Open versus closed intramedullary nailing of femur shaft fractures in adults: a systematic review and meta-analysis

Purpose
 This systematic review and meta-analysis aimed to compare the outcomes of open- versus closed-reduction and intramedullary nailing (IMN) of adult femur shaft fractures. Methods Four databases were searched from inception until July 2022 for original studies that compared the outcomes of IMN following open-reduction versus closed-reduction technique. The primary outcome was the union rate; the secondary outcomes were time to union, nonunion, malalignment, revision, and infection. This review was conducted in line with PRISMA guidelines. Results A total of 12 studies with 1299 (1346 IMN cases) patients were included, with a mean age of 32.3 ± 3.25. The average follow-up was 2.3 ± 1.45 years. There was a statistically significant difference in union rate (OR, 0.66; 95% CI, 0.45–0.97; p-value, 0.0352), nonunion (OR, 2.06; 95% CI, 1.23–3.44; p-value, 0.0056), and infection rate (OR, 1.94; 95% CI, 1.16–3.25; p-value, 0.0114) between the open-reduction and closed-reduction groups in favour of the latter. However, malalignment was significantly higher in the closed-reduction group (OR, 0.32; 95% CI, 0.16–0.64; p-value, 0.0012), whereas time to union and revision rates were similar (p = NS). Conclusion This study showed that closed-reduction and IMN had more favourable union rate, nonunion, and infection rates than the open-reduction group, yet malalignment was significantly less in the open-reduction group. Moreover, time to union and revision rates were comparable. However, these results must be interpreted in context due to confounding effects and the lack of high-quality studies. Supplementary Information The online version contains supplementary material available at 10.1007/s00264-023-05740-x.


Introduction
Femur shaft fractures are among the most common fractures encountered in orthopaedic trauma practice, with an overall incidence of 10-21/100,00 persons per year [1,2].Of those, 40% occurred due to road traffic accidents and other highenergy trauma mechanisms [1,2].
Due to the principal load-bearing role of the femur, femur shaft fractures are often associated with prolonged morbidity and extensive disability if improperly treated [3].As a result, tremendous advances in treating femoral shaft fractures have been seen, with the gold standard for treatment remaining intramedullary nailing (IMN).While the standard routine insertion of IMN is done following closed fracture reduction, this is not always feasible due to fracture complexity, equipment's availability, surgeon's experience, and patient-related factors (obesity and polytrauma); thus, an open-reduction technique might be needed to achieve proper reduction alignment in some challenging cases [4].
In addition, several studies have shown that the openreduction technique is associated with a higher risk of infection and lower union rates [5][6][7].However, some of the setbacks were low power with small sample sizes and short-term follow-up.Therefore, high-quality evidence is needed to compare outcomes properly across both groups.
The purpose of this study was to compare the clinical and radiological outcomes of open-versus closedreduction and IMN of such fractures.We hypothesize that there is no significant difference in outcomes and complication rates between patients treated with openreduction and IMN versus those treated with closedreduction and IMN.

Materials and methods
This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [8].A protocol registration was sought in advance on the International Prospective Register of Systematic Reviews (PROSPERO) with the registration number: CRD42023327089.

Search strategy: outcomes of interest
PubMed/Medline, Web of Science, Google Scholar, and Cochrane library databases were searched from inception until July 2022 with the following keywords and their derivatives: Open AND closed AND reduction AND intramedullary nail AND femur shaft AND fractures.Two authors independently screened the search results based on the title and/or abstract.If any discrepancy arose, it was resolved by a discussion with a third senior author.
The union rate was the primary outcome and was defined as the radiological bridging callus formation across three out of four cortices with a painless fracture site.Time to union, non-union, malalignment, infection, and revision rates were all used as secondary outcomes of interest.

Data extraction and items
Two independent authors used a pre-designed data collection sheet to extract data.The extracted demographic data included the first authors' surnames, study year, design and country, the mean age of patients, number of participants, number of IMN cases (open vs closed), mean follow-up period, union rate, non-union, malalignment, time to union, operative time, infection and revision rate, statistical tests, and conclusion.

Qualitative assessment (risk of bias)
Two authors assessed the methodological quality of the included studies using the Newcastle-Ottawa tool, which comprises three main elements: patient selection, comparability, and outcomes [9,10].A higher overall score indicates a lower risk of bias; a score of 5 or less (out of 9) corresponds to a high risk of bias.Rob-2 [11] tool was used to evaluate the included RCT.

