1 Introduction

Femoral shaft fractures are amongst the most common diaphyseal fractures of childhood, with an estimated annual incidence of 2–20 fractures per 100,000 children in the USA.^1–8 They convey a significant cause of morbidity, and potential mortality, in children and various methods of fixation have evolved with limited high quality evidence supporting their adoption.^7,9–11

Much debate remains in the existing literature regarding the optimal management of these injuries (12), with options varying according to patient age, size/body mass, and fracture configuration.^4,13,14 The American Academy of Orthopedic Surgeon (AAOS) states there is limited evidence to support a variety of treatment techniques, which includes intrameduallary nailing and submuscular plating, for paediatric femoral shaft fractures.¹⁰ Both techniques have their own benefits and pitfalls. However, operative treatment has been shown to reduce time to weightbearing, to provide a more predictable pattern of healing, and reduce time out of school in the paediatric population.^15,16 The National Institute for Clinical Excellence in UK suggests Titanium Elastic Nailing System (TENS) nailing for those aged 4–12 years of age, with a weight <50 kg, and submuscular plating only in those aged over 11 years and/or >50 kg in weight.¹¹ Both guidelines only have Level III evidence for those aged over 4, and treatment trend choice tends to come in and out of vogue.⁹ Furthermore, there is no literature which favours one method over another, particularly for length unstable fracture configurations. Length unstable fractures are defined as being long spiral fractures (fracture being >2x femoral diameter), or comminuted/presence of butterfly fragment.¹⁷

This paper aims to retrospectively review the patients undergoing operative fixation in our tertiary paediatric trauma centre over eight years. We hypothesize that patient outcomes following femoral fracture fixation with submuscular plating are at least equivalent to those of flexible intramedullary nail fixation. In addition, plate fixation may confer advantages in time to weightbearing and time to union making it a viable alternative for femoral fracture fixation in the school age child.

2 Materials and methods

A retrospective review of all patients who underwent operative fixation for diaphyseal femoral fractures during the period 1^st April 2009 – 1^st April 2017 was performed. Patient details were obtained by a review of a regional Picture Archiving and Communications System (PACS), identifying all patients who had received imaging of either femur in the trauma unit during this period. These were subsequently reviewed referenced to identify patients who had undergone operative fixation. These patients were then cross-referenced with operation logbooks and Theatre Management Systems to ensure data validity, and no loss of patient episode data. Trauma and Orthopaedic Research Group (TORG) approval from the Belfast Health and Social Care Trust was obtained for the current study (TORG/FORD2017/SMCDRJB).

Following patient identification, two senior consultants were asked to review preoperative injury radiographs to assess their suitability for management either with plate or flexible intramedullary nail fixation. Consultants were blinded to the final treatment outcome. If both consultants independently confirmed that the fracture pattern was suitable for either plating or nailing, the patient was deemed to have met the initial study criteria. All surgical procedures were performed by a paediatric trauma fellowship trained consultant.

A clinical note and radiographic review was performed. Pathological fractures, compound fractures, multiply injured patients, and patients who had undergone surgical fixation in other units were all excluded from the study.

The study aimed to determine if there was a difference in primary outcome measures between flexible intramedullary femoral nailing versus femoral plating. The primary outcome measures were postoperative length of stay, analgesia requirements at 24 h and 48 h post-operatively, time to full weightbearing (defined as weightbearing without walking aids) and radiological union at 12 weeks follow up. In this study, radiological union was defined as bridging callus formation on three out of four cortices on orthogonal radiographs.

The secondary outcome included the incidence of complications within 12 weeks of surgery.

Using this information, we were able to apply the criteria described by Flynn et al., to determine the post operative outcomes of both plates and TENS nailing procedures in this specific cohort.¹⁸

Statistical analysis was carried out using SPSS v.22 for Mac (IBM, Armonk NY). Data was first tested for normality using a Shapiro-Wilk test. Parametric data were analysed using a Student's t-test. Non-parametric data was analysed using ANOVA testing, and categorical data was analysed using Fisher's exact test. A p-value of <0.05 was considered statistically significant for all analyses.

3 Results

In total, 45 patients underwent operative fixation for diaphyseal femoral fractures during the study period. Of these 45 patients, 28 patients met the study inclusion criteria – the other 17 were excluded, due to either incomplete data/no note availability (5 patients), injuries as a result of polytrauma (8–4 plates and 4 nails) or because surgery occurred elsewhere and the patient was repatriated for local follow up (4 patients - 2 nails, 2 plates). The patients were evenly split in treatment modality between flexible intramedullary nail fixation and submuscular plate fixation, with 14 patients in each operative group.

