Tibia and Fibula Fractures

Motor Vehicle Accidents. Motor vehicle accidents continue to produce the highest incidence of tibial shaft fractures.⁷³ This holds true for both automobiles and motorcycles, and, indeed, tibial fractures are the most common long bone injury seen

with motorcycle accidents. Unfortunately, those tibial fractures
associated with motorcycles tend to be significantly more severe with a
higher rate of type III open injuries and even amputation (Fig. 55-4).⁶⁵,³⁰³ This higher-injury pattern is also noted to exist in pedestrians hit by an automobile, as shown in Table 55-4. Burgess et al.⁵²
reviewed 130 tibia fractures in 102 adult patients, all of whom had
been pedestrians. They noted that the vast majority of these injuries
showed a high-energy injury pattern and that 65% were Gustilo type III
open injuries. They also observed a high prevalence (30%) of bilateral
injuries.⁵²

With respect to vehicle occupants involved in motor
vehicle accidents, it may be that such safety devices as airbags may
not entirely prevent lower extremity injuries. Burgess et al.⁵¹
conducted crash site and clinical investigations into 10 patients
involved in a motor vehicle accident with a deployed airbag. They
identified a significant number of lower-extremity injuries, as well as
abdominal and thoracic injuries in these patients including six tibial
fractures.⁵¹

Violence. In the series by Court-Brown and McBirnie,⁷³
direct blows or assaults accounted for 4.5% of all tibial fractures and
occurred in younger patients. These fractures have either a highor
low-energy morphology. Levy et al.¹⁷²
reviewed 47 patients who had been assaulted by a baseball bat. They
found a significant number of complications including a ninefold higher
prevalence of compartment syndrome compared to the rest of their
population with tibial fractures.¹⁷²

Gunshot wounds also account for high-energy fractures of
the tibia. Data from firearm-related deaths suggests that the incidence
of firearm use is higher in high and upper middle income countries in
the Americas as compared to those in Europe, Asia, and Oceania with the
highest incidence being in the United States.¹⁶⁰
These can be both civilian and military, although military injuries can
also be blast type injuries from explosives such as land mines.⁶⁵
Low-velocity injuries associated with a muzzle velocity of less than
2000 feet per second are more commonly seen in civilian practice. These
generally result in less soft tissue damage than is seen with
higher-velocity arms.²⁸⁴ Some authorities argue that all high-velocity injuries should be classified as Gustilo type III injuries.²⁸⁴ With regard to civilian arms, injuries Leffers and Chandler¹⁶⁷
found that 54.3% of tibial diaphyseal fractures were comminuted OTA C
type injuries, 20% were type A, 8.6% were type B, and 17.1% only
involved one cortex.⁶⁵

Court-Brown and McBirnie⁷³
have identified soccer-related injuries as the largest contributor to
sports-related tibial shaft fractures, accounting for 80.1% of
sports-related tibial diaphyseal fractures.²⁵⁴ These fractures were typically lower-energy and

were usually Tscherne grade 0 or 1 injuries occurring in young men.²⁵⁴
These fractures can however have a higher-energy injury pattern when
they occur from a direct blow from a soccer tackle. The other commonly
reported sports-related tibial fracture is that seen with skiing.
Grutter et al.¹⁰⁹ series reported a
high incidence of skiing-related tibial fractures, many of which were
torsional lower-energy injuries. These injuries were more common in the
early 1980s as compared to the late 1980s, and this is attributed to a
change in bindings with the newer bindings allowing for release of the
ski more readily.¹⁰⁹

Orthogonal tibial views, usually anteroposterior and
lateral, as well as full-length tibial studies should be obtained to
assess the bone as well as the soft tissues. It is essential that the
knee and ankle are included in these radiographs. There are particular
associations within some fracture patterns. For example, distal third
tibial shaft fractures are associated with intra-articular fracture
lines and extensions.⁴⁴ This may
also hold true with proximal tibial diametaphyseal fractures. If there
is any suspicion that these fractures extend into the adjacent joint,
then orthogonal views of the knee and ankle should be obtained. Further
imaging of any intra-articular extension can be done using computed
tomography (CT) examination, although CT examination is generally not
indicated for purely diaphyseal fractures. In assessing the plain
radiographs, it is essential to note the location and extent of the
fracture and whether there is significant comminution within the
fracture. It is also important to examine whether there is extension
proximally into the proximal tibial articulation or distally into the
ankle joint and to see if there is a concomitant fracture of the fibula
or dislocation of either the proximal or distal tibio-fibular
articulation.²⁶³ Assessment of the
canal size to determine the suitability for intramedullary nailing can
also be done. The overall quality of the bone should be assessed with
particular reference to generalized osteopenia, any other pathologic
processes, such as tumors or infection, or the presence of previous
fractures. The soft tissues on the plain radiographs should also be
reviewed. If there is any air, this may indicate an open injury or
possibly the presence of gas producing infective organisms, depending
on the patient’s presentation.

Magnetic resonance imaging (MRI) is usually not
necessary in tibial diaphyseal fractures with or without an
intra-articular extension. However, there have been reported cases of
both knee ligament injury and meniscal damage in association with
tibial shaft fractures, and therefore an MRI examination of the knee
may be warranted if there is a clinical suspicion of knee injury.¹³³
MRI and bone scan imaging may also be necessary if it is required to
exclude a stress fracture in the presence of pain but with no obvious
evidence of a fracture on plain radiographs. If there is evidence of
any vascular compromise, referral to a vascular surgeon is necessary
and arteriography may be required.

Much was written about the nonoperative management of
tibial shaft fractures in the 1960s and 1970s. There was considerable
debate about what constituted an unacceptable level of displacement

or
malalignment. Many surgeons accepted <1 cm of shortening, 5 degrees
of valgus but no varus malalignment, 10 degrees of malalignment in the
anteroposterior plane, and 5 to 10 degrees of external rotation but no
internal rotation deformity. There is no good evidence about what
degree of malalignment or shortening is associated with functional
impairment, and it seems likely that the definitions of acceptable
malalignment arose because of the significant difficulty in achieving
an anatomic reduction with nonoperative treatment. With increasing use
of intramedullary nailing, fewer surgeons now use nonoperative
treatment for tibial diaphyseal fractures and we only advocate its use
in low-energy undisplaced fractures as shown in Figure 55-8. In 1964, Nicoll²⁰⁸
described a series of 705 cases of tibial shaft fractures treated
nonoperatively in a long leg cast (level of evidence: 4). He documented
a union time of approximately 15.9 weeks and a malunion rate of 8.6%.
Importantly, however, he described a 25% prevalence of residual joint
stiffness that increased to 70% in those who had a nonunion associated
with an open fracture. He also observed a 15.3% prevalence of infection
and a 60% prevalence of delayed union or nonunion. In the entire
series, about 5% of fractures had not united by 1 year.²⁰⁸ During the same period, Dehne⁷⁹
found that long leg casting resulted in shortening of at least half an
inch in approximately one third of patients in his series. Other
authors have described the prevalence of malunion as anywhere from
approximately 9% to 50% with the use of a long leg cast with joint
stiffness ranging from 7% to 42%.¹⁶²,¹⁶⁵,²⁶⁸,²⁸² Unfortunately, these retrospective, historical studies do not make the distinction between displaced and undisplaced fractures.

More recent literature suggests that intramedullary
nailing gives better results than cast management. In a randomized
trial of displaced tibial shaft fractures, Hooper et al.¹³⁰
randomized 62 patients to receive either an intramedullary nail (29
patients) or cast treatment (33 patients) (level of evidence: 1). They
found a significantly decreased time to union, decreased malunion rate,
and less time off work in the intramedullary nail group and concluded
that displaced tibial fractures should be treated with an
intramedullary nail.¹³⁰ Karladani et al.¹⁴⁴
enrolled 53 patients with displaced closed or Gustilo type 1 open
tibial shaft fractures into a trial and randomized 27 patients to
treatment with an intramedullary nail and 26 patients to treatment with
a plaster cast (level of evidence: 1). Fourteen fractures in the
plaster cast group showed redisplacement after reduction and required
further treatment. These 14 patients were treated with either cerclage
wires or screws and then continued cast treatment. The union time was
25 weeks for the cast managed group as compared to 19 weeks for the
intramedullary nail group.¹⁴⁴

Using the Nottingham health profile as an outcome measure, scores of
physical ability and work ability were better in the intramedullary
nail group than in the cast managed group at 3 months. In those treated
with a cast, delayed union and malunion were more common.¹⁴⁴ Bone et al.,³⁷
in a retrospective cohort study of displaced tibial shaft fractures,
treated 47 patients with a closed intramedullary nail and 52 patients
with closed reduction and casting (level of evidence: 3). They found a
shorter time to union and improved functional scores in the
intramedullary nail group using both the Short Form-36 (SF-36) as well
as the Iowa Knee Evaluation and Ankle Evaluation rating system after
approximately 4 years of follow-up.³⁷

In a systematic review of the prospective literature of tibial shaft fracture management, Coles and Gross⁶²
identified 145 fractures treated with a long leg cast from the combined
literature. They reported a 13.1% incidence of delayed union, a 4.1%
incidence of nonunion, and a 31.7% incidence of malunion.⁶² These rates were higher than those seen with either plate or intramedullary nail fixation.⁶²

Functional bracing has also been described as another
method of managing tibial shaft fractures nonoperatively. Functional
bracing, as introduced by Sarmiento, is applied after the soft tissue
swelling has resolved. In 1995, Sarmiento et al.²⁴⁷
reviewed 1000 closed tibial diaphyseal fractures that were treated with
prefabricated functional braces (level of evidence: 4). They found that
the mean final shortening was 4.28 mm, which compared to an initial
mean shortening of 4.25 mm. They suggested that final shortening does
not increase beyond the initial observed shortening.²⁴⁷
They also observed that 90% of their patients had final angular
deformity in any plane ≤6 degrees. However, they found that the
presence of an intact fibula was more likely to result in an angulatory
deformity and suggested that this pattern may not be appropriate for
functional bracing. They reported an overall incidence of nonunion of
1.1%.²⁴⁷

There was, however, a lack of validated functional
outcome scores to assess patients in the study by Sarmiento. The good
results obtained by Sarmiento have not been found by other authors. Den
Outer et al.⁸¹ documented a 40%
prevalence of nonunion and noted that one third of patients took longer
than 20 weeks to unite (level of evidence: 4). Alho et al.⁵
suggested that there were fewer complications with intramedullary
nailing which had a higher rate of excellent and good results as
compared to functional brace management (level of evidence: 4). In a
review of 103 tibial shaft fractures treated with functional bracing by
Digby et al.,⁸³ 11% had restricted ankle motion and 45% had reduced subtalar motion (level of evidence: 4).

