ORIGINAL ARTICLE

Complications of Pediatric Femur Fractures

Treated With Titanium Elastic Nails
A Comparison of Fracture Types
Ernest L. Sink, MD,* Jane Gralla, PhD,† and Michael Repine, MD*

and complications. External fixation is a simple way to achieve

Abstract: The purpose of this study was to analyze complications seen anatomic alignment for a femur fracture, but there are well-
in children with femur fractures stabilized with titanium elastic nails, reported complications, including refracture, increased time to
comparing their use in stable and unstable fracture patterns. A retro- union, and pin tract infections.5–8 Intramedullary fixation, even
spective review was performed on 39 consecutive children with femur with a trochanteric insertion, is avoided by surgeons for fear of
fractures treated with titanium elastic nails. Patients with comminuted or femoral head avascular necrosis.9 Until the recent advent of
long oblique fractures were classified as having ‘‘length-unstable’’ frac- percutaneous techniques, plating has historically required a
tures. Patients were analyzed qualitatively for any predictive factors large exposure.
or treatment variables that increased the risk of complications. There At our institution the use of TENs has expanded beyond
were 24 patients with complications (62%). Eight patients (21%) under- the treatment of simple transverse or short oblique fractures to
went unplanned surgery prior to complete fracture union. Six of the include more unstable fracture patterns. The majority of mid-
eight requiring unplanned surgery were treated for ‘‘length-unstable’’ shaft femur fractures over the study period were treated with
fractures. The complications that required unplanned surgery for either TENs regardless of the stability of the fracture pattern. The
prominent nails or loss of reduction occurred more commonly in un- purpose of this study is to review complications seen with TENs
stable non-transverse fracture patterns. The authors conclude that in and to recognize factors that increase the risk of complications
patients with ‘‘length-unstable’’ femur fractures, consideration should as it relates to the nature of the fracture. Specifically, is the
be given to methods of treatment other than titanium flexible intra- complication rate higher when TENs are used to stabilize un-
medullary nails. stable fractures?
Key Words: femur fracture, pediatric, flexible intramedullary
fixation, complications
METHODS
(J Pediatr Orthop 2005;25:577–580) Consecutive patients with femur fractures treated at our
trauma center over a 2-year period (January 2001 to January
2003) were retrospectively reviewed. The criteria for inclusion
in the study were femur fractures that underwent titanium
T he efficacy and benefit of titanium elastic nail (TEN)
stabilization of pediatric femur fractures have been well
elastic nailing and received follow-up care at our institution

documented, with minimal reported complications.1–4 For the until fracture union. Patients with pathologic fractures were

surgeon, flexible nailing is a relatively simple method to excluded.

stabilize a femur fracture. The procedure is technically easy to Fractures were classified as either ‘‘length stable’’ or
learn, the fracture is treated closed, and it is easier to maintain ‘‘length unstable.’’ Length-unstable fractures were either
comminuted or long oblique. In long oblique fractures the
anatomic alignment. length of the obliquity was twice as long as the diameter of
The choice of using TENs solves some of the compli- the femur at the level of the fracture. Comminuted fractures
cations and challenges associated with other methods of had more than one continuous fracture and a butterfly fragment.
treatment. Traction is the historic gold standard, but it results in Long spiral fracture types were recorded as oblique fractures.
a long hospital stay and use of a spica cast. It has also become The diameter of the femoral diaphysis measured by computer
a ‘‘lost art,’’ where ancillary staff are less familiar with its use on the digital radiographs was also recorded.
The surgeries were performed in a similar fashion by
eight surgeons at our institution. Medial and lateral incisions
From *Department of Orthopaedics, University of Colorado Health Science were carried out to expose the medial and lateral distal femoral
Center, School of Medicine, Denver, CO; and †Department of Pediatrics, metaphyseal flare for retrograde insertion of the two titanium
University of Colorado Health Science Center, School of Medicine, flexible intramedullary nails (Synthes USA, Paoli, PA).
Denver, CO. Specifics about the operative fixation were recorded, including
Study conducted at the Children’s Hospital Denver, Denver, CO. the nail diameter and length of the nails left outside the entry
None of the authors received financial support for this study. location. The quality of reduction achieved after surgery was
Reprints: Ernest L. Sink, MD, The Children’s Hospital Denver, 1056 East 19th
Ave., B060, Denver, CO 80218 (e-mail: [email protected]). recorded, noting varus, valgus, apex anterior (procurvatum),
Copyright 2005 by Lippincott Williams & Wilkins

