After obtaining approval from our institutional review board, we enrolled
every child six to sixteen years old who presented to our Level-1 pediatric
trauma center with a diaphyseal femoral fracture. Since the focus of this
study was children treated either with traction and a spica cast or with
titanium elastic nails, we did not initiate data collection for children
treated with other methods, such as external fixation, plate fixation,
immediate closed reduction and application of a spica cast, or fixation with a
solid antegrade nail. Children with an underlying neuromuscular disease
(cerebral palsy or myelomeningocele), a metabolic bone disorder, or pathologic
fracture were also excluded.
We were not able to randomize the treatment methods. Instead, the choice of
treatment either with traction and application of a spica cast or with nailing
was based on the preference of the attending orthopaedic surgeon. Patients
were treated by one of eight different surgeons, some of whom always chose
traction and application of a spica cast, others of whom strongly preferred
nailing, and some of whom employed both methods during the study period.
At the time of admission, an advanced practice nurse conducted an intake
interview to determine family and social considerations that might affect
treatment. Each family was asked to estimate their child's status immediately
prior to the femoral fracture and to complete the Baseline American Academy of
Orthopaedic Surgeons Pediatric Outcomes Data Collection Instrument (version
2.0), answering the 114 questions by circling the best response. A follow-up
questionnaire was administered at the time of outpatient visits at six weeks,
three months, six months, and one year. Children who were eleven years of age
or older were asked to complete the adolescent version of this
questionnaire.
All of the patients treated with traction and a spica cast had skeletal
traction for approximately three weeks. As is the policy in our institution,
the traction pin (a 4.76-mm threaded Steinmann pin) was inserted in the
operating room with the patient under general anesthesia. The traction pin was
always placed in the distal part of the femur, 2 to 3 cm proximal to the
distal femoral physis. A radiograph was made initially after the traction had
been applied, with the hip and knee flexed 90°, and then at one-week
intervals until there was sufficient callus to allow pin removal and
application of a spica cast. A one and one-half hip spica cast was used for
each patient. It was generally applied three weeks after injury and was always
applied in the operating room with the child under general anesthesia. Each
surgeon used his judgment to determine when the spica cast should be removed.
In most cases, the cast was used until six to eight weeks after the injury.
This varied somewhat on the basis of the age of the child, the degree of
healing, and the judgment of the surgeon. After the cast was removed, the
children were referred to the physical therapy department for initial gait
training. Additional visits for physical therapy were prescribed for those who
did not achieve a satisfactory range of knee motion within two to four weeks
after cast removal.
The standard technique for use of titanium elastic nails has been described
elsewhere18,19.
For implant sizing, the narrowest diameter of the femoral diaphysis is
measured, and nails that are 40% of this narrowest diameter are used. All
titanium elastic nails in this series were placed in a retrograde fashion,
through the distal part of the femur. In the first few procedures done at our
institution, the distal aspects of the nails were left too long, causing a
large amount of soft-tissue irritation. The availability of a good extraction
device and the recognition of the problem of distal soft-tissue irritation led
to a practice of cutting the distal end of the nail so that only 1 to 2 cm
protruded from the bone. This portion of the nail was not bent away from the
bone and into the soft tissues. Nails were removed when the fracture line was
no longer visible radiographically, which typically was six to nine months
postoperatively. In the one case in which this guideline was not followed, the
nails were removed early, and a refracture occurred.
At each postinjury outpatient visit, we recorded limb alignment and
rotation, quadriceps strength, the range of motion of the hip and knee, the
condition of the wound and skin, and any pain or other symptoms.
Lower-extremity lengths were determined by clinical examination. The protocol
included a scanogram at the one-year follow-up evaluation for any child with a
clinically noticeable inequality in lower-extremity length. Twenty-nine
patients had the scanogram, and the others had no clinically detectable
inequality in lower-extremity length at the time of the last follow-up.