Functional bracing and cast management require
significant attention to detail and frequent follow-up to check
fracture alignment. Interestingly, a recent economic analysis of
management strategies for closed and Type 1 open tibial shaft fractures
by Busse et al.⁵⁴ suggests that
reamed intramedullary nailing is the treatment of choice and
economically has less of a financial burden to society while casting
alone resulted in significantly higher financial costs.

From the previously mentioned international survey of
surgeons, it can be seen that intramedullary nailing can be considered
the implant of choice for tibial shaft fractures (Fig. 55-9).²⁶ Indeed, Court-Brown¹⁹
reviewed 1106 tibial shaft fractures managed by reamed intramedullary
nailing, this being the largest series to date (level of evidence: 4).
The overall infection rate following intramedullary nailing of closed
tibial shaft fractures was 1.9% and the aseptic nonunion rate was 4.4%.¹⁹ For open fractures, infection rates increased from 6.9% for Gustilo type I injuries to 16.9% for Gustilo type IIIB injuries.⁶⁶
The nonunion rates seen in those with open fractures also increased
from 12.1% for Gustilo type I injuries to 49.2% in patients with
Gustilo type IIIB injuries.⁶⁶ There
is, however, a continued debate regarding the use of reaming in
intramedullary nailing. Indeed, as seen from the international survey
of surgical preferences, there was significant decline in the use of
reamed intramedullary nailing as the severity of the open fractures
increased.²⁶

There have been a number of experimental studies
undertaken to assess the role of reaming in the biology of bone
following fracture and osteotomy. In a standardized sheep tibia
fracture model, Schemitsch et al.²⁴⁸
found that reaming nailing affected bone blood flow more than
nonreaming nailing and that revascularization of the bone took about 6
weeks in the unreamed group compared with 12 weeks in the reamed group.
However, they also showed that in the same fracture model, reaming had
no deleterious effect on the strength of the callus or union of the
fracture.²⁴⁹ Using a canine fracture model, Klein et al.¹⁵⁰
found that reaming decreased the cortical blood flow to two thirds of
the area of the cortex, and in some places this was almost full
thickness. In another canine fracture model, Hupel et al.¹³⁶
found that there was no significant difference between bone formation
and mineral apposition in those tibias that were treated with no
reaming, limited reaming, or standard reaming, although the least
damage to cortical bone was caused by limited reaming. In this model,
the unreamed nail was inserted with a tight fit. Hupel¹³⁴
also examined the effects of both a looseand tight-fitting unreamed
nail on cortical vascularity and found that a loose-fitting nail spared
cortical perfusion when compared to a tight fitting nail. At 11 weeks,
the perfusion of the cortex had increased in both groups, although the
perfusion was higher in the loose fitting nail group. Hupel et al.¹³⁵
also found that reaming increased the vascularity and perfusion of the
surrounding muscles, and they argued that reaming may increase the
extraperiosteal blood supply. The basic science work carried out by
Hupel et al.¹³⁴,¹³⁵,¹³⁶ and Schemitsch et al.²⁴⁸,²⁴⁹
suggests that the ideal implant is a small diameter nail with limited
reaming because it results in better cortical vascularity, improved
cortical porosity, the same strength of union, and the best stimulation
of the extraosseous circulation.

Reaming has been suggested not only to have local
effects on the bone but also systemic effects on the patient. There is
an argument that reaming of the tibia may potentiate the risk of fat
embolism and pulmonary complications, especially if more than one long
bone is reamed as in multiple fractures.⁶⁵,¹⁰³,²²² However, this argument is more relevant to femoral reaming, and experimental work by Christie et al.⁵⁸
has shown that most clinical problems associated with fat embolism were
confined to those who received femoral reaming as opposed to tibial
reaming, even though transesophageal echocardiography identified emboli
in 92% of reaming procedures.

Bhandari et al.²⁹
undertook a meta-analysis of reamed versus nonreamed intramedullary
nailing and its use in lower extremity long bone fractures and found
from the published randomized trials that were included in the
meta-analysis, the pooled estimate of the effect of reaming on nonunion
was an odds ratio of 0.32 (95% CI 0.17 to 0.67; p = 0.0019) (level of evidence: 1) (Fig. 55-10).
This suggests that reamed nailing would potentially eliminate two
thirds of nonunions that occur with nonreamed nailing and that,
depending on the baseline risk of patients to develop nonunion, the
number of patients that required to be treated with reamed nailing to
prevent one nonunion could potentially be anywhere from 5 to 25. The
study also found a higher rate of implant failure with nonreamed nails
than with reamed nails.²⁸ One caveat
to consider when extrapolating this data to tibial shaft fractures is
that this meta-analysis also included fractures of the femoral shaft. A
second meta-analysis conducted by Coles and Gross⁶²
also suggests that reamed intramedullary tibial nailing tends to show a
trend to a lower incidence of nonunion (8%) as compared to 16.7% with
unreamed nailing. A third systematic review and meta-analysis by
Forster et al.⁹⁴ looked at
prospective randomized studies comparing reaming and nonreaming of
tibial nailing in adults (level of evidence: 1). They identified seven
studies that were eligible for inclusion, with 291 tibial shaft
fractures being combined from the included studies. They also
identified an increased rate of nonunion with nonreamed intramedullary
nails as compared to reamed intramedullary nails.⁹⁴ Their pooled estimate of effect for nonunion was an odds ratio of 2.83 (95% CI 1.16 to 6.88; p = 0.02) for nonreamed nails.⁹⁴
That is to say, a nonreamed intramedullary nail would increase the
chances of nonunion by approximately threefold. They also found an
increased incidence of implant failure with the use of unreamed nails
compared to reamed nails, but they found no differences in malunion or
the risk of compartment syndrome using either technique.⁹⁴ This meta-analysis was limited by study heterogeneity and small sample sizes.

In another prospective randomized trial, Braten et al.⁴⁸
randomized 78 patients with tibial shaft fractures to treatment with
external fixation or a locked intramedullary nail (level of evidence:
1). They found that the time to radiographic union was not
significantly different between the groups, but there was a higher
reoperation rate in the external fixation group. They found that after
1 year there were no significant differences in knee motion, ankle
motion, or fracture site pain.⁴⁸

The study to prospectively evaluate reamed
intramedullary nails in tibial fractures (SPRINT) was a multicenter,
randomized trial comparing reamed versus nonreamed intramedullary
tibial nail insertion in 29 clinical sites in Canada, the United
States, and the Netherlands.³⁰
SPRINT aimed to enroll 1200 skeletally mature patients with open
(Gustilo types I to IIIB) or closed fractures (Tscherne Types 0 to 3)
of the tibial shaft amenable to surgical treatment with an
intramedullary nail.³⁰ Patients,
outcome assessors, and data analysts were blinded to treatment
allocation. Perioperative care was standardized, and reoperations

before
6 months were proscribed. Patients were assessed at discharge, 2 weeks
postdischarge, and at 6 weeks, 3, 6, 9, and 12 months postsurgery. A
committee blinded to allocation adjudicated all outcomes. The primary
composite outcome of reoperation included bone grafts, implant
exchanges, and dynamizations, as well as autodynamization, in patients
with fracture gaps less than 1cm postintramedullary nail insertion.
Outcome events also included operations for infection and fasciotomies
irrespective of the fracture gap.

The 1-year follow-up was completed by 1226 participants
(93%). Reoperation rates in SPRINT were far lower than previously
reported with 57 (4.6%) requiring implant exchange or bone grafting for
nonunion in comparison to 12.4% in 5 prior randomized trials. Analysis
showed that 105 (16.9%) of the reamed group and 114 (18.9%) of the
nonreamed group experienced a primary outcome event (RR = 0.90, 95% CI
0.71 to 1.15). The events appeared to differ in closed and open tibial
fractures (interaction p = 0.01). In
patients with closed fractures, 45 of 416 (11%) in the reamed group and
68 of 410 (17%) in the unreamed group experienced a primary event (RR =
0.65, 95% CI 0.46 to 0.93; p = 0.03). This
difference was largely due to differential rates of dynamization. In
patients with open fractures, 60 of 206 (29%) in the reamed group and
46 of 194 (24%) in the unreamed group experienced a primary event (RR =
1.27, 95% CI:0.91 to 1.78; p = 0.16).

SPRINT demonstrated a benefit for reamed intramedullary
nail insertion in patients with closed tibial shaft fractures largely
due to dynamization. Optimization of perioperative care and delaying
reoperation for nonunion for at least 6 months may substantially
decrease the need for reoperation in tibial fracture patients.

Tibial plating was first undertaken in the late 1800s.
However, it was rarely used in the early 1900s because of the
significant risk of sepsis and lack of good anaesthesia. It was not
until the introduction of antibiotics, good anaesthesia, and modern
surgical techniques that plating tibial shaft fractures became more
feasible.⁶⁵ This was in the middle of the twentieth century. In 1976, Ruedi, Webb, and Allgower²⁴³
published a large series of 418 acute fractures treated with the
AO/ASIF dynamic compression plate. They reported very good or good
functional results in 98.1% of closed fractures and 88.4% of open
fractures, with a nonunion rate of 1%, an infection rate of 1% in their
closed tibial shaft fractures, a nonunion rate of 5.3%, and an
infection rate of 11.6% in their open tibial shaft fractures.²⁴³ In 1982, Christensen et al.⁵⁷
also reviewed their series of 96 displaced fractures of the tibial
shaft that were treated by AO/ASIF plating techniques with rigid
internal fixation. In this series, 40% were open fractures and,
interestingly, the infection rate following surgery was 5.3% in closed
fractures and 0% in open fractures. They recorded that more than 90% of
their patients returned to work 6 months after the injury, and there
were no cases of nonunion or refracture at final review at 36 months.⁵⁷ They suggested that rigid internal fixation facilitated secondary soft tissue reconstruction.