J Pediatr Orthop Volume 25, Number 5, September/October 2005 577

Sink et al J Pediatr Orthop Volume 25, Number 5,
September/October 2005
recurvatum, and shortening. The use of postoperative immo- Minor Complications
bilization in a single leg spica was recorded. Of the 16 minor complications, 13 patients reported pain
Follow-up AP and lateral radiographs were analyzed on around the knee incisions and nail ends that often inhibited full
all postoperative visits. These radiographs were specifically range of knee motion. Eleven of these patients were noted to
analyzed for any change in alignment, fracture shortening, or have prominent palpable nails on examination. Two patients
change of the nail position. Shortening was best seen by
critical review of the lateral radiograph. Any obvious change in had postoperative wound infections after elective hardware
nail position relative to the physis (distal migration) was removal that were successfully treated with antibiotics. There
recorded. Clinical examinations were reviewed for documen- was one nonunion in the minor complication group, a 20-
tation of any clinical deformity, knee examination, prominent degree procurvatum without any obvious clinical deformity or
or painful hardware, and infection. functional deficit. There was one delayed union (.6 months)
Patients who needed unplanned surgery prior to fracture that healed without intervention in a patient with a comminuted
healing and elective hardware removal were categorized as midshaft fracture who needed the fracture site exposed to pass
having a ‘‘major complication.’’ Patients who did not need the nails. None of the patients with minor complications had
surgery to address their complication were categorized as any complaints on their final clinical follow-up.
having a ‘‘minor complication.’’ A specific reference to painful
and prominent hardware was included in minor complications. Major Complications
Also included were postoperative infections, malunion, and Eight patients required another surgery prior to fracture
delayed union. Delayed union and malunion were classified as healing. Six of the eight underwent procedures to shorten
minor in this series as neither required surgery to address the
complication, which was measured radiographically but not or remove extremely prominent or exposed nails around the
seen clinically. knee. All of these six patients had either long oblique or
Once the review was complete, the patients with minor comminuted length-unstable fractures. As the fractures short-
and major complications were analyzed for factors either in ened or angulated in the early postoperative period, the nails
fracture stability or treatment that might increase the risk of migrated distally, becoming prominent or exposed (Fig. 1).
complications. Comparisons where performed between stable The other two patients in this group had transverse fractures.
and unstable fractures. Statistical comparisons for categorical One developed thigh compartment syndrome requiring
variables were made using the chi-square test or Fisher exact fasciotomy and plating. The final patient, with a transverse
test when 25% or more of the expected cell frequencies were fracture, was stabilized with two 4.0-mm nails placed in the
less than 5 and the Cochran-Armitage test for trend. All canal with the convex contour of the rod directed in the same
statistical analyses were performed using SAS version 8.2. plane, causing an early angulation of 16 degrees and neces-
sitating application of an external fixator the following day.
None of the eight patients with major complications had
RESULTS any long-term morbidity necessitating further treatment at the
Sixty-three patients were surgically treated for femur time of their last follow-up.
fractures in the 2-year study period. Fourteen patients did not
have complete follow-up. Ten of the remaining patients had Stable Versus Unstable Fractures
a procedure other than TEN stabilization. Thus, 39 patients The complication rate for stable transverse fractures was
(39 femurs) met the criteria for review. The indications for 50% (12/24), and only two patients underwent surgery for their
treatment other than TENs in the remaining patients were complication. The complication rate for the unstable fractures
varied. Five patients underwent submuscular bridge plating was 80% (12/15), and 40% (6/15 patients) needed surgery to
in unstable femur fractures. These cases where performed in address the complication. The overall complication rate for
the last 6 months of the study period after initial data stable versus unstable fractures did not reach statistical signifi-
concluded that the complication rate was higher in unstable cance (P = 0.06). The risk of unplanned surgery was greater
fractures treated with elastic nailing. External fixation was for the treatment of unstable fractures (2/24 stable fractures vs.
used in three patients, two of these because of fracture 6/15 unstable fractures, P = 0.025). There was a statistically
comminution. A locked nail was used in two patients, one significant trend toward more serious complications among
because of comminution and one because of patient size. Thus, patients with unstable fractures (Fig. 2, P = 0.006).
Ten of the 15 patients in the unstable fracture group had
until the use of bridge plating, in only three patients over either fracture shortening or angulation, compared with only
2 years, treatment other than TENs was chosen because of 3 of 24 patients in the transverse fracture group (P = 0.0005).
fracture instability. Mean follow-up was 11 months (range 3– There were nine patients with unstable fractures who did not
29 months). Twenty-four patients were treated for length- have a major complication; however, six of these patients did
stable fractures and 15 for length-unstable fractures. have minor complications, and four of these were due to
In this series 24 patients (62%) had complications. There fracture shortening resulting in symptomatic nail prominence.
were 16 minor complications (41%) that did not require Leaving the nail ends past the physis increased the
unplanned surgery. Eight patients (21%) had a major compli- probability of complaints of nail prominence: 13 of 14 (93%)
cation that resulted in an unplanned surgery prior to complete of these patients had pain compared with 6 of 24 (25%) in
fracture union. One of these eight patients required two whom the nails were cut along the metaphyseal flare proximal
separate surgeries. to the physis (P , 0.0001). When the combined nail diam-
eter was less than 80% of the femoral canal diameter, the

