Supracondylar Fractures of the Distal Humerus

All patients with a history of a fall onto an
outstretched hand as well as pain and inability to use the extremity
should have a thorough radiographic evaluation, generally including
anteroposterior (AP) and lateral views of the entire upper extremity.
Comparison views rarely are required by an experienced physician, but
occasionally may be needed to evaluate an ossifying epiphysis. A true
AP view of the distal humerus, rather than of the elbow, allows more
accurate evaluation of the distal humerus and decreases the error in
determining angular malalignment in the distal humerus. The lateral
film should be taken as a true lateral with the humerus held in the
anatomic position and not externally rotated (Fig. 14-12). Oblique views of the distal humerus (Fig. 14-13)
occasionally may be helpful when a supracondylar fracture or occult
condylar fracture is suspected but not seen on standard AP and lateral
views, but should not be routinely ordered to evaluate an elbow injury.

Initial radiographs may be negative except for a posterior fat pad sign. A series by Skaggs et al.¹⁸⁴
of 34 patients with traumatic elbow pain and a posterior fat pad sign
but no visible fracture found that 18 (53%) had a supracondylar humeral
fracture, 9 (26%) had a fracture of the proximal ulna, 4 (12%) had a
fracture of the lateral condyle, and 3 (9%) had a fracture of the
radial neck.

Two main radiographic parameters are used to evaluate
for the presence of a supracondylar fracture. The anterior humeral line
should cross the capitellum through the middle third on a true lateral
of the elbow (Fig. 14-14). In an extension-type
supracondylar fracture, the capitellum is posterior to this line. The
Baumann angle, also referred to as the humeral capitellar angle, is the
angle between the long axis of the humeral shaft and the physeal line
of the lateral condyle (normal range, about

9 to 26 degrees) (Fig. 14-15). A rule of thumb is that a Baumann angle of at least 10 degrees is acceptable; a decrease in the Baumann angle is a sign that a fracture is in varus angulation.

If the AP and lateral views show a displaced type II or
III supracondylar fracture but do not show full detail of the distal
humeral fragment, we usually obtain further x-ray evaluation in the
operating room with the patient anesthetized. Repeat trips for
radiographs evaluation in the emergency setting generally result in
increased pain without significant improvement in radiographic quality.
Detailed radiographs need to be obtained at some point, however, to
define the fracture anatomy with particular emphasis on impaction of
the medial column, supracondylar comminution, and vertical split of the
epiphyseal fragment. T-condylar fractures (Fig. 14-16)
can initially appear to be supracondylar fractures, but these generally
occur in children over 10 years of age in whom supracondylar fractures
are less likely.

In a young child, an epiphyseal separation²¹⁴
can mimic an elbow dislocation. In an epiphyseal separation, the
fracture propagates through the physis without a large metaphyseal
fragment. This fracture occurs in very young children with primarily
chondral epiphyses. On physical examination, the patient appears to
have a supracondylar fracture with gross swelling about the elbow and
marked discomfort. The key to making the diagnosis and differentiating
this injury from an elbow dislocation is the alignment of the
capitellum with the radial head. Sometimes, a supracondylar fracture
with a small metaphyseal fragment can mimic a lateral condylar fracture
(Fig. 14-17). In such cases, more data are
required to initiate treatment. An arthrogram may be helpful to
determine the extent of the elbow injury. In selected patients,
magnetic resonance imaging (MRI) or ultrasonography²¹⁴ may also aid in evaluating the injury to the unossified epiphysis.

For fractures with displacement that require reduction,
initial splinting with the elbow in approximately 20 to 40 degrees of
flexion provides comfort and allows further evaluation. Tight bandaging
or splinting should be avoided, as should excessive flexion or
extension, which may compromise the vascularity of the limb and
increase compartment pressure.¹⁹,¹²⁹
The arm should then be gently elevated. A careful examination of the
neurologic and vascular status is vital in all patients with a
supracondylar fracture, as well as an assessment of the potential for
compartment syndrome. The remainder of the limb should be assessed for
other injuries, and radiographs should include any area that is tender,
swollen, or lacks motion.