In a randomized trial, Van der Linden and Larsson²⁸²
evaluated the treatment of displaced tibial shaft fractures by
comparing plating with nonoperative management. They found that there
were more complications in the plating group, but healing time and
anatomic results were better when compared to nonoperatively managed
patients.²⁸² They suggested,
however, that open fractures were not suitable for plate fixation. They
found the most malalignment in the nonoperative group occurred at the
distal end of the tibia.²⁸²

The use of AO techniques requires anatomic reduction and
rigid fixation with a dynamic compression plate to allow the bone to
heal by primary direct bone healing. This is much more easily
accomplished in simple fracture patterns where direct compression with
the plate can be obtained. However, this necessitates a significant
amount of soft tissue stripping of the fracture in addition to
periosteal stripping, both of which are avoided with the use of an
intramedullary nail for diaphyseal fractures. There is also concern
that full contact plates rigidly applied to the bone may result in some
bone resorption beneath the plate. This was initially thought to be the
result of stress shielding, but it is now thought to be due to
localized dysvascularity from plate compression of the periosteum.⁵⁰,⁶⁴,¹⁴⁹,²²⁵,²²⁶,²⁹⁵ Because of this problem, plates were developed that had less contact with the bone.⁹²,¹⁴¹,¹⁵⁴
Since plating comminuted fractures with standard compression plating
techniques is technically challenging and results in significant
stripping of the soft tissues, biologic plating techniques were
developed whereby the plate spans the zone of injury with fixation
proximally and distally. To facilitate this indirect reduction
techniques including the use of fluoroscopy, bone distractors or
external fixators for reduction and ball-spike pushers and sharp bone
holding forceps have been developed.¹⁵³,¹⁷¹
These changes in surgical technique have taken into account the amount
of soft tissue damage associated with open fracture surgery. Further
surgical changes include supraperiosteal plating techniques as well as
submuscular and minimally invasive plating techniques. Oh et al.²¹⁴
have reported on the use of percutaneous plating techniques in unstable
tibial shaft fractures They found that of 24 unstable tibial fractures,
22 united without secondary surgery, and there was only one significant
angular malalignment with no infection being seen.

Minimally invasive plate osteosynthesis techniques have
recently been applied to fractures of the proximal distal tibia and
have been used in conjunction with newer designs of locked plates.
Locked plating means that the screw head is locked to the plate by
either a threading mechanism or other type of device. This effectively
creates a fixed angle between the screw and the plate. There is a
significant risk of malunion in the treatment of proximal and distal
tibial diametaphyseal fractures with all treatment methods including
intramedullary nailing and nonoperative management. Recently, the use
of locked plate fixation utilizing minimally invasive techniques has
been put forward as one way of maintaining alignment in proximal and
distal tibial fractures.⁴⁰,²¹⁴
Locking the screw to the plate creates multiple fixed angle points of
fixation whose mode of failure differs from conventional plating.²⁸⁵
Conventional plate/screw constructs allow for some toggle at the screw
plate junction. Stability is provided by the compression of the plate
to the bone, the fixation of the screws into bone, as well as
compression at the fracture.²⁸⁵ The screws can fail sequentially, allowing the plate to come off the bone.

In contrast, locked plate fixation provides a rigid
construct and the locked screws must fail together with the screws
pulling out of the bone at the same time.²⁸⁵
This type of construct has been known for a number of years to increase
the pull-out strength of the screws and indeed the Schuli nut was
initially developed to lock the screw to the plate and thus increase
pullout strength. It therefore represented an early form of low
contact, locked screw/plate fixation.¹⁵¹
This increase in pullout strength is desirable in weakened bone such as
may be encountered with osteopenia or osteoporosis. Also, the fixed
angle of the screws into the metaphyseal portion of the bone provides
bicolumnar support, transferring load from both sides of the proximal
or distal tibia to the shaft. While intramedullary nailing is advocated
as the treatment of choice for diaphyseal fractures, not all proximal
or distal fractures permit this technique, and locked plating may be a
useful technique for these fractures.

While fixed angular plating systems have been used for
intra-articular proximal tibial fractures as well as intra-articular
distal tibial fractures, their use in OTA 41-A fractures is promising
as they not only help to provide bicolumnar support to the joint, but
they can be placed using minimally invasive techniques and they can
span large areas of comminution as shown in Figure 55-11. Ricci et al.²³⁶
have documented a series of 38 patients with proximal tibia fractures,
18 of which were OTA type A fractures. They were treated using a
minimally invasive approach.²³⁶ All of the fractures united in satisfactory alignment.

Nork et al.²¹⁰
published the results of a series of 37 proximal tibial shaft fractures
treated with an intramedullary nail (level of evidence: 4). The average
distance from the articular surface was 67.8 mm. They found acceptable
alignment in 34 out of 37 (91.9%) of their fractures, with three
patients showing angular

malalignment.
They suggested that multiple techniques were needed to obtain and
maintain reduction. These techniques included the use of unicortical
plates as well as a femoral distractor.²¹⁰
In a similar prospective study of 45 patients with proximal third
tibial shaft fractures treated with an intramedullary nail, Vidyadhara
and Sharath²⁸³ reported seven
(15.6%) cases of malunion. Bonegrafting procedures were done in three
cases to obtain union. They reported good functional outcomes in 96% of
patients using the lower extremity functional score. Their
recommendations include using an intramedullary nail with a high
proximal bend and static interlocking of proximal screws (level of
evidence: 4).²⁸³ Krettek et al.¹⁵⁶,¹⁵⁷
have described the technique of using blocking screws in 23 proximal
and distal fractures. They found a mean loss of reduction of 0.5
degrees in the frontal plane and 0.4 degrees in the sagittal plane.¹⁵⁶,¹⁵⁷
Blocking screws are placed so as to effectively reduce the size of the
tibial canal either proximally or distally and thereby guide both the
guide wire and the nail into an acceptable position. They can be placed
in any plane but are usually placed in either the sagittal or coronal
planes. Techniques for preventing or minimizing malalignment when using
an intramedullary nail for extra-articular proximal diaphyseal
fractures will be discussed in the authors’ preferred method of
treatment section.

In an effort to evaluate operative techniques in the management of extra-articular proximal tibial fractures, Bhandari et al.²⁴
conducted a systematic review of the literature. At the time of their
review, there were three prospective case series evaluating
intramedullary nailing, three prospective case series evaluating
plates, and 14 retrospective case series evaluating plates, nails, or
external fixation. Based on these studies, there was weak evidence
(Grade C) to help guide treatment. Intramedullary nailing showed a
trend towards lower infection rates at 2.5% (95% CI, 0.1% to 3%)
compared to plating at 14% (95% CI, 8% to 23%) and external fixation at
8% (95% CI, 4% to 15%) (Table 55-6).²⁴
The rates of nonunion and malunion were similar in all groups, and all
groups had overlapping confidence limits. Nonunion rates for
intramedullary nailing were 3.5% (95% CI, 1.7% to 7%) compared to 2%
(95% CI, 0.3% to 8%) for plating and 8% (95% CI, 4% to 15%) for
external fixation.²⁴ Malunion rates
for intramedullary nailing were 20% (95% CI, 1.5% to 26% as compared to
10% (95% CI, 5% to 18%) for plating and 4% (95% CI, 1.5% to 10%) for
external fixation.²⁴

Intramedullary nailing of distal diametaphyseal
fractures often presents the surgeon with some technical difficulties.
This is due to the anatomy of the distal metaphyseal flare, the
proximity of the fracture to the joint, and the technical difficulties
associated with distal nail fixation. There is also a known association
of posterior malleolar fractures with distal tibial spiral fractures (Fig. 55-12).³⁹,¹⁶¹ This was initially described by Bostman⁴⁴ who noted a 0.6% prevalence of this fracture. However, this association has been reported to vary from 25% to 48%.⁸⁷,⁸⁸,⁸⁹

Im and Tae¹³⁷
reported on a randomized controlled trial of 64 consecutive distal
diametaphyseal fractures treated with either an intramedullary nail or
plate fixation and found that union rates were similar between groups
(mean 18 weeks for intramedullary nail and 20 weeks for plate fixation)
and that the Olerud and Molander functional ankle scores were very
similar between the groups at 2 years. However, they did find that the
intramedullary nail group had increased ankle dorsiflexion and the
plate fixation group had six superficial infections and one deep
infection as compared to one superficial infection in the
intramedullary nail group (p = 0.03) (level of evidence: 1).¹³⁷ More recently, Nork et al.²⁰⁹ reviewed their series of 36

tibial fractures, involving the distal 5 cm of the tibia, treated with
a reamed intramedullary nail (level of evidence: 4). They assessed
functional outcomes at 1 to 2 years and showed that there is continuing
improvement up to 2 years after the operation, but patients were still
worse by approximately one standard deviation as compared to population
norms when using the Musculoskeletal Functional Assessment Score.²⁰⁹
The advent of newer nail designs with very distal cross locks and
orthogonal cross locking screws, including two lateral and one
anteroposterior screws, help with distal fixation. In addition, close
attention to nailing technique including accurate guide wire placement
in the center of the plafond and close scrutiny of the ankle joint for
intra-articular fracture lines may help prevent previous problems seen
with intramedullary nailing of distal metaphyseal fractures. Recently,
Egol et al.⁸⁶ have suggested that
adjunctive fibular plating in distal tibial fractures treated with an
intramedullary nail may maintain fracture alignment better than those
treated with only an intramedullary nail (level of evidence: 3).
Biomechanically, the addition of a fibular plate has been shown to
increase the resistance to torsional forces.¹⁹⁹

Oh et al.²¹⁴ have
reported on percutaneous plating techniques in unstable tibial shaft
fractures and found that of 24 unstable tibial fractures, 22 united
without secondary surgery (level of evidence: 4). They used a narrow
limited contact dynamic compression plate and had one significant
rotational malalignment of more than 10 degrees and no angular
deformity of more than 5 degrees of varus or valgus.²¹⁴ Maffulli et al.¹⁷⁸
also looked at 20 of their patients in whom a percutaneous plating
technique was used (level of evidence: 4). They found that with
traditional nonlocking plates, the majority of patients had excellent
and good results, although seven patients reported stiffness around the
ankle. They also reported few soft tissue complications.¹⁷⁸

A systematic review of 1125 fractures was undertaken by Zelle et al.³⁰²
They identified only retrospective observational studies, and of those
that were included, they recorded a nonunion rate in those treated with
an intramedullary nail of 5.5% (95% CI, 3.78% to 8.1%) whereas in those
patients treated with a plate, the rate was 5.2% (95% CI, 2.4% to
10.9%). In the nonoperatively managed group, the nonunion rate was 1.3%
(95% CI, 0.7% to 2.7%) (Table 55-7). The
malunion rate was similar in all three groups, ranging from 13.1% (95%
CI, 8% to 20.8%) in those treated with a plate to 16.2% in the
intramedullary nail group (95% CI, 16.0% to 20%). They suggest that the
inferences that can be made from these observational studies are
limited due to patient numbers. Further large scale randomized
controlled trials would be necessary to help ascertain which technique
is preferred. Indeed these trials are currently ongoing.