578 q 2005 Lippincott Williams & Wilkins

J Pediatr Orthop Volume 25, Number 5, September/October 2005 Complications of Pediatric Femur
Fractures

FIGURE 1. A, AP and lateral radiographs of an 8-year-old with a comminuted femur fracture. B, Postoperative AP and lateral
radiographs after fracture stabilization with two 3.0-mm titanium flexible nails. The patient was placed into a single leg spica cast.
C, Six-week follow-up radiographs. The cast was removed, and the medial nail was extremely prominent and the lateral nail was
exposed. Shortening of the fracture is apparent on the lateral radiograph. D, Postoperative radiograph after the patient underwent
an unplanned surgery for prominent and exposed nail shortening.

complication rate was higher (Fig. 3, P = 0.008), with 7 of

20 (35%) having a major complication compared with 1 of
19 (5%) when the canal fill was at least 80%. The routine use
of a postoperative single leg spica cast in this series did not
avoid the occurrence of complications: 19 of 29 (66%) patients
had a complication with a cast compared with 5 of 10 (50%)
without a cast (P = 0.89). The major complication rate was also
not statistically different: 6 of 29 (21%) with a cast versus 2
of 10 (20%) without a cast (P = 0.68).

DISCUSSION
This review was initiated to analyze our complications
with the expanded use of TENs to include most femur
fractures treated surgically over a 2-year period, including
unstable femur fractures. The rate of complications was 62%
in this series, and eight patients (21%) underwent unplanned
surgery prior to complete fracture healing. The rate of com-
plications and the rate of unplanned surgery were higher when
stabilizing comminuted or long oblique fractures.
FIGURE 2. Type of complication by fracture stability.