Radiographs were reviewed for alignment, callus formation, and the status of
the implants.
We recorded several important recovery milestones: time to walking with
aids, time to independent walking, time absent from school, and time until
full activity was allowed. The clinical end point was defined as a healed
fracture with a return to full activity. This usually occurred at
approximately one year. Children with complications were followed until the
complications had resolved.
Complications such as unacceptable alignment, inequality in the lengths of
the lower extremities, refracture, an unplanned reoperation, and skin or wound
problems were recorded. Our age-based guidelines for acceptable alignment and
shortening at the time of union, based on the work by Kasser and
Beaty1, are
presented in Table I.
Results from the American Academy of Orthopaedic Surgeons Pediatric
Outcomes Data Collection Instrument (version 2.0) were entered into a database
that used the recommended formulae to calculate the global scores as well as
subscores for transfers, sports, pain and comfort, expectations, and
happiness. The Student t test was used to compare the recovery milestones and
outcome data between the two treatment groups; significance was set at p <
0.05.
The charges for treatment were calculated by tallying the hospital charge
data of each patient from the injury until the time of final follow-up.
Charges were categorized into room charges (stemming from the injury as well
as any inpatient rehabilitation) and surgical charges (including those for
fracture treatment, implant removal, and surgical management of any
complications).
Eighty-three children were enrolled over a three-year period. Thirty-five
children (thirty-five fractures), with a mean age of 8.7 years (range, six to
sixteen years), were treated with skeletal traction and application of a spica
cast, and forty-eight children (forty-nine fractures), with a mean age of 10.2
years (range, six to sixteen years), were treated with titanium elastic nails.
All fractures healed, and no child sustained a complication that was expected
to cause permanent disability. On the basis of the criteria in
Table I, eighty children had
acceptable alignment at the time of union. Three children treated with
traction and application of a spica cast had unacceptable angulation or
inequality between the lengths of the lower extremities at the time of union:
two had 1.6 cm of shortening on the injured side, and the other patient had a
29° varus angular deformity as well as a 1-cm length inequality
(Figs. 2-A through 2-D). No
patient treated with nailing had angulation or a length inequality that
exceeded the guidelines. The largest inequality in lower-extremity length in
the group treated with nailing was 1.0 cm. Several more years of follow-up
will be necessary before the precise final difference in the lower-extremity
lengths of each patient can be determined.
There were significant differences between the two treatment groups with
regard to the recovery milestones (Table
II). Compared with the children treated with traction and a spica
cast, those treated with titanium elastic nails had shorter hospitalization,
walked with support sooner, walked independently sooner, and returned to
school earlier. These differences were significant (p < 0.0001).
We were unable to assemble sufficient outcome data from the questionnaire
to allow valid statistical comparison (Fig.
3). Many parents refused to complete the forms, skipped questions,
or missed appointments at key intervals, all of which limited the data
collection. With the two different questionnaires (child and adolescent), two
treatment methods, and four narrowly defined recovery intervals, the number of
forms completed proved insufficient for statistical analysis. However, the
available questionnaire data showed a trend toward better function, especially
in the first six months after injury, for those treated with titanium nails.
By one year after the injury, there was little difference between the two
groups. Some children who had been treated with nailing had a temporary
reduction in function following nail removal, while restrictions were in
place, usually at six to twelve months postoperatively.
A complication occurred in twelve (34%) of the thirty-five children treated
with skeletal traction and application of a spica cast. Three children had an
inequality between the lengths of the lower extremities or angulation that
exceeded what we considered to be acceptable results. In addition, two
patients had a substantial loss of reduction; one was treated with
remanipulation and application of a spica cast, and the other was treated with
closed reduction and internal fixation with titanium elastic nails. A 5 by
4-cm pressure ulcer developed inside the cast of a nine-year-old girl, who
also had 18° of varus angulation and a 15-mm inequality between the
lengths of the lower extremities at eleven months after the injury. Two
patients sustained a refracture. The other four complications included three
pressure ulcers and one case of severe knee stiffness requiring manipulation
with the patient under anesthesia.