While the use of external fixation has been declining in
the operative treatment of closed tibial shaft fractures, there is
still a significant proportion of surgeons who opt to use external
fixation in the management of the severely traumatized limb. This is
usually in open tibial shaft fractures, fractures associated with
compartment syndrome, or in the polytraumatized patient who is deemed
too hemodynamically unstable to undergo definitive fixation (Fig. 55-13).
External fixation frames in tibial shaft fractures can take the form of
uniplanar monoaxial frames that use Schanz pins above and below the
fracture site, multiplanar external frames that use configurations that
place the Schanz pins in different planes, which are usually
orthogonal, and the use of circular tensioned fine wire external
fixation frames, or a combination of both tensioned fine wire circular
frames and half pin fixation. This is often referred to as hybrid
external fixation.

Much biomechanic work has been undertaken on uniplanar,
multiplanar, and Ilizarov type ring fixators. Many factors such as
number of pins proximal and distal to the fracture site, the distance
of the pins from the fracture site, the use of double stacking cross
bars and multiplanar cross bars, and the inclusion of one, two, or
three rings proximal and distal to the fracture site all have a bearing
on fracture stability when assessed biomechanically.⁷⁰,⁷⁶,¹⁰⁰,¹⁵⁵,³⁰⁴
However, it is not known which configuration provides the optimal
stability for fracture healing. Arguably, the most important detail in
this regard is the quality of fracture reduction and amount of cortical
apposition irrespective of the fixator used.

Bhandari et al.²⁷
have conducted a systematic review and meta-analysis of the literature
regarding the management of open fractures. They identified comparative
studies which had evaluated clinical outcomes following external
fixation, plating and external fixation, and unreamed intramedullary
nailing of open tibial shaft fractures. One quasirandomized trial by
Bach and Hansen²⁰ compared plate fixation with external fixation. They found a statistically significant reduction in the rates of

reoperation with external fixation compared to plate fixation (6.7% compared to 50%).²⁰
However, all of the fractures healed in their study and the
meta-analysis by Bhandari et al. suggested that external fixation did
not significantly alter the risk of nonunion (RR = 0.52, 95% CI,
0.21%-1.28%), deep infection (RR = 0.39, 95% CI, 0.13%-1.11%), failure
of fixation (RR = 0.58, 95% CI, 0.1%-3.2%), or malunion (RR = 2.6, 95%
CI, 0.29%-23.5%). The trend, however, was for increased nonunion, deep
infection or fixation failure in the plate group, and a trend toward
increased malunion in the external fixation group.²⁰,²⁷
In a comparison of unreamed nailing and external fixation, the pooled
estimate of effect favored the use of unreamed nails over external
fixators in the management of open fractures, with the use of unreamed
nails significantly reducing the risk of reoperation (RR = 0.51, 95%
CI, 0.37%-0.69%) (Table 55-8). The use of an
unreamed nail reduced the risk of reoperation by 49% when compared with
external fixation. This means that in low-risk patients, an orthopaedic
surgeon would need to treat 41 patients with an unreamed nail instead
of an external fixator to avoid a single reoperation. In high-risk
patients, surgeons would have to treat three patients with an unreamed
nail instead of an external fixator to avoid a single reoperation.²⁷
This provides strong evidence that unreamed nails provide benefit in
the treatment of open tibial shaft fractures when compared with
external fixation.

In a more recent study of the treatment of type IIIA
open fractures of the tibial shaft, patients were randomized to receive
either an Ilizarov external fixator (32 patients) or an unreamed tibial
nail (29 patients).¹³⁸ This was a
pseudorandomized study that had no concealment of allocation and for
this reason would be considered level 2 evidence. They noted a
statistically significant shorter time to healing in the Ilizarov
external fixation group (19 weeks) as compared to the unreamed tibial
nail group (21 weeks) (p = 0.039). They
found a higher need for reoperation to obtain bony union in the
unreamed tibial nail group but also noted a significant incidence of
pin tract infections and joint contracture with the Ilizarov fixator.¹³⁸
There were no differences in the rates of malunion. They suggest that
the use of either of these methods of fixation should be made on a case
by case basis.¹³⁸

In a systematic review of open reduction and plate fixation of open tibial shaft fractures, Giannoudis et al.¹⁰⁴
identified 417 citations with their search strategy and 11 of these met
their inclusion criteria. These 11 studies reported on 492 tibial shaft
fractures. All except for one were retrospective case series, the one
exception being a quasirandomized trial. They found that the union rate
ranged from 62% to 95%, the infection rate ranged from 4% to 35%, and
the reoperation rate ranged from 8% to 69%.¹⁰⁴
The majority of the studies used standard 4.5 mm dynamic compression
plates, and one study assessed the use of biologic plate fixation. They
suggest that plating may be a viable option in the treatment of open
fractures, although this would have to be tested in a large randomized
controlled trial.

As we have noted from the surgical preference survey data,²⁶
the preference for intramedullary nailing of open tibial shaft
fractures (specifically type IIIA, B, and C) decreases, and external
fixation is more frequently used for primary treatment in the presence
of significant segmental bone loss and contamination. In damage control
surgery in the severely injured patient, external fixation is often the
treatment of choice because of its rapid application.¹²⁸,²²⁰

A recent systematic review and meta-analysis of
intramedullary nailing following external fixation has reported on 9
studies (n = 268 patients) where there was a planned conversion from an
initial external fixator to an intramedullary nail and 12 studies (n =
236 patients) where they reported on the use of intramedullary nailing
as a reconstructive procedure. The protocols in these studies usually
suggested removing the external fixation with curettage, débridement,
and irrigation of pin sites and adjunctive antibiotic coverage.

Data from the studies suggests that decreased duration
of external fixation frame results in an 83% reduction in the risk of
infection (p < 0.001) when compared a period of external fixation that exceeds 28 days (Table 55-9).
However, union rates with subsequent intramedullary nailing are up to
90% (95% CI, 88%-93%). Casting following external fixation was found
not to significantly reduce the risk of infection, but it significantly
increased the rate of nonunion.³³

Traumatic bone loss is most commonly seen in the tibia because of its subcutaneous location.¹⁴⁷,¹⁷³,²³⁸
While the overall incidence of fractures associated with bone loss is
generally low, the literature suggests that tibial fractures can
account for up to 68% to 79% of these fractures.¹⁴⁷,²⁹⁹ While cases of reimplantation of large segments of bone have been described in adolescents,¹⁸³ this is not advocated in the majority of open fractures.

The OTA classification of bone loss consists of three types, as shown in Figure 55-14.
Type 1 bone loss involves less than 50% of the bone diameter, type 2
involves more than 50% of the bone circumference, and type 3 involves
segmental loss.⁶⁵ Bone loss is not uncommon. It can occur in 11% to 12% of open fractures, and most commonly occurs in the tibia.¹⁴⁷
Published guidelines exist for the various reconstructive options.
However, due to the prevalence of tibial fractures associated with
segmental bone loss, the evidence supporting these guidelines is
retrospective and is in most instances is considered level 4.¹⁴⁷,²³⁸,²⁹⁹ However, Robinson et al.²³⁸
suggest that critical size defects less than 6 cm are amenable to bone
grafting techniques, and those more than 6 cm require other
reconstructive options. Thus, treatment for segmental bone loss must
necessarily be individualized and be based on patient’s values and
preferences as well as on surgical resources and technical issues. We
will discuss further reconstructive options in the section on
nonunions, however we use critical cut-off values for defects of 1 to 6
cm and defects more than 6 cm. Defects up to 6 cm are treated with an
intramedullary nail, if possible, and a later bone graft procedure is
undertaken after first ensuring adequate soft tissue coverage. With
defects more than 6 cm, careful planning is needed, the options being
nailing, with lengthening over the

nail,
immediate fine-wire tensioned frame application with later bone
transport, or immediate shortening and subsequent lengthening or
external fixation with planned free fibula transfer. Patient factors to
be considered in the decision-making process are the patient’s age,
preinjury functional status, comorbid medical conditions such as
diabetes and peripheral vascular disease, smoking status, and
concomitant injuries. Patient values and preferences include the
ability to psychologically deal with potentially multiple operations
including both soft tissue and bony reconstruction, or potentially
coping with months in an external fixation device depending on which
treatment is undertaken. Surgical issues include the condition of both
soft tissue and bone donor sites and the availability of surgical
expertise in using microvascular tissue transfer techniques or complex
fine-wire external fixation frames.