Quantitative analysis (meta-analysis)
A meta-analysis of eligible studies using R (version 4.0.2,R Core Team, Vienna, Austria, 2020) using the meta package (i.e., forest_meta, metacont, metabin, and metabias functions) was performed.Odds ratios (OR) and their associated 95% confidence intervals were expressed for dichotomous variables (e.g., rate of union).For continuous variables, standardized mean differences (SMD) and their associated standardized errors and deviation values were calculated for all eligible studies.In studies that have only provided median values (± range) or isolated mean values, their standard deviation was imputed per the guidelines of Cochrane (refer to Chapter 7.7.2014).Heterogeneity among effect sizes was evaluated using the I-squared statistic.Definitions for heterogeneity were adapted from the Cochrane handbook (> 25% mild, 25-50% moderate, > 50% severe).Both a funnel plot and Egger's test of asymmetry were utilized to assess publication bias.Due to the low heterogeneity for the dichotomous variables, a common-effects model was utilized for the included studies; otherwise, a random-effects model was deployed.

Search results
Rayyan AI website was used to manage the literature search results [12].Searching the databases yielded 197 articles, and after removing 23 duplicates, 174 records were screened by title and abstracts, of which 151 were excluded.A total of 23 papers were eligible for a full-text review.As a result, 12 studies met the eligibility criteria and were included in the qualitative and quantitative syntheses.The PRISMA flowchart is displayed in Fig. 1.

Studies characteristics
Twelve studies (1331 femur shaft fractures) were included in this meta-analysis, with a mean patient age of 32.3 ± 3.25.The mean follow-up was 2.3 ± 1.45 years.One randomized clinical trial was included, six studies were prospective cohorts, and five were retrospective.Ten studies were used to analyse union rates, eight studies for time to union, malalignment, and revision, and nine studies reported on non-union and infection rates.The characteristics of the included studies are summarized in Table 1.

Quality assessment (level of evidence (LoE) and risk of bias)
Based on the OCEBM criteria [24], one study was level 2a, six studies were level 2b, and five were level 3a (Table 1), with  an overall grade B of recommendation assigned to the review [25].The Newcastle-Ottawa scores of all 11 observational studies ranged from 4 to 7, with an average of 6 ± 1, indicating an acceptable overall risk of bias (Table 2).However, using Rob-2 assessment tool, the RCT by Chaudry et al. had a high overall risk (Table 3).While five (41%) of the included studies reported high bias risk, the remaining nine (59%) were of fair quality.A summary of the qualitative assessment, according to the Newcastle-Ottawa scale, is shown in Table 2.

Union rate
A total of ten articles investigated rate of union among patients undergoing open-and closed-type reductions.Our analysis demonstrated that patients undergoing openreduction surgeries are associated with a lower probability of union than their closed counterparts (OR, 0.66; 95% CI, 0.45-0.97;p-value, 0.0352) (refer to Fig. 2).

Time to union
Eight articles provided relevant data on the time taken for completion of union.Heterogeneity among articles was high (I 2 , 89.7%; p-value, < 0.001); thus, a randomeffects model was utilized.Differences in time to union were insignificant among participants undergoing openand closed-reductions (SMD, 0.41; 95% CI, − 0.11 to 1.21; p-value, 0.0906) (refer to Fig. 3).

Non-union rate
In terms of direct non-union rates, a total of nine studies reporting on incidence of non-union across open-and closed-reduction surgeries.Participants undergoing openreduction were associated with less favourable outcomes compared to closed-reduction in terms of non-union rates (OR, 2.06; 95% CI, 1.23-3.44;p-value, 0.0056) (refer to Fig. 4).

Infection rate
Probability of post-operative infections among patients undergoing open reduction was significantly higher than that of their closed-reduction counterparts (OR, 1.94; 95% CI, 1.16-3.25;p-value, 0.0114) (refer to Fig. 6).

Publication bias
Egger's test of bias demonstrated insignificant publication bias for dichotomous variables (p-value, 0.6913) (refer to Supplementary Funnel figures 1 through 5).

Discussion
The main findings of this systematic review and metaanalysis were that union rate, nonunion rate, and infection rate were more favourable in the closed-reduction group, whereas the malalignment was more favourable in the open-reduction group.However, time to union and revision rates were comparable.A meta-analysis summary of the 6 main variable outcomes comparing open-versus closed-reduction methods is shown in Contrary to several studies which showed no significant difference in union rate [16,17,19,20], our analysis demonstrated that patients with femur shaft fractures treated with open reductions IMN had a significantly lower union rate when compared to those treated with closed reduction (OR, 0.66; 95% CI, 0.45-0.97).The violation and disruption of the fracture haematoma and its positive role in bone healing might explain this significant finding [26].Similarly, a significantly higher non-union rate was observed in the open-reduction group (OR, 2.06; 95% CI, 1.23-3.44).Non-union was consistently defined across the included studies as the radiographic persistence of a radiolucent line without progression of callus formation, along with pain at the fracture site at a minimum of six months after the surgery.This universal agreement on defining nonunion enabled us to pool these studies and compare this outcome with very low heterogeneity (I 2 = 0%) and statistical significance.Telgheder et al. [21] retrospectively studied 107 patients with traumatic femur shaft fracture who underwent intramedullary nail preceded by either closed or open reduction; they reported a comparable mean time to union in closed reduction and open reduction and combined groups of 5.4 months, 6.2 months, and 5.6 months, respectively (p = NS).This present study supports this finding as there was no difference in time to union among both groups (SMD, 0.41; 95% CI, − 0.11 to 1.21, p = NS).
Karaman et al. [27] reported a 41.7% incidence of rotational malalignment of more than 10° following closed reduction and intramedullary nailing of femur shaft fracture on CT; these patients were symptomatic and had significantly lower functional scores compared to those without malalignment.Open-reduction techniques can aid in the precise and anatomical restoration of fracture fragments.Similarly, our meta-analysis showed that open reduction reduces the risk of malalignment to less than one-third compared to the closed-reduction group (OR, 0.32; 95% CI, 0.16-0.64,p < 0.05).