Ten (21%) of the forty-eight children treated with titanium elastic nails
had a complication; irritation at the nail entry site, which occurred in eight
children (17%), was the most common problem. The irritation was minor in six
children. However, in two who had been treated early in the series, the distal
end of the nail was very long and bent excessively
(Fig. 4), leading to subsequent
wound breakdown, but neither deep infection nor osteomyelitis developed. These
children were treated with nail removal, and the final result was
satisfactory. There were two other complications in the nailing group. A
fourteen-year-old boy with a midshaft transverse fracture sustained a
refracture because the nails were removed only six and one-half weeks after
implantation. At six weeks following nail removal, he complained of pain after
a fall and radiographs revealed angulation through the fracture callus. He was
treated with closed reduction and repeat nail implantation. The other
complication involved a thirteen-year-old boy who fell a few days after
surgery and bent the nails (Figs. 5-A, 5-B,
5-C, 5-D, 5-E,
5-F). He had a repeat closed reduction in the operating room and
was non-weight-bearing, in a wheelchair, when discharged. He was lost to
follow-up for ten months. When he subsequently returned, he was still using
the wheelchair and he reported pain in the thigh. Radiographs showed a delayed
union, which was treated with external fixation, with a satisfactory final
result. No child treated with nailing had >1 cm of inequality in the
lower-extremity lengths, none lost rotational alignment, and there were no
complications related to implant removal (beyond the refracture following
premature nail removal noted above).
Of the problems and complications encountered in the two treatment groups,
only skin irritation and joint stiffness in the patients treated with skeletal
traction and a spica cast would be considered intrinsic to the treatment
method. All of the other problems and complications would have been
preventable by use of optimal techniques and postoperative management.
There was no significant difference between the total hospital charges for
treatment with titanium elastic nails and those for skeletal traction and
application of a spica cast (Table
II). The room charges were significantly higher for the treatment
with traction and a cast (p < 0.001), but the surgical charges were
significantly higher for the nailing (p < 0.0001).
Until recently, skeletal traction and application of a cast was the
preferred method for treatment of diaphyseal femoral fractures in children and
young adolescents2.
This method stood the test of time because it was relatively conservative and
permanent complications impairing future function were rare. However,
orthopaedists increasingly have tried a variety of methods to avoid prolonged
immobilization2,5,18-22.
Although a spica cast is safe and effective for many patients, operative
stabilization is particularly suited for children who sustain multiple
injuries from high-energy
trauma7, children
with a head injury or
spasticity23, and
children older than ten years of
age24. Recent
studies have also increased our awareness of the psychosocial and economic
effects of spica cast immobilization on children and their
families24-26.
The ideal device for the treatment of most femoral fractures in children
would be a simple, load-sharing internal splint that allows mobilization and
maintenance of alignment and extremity length until bridging callus forms. The
device would exploit a child's dense metaphyseal bone, rapid healing, and
ability to remodel, without risking damage to the physes or the blood supply
to the capital femoral epiphysis. Both Ender nails and titanium elastic nails
offer these features. Heinrich et
al.17 reported
excellent results and no serious complications in seventy-eight children
treated with Ender nails, which are widely available and used in many centers
to manage femoral fractures in children. For more than two decades, French
surgeons have used titanium elastic nails to achieve stable intramedullary
fixation15,16,27-30.