Overall, 21/28 (75%) of patients were male. There was no difference in the proportion of males in the two treatment groups. Mean age for the plate fixation group was significantly lower compared to the nailing group (7.7 ± 1.8 Vs 9.7 ± 1.9 years; p = 0.008).

Post operative length of stay was shorter for the plate fixation group by 1.5 days, but this failed to reach statistical significance (6.3 ± 2.1 Vs 7.8 ± 3.0; p = 0.134). Fig. 1.

Close

Fig. 1 Boxplot for the post-operative length of stay.

Export to PPT

Analgesia requirements at 24 and 48 h were higher in the nailing group, but this failed to reach statistical significance (at 24 h 0.61 ± 0.25 Vs 0.50 ± 0.29; p = 0.2 and 48 h 1.12 ± 0.97 Vs 0.82 ± 0.45; p = 0.316). Fig. 2.

Close

Fig. 2 Analgesia requirements at 24 and 48 h.

Export to PPT

Radiographic union, determined by bridging callus on 3 out of 4 cortices on orthogonal radiographs, was higher at 12 week review in the plate group. In the plate group, all patients had radiological union at 12 weeks, compared to 10/14 (71.4%) in the nailing group. This failed to reach significance (p = 0.098).

Reoperation was performed for removal of metal in 11/14 TENS patients, after radiological union. Four underwent unscheduled re-intervention for nail migration,² soft tissue irritation¹ or end cap problems¹ prior to union was established. Three patients had their metalwork left in situ.

Reoperation was performed in 12/14 plate patients to remove the plate after radiological union was achieved. No unscheduled re-operation was performed for the plate group. However, one patient had a residual limb length discrepancy with overgrowth on the affected side requiring contralateral epiphysiodesis 12 months after initial fixation. Table 1 summarises the comparisons between the two treatments.

At a minimum follow-up of 2 years (Mean 5.3 ± 2.5 nails Vs 5.3 ± 2.4 plates) no patient had undergone any further corrective surgery.

Flynn described 4 criteria by which to determine post operative outcomes. These criteria included leg length discrepancy, angular deformity, pain and the occurrence of complications, and were graded as excellent, satisfactory or poor.¹⁸ There was a tendency for plates to have better overall outcomes compared to nails (Excellent 10 Vs 6, satisfactory 3 vs 7, poor 1 Vs 1), but there was no statistically significant difference p = 0.32). Table 2 summarises the data.

4 Discussion

Paediatric femoral fractures have a bimodal incidence, occurring in <8 year olds due to sporting injuries, and >13 year olds due to road traffic collisions.⁴ Incidence rates vary widely, from 2 to 20/100,000,^3–8 with higher rates of 2.5–4.7 times in male patients, and moreso in lower socioeconomic classes.^4–6 In the UK specifically, Talbot and colleagues demonstrated an overall incidence of 5.8/100,000 children aged <16 years, with an overall 2.5x higher incidence in boys.³

Studies reporting treatment modalities and outcomes are widely varied, and heterogeneous due to the different treatment options, childrens age and weight, fracture stability and configuration.^3,4,19,20 Recent evidence suggests that the overall incidence of these injuries is in the decline, therefore making large scale prospective robust studies difficult to establish^3–8 In the UK, no clear guidance exists on the optimum treatment for these injuries, with any recommendations coming level 4 or 5 evidence.¹⁴ In the US however, the AAOS demonstrates that there is only limited evidence for various treatment modalities for those aged >11 years (AAOS). However, rigid intramedullary nails should be avoided in the 5–12 year old age group because of the increased risk of AVN.¹⁸

TENS nailing is a good treatment option with perceived benefits of small wounds and minimal soft tissue injury for their insertion. Bridge plating allows healing with minimal disruption of the soft tissue envelope and can be performed in larger, heavier paediatric patients with or without unstable fracture configurations. Additional benefits include its use for narrow medullary canals in patients not amenable to conservative treatment modalities.

