Comparing the Intramedullary Nail and Extramedullary Fixation in Treatment of Unstable Intertrochanteric Fractures

Treatment options for unstable intertrochanteric fractures include intramedullary nail and extramedullary fixation, although evidence regarding the most appropriate treatment for such fractures remains controversial. Our hypothesis was that there would be no obvious differences in mortality rates, functional outcomes and complications between the two groups. We therefore conducted a meta-analysis to compare the relative advantages of intramedullary nail and extramedullary fixation. A total of 10 randomized controlled trials including only patients with unstable intertrochanteric fractures were included in the final analysis. We found that no statistically significant difference in one-year mortality was observed between the two groups (RR: 0.78, 95% CI: 0.55–1.10, p = 0.160). Analysis of exact p values from five included studies indicated that functional outcomes were markedly better for patients of the intramedullary nail group when compared with those of the extramedullary fixation group (p = 0.0028), although evidence remains controversial. Higher incidences of all complications were noted for extramedullary fixation (RR:1.48, 95% CI: 1.12–1.96, p = 0.006). However, no significant differences in implant-related complications were observed between the two groups (RR:1.20, 95% CI: 0.73–1.97, p = 0.475). Therefore, comparing with extramedullary fixation, the intramedullary nail method would be more reliable and should be encouraging for unstable intertrochanteric fractures.

Mortality. The incidence of mortality was reported in all 10 studies. Five studies 17,[21][22][23]25 reported higher mortality for the extramedullary fixation group than the intramedullary nails group, although the opposite result was noted in three other studies 16,18,20 . No deaths were reported in the remaining two studies, which were excluded from this portion of analysis 19,24 . Chi-square, I-square, and L' Abbé analyses indicated no statistical heterogeneity (χ 2 = 4.01,P = 0.778, I 2 = 0.00%) (Fig. 3A), and data pooled using a fixed-effects model indicated no statistically significant difference between the two groups (RR:0.78, 95% CI: 0.55-1.10, p = 0.160,) (Fig. 3B).  , and timed two-minute walk test scores were also evaluated in some studies 26,27 . Exact p values were reported in five of the included studies 16,18,20,21,25 , while the remaining studies only specified whether results were statistically significant. Functional outcomes were markedly better in the extramedullary fixation group than in the intramedullary nail group in three studies 16,18,22 . A P-value <0.05 was considered statistically significant in seven studies 17,[19][20][21][23][24][25] . Given that there is no universal functional scoring system for measuring postoperative function and the limited number of exact p values, only five studies were included in this portion of the meta-analysis. Significant differences were noted between the intramedullary nail and extramedullary treatment groups of these studies (p = 0.0028), although these results remain questionable, as four of the five remaining studies reported no significant difference between the two groups. Therefore, it is necessary to establish a universal system for the assessment of postoperative function in patients with unstable intertrochanteric fractures.
Complications. All  Publication bias. We assessed publication bias using Begg's test and Egger's test. The funnel plot for the meta-analysis of mortality for intramedullary nail versus extramedullary fixation was largely symmetric (P Begg = 0.902, P Egger = 0.567, Fig. 5A,B). Similar results were observed for complication rates (P Begg = 0.210, P Egger = 0.137, Fig. 5C,D). In addition, we performed the Duval and Tweedie nonparametric "trim and fill" method of accounting for publication bias in meta-analysis, which indicated no publication bias for implant-related complications among these studies (Fig. 5E). The pooled estimate of the fixed-effects model was −0.326 (95% CI: −0.857-0.204), while the pooled estimate of the random-effects model was −0.283 (95% CI: −1.061-0.496).