Ligier et al.15
stated that "Ender nails are not elastic enough for treating
children," and reported excellent results, with no major angular or
rotational malunions, after treatment of 123 fractures with titanium elastic
nails. Two fractures healed with 2 cm of lower-extremity length inequality,
and there was a 10% prevalence of skin ulceration caused by the prominent
distal portion of the nail at the insertion site. Recently, titanium elastic
nails have become widely available in North America. An initial multicenter
study19 showed
excellent or satisfactory results in fifty-seven of fifty-eight cases treated
with such nails. No child lost rotational alignment in the postoperative
period. Irritation of the soft tissue near the knee by the nail tip occurred
in four patients and led to a deeper infection in two of them. Although some
surgeons feel strongly that titanium nails are more suitable for the treatment
of fractures in
children15, we are
aware of no study directly comparing the two methods. We have no data from our
study to support the notion that titanium nails are superior to Ender
nails.
This study was designed to evaluate outcomes prospectively in the first
year following treatment of a diaphyseal femoral fracture in children between
six and sixteen years of age. Other studies have addressed this important
patient
group3,31,32,
but none have done so in a comprehensive way with a prospective, cohort
outcome design. Our goal was to consider not only clinical and radiographic
results, but also psychological, social, and cost issues.
This study had several limitations. First, we could not randomize the
treatment method. Although the patients and fracture types were similar, we
cannot exclude the possibility of a selection bias favoring one type of
treatment. The children treated with the titanium elastic nails were older
(10.2 compared with 8.7 years of age) and thus were subject to more stringent
criteria for alignment and lower-extremity length inequality
(Table I). Second, because this
was a single-center study, the results should be generalized with caution.
Other surgeons may find different results or complication rates with the two
treatment methods. Third, we had difficulty gathering all possible data with
the American Academy of Orthopaedic Surgeons Pediatric Outcomes Data
Collection Instrument. The nature of our study required patients and families
to complete the forms at four specific times in the first year after injury.
If the family missed follow-up appointments or refused or failed to complete
the form, time-sensitive data were lost. Finally, we were unable to achieve
our initial goal of determining the comprehensive cost of treatment. We could
only record hospital charges and allow them to serve as a proxy for the entire
cost of treatment. Several previous studies in which cost was
considered3,31-33
were also flawed, either because all costs (including those for physical
therapy, transportation, all outpatient visits, and radiographs) were not
considered or charge data were used, an approach that is known to be
inherently flawed and difficult to generalize from. Despite the prospective
design of our study, we too were unable to acquire useful cost data. With so
many different insurance arrangements in our region, we could not capture
outpatient costs. Specifically, we hoped to be able to quantify the total cost
associated with the prolonged immobilization of a school-aged child. For
example, a child treated with skeletal traction and a spica cast often
requires ambulance transportation to outpatient visits and physical therapy
before and after removal of the cast. Also, there are costs to families and
society, such as missed work for a parent caretaker or tutoring costs during
months of missed school. These costs are difficult to quantify and may be much
more apparent to the child and the family than to the orthopaedic surgeon.
The focus of this study was the treatment and recovery period. We did not
attempt to determine the long-term results of treatment with titanium elastic
nails or skeletal traction and a spica cast, since the long-term result after
a diaphyseal femoral fracture in this age-group is almost always satisfactory,
regardless of which of several methods is used for treatment. With the focus
on the first year, we cannot comment on the exact final lower-extremity length
and alignment of each child. In this series, all of the fractures healed and
it is unlikely that any of the patients with problems or complications will
experience a permanent loss of function.
Orthopaedic surgeons have been motivated to consider alternatives to
skeletal traction and application of a spica cast to avoid the adverse
physical, psychological, and social consequences of prolonged immobilization
of a school-aged child. The results of this prospective study support the
recent empiric observations and published retrospective studies indicating
that children treated with titanium elastic nails achieve recovery milestones
significantly faster than do those treated with traction and a cast. Hospital
charges for the two treatment methods are similar, and when we performed the
procedure the complication rate associated with titanium elastic nails
compared favorably with that associated with skeletal traction and application
of a spica cast.
Note: The authors acknowledge Joann D'Italia, MSN, CRNP, for her
help with the cost data, and Beverly Teti, CRNP, for her help gathering
clinical data at follow-up visits.