Most supracondylar humeral fractures in children can be
treated with closed reduction and pinning. Unless the fracture is open,
a closed reduction should be initially attempted in all fractures,
including type III fractures. With the patient supine and the child’s
arm on a radiolucent arm board, the fracture is first reduced in the
frontal plane and reduction is checked with fluoroscopy. The elbow is
then flexed while the olecranon is pushed anteriorly to correct the
sagittal deformity. Restoration of the Baumann angle (which is
generally >10 degrees) on the AP view, intact medial and lateral
columns on oblique views, and the anterior humeral line passing through
the middle third of the capitellum on the lateral view are indications
of a successful reduction. Because of the amount of rotation present at
the shoulder, some rotational malalignment in the axial plane can be
tolerated at the fracture site, but any rotational malalignment can
compromise fracture stability, so, if present, stability of the
reduction should be carefully evaluated, and a third pin probably
should be used.

Reduction is held with two or three Kirschner wires
(Kwires), as discussed later in this chapter, and the elbow is
immobilized in 50 to 60 degrees of flexion, depending on the amount of
swelling and vascular status. A considerable gap in the fracture site
or an irreducible fracture with a rubbery feeling on attempted
reduction may be signs that the median nerve and/or brachial artery is
trapped in the fracture site and open reduction is indicated (Fig. 14-18).

Clinical and X-Ray Outcomes. Good clinical outcomes in
large number of patients have been reported with both pinning
techniques. Although many of these reports are case series, two

prospective randomized clinical trials have been reported. Kocher et al.¹¹⁰
showed no statistically significant difference between the two
treatment groups in any radiograph or clinical outcome measures in
their prospective randomized clinical trial comparing lateral and
crossed-pinning techniques in the treatment of 52 type III
supracondylar humeral fractures. Another prospective randomized study
by Blanco et al.²⁶ compared 57
patients with type III supracondylar fractures treated with
lateral-entry pinning to 47 treated with crossed-pinning and found no
statistically significant difference in the radiographic outcomes.

Ulnar Nerve Injury. Iatrogenic injury to the ulnar nerve
with use of crossed pins has been reported to be as low as 0%; however,
the two largest series of supracondylar fractures have shown the
prevalence to be 5% (17 of 345) and 6% (19 of 331).³³,⁴¹,⁹⁴,¹²⁷,¹⁶⁵,¹⁶⁹,¹⁸³,²¹¹ Others have reported that these iatrogenic injuries are more frequent.¹⁶⁹,²⁰⁶ In 1977, Arino et al.¹²
recommended the use of two lateral pins to avoid injury to the ulnar
nerve, and a recent meta-analysis seems to support this recommendation:
iatrogenic ulnar nerve injury occurred in 40 of 1171 (3.4%) patients
with medial and lateral crossed pins and in only 5 of 738 (0.7%) of
those with lateral-entry pins. Although iatrogenic ulnar nerve injuries
usually resolve, several permanent iatrogenic ulnar nerve injuries have
been described.¹⁶¹,¹⁶⁵,¹⁸³

In an evaluation of 164 children (328 ulnar nerves), Zaltz et al.²¹¹
identified a group of children who may be at particular risk of ulnar
nerve injury during fracture reduction: those whose ulnar nerves sublux
onto the medial epicondyler and especially those whose nerves dislocate
anterior to the medial epicondyle. In 61% of children younger than 5
years of age, the ulnar nerve migrated over, or even anterior to, the
medial epicondyle when the elbow was flexed more than 90 degrees.²¹¹ Wind et al.,²⁰⁶ however, identified ulnar nerve subluxation in fewer than 2% of 22 children with supracondylar fractures. Flynn et al.⁶⁵
recommended palpating the point of the medial epicondyle and inserting
the pin anterior to avoid the nerve; however, Wind et al.²⁰⁶
found that the location of the ulnar nerve was not determined
accurately enough by palpation to allow blind medial pinning. They
found an average difference of almost 2 mm between the predicted and
the actual locations of the nerve. Making an incision over the medial
epicondyle to make certain the ulnar nerve is not directly injured by a
pin may not ensure protection of the nerve,¹⁸³ but Weiland et al.²⁰⁰
reported no iatrogenic ulnar nerve injuries in 52 patients with crossed
pins in whom a small medial incision was used. Green et al.⁸⁰
reported only one iatrogenic nerve injury (1.5%) in 65 patients treated
with two lateral pins and one medial pin inserted with a mini-open
technique. In their systematic literature review, Brauer et al.³¹
found the probability of iatrogenic ulnar nerve injury to be five times
higher with crossed pins than with lateral-entry pins; however, when
only the two prospective studies were included there was no significant
difference.