q 2005 Lippincott Williams & Wilkins 579

Sink et al J Pediatr Orthop Volume 25, Number 5,
September/October 2005
nails in unstable fractures aided by placing a screw into the
distal nail end to prevent shortening. Although stainless steel
nails may have been more successful for unstable fractures,
titanium nails have increased in popularity over the stainless
steel model due to the ease of insertion.
In this series we specifically compared results in stable
and unstable fracture patterns. To our knowledge no study has
found fracture stability as an independent factor for com-
plications. This study was a consecutive series where nails
where used in the majority of fractures. There where only three
cases where the fracture pattern dictated other forms of
fixation (except during the last 6 months of the study period,
FIGURE 3. Combined nail canal fill by type of complication. where submuscular bridge plating was used after review of the
first 18 months). Our results have directly decreased the use of
TENs in unstable fractures at our center.
In the study period, 15 patients with comminuted or long In summary, complications resulting in unplanned surgery
oblique fractures were stabilized with TENs. The complication were seen more commonly in the unstable fracture group.
rate encountered in these fractures (12/15 [80%]) was greater Although the complications led to little or no long-term mor-
than that seen in transverse femur fractures (12/24 [50%]) but bidity, they can complicate treatment for the surgeon and
did not reach statistical significance. However, the types of families. For the more unstable comminuted or long oblique
complications requiring unplanned surgery (major complica- femur fractures, the surgeon should consider methods of stabi-
tions) were statistically significant. Ten of the 15 patients in lization other than TENs.
this unstable fracture group had either fracture shortening or
angulation, compared with only 3 of 24 patients in the trans- REFERENCES
verse fracture group. 1. Carey TP, Galpin RD. Flexible intramedullary nail fixation of pediatric
femoral fractures. Clin Orthop. 1996;332:110–118.
The efficacy of TENs has been well documented, yet the 2. Flynn JM, Hresko T, Reynolds RA, et al. Titanium elastic nails for
complications of their use have only recently been receiving pediatric femur fractures: a multicenter study of early results with analysis
attention. Flynn et al2 reported 19 ‘‘problems,’’ which included of complications. J Pediatr Orthop. 2001;21:4–8.
malalignment in 7, soft tissue knee irritation by nail promi- 3. Heinrich SD, Drvaric DM, Darr K, et al. The operative stabilization of
pediatric diaphyseal femur fractures with flexible intramedullary nails:
nence in 4, and nail backout in 1 patient; there were 9 patients a prospective analysis. J Pediatr Orthop. 1994;14:501–507.
with comminuted fractures, but no discussion as to whether the 4. Linhart WE, Roposch A. Elastic stable intramedullary nailing for unstable
complications were more common in these fractures. femoral fractures in children: preliminary results of a new method.
Luhmann et al10 found 21 complications in 43 femur fractures J Trauma. 1999;47:372–378.
(49%). The majority of complications were associated with 5. Blasier RD, Aronson J, Tursky EA. External fixation of pediatric femur
fractures. J Pediatr Orthop. 1997;17:342–346.
nail prominence and pain. They found that the complications 6. Gregory PT, Penvy T, Teague D. Early complications with external
encountered may be lessened by using the largest possible nail fixation of pediatric femoral shaft fractures. J Orthop Trauma. 1996;10:
diameter and leaving the nail protruding from the bone less 191–198.
than 2.5 mm. Their findings were repeated in our series. We 7. Miner T, Carroll KL. Outcomes of external fixation of pediatric femoral
shaft fractures. J Pediatr Orthop. 2000;20:405–410.
found that if the combined diameter of the nails was at least 8. Skaggs DL, Leet AI, Money MD, et al. Secondary fractures associated
80% of the canal diameter, the incidence of major and minor with external fixation in pediatric femur fractures. J Pediatr Orthop. 1999;
complications was less. We found it difficult to obtain an 19:582–586.
accurate measurement of the residual nail left outside the 9. Buford D Jr, Christensen K, Weatherall P. Intramedullary nailing of
cortex. We did find that if the nails ends were left distal to the femoral fractures in adolescents. Clin Orthop. 1998;350:85–89.
10. Luhmann SJ, Schootman M, Schoenecker PL, et al. Complications of
physis, complaints of nail prominence and pain were greater. titanium elastic nails for pediatric femoral shaft fractures. J Pediatr
The successful use of flexible nails in comminuted Orthop. 2003;23:443–447.
fractures has been reported. Rathjen et al11 had good results 11. Rathjen KE, De la Garza DJ. Flexible intramedullary fixation for difficult
with the use of stainless steel flexible nails in ‘‘difficult femoral femoral shaft fractures. Pediatric Orthopaedic Society of North America.
shaft fractures’’ including comminuted and oblique fractures. 2000.
Linhart and Roposch4 discussed the use of the stainless steel

580 q 2005 Lippincott Williams & Wilkins