Common effect model Random effects model
Heterogeneity: I 2 t 2 = 0%, = 0, p = 0.45  Infection rates following intramedullary nails of femoral shaft fractures are generally low, ranging from 1 to 3.8% [28].Out of 118 enrolled patients, Nandhimandalam et al. [23] reported 4 cases of superficial infection, 2 in each arm and only one with deep infection in the open group, results which were statistically insignificant.These findings were also replicated by subsequent studies [20,21].Haq et al. [22] compared the infection rate in closed and open groups and reported a significantly higher risk with open reduction (6.4% vs 24%).However, pooling various studies demonstrated a twice higher risk of infection in patients with open reduction than those who underwent closed reduction (OR, 1.94; 95% CI, 1.16-3.25,p < 0.05).The increased surgical time, peri-operative antibiotics use, and direct manipulation of soft tissue through open reduction can contribute to this increased infection risk.However, these confounding factors were not clearly mentioned in the included studies.Thus, a future prospective study matching these confounders is warranted.
Although the revision rate difference was insignificant between the two groups, with a higher probability of revision in the closed-reduction group (OR, 0.85; 95% CI, 0.53-1.35), it is worth mentioning that the revision is done for all causes necessitating surgery, including but not limited to deep infection, nonunion, and malalignment [13,21,29].
To the best of our knowledge, this systematic review was the first to analyze and compare open-versus closed-reduction techniques in IMN treatment of femur shaft fractures.However, several limitations must be acknowledged.First, some of the included studies were of low quality, reducing the confidence in the results, including an RCT with a questionable randomization method and an apparent selection bias.Regardless, these studies remain the highest level of available evidence in the literature.Another weakness was the inadequate reporting of some confounding factors, such as the type of IMN used and baseline comorbidities which were not clearly stated in all studies.Also, several comparison outcomes, like operative time and radiation exposure, were either missed or poorly reported.Thus, future prospective studies are needed to adjust for these confounders.

Conclusion
Closed-reduction IMN was associated with more favourable union rate, nonunion rate, and infection rate compared to the open-reduction group, whereas malalignment was significantly less in the open-reduction group.However, time to union and revision rates were comparable.This finding should be applied in context due to potential confounding factors and the lack of high-quality studies.

Fig. 1
Fig. 1 PRISMA flow diagram of record identification, screening, and selection in metaanalysis

Table 3
Rob-2 risk of bias assessment tool used to evaluate included RCTs H, high bias; S, some bias

Table 4 .
The intramedullary nail is considered the treatment of choice for nearly all femur shaft fractures.Ideally, fracture reduction is achieved by closed-reduction means; however, open reduction might be inevitable in certain situations where we have soft tissue interposition, severely Forest plot comparison of the union rate between open and closed groups.OR, odds ratio; CI, confidence interval; Event, number of united fractures; total, all treated fractures Forest plot comparison of non-union between open and closed groups.OR, odds ratio; CI, confidence interval; Event, number of nonunions; total, all treated fractures Forest plot comparison of infection rate between open and closed groups.OR, odds ratio; CI, confidence interval; Event, number of infections; total, all treated fractures comminuted fracture, and associated injuries, and in obese or muscular patients.Many studies have compared the outcome of closed versus open reduction in femur shaft fracture, with some authors claiming that the outcome is comparable between each group.
3eightFig.3Forestplot comparison of the time to union between open and closed groups.CI, confidence interval; SD, standard deviation; CI, confidence interval; SMD, standardised mean difference; total, all treated fractures Forest plot comparison of revision rate between open and closed groups.OR, odds ratio; CI, confidence interval; Event, number of revisions; total, all treated fractures

Table 4
Meta-analysis summary of main outcomes comparing open-versus closed-reduction methods Bold emphasis statistically significant values (p-value < 0.05) OR, odds ratio; MD, mean difference; CI, confidence interval; Effect, common-effect model