Comparing Intramedullary Nailing, External Fixation and External Fixation followed by Intramedullary Nailing as management for open fractures of the tibial shaft: a Systematic Review revIew

Methods and Findings: Systematic review of the literature was conducted on the following databases: PubMed and VHL from 2000 to 2013 aiming to compare Intramedullary Nailing (IM Nailing), External Fixation (EF), and EF followed by IM Nailing in the treatment of open fractures of the tibial shaft. After analysis by inclusion criteria 24 articles met the eligibility criteria. The healing time was under 32 weeks, and the nonunion and defective healing rates were under 13.35% and 16.6%, respectively. The infection rate ranged from 3% to 53% for patients treated with EF as definitive management; from 0% to 22% for those treated with IM Nailing as definitive management; and from 0% to 16.7% for those who underwent EF followed by IM Nailing.


Introduction
The open fractures of the tibial shaft usually result from high-energy trauma and have extensive soft tissue damage associated [1]. The National Center for Health Statistics reports an annual incidence of 492,000 fractures of the tibia and fibula per year in the United States [2].
The treatment of open tibial fractures is often a dilemma because it requires particular caution and individual assistance for each case. Although the fixation method for open tibial fractures is controversial, external fixation (EF) is considered the primary emergency procedure. The main advantages of EF are: quick installation; little implant in the injury; less surgical trauma of the damaged soft tissue; easy inspection of the injury; management with little pain; and good stability of the fracture [1].
However, Intramedullary Nailing (IM Nailing) has been increasingly adopted and has showed to be an effective primary definitive fixation method of the open tibial shaft fractures up to the grade IIIA, Gustilo et al. [2,3]. Besides, some authors present another strategy which is based upon changing the external fixation of open tibial fractures to IM Nailing up to the second week after the initial trauma with lower infection risk [4] and high healing rates [5].
All treatments have their particular disadvantage, immediate unreamed interlocking nailing (immediate nailing) for Gustilo type IIIB open tibial fractures has the risk of deep infection. External fixation is associated with delayed union, nonunion, malunion, and ankle joint stiffness [6]; External fixation followed by delayed interlocking nailing (delayed nailing) is associated with intramedullary and pinsite infection [7].
The treatment of open fractures presented in the literature lies on 3 main management options: EF as definitive treatment, IM Nailing as definitive treatment, and EF followed by IM Nailing. Due to the importance of the open tibial shaft fracture and its correct handling, the present study was ba-sed on the following guiding question: Is there something new we can learn analyzing studies about the different treatments (Intramedullary Nailing, External Fixation, and External Fixation followed by Intramedullary Nailing) of open fractures of the tibial shaft in the period 2000-2013? This review highlights the importance of the open tibial shaft fracture and its correct handling as a fundamental dimension to be considered in patients suffered serious accidents and are received in emergencies. Thus, our objective was to evaluate the current evidence concerning to different aspects (hospital stay, healing time, nonunion, defective healing and infection rates) of three treatments -IM Nailing, EF and EF followed by IM Nailing -and formulating a systematic review with emphasis on the primary fixation methods for this comorbidity. Our hypothesis is that, despite the growing interest toward the theme, EF provides better outcomes than clinical indicators than the other two treatments. Thereby meriting greater theoretical contributions subsidized by clinical multicenter studies as well as research of recognized statistical support as metaanalysis.

Methods
This systematic qualitative review of the literature was made by collecting manuscripts from the following database: Biblioteca Virtual de Saúde (BVS) and PubMed. The qualitative approach was chosen because, as for the use of quantitative methods such as meta-analysis, the information needed for evenly comparing the sample is not available in all manuscripts -such a fact would limit the analysis to a small number of studies. The period of the literature studied goes from January 1 st , 2000 to December 1 st , 2013. The reason to limit the search to the 2000-2013 period is to analyze the main open tibial shaft fixation methods in the last 13 years. The search was conducted using the following descriptors: # 1 "Tibial Fracture" (MeSH); # 2 "Open Fracture" (MeSH); # 3 "External Fixators" (MeSH); # 4 "Intramedullary Nailing" (MeSH); And their Portuguese correlatives: # 5 "Fraturas Expostas" (keywords); # 6 "Haste Intramedular" (keywords); # 7 "Fixador Externo" (keywords); # 8 "Fraturas de tíbia" (keywords); The data compilation took place during February 2014. The selection of the manuscripts occurred primarily by the analysis of the titles of the abstracts. Then, the articles identified by the search strategy were assessed independently by the authors, according to the following inclusion criteria: (1) original unabridged articles from the CAPES (Coordenação (1) unoriginal studies such as letters to the editor, reviews, systematic reviews, and editorials; (2) studies whose samples were animals; (3) articles whose case analysis involved pediatric patients. The manuscripts repeated in more than one of the databases were accounted only once.
Subsequently, each selected article was thoroughly read and the data important to this review were collected and organized in a spreadsheet containing: Authors, Year, Type of Treatment, Sample Number (open fractures), Hospital Stay, Healing Time, Nonunion, Defective Healing, and Infection Rates (PICOS) ( Table 1). The data were extracted independently by two researchers and the differences analyzed by a senior researcher in the area. Some studies reported tibia shaft fractures in children or animals. These were not compiled/ tabulated because they did not agree with the selected main theme in this research.
The search in the BVS database took 3 steps: a) descriptors #5 AND #6 were used, resulting in 17 articles; b) descriptors #6 AND #7 were cross-checked, resulting in 8 articles; c) descriptors #5 AND #8 were used and 280 manuscripts were found.
This was a literature review therefore no patient recruitment was involved. In this sense, ethical approval was not necessary.
According to the strategy adopted, 489 articles were initially found (Figure 1). After a thorough analysis by the inclusion criteria, 24 articles met the eligibility criteria. Table 1 presents the results obtained in this systematic review from the manuscripts constituting the final sample. The articles were organized according to the fixation strategy. The papers containing more than one management strategy may be analyzed more than once.