Even if the ulnar nerve is not penetrated by the pin, placement of a medial pin adjacent to the nerve can cause injury. Rasool¹⁶⁵
described iatrogenic ulnar nerve injuries in 6 patients in whom early
exploration found direct penetration of the nerve by the pin in 2,
constriction of the cubital tunnel in 3, and fixation of the nerve
anterior to the medial epicondyle in 1. In a series of 345
supracondylar humeral fractures treated by percutaneous pinning, Skaggs
et al.¹⁸³ found that ulnar nerve
injury occurred in 4% of patients with medial pins placed without
hyperflexion and in 15% of those in whom the medial pin was placed
while the elbow was hyperflexed. No iatrogenic ulnar nerve injuries
occurred in the 125 fractures treated with lateral-entry pins alone.
These studies indicate that if a medial pin is used, the lateral pin(s)
should be placed first, then the elbow should be extended and the
medial pin placed without hyperflexion of the elbow. Avoidance of a
medial pin is the simplest way to avoid iatrogenic ulnar nerve injury:
no iatrogenic nerve injuries occurred in a series of 124 consecutive
fractures stabilized with lateral-entry pins, regardless of
displacement or fracture stability.¹⁸²

Construct Stability. Another issue with pin
configuration is stability, which remains somewhat unclear because some
previous biomechanical studies of stability of various pin
configurations have not used currently recommended configurations. For
example, two studies that evaluated the torsional strength of pin
configurations and found crossed-pins to be stronger than two lateral
pins,¹⁵¹,²¹³
tested two lateral pins placed immediately adjacent to each other and
not separated at the fracture site as is currently recommended.¹⁷⁴,¹⁸² Lee et al.,¹²³
on the other hand, found that two divergent lateral pins separated at
the fracture site were stronger than crossed pins in extension loading
and varus but the configurations were equal in valgus (Fig. 14-19).¹²³
They attributed the greater strength with lateral pins to the location
of the intersection of the two pins and greater divergence between the
two pins, which allowed some purchase in the medial column as well as
the lateral column (see Fig. 14-6).

Intraoperative testing of the stability of lateral-entry
pin fixation has been advocated. In a study of 21 children with type
III fractures, two lateral-entry pins were inserted after reduction and
stability was assessed by comparing lateral fluoroscopic

images in internal and external rotation.²⁸
If the fracture remained rotationally unstable, a third lateral-entry
pin was inserted and images were obtained. A medial pin was added only
if instability was demonstrated after the insertion of three lateral
pins. Rotational stability was achieved with two lateral-entry pins in
6 patients, with three lateral-entry pins in 10 patients, and with an
additional medial pin in 5. No patient required a reoperation using
this protocol. The authors concluded that supracondylar fractures that
are rotationally stable intraoperatively after pin fixation are
unlikely to displace postoperatively. It is notable that they found
only a small proportion (26%) of these type III fractures to be
rotationally stable after fixation with two lateral-entry wires.²⁸

A systematic review of 35 articles reported loss of
reduction in none of 849 fractures treated with crossed pins and in 4
of 606 (0.7%) of fractures treated with lateral-entry pins.³¹
Deformity, defined as a carrying angle loss of more than 10 degrees or
cubitus varus of more than 5 degrees, occurred in 29 (3.4%) of those
with crossed pinning and 36 (5.9%) of those with lateral-entry pins.
However, this article did not include the largest series in the
literature that favored lateral-entry pins.³¹