Débridement, including surgical irrigation, is the
crucial first step in the management of open tibial wounds.
Unfortunately, the débridement may result in increasing the amount of
segmental bone loss as devitalized bone must be removed as part of the
procedure. Intramedullary nailing has been put forward as the technique
of choice, when technically possible, for the treatment of tibial
fractures associated with bone loss. The use of unreamed statically
locked nails is supported to an extent by the recent SPRINT randomized
controlled trial.³⁰ Those patients
who had an open fracture treated with an unreamed nail showed a trend
toward fewer outcome events than those who received a reamed
intramedullary nail (RR = 1.27,95% CI, 0.91%-1.78%; p
= 0.16) (level of evidence: 1). Intramedullary nailing confers the
benefits of minimally invasive stabilization with soft tissue
dissection which is essentially limited to the débridement. It also
provides stable internal fixation, allows access to the soft tissues
for wound management and possible free tissue transfer if deemed
necessary, and allows for potential lengthening over the nail with
external fixation if desired.¹⁴⁷
However, the problems associated with initial fixation include
obtaining the correct alignment in the face of significant bone loss.
It may be necessary to template length and rotation from the
contralateral limb or potentially the ipsilateral fibula if there is
enough of it. The use of intramedullary nailing, however, is limited to
mainly

diaphyseal
fractures or metadiaphyseal fractures that allow for placement of
proximal and distal cross-locking screws and usually for defects less
than 6 cm. In the face of large proximal or distal defects that
preclude intramedullary nailing, temporary or definitive external
fixation may become necessary. External fixation in the form of
tensioned fine-wire frames confers the advantage of allowing bone
transport at a later date or facilitating immediate shortening with
later lengthening. However, it increases the complexity of wound
management and potential free tissue transport. There is the risk of
pin site infection and the problems associated with a prolonged period
of frame application that may not be desirable in some patients.
Definitive plate fixation after the initial acute management, while
providing stability to the bone, may necessitate further soft tissue
dissection and may therefore preclude later bone transport and limit
bony reconstruction to bone grafting or free bone tissue transport
techniques.⁶⁵,⁷⁴,¹⁴⁷

A common complication associated with intramedullary tibial nailing is anterior knee pain. Court-Brown et al.⁶⁹
have reported on a series of 169 patients treated with an
intramedullary nail. They found that anterior knee pain was present in
56.2% of patients and seemed to affect younger more than older patients
(level of evidence: 4).⁶⁹ A
significant number of the patients with anterior knee pain had pain
with kneeling (91.8%) and pain at rest (33.7%). They found also found
that the pain resolved or improved with nail removal.⁶⁹

Keating et al.¹⁴⁶
also reported on the prevalence of knee pain following intramedullary
nailing, and in a consecutive series of 110 tibial shaft fractures they
found that 57% of patients developed anterior knee pain (level of
evidence: 4). They also found no significant correlation between
proximal protrusion of the nail and knee pain, but they suggested that
there might be an association with a patellar tendon splitting approach
as compared to a parapatellar insertion with a higher number of
patients (77%) having pain with the patellar tendon splitting approach
as compared to the paratendinous approach (50%). They found that 80% of
their patients required nail removal and that the majority had their
pain either completely or partially relieved.¹⁴⁶

In a prospective randomized trial, Toivanen et al.²⁶⁹
randomized 50 patients to receive an intramedullary nail through either
a paratendinous or transtendinous approach. They then followed these
patients for an average of 3 years (level of evidence: 1).²⁶⁹
They found no significant difference in anterior knee pain between the
transtendinous to the paratendinous approach (RR = 1.06, 95% CI,
0.81%-1.39%); however, they reported a high incidence of anterior knee
pain with both techniques (67% with the transtendinous approach and 71%
with the paratendonous approach).²⁶⁹

In a study using ultrasound to assess the patellar tendons of patients who had had an intramedullary nail, Vaisto et al.²⁸¹
compared those with anterior knee pain and those without. They found no
significant differences in the ultrasonographic appearance of the
tendon and specifically found no differences in patellar tendon blood
circulation, calcification of the patellar tendon, or granulation
tissue.²⁸¹ This suggests that the etiology of anterior knee pain following intramedullary nailing is still unclear.²⁸¹
A follow-up study by the same authors found that anterior knee pain
symptoms in a significant number of patients may disappear between 3
and 8 years.²⁷⁹ Interestingly, those
patients in whom anterior knee pain was still present at 8 years had a
lower functional knee score and quadriceps weakness (level of evidence:
4).²⁸⁰ Bhattacharyya et al.³⁴
assessed the effect of nail prominence on anterior knee pain and
suggested that anterior nail prominence was associated with pain at
rest, while superior nail prominence was associated with pain while
kneeling and walking (level of evidence: 4). They suggested placing the
nail just deep to the bone.³⁴

The etiology of anterior knee pain following
intramedullary nailing is as yet unclear; however, it is seen to
decrease in the majority of patients up to 8 years following the
procedure.

The most common neurologic injury following intramedullary tibial nailing is injury to the peroneal nerve. Koval et al.¹⁵²
documented the prevalence of neurologic injury to be approximately 30%
in a retrospective review of 60 patients treated with a reamed
intramedullary nail, but they stated that the majority were minor
sensory neurapraxias with 89% of these being transient and resolving
within 3 to 6 months. However, 2 patients in their series continued to
have a nerve deficit at 1-year following the procedure (level of
evidence: 4).¹⁵²

In a prospective series of 208 patients with a tibial
shaft fracture treated with a reamed intramedullary nail, 5.3% were
seen to develop peroneal nerve dysfunction postoperatively.²³⁹
The mean age of patients in whom this developed was 25.6 years and most
of the patients had had a closed fracture caused by a varus force with
no evidence of compartment syndrome.²³⁹
The majority showed good recovery after 3 to 4 months, but 3 of 11
patients still had residual symptoms resulting in weakness of their
extensor hallucis longus.²³⁹ Williams²⁹⁴
also found that 19% of patients with tibial fractures treated with a
reamed intramedullary nail had peroneal nerve lesions. It has also been
suggested

that oblique proximal cross screws can damage the peroneal nerve.⁸⁴

It is important to maintain good surgical technique when
inserting oblique proximal cross screws and to avoid placing the drill
and screws too far through the distal cortex. Fluoroscopic imaging can
be used to check that the screws are not placed close to the fibular
neck.⁸⁴

Leunig and Hertel¹⁷⁰
described thermal necrosis in 3 patients, all of whom had narrow
intramedullary canals (level of evidence: 4). They found that after
reamed intramedullary nailing, these patients developed pretibial
blistering and further went on to develop osteomyelitis. They
attributed this to thermal necrosis from tight reaming of the canal.
Thermal necrosis has also been implicated in the development of
postoperative infection following intramedullary nailing in another
series of reamed nails.²²⁷

Giannoudis et al.¹⁰²
have assessed thermal necrosis in vivo in adult subjects. They used
temperature probes while reaming and found that in canals that were
either 9, 10, or 11 mm in diameter, the amount of reaming did not
produce a clinical problem. However, they did note a significant rise
in temperature in those canals that were 8 mm in diameter once they
were reamed to at least 10 mm.¹⁰² To
minimize the risk of thermal necrosis, it is suggested that a
tourniquet should not be used and reaming should not be forceful.

In an international survey of the management of open fracture wounds, Petrisor et al.²²⁸
surveyed 1764 surgeons, 984 of whom responded. About 70% of surgeons
favoured irrigation with normal saline, whereas bacitracin was
routinely added to the saline by 16.8% of surgeons. The majority of
surgeons (71%) used low pressure when delivering irrigation solutions
to the wound, but there was, in fact, variation in surgeon opinions as
to what constituted low or high pressure.

In a randomized controlled trial, Anglen⁸
enrolled 400 patients with 458 open fractures and randomized them to
either receive a Castile soap solution or normal saline with
bacitracin. He found no significant difference between these irrigation
solutions, noting that the infection risk with soap was 13% compared
with 18% for bacitracin solution (p = 0.2).⁸
The confidence intervals were wide, but the point estimate suggested a
trend toward soap solution being better in reducing the risk of
infection (RR reduction of 28%) (level of evidence: 1).⁸

With regard to irrigating pressures, experimental data
suggests that high pressure lavage is effective at removing debris and
bacteria from contaminated open wounds. However, this may be achieved
at the expense of potentially further damaging the bone ends, as well
as displacing bacteria further into the intramedullary canal.²²,²³,¹⁶⁶
There is also some experimental data to suggest that there are effects
at the cellular level from high pressure lavage. Callous strength in a
rat femur fracture model is less in those treated with high pressure
lavage. In addition, high pressure promotes stem cell differentiation
towards an adipocyte cell type rather than osteoblast cell type.³¹

Currently, a large multicenter international randomized
controlled trial (fluid lavage in open fracture wounds [FLOW])
assessing open fracture irrigation techniques is underway. It is
comparing normal saline solution with Castile soap as well as comparing
high pressure lavage, low pressure lavage, and bulb syringe pressure in
a 3 × 2 factorial design.

Systemic Antibiotics. A systematic review and
meta-analysis of the use of antibiotics for preventing infection in
open limb fractures has been done by the Cochrane group.¹⁰⁵
They reviewed seven studies that provided data on 913 patients. They
found a reduction in the rates of early infection with the use of
antibiotics (RR = 0.41, 95% CI, 0.27%-0.63%). This translates to an
absolute risk reduction of 0.08 (95% CI, 0.04%-0.12%) and a number
needed to treat of 13 (95% CI, 8%-25%). This means that for every 13
patients treated with antibiotics, 1 infection can be prevented. The
number could be as low as one for every 8 patients or as high as one
for every 25 patients. Even though these studies were of low quality
and lacked the power to find significant differences, this represents
the highest level of evidence available. Unfortunately, there was no
clear consensus as

to which type of antibiotic to use, which route to use, or for how long.

