 (0.52–4.42)||1.64 (0.51–5.32)|
| Age||0.98 (0.94–1.01)||-|
| AO/OTA type 42-C||2.00 (0.72–5.53)||-|
| Gustilo type III||2.40 (1.01–5.68)b||-|
| Provisional plate used||1.59 (0.66–3.84)||1.24 (0.46–3.35)|
Provisional Plate Use and Complicationsa
|Complication||Provisional Plate Use, No.||Adjusted Odds Ratiob (95% Confidence Interval)|
|Infection||7 (20%)||11 (15.1%)||1.64 (0.51–5.32)|
|Anyc||12 (34.3%)||18 (24.7%)||1.24 (0.46–3.35)|
Provisional Plate Retained Versus Removed and Complicationsa
|Complication||Provisional Plate, No.||Adjusted Odds Ratiob (95% Confidence Interval)|
|Infection||5 (25%)||2 (13.3%)||0.43 (0.07–2.69)|
|Anyc||8 (40%)||4 (27%)||0.55 (0.12–2.46)|