Bloom et al.²⁷
recommended consideration of a three-pin pattern, either three
laterally divergent pins or two lateral pins and one medial pin, for
supracondylar humeral fractures with a less than complete anatomic
reduction. In their biomechanical analysis, three lateral divergent
pins were as strong as crossed-pinning and both were stronger than two
lateral divergent pins.²⁷ Another
biomechanic study of simulated fractures with medial comminution
determined that three lateral divergent pins provided torsional
stability equal to that of standard medial and lateral crossed-pinning.¹²⁰
These biomechanical studies support clinical recommendations that three
lateral-entry pins provide the most stable fixation of type III
fractures.¹⁷⁴,¹⁸²

Technical Points. In a series of 124 consecutive fractures treated with lateral-entry pins, Skaggs et al.¹⁸²
found no malunions or loss of fixation. From this successful series,
combined with a failure analysis of eight fractures outside of this
series, they concluded that the important technical points for fixation
with lateral-entry pins are: (i) maximizing separation of the pins at
the fracture site, (ii) engaging the medial and lateral columns
proximal to the fracture, (iii) engaging sufficient bone in both the
proximal segment and the distal fragment, (iv) maintaining a low
threshold for use of a third lateral-entry pin if there is concern
about fracture stability or the location of the first two pins, and (v)
using three pins for type III fractures.¹⁸² Gordon et al.⁷⁸
also recommended adding a third pin for type III fractures when
intraoperative stress indicated lack of stability with 2 pins. Lee et
al.¹²² reported 92% excellent
clinical results in 61 consecutive types II and II fractures treated
with three lateral pins. None of their patients had loss of reduction,
cubitus varus, hyperextension, loss of motion, or iatrogenic nerve
injury; none required additional surgery, and only one patient had a
minor pin track infection.¹²²

Sankar et al.¹⁷⁴
studied eight supracondylar humeral fractures in which reduction was
lost and determined that loss of fixation in all was due to technical
errors that could have been identified on the intraoperative
fluoroscopic images and could have been prevented with proper
technique. Three types of pin-fixation errors were identified: (i)
failure to engage both fragments with two pins or more, (ii) failure to
achieve bicortical fixation with two pins or more, and (iii) failure to
achieve adequate pin separation (>2 mm) at the fracture site.¹⁷⁴

Open reduction is indicated for open fractures,
fractures in which closed reduction fails, and fractures associated
with a dysvascular limb. Although earlier authors expressed concerns
about elbow stiffness, myositis ossificans, ugly scarring, and
iatrogenic neurovascular injury with open reduction, several more
recent reports have shown low rates of complications with open
reduction. Weiland et al.²⁰⁰
reported that, in 52 displaced fractures treated with open reduction
through a lateral approach, 10% had a moderate loss of motion but none
had infection, nonunion, or myositis ossificans. Fleuriau-Chateau et
al.,⁶³ in a series of 34 patients
treated with open reduction through an anterior approach, reported a 6%
(2 of 34) unsatisfactory loss of motion but no infection, myositis
ossificans, malunion, or Volkmann contracture. Reitman et al.¹⁶⁷
reported that 78% (51) of 65 patients treated with open reduction
(through either a medial or a lateral approach) had excellent or good
results according to the criteria of Flynn et al.⁶⁵; 4 patients had loss of motion. Ay et al.¹⁴ found no loss of motion or clinical deformity in 61 patients treated with open reduction. Kaewpornsawan⁹⁹
compared closed reduction and percutaneous pin fixation with open
reduction (through a lateral approach) in 28 children and found no
differences in the frequency of cubitus varus, neurovascular injury, or
infection; range of motion, union rate, and results according to the
criteria of Flynn et al.⁶⁵ also were not significantly different.