Local Antibiotics. The technique of local antibiotic
delivery using antibiotic impregnated cement beads has been developed
as an adjunct to systemic antibiotic therapy. Experimental data
suggests that antibiotic beads impart high local levels of antibiotic
to the bone and surrounding tissues without significantly affecting
systemic antibiotic levels.²¹,¹²⁷,²⁸⁶ Henry et al.¹²⁷
retrospectively reviewed 404 open fractures in 339 patients who
presented with all types of open injuries, with 30.7% being Gustilo
type III. They found a 2.7% incidence of acute wound infection with
antibiotic beads and systemic antibiotics compared with 11.4% with
systemic antibiotics alone (level of evidence: 4).¹²⁷ A further study by Ostermann et al.²¹⁶
compared 157 open fractures treated with prophylactic antibiotics to
547 open fractures treated with antibiotic bead pouches in addition to
prophylactic systemic antibiotics (level of evidence: 4). They found a
4.2% prevalence of acute or chronic infection compared to 17% with
prophylactic antibiotics alone.²¹⁶ Keating et al.¹⁴⁵
reviewed their series of 81 open tibial fractures treated with
intramedullary nails (level of evidence: 4). Their protocol did not
include the use of antibiotic bead pouches as they were introduced
later in their study. The found a 4% prevalence of deep infection after
this protocol change compared with 16% before it.¹⁴⁵

There is no clear consensus on the appropriate timing
for the closure of open tibial fracture wounds. Immediate closure of
open tibial wounds has been shown to result in a higher incidence of
infection and possibly nonunion.²⁴⁵ Russell et al.²⁴⁵
reviewed all types of open tibial injuries (41% type 1, 39% type 2, and
20% type 3) and found the deep infection rate to be 20% after primary
closure and 3% after delayed closure (level of evidence: 4). However,
it may be that some conditions may allow for primary closure as
discussed by Rajasekaran.²³⁴ These
include an adequate débridement performed within 12 hours, no
significant skin loss due to either injury or débridement, skin closure
without tension, no farmyard contamination, and no vascular
insufficiency. This approach has been substantiated by DeLong et al.,⁸⁰ and more recently by Hohmann et al.,¹²⁹
who suggest that primary wound closure may be safe in selected cases
which include low-energy uncomplicated open fracture wounds. The
adequacy of the débridement and skin perfusion and the importance
tensionless skin apposition cannot be overstated when considering a
wound for primary closure. Advantages of this approach include possible
prevention of nosocomial infection and decreased length of hospital
stay.¹²⁹

With regard to the timing of delayed closure, an
international survey of orthopaedic surgeons suggested that 94% of
surgeons believed that open fracture wounds should be closed in less
than 7days.²⁹ Breugem et al.⁴⁹
have conducted a systematic review of the literature in an effort to
determine if there are evidence based guidelines for the timing of
wound closure. They identified 10 studies for inclusion, each with
methodologic flaws, with a mixture of retrospective and prospective
studies and because of this were unable to pool any data. They suggest
that if the reviewed literature indicates that early aggressive
débridement and closure of the open fracture wound between 3 to 5 days
reduces the risk of osteomyelitis and delayed union. They did not,
however, present an estimate of this risk because of nonpooling of the
data. They also say that a lack of clearly defined parameters regarding
wound grading and outcomes limits the inferences that can be made from
these studies.

The use of negative pressure wound therapy has been
increasing in the management of acute traumatic wounds. Basic science
data presents biologic rationale for the use of negative pressure wound
therapy in that it has been shown to increase granulation tissue,
enhance wound blood flow, and aid in the clearance of some bacterial
species.¹³,¹⁴,²⁰⁰,²⁰¹,²⁰²
Clinical data on negative pressure wound therapy in acute wound
management is increasing; however, the majority of randomized
controlled trials have been conducted on chronic wounds. A recent
systematic review and meta-analysis of randomized controlled trials has
been carried out by Gregor et al.¹⁰⁸
It identified 17 published studies for inclusion as well as 19
unpublished trials, of which 5 were stopped early. Of the 17 included
studies in the review, 7 were randomized controlled trials, which
included 1 study on acute open wounds, and 10 were nonrandomized
trials. Of the 19 ongoing trials, 2 were related to high-risk fractures
and open fractures.¹⁰⁸,²⁶⁰,²⁶¹
Their analysis showed that the pooled estimate of effect on changes in
wound size favored negative pressure wound therapy (standardized mean
difference of -1.40; 95% CI, -2.07 to -0.054).¹⁰⁸
Since most of these trials were for chronic wound closure, some
extrapolation to the acute setting is necessary. However, given that
initial basic science data provides a biologic rationale and clinical
evidence suggests a benefit as regards wound size, the continued use of
negative pressure wound therapy is promising. Further randomized trials
in the setting of acute trauma, measuring both wound closure and
functional outcome, are necessary to help delineate the indications for
its use. However, we have observed a decrease

in
the use of free flaps for open wounds in their institution secondary to
the use of negative pressure wound therapy. Once an appropriate
granulation bed of tissue has been created, less aggressive plastic
surgical techniques, such as split thickness skin grafting or full
thickness grafting, may be employed.

While the use of negative pressure wound therapy may be
increasing, there are still wounds where it is inappropriate to use
this technique. These include those with exposed bone and tendon; these
being most commonly seen at the distal aspect of the tibia where
rotational flap muscle coverage is not generally feasible. Demographic
data from the Lower Extremity Assessment Project (LEAP) concerning
those patients who required soft tissue showed that the patients were
commonly young, healthy males who were involved in a motor vehicle or
motorcycle accident.²³² In this
cohort, free flaps were more commonly used for distal tibial fractures.
The authors suggested that the use of a free flap as compared to a
rotational flap was less likely to result in wound complications that
may require further operative intervention.²³²
Free muscle flaps have been used in the past for reconstruction,
although, recently fasciocutaneous flaps have become more popular.
Yazar et al.³⁰⁰ found no significant
difference in functional outcomes between the use of free muscle flaps
or fasciocutaneous flaps for reconstructing open distal third tibial
shaft fractures.

Tibial diaphyseal fractures are associated with the development of compartment syndrome.¹⁸⁹,¹⁹² Epidemiologic data by McQueen et al.¹⁹²
showed that in 164 patients with acute compartment syndrome presenting
to the Edinburgh Trauma Unit, 69% of these were associated with a
fracture, and in half of those, the fracture was in the tibial
diaphysis. Compartment syndrome was more common in younger men under
the age of 35, and the highest incidence was observed in adolescents.¹⁹²
They also found that the overall prevalence of compartment syndrome
following tibial shaft fractures was 4.3%.192 Further literature
suggests that the prevalence of compartment syndrome associated with a
tibial shaft fracture may range from 1% to 10%.⁴⁷,¹⁹²,²⁷⁴ Compartment syndrome is discussed in detail in Chapter 27.

Many studies have addressed the issue of diagnosis in
compartment syndrome. Diagnostic criteria include clinical parameters
such as pain, pain with passive stretch of muscles within the
compartment, and paresthesiae, while other criteria are based on
measurements such as the actual compartment pressure or the ΔP value,
this being the diastolic pressure minus the compartment pressure. There
has been some discussion about the value of these diagnostic
parameters. Ulmer²⁷⁴ undertook a
systematic review of diagnostic studies and identified 104 articles
from the literature which addressed the diagnosis of compartment
syndrome. Because of insufficient data and retrospective study designs,
which have an inherent risk of introducing bias, only four studies were
eligible for inclusion, and these studies provided the data to
calculate positive and negative predictive

values.²⁷⁴ Ulmer²⁷⁴
found that the clinical signs of compartment syndrome, namely pain,
pain with passive stretch, paresthesiae, and paresis had a low
sensitivity and high specificity. That is if compartment syndrome was
not present, there was a significant chance that the signs were not
present. He went on to use likelihood ratios to determine the posttest
probabilities (using odds ratios) of having compartment syndrome with
one, two, three, or four of these clinical signs. Using a prevalence
rate of 5% for compartment syndrome associated with a tibial shaft
fracture, this being the median value from the collated study data, he
determined that with one clinical sign the probability of compartment
syndrome is approximately 25%, with two clinical signs it rises to 68%,
and with three and four clinical signs, the probability of compartment
syndrome rises to 93% and 98%, respectively.²⁷⁴

It is also impossible to diagnose compartment syndrome
clinically in unconscious patients or in patients on ventilators. It
may also be difficult to make the diagnosis in alcoholics and drug
addicts.¹⁸⁹ A good argument can be
made for monitoring those patients who are potentially at risk for
developing compartment syndrome or who are not able to respond
appropriately to pain or clinical examination.¹⁸⁹,²⁴¹
Monitoring of compartments can be done either continuously or as a
single measurement. There are commercially available products that
allow for single use compartment pressure monitoring with a digital
manometer and also allow for indwelling catheters to be placed. Other
manometers can be used including an arterial line transducer as well as
Whitesides’ technique, which uses a mercury manometer.²⁹¹ Examples of pressure measurement devices are shown in Figure 55-38.

Different types of insertion catheters include side-port needles, slit catheters, and regular straight needles.¹⁹¹,²⁴¹,²⁹⁰ Boody and Wongworawat³⁸
have assessed three needle configurations and three pressure measuring
devices. They suggest that slit catheters and needles with a side-port
are more accurate than a straight needle, and that an arterial line
transducer and the Stryker device are more accurate than the Whitesides
technique.³⁸ The most accurate
device in their study was the arterial line manometer with a slit
catheter. They also found that straight needles tended to overestimate
the pressure measurements.³⁸ These results are comparable to those found by Moed and Thorderson,¹⁹⁸
who reported that straight needles overestimated compartment pressures
by 18 to 19 mm Hg when compared to slit catheters or side-ported
needles.

The question arises as to whether the absolute pressure or the ΔP should be used to diagnose compartment syndrome.¹⁹¹,²⁹¹
Experimental canine models have suggested that there is an increased
risk of muscle damage with absolute compartment pressure levels that
exceed 30 mm Hg.¹¹⁹ Some authors have recommended values as high as 45 mm Hg as the absolute pressure threshold for fasciotomy.²⁰³ However, Whitesides²⁹¹
suggested that the pressure threshold for fasciotomy should be when the
pressure difference between the compartment pressure and the diastolic
pressure is less than 30 mm Hg, the ΔP. It has been recommended that
the ΔP be used because the use of the absolute measurement of 30 mm Hg
results in an excessively high number of fasciotomies.¹⁹¹,²⁹¹
Indeed, in a prospective study of 116 patients with continuous
monitoring, it was found that in the first 12 hours of monitoring, 53
patients had absolute pressures over 30 mm Hg, 30 patients had
pressures over 40 mm Hg, but only 1 patient had a ΔP less than 30mm Hg,
and he had a fasciotomy (level of evidence: 4).¹⁹¹
Over the next 12 hours of the study, only 2 patients had a ΔP less than
30 mm Hg and had fasciotomies, whereas 35 patients had absolute
pressures over 30 mm Hg with 7 having pressures over 40 mm Hg. None of
the patients in this series developed any sequelae of compartment
syndrome.¹⁸⁹,¹⁹¹ The authors suggested that the use of an absolute decompression threshold of

30 mm Hg would have resulted in a 43% rate of fasciotomy.¹⁹¹
This is not insignificant as a significant proportion of patients will
have long-term sequelae and morbidity from the fasciotomy including
pain, sensation changes around the incision, scar tethering, and muscle
herniation.⁹³ In another study of 25
patients with tibial diaphyseal fractures complicated by acute
compartment syndrome, of which 13 of the patients had undergone
continuous monitoring of compartment pressures, the investigators found
that the delay from injury to subsequent fasciotomy in the monitored
group was 16 hours compared with 32 hours in the nonmonitored group
(level of evidence: 4).¹⁸⁹,¹⁹¹ They recommend continuous compartment monitoring of all patients with tibial shaft fractures.