Discussion
Unstable intertrochanteric fractures are difficult to manage 2 . Several fixation devices have been developed to overcome the difficulties encountered in the treatment of such fractures, including extramedullary (DHS, CHS, PFLCP, AMBI, SHS) and intramedullary (ITN, PFNA, GN) devices 28 . However, researchers have reported conflicting results regarding differences in mortality, functional outcomes, and complications between intramedullary nail and extramedullary fixation 11,14,29 . We therefore conducted a meta-analysis to determine whether evidence-based support for an optimal fixation device exists for unstable intertrochanteric fractures. In order to provide the best current evidence on the treatment of unstable intertrochanteric fractures, we included only studies that compared intramedullary nail and extramedullary devices. We identified 10 relevant randomized controlled trials and subsequently compared mortality rates, functional outcomes, and complications between patients treated with intramedullary nail and extramedullary fixation devices.
Johnell O concluded that hip fracture is a significant cause of morbidity and mortality worldwide 30 . Our study demonstrated no statistically significant difference in one-year mortality between intramedullary nail and extramedullary fixations and other research has shown that the advanced age and medical comorbidities led to a mortality rate of almost 10% within the first year after the fracture occurred 31 . Surgical management of unstable intertrochanteric fractures has evolved over the past few decades in a quest to improve functional outcomes in this patient population. when considering patients' postoperative function of the two kinds of fixations, a literature review by I.B. Schipper suggested both intramedullary and extramedullary fixation offer clinical advantages for unstable trochanteric femoral fractures 4 . Another meta-analysis recommend the intramedullary nail technique for the treatment of unstable femoral intertrochanteric fractures due to better functional outcomes and reduced blood loss 28 . Our study pointed to exact p values from five studies indicated that functional outcomes were markedly better for intramedullary nail than extramedullary fixation (p = 0.0028). However, the lack of universally accepted functional scoring systems and low number of studies included in this analysis indicate that further investigation is required.
There are multiple factors that influences these two treatment options such as type of implant, type of fracture, patients age, co-mobidities, bone quality, time of treatment. Suitable implant selection according the type of fracture is an important factor to reduce the incidence of complication for unstable fracture in aged patients 32 . For the treatment of intertrochanteric fractures, intramedullary nail and extramedullary fixation, but each has advantages and disadvantages. On the one hand, Intramedullary fixation has multiple theoretical advantages for the treatment of unstable fracture patterns, such as biomechanical advantages, simple manipulation, small exposure, less complications, earlier mobilization. But they would be subject to the diameter of the medullary cavity and were inferior to extramedullary fixation in controlling rotational stability. Worse still, the endosteum blood supply was destroyed in the process of reaming. On the other hand, extramedullary fixation like the dynamic hip screw, whose advantage was interfragmental compression effect with a high union rate, and a minimally invasive technique was used to reduce soft tissue stripping and blood loss. But the extramedullary fixation had a higher incidence of varus collapse, medialization of the distal fragment and cut out of the femoral head screw in the treatment of unstable intertrochanteric fractures, which combined with lateral wall or posteromedial comminution, fractures with reverse obliquity patterns 33 . Honestly speaking, there was still no significant difference in the strength of fixation of stable and unstable infertrochanteric fractures between intramedullary nail and extramedullary fixation, although the intramedullary nail more rigid fixation 34 . Intertrochanteric fractures classified as unstable (AO/ASIF classification: 31-A2/31-A3), however, have a higher risk of complications and mechanical failure in comparison with stable fractures. Recent data have suggested that some unstable fracture patterns, such as reverse obliquity, and highly comminuted, could benefit from intramedullary nailing 35 . Comorbidities like osteoporosis may lead to technical problems during the procedure and complications sustained screw cut-out,  loss of reduction, delayed union, malunion or nonunion, and various deformities of the femur, Because of the biome-chanical advantage and unique design (ITN and PFNA) for osteoporosis, the intramedullary nail appears to be a reliable implant in the management of intertrochanteric fracture in elderly patients with primary osteoporosis 36 . Higher operative time could result in more blood loss and higher infection rate, therefore, in order to reduce operative time, we had to choose to stick with their most-familiar implant system 37 .
More recent studies have reported little difference in complication rates and ambiguous clinical outcomes between intramedullary nail and open reduction/internal fixation (ORIF) surgical methods, making it difficult to determine the ideal implant due to risks and benefits associated with each device 8,38,39 . We observed a higher incidence of complications for extramedullary fixation than for intramedullary nail, which may be due to the biomechanical advantages of intramedullary fixation 34 . We also carried out a subgroup analysis based on implant-related complications, such as Femoral shaft fracture, Non-union, Cut-out, Migration of screw, Breakage of Implant and so on. Nevertheless, no significant differences in implant-related complications were observed between the groups. The main reasons responsible for the implant-related complications are such iatrogenic factors as biomechanically unsuitable position, unskillful surgical technique and improper post-operative instruction for functional exercise 40,41 .
The present meta-analysis, however, is limited in that few large-scale, multi-center RCTs specified for unstable femoral intertrochanteric fractures were included. Many trials included both stable and unstable fractures were not taken into account, only 10 published studies could be used for specific analysis of results in unstable fractures. Moreover, In our research, only five studies were included in this portion of the meta-analysis to evaluate the function. This fact limits the validity considerably. A significant difference considering all included studies was not possible. Besides, many trials failed to analyse results according to fracture type, patients age, co-mobidities, bone quality and time of treatment. As a result, Future large-scale studies should therefore aim to establish a universal standard for evaluating the efficacy of both treatments in this patient population. Similarly, evidence suggesting that patients treated with intramedullary nail experience better functional outcomes remains questionable, further supporting the need for a universal tool for the assessment of postoperative function. However, more conclusive evidence suggests that intramedullary nail is associated with fewer complications than extramedullary fixation, Therefore, the intramedullary nail method would be more reliable and should be encouraging for unstable intertrochanteric fractures.

Materials and Methods
Search strategy. We searched PubMed, Embase, Web of Science, and CBM databases using combinations of the following keywords: "unstable intertrochanteric fractures", "intramedullary nail and extramedullary fixation", "PFNA", "PFN", "Gamma nail", "InterTan", and "DHS", "CHS", "PFLCP", randomized controlled trials" (last update on December 31, 2016). Reference lists for identified reports were also retrieved and reviewed for other potentially relevant studies. All studies were carefully evaluated for repeated data. Criteria used to define duplicate data included study period, hospital, treatment information, and any additional inclusion criteria.
Data extraction and quality assessment. Two authors independently extracted the data from all eligible articles, and any disagreements were resolved by discussion and consensus among the authors. Information retrieved for each study included author names, year of publication, original country, methods, number of patients, functional outcomes (clinical assessment scores) and associated p values, mortality, complications, implant-related complications. We also evaluated the potential for bias in all included studies. Evaluation criteria and methods followed the Cochrane Collaboration's proposal. Statistical software Stata 12.0 (StataCorp LP, College Station, TX, USA) was used to assess the risk of bias.
Statistical analysis. We evaluated differences in outcomes between intramedullary nail and extramedullary fixation by calculating the pooled relative risk (RR) and corresponding 95% confidence intervals (CI). Heterogeneity was assessed using chi-square and I-square tests. A fixed-effects model was used when there was no significant heterogeneity among the included studies (I 2 ≤ 50%, P > 0.10). A random effects model was used when an obvious heterogeneity was observed among the included studies (I 2 > 50%, P < 0.10). L' Abbé plots also demonstrated that there was no significant heterogeneity. Begg's funnel plots and Egger's test were used to assess the possibility of publication bias. Sensitivity analyses were also performed to assess the stability of the pooled effects. We performed statistical analysis with Stata version 12. A two-tailed P value less than 0.05 was considered statistically significant [42][43][44] .