Approach for Open Reduction. Although most orthopaedists
are more familiar with lateral and medial approaches, we prefer a
direct anterior approach, especially in those with neurovascular
compromise. An advantage of the anterior approach is that it allows
direct visualization of the brachial artery and median nerve as well as
the fracture fragments. The relatively small (5 cm) transverse incision
along the cubital fossa produces a scar that is much more cosmetically
acceptable than that of the lateral approach, and scar contraction
limiting elbow extension is not an issue. In addition, this approach
does not disrupt the posterior periosteal hinge, which is useful for
fracture reduction. A series of 26 patients treated with the anterior
approach showed results equivalent to those of the traditional lateral
or combined lateral and medial approach in terms of malunion, the
criteria of Flynn et al.,⁶⁵ and range of motion.

The posterior approach is associated with high rates of
loss of motion and osteonecrosis caused by disruption of the posterior
end arterial supply to the trochlea of the humerus,³²,²⁰⁹ and it generally is not recommended for open reduction of supracondylar humeral fractures in children (Fig. 14-20).

Decreased Frequency of Complications. Open reduction has
been increasingly accepted because there are relatively few
complications with this method. Surgical experience¹¹,¹³,³⁸,⁴⁸,⁶³,⁶⁹,⁷⁵,¹¹²,¹⁵⁰,¹⁶⁷,²⁰⁹ has dispelled the fears of infection, myositis ossificans, and neurovascular injury.⁷³,¹⁷⁶,¹⁸⁶,¹⁹⁹ In a combined series of 470 fractures treated by open reduction, the incidence of infection was 2.5%, all of which resolved.⁷,²²,³⁶,⁵²,⁷⁴,⁷⁷,⁸¹,⁸⁶,¹⁰⁶,¹¹⁴,¹⁴⁶,¹⁵⁷,¹⁶³,¹⁷⁸,¹⁹⁸,²⁰⁰ The incidence of neurovascular complications from the procedure itself was essentially zero.

Four patients with myositis ossificans (1.4%) were reported, all in a single series.⁷⁷

The most frequent complication of surgical management
appears to be a loss of range of motion. One of the reasons given in
the past for loss of motion was the use of a posterior approach. It has
been stated that approaching the fracture through the relatively
uninvolved posterior tissues induces added scarring, leading to
stiffness. In earlier reported series using a posterior approach, loss
of range of motion was significant. Use of the anterior approach has
resulted in lower stiffness rates and complications similar to those of
closed treatment. Residual cubitus varus occurred in as many as 33% of
patients in some of the earlier series,⁸,⁵¹,⁷⁷,²⁰⁰
most due to inadequate surgical reduction. When good reduction was
obtained, the incidence of cubitus varus deformity was low. Surgical
intervention alone does not guarantee an anatomic reduction; the
quality of the reduction achieved at the time of surgery is important.

Lal and Bahn¹¹⁷
reported that delayed open reduction, 11 to 17 days after injury, did
not increase the frequency of myositis ossificans. If a supracondylar
fracture is unreduced or poorly reduced, delayed open reduction and pin
fixation appear to be justified. Ağş⁴ showed that delayed open reduction and pinning can be safely accomplished after skeletal traction and malreduction.

Open supracondylar fractures generally have an anterior
puncture wound where the metaphyseal spike penetrates the skin. Even if
the open wound is only a small puncture in the center of an anterior
pucker, open irrigation and débridement are indicated. The anterior
approach, using a transverse incision with medial or lateral extension
as needed, is recommended. The neurovascular bundle is directly under
the skin and tented over the metaphyseal fragment, so care should be
taken even as the skin is incised. The skin incision can be extended
medially proximally and laterally distally. Often, only the transverse
portion of the incision is required, which gives a better cosmetic
result. The brachialis muscle usually is transected because its muscle
belly inserts on the coronoid attachment and is highly vulnerable to
trauma from the proximal metaphyseal fragment. The fracture surfaces
are examined and washed, and a curette is used to remove any dirt or
entrapped soft tissue. Once the débridement and washing are complete,
the fracture is stabilized with K-wires. All patients with open
fractures also are treated with antibiotics, generally, cephalothin for
Gustilo types I, II, and IIIA injuries, with the addition of an
aminoglycoside for types IIIB and IIIC fractures.