More recently, Harris et al.¹²¹
randomized 200 extra-articular tibial shaft fractures into monitored
and unmonitored groups for compartment pressure measurement. The
monitored group received continuous compartment pressure monitoring for
36 hours (level of evidence: 1).¹²¹
They had five cases of compartment syndrome in the nonmonitored group
and no cases in the monitored group of patients; however, within the
monitored group of patients, there were 18 patients with a ΔP of less
than 30 mm Hg. None of these patients developed compartment syndrome or
late sequelae. They found that in the awake and alert patient, they
were able to diagnose compartment syndrome in an appropriate time using
clinical signs.¹²¹ They suggest that in the awake and alert patient, continuous compartment pressure monitoring may not be necessary.

However, a strong case can be made for the use of
continuous compartment pressure monitoring in those patients who are of
significant risk of developing compartment syndrome, such as adolescent
males, and those who are unable to communicate pain. This includes
patients with head injury and ventilated patients as well as alcoholics
and drug addicts. One caveat is that the diastolic blood pressure in
patients under anaesthesia may be lower than in the same patients when
measured preoperatively and postoperatively. This was observed in a
prospective cohort study looking at a consecutive series of 242
patients with tibial shaft fractures undergoing intramedullary nailing
(level of evidence: 2).¹⁴³,²⁷¹
The authors state that the intraoperative ΔP may be low when compared
with values recorded when the patient is awake, and it is suggested
that the decision to perform a fasciotomy should be delayed until the
patient is awake and has had serial compartment pressure monitoring.¹⁴³
One other factor to consider when monitoring patients is that the
highest pressures are obtained close to the fracture site, and the
recommendation is to use the highest recorded pressure.¹²⁴

Other methods of diagnosing acute compartment syndrome
have been studied to determine their relative effectiveness. These
include MRI, pulsed phase-locked loop ultrasound, scintigraphy, laser
Doppler flowmetry, and near-infrared spectroscopy.³,¹¹,²⁵²,²⁹³
These modalities all carry the benefits of noninvasive monitoring, but
they have mixed degrees of sensitivity and specificity due to the
traumatic nature of compartment syndrome associated with tibial shaft
fracture. Some techniques also require specialized equipment and
expertise which limits their effectiveness and overall
generalizability. Most of these have been used for chronic exertional
compartment syndrome and while they show promise, their routine use in
the traumatized limb is still to be determined.

Experimental evidence has suggested that intramedullary nailing is associated with an increase in compartment pressure.¹⁹⁷ Some surgeons attribute this to the process of reaming.²⁶⁷
However, clinical evidence suggests that the compartment pressure may
be raised by the reduction of the fracture and passage of the nail as
opposed to differences in reaming technique.¹⁹¹,²⁰⁵ McQueen et al.¹⁹⁰
continuously monitored 67 patients who were undergoing intramedullary
tibial nailing for fracture. They found no significant correlation
between the severity of injury or delay to surgery and the
intraoperative compartment pressures. The overall incidence of
compartment syndrome was 1.5%.¹⁹⁰ Nassif et al.²⁰⁵
conducted a randomized trial comparing the effect of reamed and
unreamed intramedullary nailing on intraoperative compartment
pressures. They found that highest pressures occurred during reaming in
the reamed group and during nail insertion in the unreamed group.
However, there was no significant difference in pressures between the
two groups.²⁰⁵ They also found that
even though ΔP values were greater than 30 mm Hg after nail insertion,
compartment syndrome did not occur in any patient.²⁰⁵ Tornetta and French²⁷¹
have also documented that intracompartmental pressures are raised
during manual traction and nail insertion but that they decreased
postoperatively. They suggest that perioperative continuous monitoring
in this instance is not required.²⁷¹
Data from the SPRINT trial showed no significant difference in the
rates of fasciotomy following reamed and unreamed tibial nailing.³⁰

The clinical signs of an established nonunion include
pain, motion, and swelling at the fracture site. An examination of the
overlying skin and soft tissues should be undertaken looking for any
draining sinuses or other clinical signs of infection as well to
ascertain how well the soft tissues may be able tolerate further
surgery. It is also important to assess the overall alignment of the
leg for angular and rotational malalignment and limb-length
discrepancies. A thorough examination of the knee, ankle, and hindfoot
joints should be carried out to assess range of motion and to look for
contractures. Radiologically, the nonunion may present as a persistent
fracture line or with a lack of bridging callus in of three out of the
four cortices on the anteroposterior and lateral radiographs.²⁸⁹ The radiographs may also show hardware breakage or failure, as shown in Figure 55-41.
Malalignment may also be present particularly if the fracture was
originally treated nonoperatively. Further imaging with CT scans or
bone scans may also be useful.

The literature suggests that there are a number of
factors that help to predict nonunion. These are fracture morphology,
mechanism of injury, degree of soft tissue injury, diabetes, alcohol
consumption, smoking history, surgical delay, the use of steroids or
anti-inflammatory medication, vasculopathy, age, and gender.²¹⁵,²⁴⁷,²⁷³ In an observational study to determine predictors of reoperation, Bhandari et al.³²
identified 200 patients with tibial shaft fractures and assessed 20
possible prognostic factors. They used both univariate and multivariate
regression modeling techniques and suggested that three variables
predicted the risk of reoperation. These were the presence of an open
fracture wound (RR = 4.32, 95% CI, 1.76%-11.26%), lack of cortical
continuity between the fracture ends following fixation (RR = 8.33, 95%
CI, 3.03%-25%), and the presence of a transverse fracture (RR = 20.0,
95% CI, 4.34%-142.86%) (level of evidence: 3) (Table 55-11).
The authors suggest that using these three variables, one can
categorize patients into four different risk groups. These are low
risk, where there are no risk factors and a 4.4% prevalence of
nonunion, moderate risk, where there is one risk factor and an 18%

prevalence,
high risk, with two risk factors and a 47% prevalence, and very high
risk, with three risk factors and a 94% prevalence of requiring
reoperation to promote bone healing. The only modifiable risk factor
suggested in the multivariate analysis was cortical continuity.³²

Further work on predicting nonunions was conducted by
Audige et al., who attempted to construct a path analysis of prognostic
factors that were associated with delayed healing or nonunion.¹⁷
Their path analysis and multivariate model of prognostic factors
determined that the most important factors were the extent of the skin
injury, the presence of a postoperative fracture gap, and a distal
location of the fracture in the shaft.²²²
Interestingly, they identified associations between some of the
prognostic variables. For instance, surgical treatment using either a
nail, external fixator, or plate was potentially influenced by the
fracture type as well as the skin injury.¹⁷
In this path analysis, the skin injury severity was shown to have the
most significant effect on the risk of delayed or nonunion with an open
skin injury of greater than 5 cm in length resulting in a RR of 5.7
(95% CI, 3.0%-10.9%) or approximately a sixfold increase in the risk of
delayed union or nonunion when compared with no skin injury with a RR
of 1.0. It is interesting to note that a closed skin injury resulted in
an approximately threefold increase in the risk of nonunion (RR = 2.8,
95% CI, 0.85%-8.9%). This is in keeping with previous work by Gaston et
al.,⁹⁹ which suggests that the
Tscherne classification of closed fractures correlates with functional
outcome, and the presence of an open wound correlates with the time to
union. Thus, from path analysis work there would seem to be range of
predicted risk of nonunion from the lowest predictive risk which would
include no skin injury, a midshaft fracture, and no postoperative
fracture gap, to a 70% risk in patients with a distal tibial shaft
fracture, an open skin injury greater than 5 cm in length, and a
postoperative fracture gap.¹⁷

Independent predictors of nonunion include the extent of
skin and soft tissue damage, a postoperative facture gap, and distal
fractures with a higher-energy fracture morphology.

Aseptic Nonunion.

Nonunions without Bone Loss.
Intramedullary nailing is advocated by many authors for the treatment
of aseptic hypertrophic but otherwise aligned nonunions. Megas et al.¹⁹³
looked at 50 patients with aseptic tibial nonunion treated with a
reamed intramedullary nail (level of evidence: 4). The group consisted
of 36 patients initially treated with external fixation, 6 with a
plate, 1 with a previous intramedullary nail, and 7 with nonoperative
treatment. Sixteen of these patients required opening of the nonunion
site to remove hardware. All of the fractures united within 6 months of
the intramedullary nailing procedure.¹⁹³
The intramedullary nailing in those who had previously been treated
with external fixation was performed 10 days after removal of the
external fixator.¹⁹³ Sledge et al.²⁵⁷
also used reamed intramedullary nailing in 51 patients with nonunion
who were initially treated with a number of different procedures. Ten
patients required bone grafting and 33 were treated with a partial
fibulectomy. Ninety-six percent of their patients united in an average
of 7 months (level of evidence: 4).²⁵⁷ Alho et al.⁶
also analyzed 25 nonunions treated with a reamed intramedullary nail.
All of their nonunions united, but they did require to bone graft 1
patient (level of evidence: 4).

The technique of intramedullary nailing of tibial
nonunions is slightly more demanding as passage of the nail past
through fibrous nonunion site may be difficult. The canal may need to
be opened at the nonunion site with an awl being passed proximally and
distally from the nonunion to allow for the subsequent passage of the
power reamers and the nail. It is important that the initial power
reamer is an end cutting reamer to facilitate passage through the
nonunion. If there is any tibial malalignment, a fibular osteotomy will
probably be required to permit anatomic alignment of the tibia. This is
especially true of distal tibial fractures (Fig. 55-42). It is also important to remove all previous hardware as this may impede placement of the nail.

Exchange Nailing. In a series of 1106 tibial shaft fractures treated with an intramedullary nail, Court-Brown⁶⁶
found a nonunion rate of 4.4% in closed tibial shaft fractures. In open
tibial fractures, the nonunion rate varied with the Gustilo grade, but
it was noted to be as high as 38% in Gustilo IIIB fractures. Most of
the nonunions were treated by exchange nailing, which is the practice
of removing the previously placed intramedullary nail, overreaming the
canal, and placing a larger nail into the tibial shaft.⁶⁶
Usually, the nail that is placed is 1 to 2 mm larger in diameter than
the previously placed intramedullary nail, depending on the size of the
original nail. Cross screws are rarely required as the fibrous union
provides stability, but if they are used, the nail is usually locked
dynamically with a screw being in the bone end of the nail closest to
the nonunion.