A change in preference towards TENS nailing has been advocated by some authors, owing to favourable biomechanics with lower rates of avascular necrosis versus solid intramedullary nails,¹⁸ whilst promoting faster weight bearing and shorter hospital inpatient stays,^3,6,14,21 This is specifically advantageous in younger (6–12 year olds), lighter (<50 kg) patients, with length stable fractures.^14,17,22,23 A growing body of evidence and understanding of the nuances associated with difference fixation options, their risks and complications, continues to evolve and complication rates of up to 75% are noted to occur mainly as a direct consequence of surgical inexperience with TENS.²⁰ These are commonly due to coronal and sagittal plane angular deformities (up to 39%), and rotational malunion (up to 11%).^24–26 Leg length discrepancy, either lengthening or shortening of the affected limb, can also occur in up to 58% of cases.²⁵ However, the majority of malunions, residual deformities and limb length discrepancies are not clinically significant.^25,26 These deformities are more common in overweight children, and as such TENS nailing has a weight limit “ceiling” of 50 kg,^24,27–31 and use in children heavier than this increases the risk of sagittal and coronal plane deformity.³²

These complications have led to a resurgence in operative fixation using a variety of plating systems.^33,34 Proponents of this report more rapid healing, earlier radiological union rates, with plate failure being a rare, if ever, event.^33–36 Biomechanically, plate fixation has shown favourable characteristics over TENS nailing.³⁷ Additionally, these can be utilised in heavier patients above the 50 kg threshold for TENS, and have a lower reported complication rate of 4–8%,^8,22,24 and whilst some argue more extensive periosteal stripping with plating, present day low contact bridge plating reduces this to a minimum.³⁸

Sutphen reported on 344 paediatric femoral fractures and compared rigid nailing, TENS and plating, with a minimum follow-up of 12 weeks as used in the current study. They reported a statistically significant shorter mean time to union (6 Vs 8 weeks), faster mobilization (7 Vs 12 weeks), lower rates of malunion, reduced hardware irritation and lower complication rates with plate fixation compared to TENS.³⁹

Flynn described a scoring system to determine outcomes based on the fixation technique for diaphyseal femoral fractures in children,¹⁸ which was based on 4 parameters; limb length inequality, malalignment, pain and presence of complication. Li et al. compared statistically similar groups of TENS and plate patients with subtrochanteric fractures. They demonstrated faster times to mobilization, better Flynn outcome scores and equivocal union rates were observed in plates compared to TENS.⁴⁰ The authors also reported complication rates in TENS nailing to be 4x higher than plating (48% vs 14%; p = 0.008), particularly in length unstable fracture configurations.

Both TENS and plating can be used in the treatment of diaphyseal femoral fractures in the 5–12 year old paediatric population.^8,22 The outcomes reported in the literature are favourable with both techniques, but subtle nuances may favour one method, with the potential advantages as outlined above, over the other on an individualised patient basis. Sink reported a change in institutional practice away from TENS toward bridge plating as a result of increased complications with TENS, with reports stating that 75% of complications occur as a result of lack of surgical experience.^20,24 The authors reported a shift towards plating particularly for non-transverse (i.e. length unstable fracture configurations), overweight patients (>49 kg) who were older. This resulted in an overall reduction in all complication rates.¹²

In the current study, there was equivalent length of stay, time to union and time to weight bearing in the plate fixation group, with a tendency toward shorter lengths of stay, reduced analgesic requirements and fewer complications. The plate group tended to be younger, however this is in agreement with a large epidemiological study of UK practice.³ The total Flynn outcome scores indicated a good/excellent post-operative result in >90% of all cases, and whilst plating had a higher proportion of excellent outcomes (10/14; 71.4% Vs 6/14; 48.9%) these failed to reach statistical significance.

The study is limited by the small numbers and its retrospective, non randomised design. Whilst not significantly powered to determine significant differences in this cohort, the findings are still applicable to the wider paediatric population based on the conclusions agreeing with current literature. We feel the results are applicable to the non-paediatric specialist centre that may deal with these injuries, lacking in TENS expertise, and/or not have TENS nailing systems available. Plating is therefore a reasonable option, technically easier procedure and is associated with good outcomes.^{12,13,24,39,41} As yet, neither technique has demonstrated clear superiority, and large scale studies are lacking in data, up to 25% of required information,^3,4 in order to derive this answer. Large multicentre studies, ultilising a standardized treatment algorithm and robust data collection is required, however this will be difficult due to the decreasing incidence of these injuries,⁷ significant bias (patient, demographic, surgical/technical, and geographical) and the type of health care system in place.

5 Conclusions

TENS nailing can be used effectively in 6–12 year old children with length stable diaphyseal femoral fractures. However, some paediatric femoral fractures may benefit from plate fixation. These include overweight patients, with length unstable fractures. Plate fixation can be easily performed, with good outcomes and minimal complications. The lack of a standardised management algorithm and a low incidence of these injuries requires further evaluation with a long term prospective, multicenter randomised control trial to fully determine the best treatment option and the variables that impact on clinical outcomes.

Ethical approval

Trauma and Orthopaedic Research Group (TORG) at Belfast Health and Social Care Trust (Reference TORG/FORD2017/SMCDRJB).