Traction rarely is used as definitive treatment for
supracondylar fractures in children in most modern centers. Indications
for traction may include severe comminution, lack of anesthesia,
medical conditions prohibiting anesthesia, lack of an experienced
surgeon, or temporary traction to allow swelling to decrease. Devnani⁵⁸
reported using traction in the gradual reduction of 28 fractures with
late presentation (mean of 5.6 days), although 18% of these children
required a corrective osteotomy for malunion. Cubitus varus has been
reported after traction treatment in 9% to 33% of patients in some
series,⁹¹,¹⁶⁰ while others have reported excellent results.⁵¹,⁷¹,¹⁸⁷
Because of the excellent results obtained with closed reduction and
pinning and the brief hospital stay required after this procedure
(usually no more than one night), it is difficult to justify the 14 to
22 days of hospitalization required for traction treatment. Advocates
of traction in the treatment of supracondylar fractures have reported
that the use of overhead traction with use of an olecranon wing nut¹⁵,¹⁵⁵,²⁰⁸ gives superior results to sidearm traction (Fig. 14-21).

Type I Fracture (Nondisplaced). In general in a type I
fracture, the periosteum is intact with significant inherent stability
of the fracture. A type I fracture may become apparent only with repeat
radiographs at 1- to 2-week follow-up after presentation with elbow
pain and initial radiographic findings limited to a posterior fat pad
sign. Periosteal reaction in the distal humerus may be all that is
visible on radiograph.

Simple immobilization with a posterior splint applied at
60 to 90 degrees of elbow flexion with side supports is preferred by
some.³⁹,²⁰³
We prefer a carefully constructed, nonconstrictive, circumferential
cast to avoid skin problems, blistering, and the tourniquet effect of
elastic wraps left in place for a few weeks. The arm always looks
better when it comes out of an adequatelypadded well-made cast than
when it comes out of splint with an elastic wrap. If there is no
significant swelling about the elbow, circumferential casts can be
used, but the parents must be educated about elevation and the signs
and symptoms of compartment syndrome. The elbow should not be flexed
more than 90 degrees. Using Doppler examination of the brachial artery
after supracondylar fractures, Mapes and Hennrikus¹²⁹
found that flow was decreased in the brachial artery in positions of
pronation and increased flexion. Before the splint is applied, it
should be confirmed that the pulse is intact and that there is good
capillary refill with the amount of elbow flexion intended during
immobilization. A sling helps decrease torsional forces about the
fracture.

Radiographs are obtained 3 to 7 days after fracture to
document lack of displacement. If there is any evidence of distal
fragment extension, as judged by lack of intersection of the

anterior humeral line with the capitellum, the fracture has changed into a type II fracture and should be treated as such.

An acceptable position is determined by the anterior
humeral line transecting the capitellum on the lateral radiograph, and
a Baumann angle of more than 10 degrees or equal to the other side. The
duration of immobilization for supracondylar fractures is 3 weeks,
whether type I, II, or III. In general, no physical therapy is required
after this injury. Patients may be seen 4 weeks after immobilization is
removed to ensure that range of motion and strength are returning
normally. Because the outcomes of type I fractures are predictably
excellent if alignment is maintained at the time of early healing,
follow-up visits after cast removal are optional depending on family
and medical circumstances.

The initial radiograph is a static representation of the
actual injury that may involve soft tissue disruption much greater than
might be expected from the minimal bony abnormality. Excessive
swelling, nerve or vascular disruption, and excessive pain are
indicative of a more significant injury than a type I fracture, in
which case periosteal disruption may render this fracture inherently
unstable.