Court-Brown⁶⁶ found
that exchange nailing was successful in 89.5% of nonunions in closed
tibial shaft fractures (level of evidence: 4). In a earlier study,
Court-Brown et al.⁷¹ had looked at
33 tibial nonunions treated with an exchange nail and found that
approximately 88% united after one exchange nail, and the rest united
after a second exchange nailing procedure (level of evidence: 4).
Complications after exchange nailing included a higher incidence of
wound infection (12.1%) when compared with primary nailing, with 1
patient developing deep infection.⁷¹ Templeman et al.²⁶⁶
reported an 11% infection rate in their series of 28 patients with
aseptic non-union treated with exchange nailing. Two of their three
infections occurred in patients who had had Gustilo IIIB fractures
(level of evidence: 4).²⁶⁶ However,
they had a 93% union rate after the exchange nail with the other two
fractures healing after a second bone graft procedure. Analysis of
these two studies shows that the contraindications to exchange nailing
include bone loss over 2 cm in length and involving more than 50% of
the circumference of the tibia and active infection.⁷¹,²⁶⁶

A treatment algorithm has been put forward by Keating et al.¹⁴⁷ and by Robinson et al.²³⁸
These studies suggest that a 6 cm bone is the practical limit for the
use of autogenous bone graft techniques. However, it has been shown
that defects of less than 2 cm associated with less than 50%
circumferential bone may heal with exchange nailing.⁷¹
Broadly, we are in agreement with these view but we will bone graft
defects greater than 1 cm, and we reserve exchange nailing for those
defects less than 1 cm. Defects of less than 4 to 6 cm are amenable to
treatment with autogenous bone grafting after the wound has healed, and
there is no clinical, biochemical, or radiographic evidence of
infection.

Recent work on osteoinductive proteins, such as bone
morphogenetic protein (BMP)-2 or BMP-7, suggests that there may be role
for their use as inductive agents. However, they require the use of a
conductive carrier, usually an allograft, and have in some instances
shown equivalency but not superiority to autograft.¹⁴² Jones et al.¹⁴²
compared autogenous bone grafting and BMP-2/freeze-dried cancellous
allograft for tibial defects that averaged 4 cm in length. They
assessed 30 patients and found that 10 patients treated with autograft
and 13 patients treated with BMP-2/allograft healed without any further
interventions. They suggest that BMP-2/allograft composite is both safe
and effective for treating this size of tibial defect (level of
evidence: 2).¹⁴² However the cost-effectiveness of the use of BMPs in healing segmental tibial defects is still being debated.⁷

When defects are larger than 6 cm, individualized
treatment plans are necessary. The main treatment choices include the
use of repeated bone graft or vascularized bone transport techniques
compared to techniques incorporating the Ilizarov or other subsequent
lengthening, or bone transport techniques either with the frame or over
a nail (Figs. 55-43 and 55-44).¹⁹,⁷⁷,¹⁴⁷,¹⁸⁸,²⁹⁹

Cierny and Zorn⁵⁹
assessed 44 patients in a nonrandomized fashion who had tibial
segmental defects from 4 cm to 12 cm. Twenty-one patients were managed
with Ilizarov distraction, and 23 patients were managed with massive
cancellous grafts and tissue transfers. They found consolidation in 71%
of the Ilizarov group and 74% in the bone graft group; however, the
major complication rate in the bone graft group was 60% compared to 33%
in the Ilizarov group.⁵⁹ They also
found that the Ilizarov technique was better for compromised hosts and
was less expensive. However, they comment that early patient
independence and the need for minimal patient cooperation makes
“conventional” treatment more appealing with defects less than 6 cm.⁵⁹ Paley and Maar²²¹
assessed their series of 19 patients, with a mean tibial defect of 10
cm, treated with Ilizarov bone transport techniques. Union was achieved
in all patients, but 10 patients required bone grafting or débridement
of the bone ends at the docking site and the average time in the
Ilizarov frame was 16 months.²²¹ Similar results using the Ilizarov technique have been found by other authors.²,³⁶

A number of authors have attempted to classify infected tibial nonunions. Jain and Sinha¹⁴⁰
categorized them into two groups. Type A being infected nonunions of
the long bone with nondraining or quiescent infection with or without
an implant.¹⁴⁰ Type B cases were an infected nonunion of the long bone with a draining or active infection.¹⁴⁰
They then subdivided both these categories based on the length of the
fracture gap. Subtype 1 was a nonunion with a bone gap smaller than 4
cm and subtype 2 was a nonunion with a bone gap longer than 4 cm.¹⁴⁰

Court-Brown et al.⁷²
reviewed 459 tibial shaft fractures treated by reamed nailing and
classified infection into three groups. Group 1 patients were those who
presented early with pain, erythema, and swelling at the fracture site
but no evidence of a pyogenic collection.⁷² Group 2 consisted of patients who presented early but had evidence of pyogenic collection.⁷² Group 3 included patients who presented late and had developed a pyogenic discharge.⁷² Zych and Hutson³⁰⁵
also separated patients with infection into three subgroups: acute,
subacute, and chronic phases. They defined the acute phase as being
within 2 weeks of injury. Patients presented with pain and fever, but
radiographs usually showed no changes. In the subacute group, the
patients presented more than 2 weeks after nailing, and they often
presented with a pyogenic collection. Radiologic changes were common.
In the chronic group, patients presented with chronic infection. Zych
and Hutson³⁰⁵ subdivided this group
into those who had a persistent nonunion and those in whom the nonunion
had united but the infection persisted.

The diagnosis of infected nonunion is often based on
clinical grounds. Patients often present with erythema either around
skin incisions or related to underlying hardware. There may also be
soft tissue fluctuance or a draining sinus. It is important to ask
about constitutional symptoms such as fevers, chills, sweats, and
general malaise. Current imaging, in the form of nuclear medicine
scanning, is usually undertaken. Two scans are often used: the
technetium labelled phosphate bone scan and the indium-111-labelled
white cell scan. Other imaging techniques that are useful in the
diagnosis of infection are discussed in Chapter 24.

It has been shown that the combination of technetium-99m
methylene diphosphonate bone scans and indium-111-labelled white cell
scans have a high sensitivity and specificity for diagnosing infection
in the presence of fracture nonunion. Indeed, they have been shown to
range from sensitivities of 60% to 86% and specificities of 84% to 97%.¹⁴⁸,²⁰⁷
Using these numbers, we can calculate the likelihood ratio, given a
positive test, as being approximately 2.5 to 7. However, there are
situations that may result in a false-positive test and these include
metaphyseal sites of fracture that are adjacent to a joint with
arthritis, instability in the delayed or nonunion site, recent surgery,
as well as conditions such as heterotopic ossification or myositis
ossificans.²⁰⁷,²¹⁹
White cell-labelled imaging studies may also be combined with bone
marrow imaging. It has been suggested that this may increase test
accuracy up to 90%.²¹⁶,²¹⁷
This is due to the fact that fractures, hardware, recent surgery, or
neuropathic joints may lead to alterations in the bone marrow.²¹⁷,²¹⁸
Although these tests are not necessary in every case of infected tibial
nonunion, they may be valuable diagnostic aids when the clinical or
radiographic signs are not clear.

The risk of infection increases in relation to the
severity of soft tissue damage. Some of the potential risk factors for
developing infection following intramedullary nailing include
inappropriate fasciotomy closure, exchange nailing, and thermal
necrosis.⁶⁶,²²⁷ Marginal necrosis of skin flaps after plastic surgical reconstruction of soft tissue defects may also cause bone infection.²²⁷ Jain and Sinha,¹⁴⁰
based on their classification, recommend single stage débridement with
bone grafting and fracture stabilization for type A1 infected nonunions
and a staged protocol for type B1 infected nonunions. They suggest
distraction osteogenesis is the procedure of choice for both A2 and B2
nonunions.¹⁴⁰

Court-Brown et al.,⁷²
from their classification system, recommend treating Group 1 patients
with high dose intravenous antibiotics followed by oral antibiotics for
6 to 8 weeks. In Group 2 patients, treatment with exchange nailing with
reaming of the intramedullary canal was undertaken. Union occurred in
these patients, but was prolonged compared to the Group 1 patients. The
Group 3 patients required extensive débridement as well as
reconstruction of the bone and soft tissues.⁷² Zych and Hutson³⁰⁵
suggest that acute phase patients are treated with appropriate
antibiotics. Their subacute phase patients required surgical
débridement with exchange nailing, and their chronic phase patients
generally required bone and soft tissue débridement with subsequent
bone and soft tissue reconstruction. The most common pathogen in their
series was Staphylococcus aureus, which occurred in 64% of patients.³⁰⁵

In an attempt to identify the highest level of evidence in the management of infected long bone nonunions, Struijs et al.²⁶² conducted a systematic review of the literature. They identified

16 case series (level of evidence: 4) that discussed a one stage
revision. Thirteen of these studies included fractures of the tibia.
They also identified 18 case series (level of evidence: 4) that
assessed the use of a two stage revision procedure for infected
nonunion, of which 14 case series discussed fractures of the tibia.²⁶²
The rest of the series and the review discussed fractures of the femur.
Interestingly, both protocols of management resulted in union rates and
persistent infection rates that were similar. With regard to one stage
surgery, there was low quality evidence from observational case series
suggesting a union rate of 70% to 100% and a persistent infection rate
of 0% to 55%. One subgroup of the one stage surgery group included
patients who were treated with débridement and bone transport using an
Ilizarov frame This technique was associated with union in 70% to 100%
and persistent infection in 0% to 55%.²⁶²
In the two stage revision technique, there was low quality evidence
(Grade C) from the observational case series that suggested union rates
of 66% to 100% with persistent infection rates of 0% to 60%. In the
subgroup of débridement and planned secondary bone grafting, union
rates were 75% to 100% with persistent infection from 0% to 60%. The
subgroup of those who received débridement, antibiotic beads, and
planned secondary fixation had union rates of 93% to 100% with
persistent infection in 0% to 18%.²⁶²
They concluded that there were significant limitations in all of the
case series as well as significant heterogeneity between studies which
made it difficult to suggest clear recommendations based on the current
literature.²⁶²