Studies included in the review (n=24)
The healing time for EF followed by IM Nailing was mentioned in 3 out of the 5 selected studies: 19

Discussion
The lack of studies on three types of management for open tibial shaft fractures is noticed, mainly as for external fixation followed by IM Nailing. The articles occasionally found do not assess the same variables in the results, making the comparison between the techniques difficult. In some cases [5,6,12,19], the sample was really small, limiting the analysis of the results.

Hospital Stay
Five studies [4,8,11,18,20] reported on hospital stay, which ranged from 1 to 3 weeks, considering all types of management for open tibial shaft fracture. The shortest hospital stay was 8.6 days [8] using EF as isolated management; and the longest one was 22 days [4] using EF followed by IM Nailing. It must be emphasized that the minority of the studies reported on hospital stay. Kaftandziev et al. [11], for instance, compares the use of EF and unreamed IM nailing as definitive management in the treatment of open tibial shaft fractures grades IIIA and IIIB, according to Gustilo's classification. The group treated with IM Nailing as definitive management showed shorter average hospital stay when compared to the group treated with EF as definitive management (17.6 versus 21.6 days); however, the authors do not report whether this difference is statistically significant. Moreover, in this study, the authors do not make it clear what criterion was used for choosing either IM Nailing or EF. It is known that, in more severe cases, EF is usually the option. This might explain the longer hospital stay for the group treated with EF.

Conversion Time
In the studies on EF followed by IM Nailing the conversion time for IM Nailing varied considerably. When choosing the treatment for a tibial shaft fracture, several factors must be taken into account, including hospital stay, treatment costs, and the infection risk. With EF followed by IM Nailing, there is an increase in the hospital stay, number of surgeries, use of implants, and other medical equipment and products, such as medicines and dressings. An increase of the infection risk by the IM Nailing is possible when the conversion is delayed, because, when external fixation is kept for a long time, there may be contamination and pin-tract infection and, consequently, medullary canal infection and the IM Nailing tract. The use of IM Nailing followed by EF still needs more studies for safe, efficient procedures to be determined, besides the assessment of its actual cost-benefit relationship.

Healing Time
In the studies on EF followed by IM Nailing, Ueno et al. [6] did not show any statistically significant difference in the healing time between the group treated by IM Nailing as definitive management and the one treated by EF followed by IM Nailing.
The general observation of the aforementioned data suggests that the groups treated by EF as definitive management show shorter healing time. However, the heterogeneity between the studies, particularly regarding the characteristics of the patient and more specifically of the fracture to be managed, as well as the applied techniques, do not allow a deeper analysis of the issue.

Nonunion
One paper [20] on IM Nailing as definitive management mentioned that there were no nonunions. In this work, unreamed nails were applied to 19 tibial shaft fractures and all of them healed. However, 6 fractures took over 8 months to heal, what is considered nonunion in many other studies. Besides, 10% of the fractures had delayed healing (over 6 months, according the paper), and they had to be dynamized and/or subjected to other procedures (reaming of the shaft, bone graft). In the other studies on IM Nailing as definitive management [2,6,10,11,16,17,19,[20][21][22][23][24] the nonunion rate was between 0% [6] and 11.1% [11].
In the works on EF as definitive management, two studies [8, 10] affirm there were no nonunions. These papers do not mention the time determined as diagnostic criterion for nonunion. Inan et al. [8] reported 12.5% of delayed healings, but no fracture took more than 32 weeks to heal, whereas Wani et al.
[10] affirm that the healing time ranged between 12 and 33 weeks. Three studies on external fixation followed by IM Nailing mentioning nonunion rate were found: 0% [1], 25% [7] and 36% [6]. It is noteworthy that there were differences, even though small ones, as for the criterion used by some authors to diagnose nonunion.

Defective Healing
Park et al. [6] reported that there was not defective healing in their intramedullary nailing sample, but they do not mention the criteria adopted. The defective healing rate ranged from 6.5% [24] to 15.9% [19] in the other studies on IM Nailing, except for Tielinen et al. [20], in which it was 36.8%. In this paper, 18 Gustilo IIIB and 1 Gustilo IIIC fractures were managed by unreamed IM Nailing and developed complications. Nine of those patients progressed with delayed healing (>24 weeks), needing change of the nail, bone graft, or dynamization of the shaft. Surprisingly, no fracture ended up misaligned. However, 7 fractures shortened 1-2cm, and 2 of them showed external rotation over 10º. Six studies on EF as definitive manage-

Conclusion
There are few cases in the literature about open tibial shaft fracture managed by EF followed by IM Nailing. However, a significant sample of patients and articles on IM Nailing and EF as definitive management was found with better results for these groups. Besides, there is no standard procedure for carrying out research into all 3 types of management (EF as definitive management, IM Nailing as definitive management, or EF followed by IM Nailing), which makes the comparison difficult.
In spite of the methodological limitations previously discussed, which make more adequate comparison impossible, it is noticed that the cases management by EF as definitive management have presented shorter healing time for the analyzed samples. The nonunion cases that showed greater percentages were those of EF followed by IM Nailing. The studies on IM Nailing as definitive management found a greater number of defective healings. The shortest hospital stay, although not described by many studies, was indicated by the sample managed with EF as definitive management. There is the need to encourage research comparing the three therapeutic proposals aiming better knowledge of their advantages and disadvantages for optimization of the management of open tibial shaft fracture.