Type II Fracture (Hinged Posteriorly, with Posterior
Cortex in Continuity). This fracture category encompasses a broad array
of soft tissue injuries. Careful assessment of the soft tissue injury
is critical in treatment decision-making. Because the posterior cortex
is in continuity, good stability should be obtained with closed
reduction. Significant swelling, obliteration of the pulse with
flexion, neurovascular injuries, excessive angulation, and other
injuries in the same extremity are indications for pin stabilization of
most type II fractures.

Currently, the treatment of choice for type II fractures
is operative stabilization rather than cast immobilization. The distal
humerus provides only 20% of the growth of the humerus and has little
remodeling potential. The upper limb grows approximately 10 cm during
the first year of life, 6 cm during the second year, 5 cm during the
third year, 3.5 cm during the fourth year, and 3 cm during the fifth
year of life.⁵⁹ In toddlers (<3
years of age), some remodeling potential is present so nonoperative
treatment may be appropriate for a type II fracture in which the
capitellum abuts the anterior humeral line but does not cross it. In a
child who is 8 to 10 years old, however, only 10% of growth of the
distal humerus remains, so adequate reduction is essential to prevent
malunion.

Two studies support the initial treatment of type II fractures with closed reduction and casting. Hadlow et al.⁸⁴
noted that if pinning of all type II fractures in their series had been
done, 77% (37 of 48) of patients would have undergone an unnecessary
operative procedure; however, 23% (11 of 48) of the patients required
later operative reduction because they lost reduction after closed
reduction; 14% (2 of 14) who were followed had poor outcomes by the
Flynn criteria.⁶⁴ Parikh et al.,⁶⁴
in a retrospective review of 25 elbows treated with closed reduction
and casting, found similar results: 28% (7 of 25) loss of reduction,
20% (5 of 25) delayed surgery, and 2% (2 of 25) unsatisfactory outcomes
according to the Flynn criteria.

In contrast, in their consecutive series of 69 children
with type II fractures treated with closed reduction and pinning,
Skaggs et al.¹⁸² found no
radiographic or clinical loss of reduction, no cubitus varus, no
hyperextension, and no loss of motion. No patient had an iatrogenic
nerve palsy, and none required additional surgery.¹⁸²
In a review of 191 consecutive type II fractures treated with closed
reduction and percutaneous pinning, there were four (2%) pin track
infections, of which three were treated successfully with oral
antibiotics and pin removal and one required operative irrigation and
débridement

for
a wound infection not involving the joint. No nerve or vascular
injuries, no loss of reduction, and no delayed unions or malunions were
reported.¹⁸⁵

Avoiding compartment syndrome is another reason for
choosing operative treatment of these injuries. The amount of
hyperflexion needed to maintain reduction of unpinned type II fractures
predisposes these patients to increased compartment pressures.¹⁹ Mapes et al.¹²⁹
used Doppler examination to determine that pronation and increased
flexion caused decreased flow in the brachial artery, leading them to
recommend a position of flexion and supination for “vascular safety.”
Pinning of these fractures avoids immobilization with the elbow
markedly flexed. In general, for any fracture that would require elbow
flexion of more than 90 degrees to hold reduction, pins should be used
to hold the reduction, and the elbow should be immobilized in less
flexion (usually about 45 to 70 degrees). If pinning is chosen, two
lateral-entry pins¹⁷⁴,¹⁸¹,¹⁸³,¹⁹⁵
through the distal humeral fragment, engaging the opposite cortex of
the proximal fragment, generally are sufficient to maintain fracture
alignment (Figs. 14-22 and 14-23), although three pins can be used for added stability.¹⁸²
Because the posterior cortex and the intact periosteum provide some
degree of inherent stability, crossed-pinning of type II fractures
generally is not needed. The techniques for crossed and lateral pinning
are described later in this chapter. If pin stabilization is used, the
pins are left protruding through the skin and are removed 3 to 4 weeks
after fixation, generally without the need for sedation or anesthesia.

Type III Fractures. For any child who presents to the
emergency room with the elbow in either extreme flexion or extension,
the arm is carefully placed in 30 degrees of flexion to minimize
vascular insult and compartment pressure. In a type III fracture, the
periosteum is torn, there is no cortical contact between the fragments,
and soft tissue injury may accompany the fracture. Careful preoperative
evaluation is mandatory. If circulatory compromise is indicated by
absent pulse and a pale hand or if compartment syndrome is suspected,
urgent reduction and skeletal stabilization are mandatory. Most type
III fractures require operative reduction and pinning.

Cast immobilization techniques have been described as
well. Because type III fractures are inherently unstable, the elbow
must be held in extreme flexion to prevent the distal fragment from
rotating; these fractures tend to rotate with flexion of less than 120
degrees.¹⁴¹ A figure-of-eight cast (Fig. 14-24)
can be used to maintain flexion of at least 120 degrees. If the arm can
be flexed to 120 degrees with an intact pulse, casting can be

used
as primary treatment. Usually, however, severe swelling prevents the
elbow from being kept in hyperflexion or compartment syndrome could
result. Alburger et al.⁶
reported that using skin traction initially until swelling decreased
allowed successful casting without pinning. In most series,⁴⁸,¹¹⁵,¹⁵⁸,¹⁹⁸
the results of type III fractures treated with closed reduction and
cast immobilization are not as good as the results of pinning. Hadlow
et al.,⁸⁴ however, suggested that
selective use of casting is beneficial, reporting that in their series,
61% of type III and 77% of type II fractures were successfully treated
without pinning.

When a cast is used as primary treatment, it should be
worn for 3 to 4 weeks. Although a number of historic series used
casting as primary treatment, most recent reports favor pinning of this
fracture because of concerns about vascular compromise, compartment
syndrome, and malunion. It must be emphasized that flexion of the elbow
of 90 degrees or more with a type III supracondylar fracture
significantly increases the risk of compartment syndrome and should
rarely, if ever, be done if modern operative facilities and an
experienced surgeon are available.¹⁹ Traction may be a safer alternative.

The Special Case of Medial Column Comminution.
Although not as markedly displaced as most type III fractures,
fractures with medial comminution usually require operative reduction
because collapse of the medial column will lead to varus deformity in
an otherwise minimally displaced supracondylar fracture (see Fig. 14-7).⁵⁵ De Boeck et al.⁵⁵
recommended closed reduction with percutaneous pinning for fractures
with medial comminution even with minimal displacement to prevent
cubitus varus. In their retrospective review of 13 patients with medial
comminution, none of the 6 who had operative fixation had cubitus
varus, while 4 of 7 treated nonoperatively developed cubitus varus.

Type IV Fractures. Open reduction has traditionally been recommended for this extremely unstable fracture, but Leitch et al.¹²⁵
described closed reduction in 9 patients. Their technique involves
placing two K-wires into the distal fragment, reducing the fracture in
the AP plane, and verifying the reduction with fluoroscopy. Then the
fluoroscopy unit, rather than the arm, is rotated to obtain a lateral
view (Fig. 14-25). The fracture is reduced in
the sagittal plane, and the K-wires are driven across the fracture
site. All fractures treated with this technique united with no cubitus
varus, malunion, loss of motion, or additional operative treatment.
Because of the limited number of these multidirectionally unstable
fractures in their series, neither the need for open reduction nor the
true complication rates can be determined.

Approximately 10% to 20% of patients with type III supracondylar fractures present with an absent pulse.⁶⁰,¹⁵⁷,¹⁷⁵,¹⁷⁷
The presence of a pulse and perfusion of the hand should be documented.
Perfusion is estimated by color, warmth, and capillary refill.
Capillary refill alone can be deceiving; after wrapping a rubber band
around a finger, there is instant venous capillary refill, so this must
be differentiated from arterial capillary refill.

A relatively recent study by Choi et al.⁴⁴
demonstrated which patients are likely to need arterial repair or to
develop a compartment syndrome based on preoperative presentation. Of
1255 patients with operatively treated supracondylar humeral fractures,
33 presented with absent distal pulses. The key difference in the
outcome of these patients was whether the hand was well-perfused at
time of presentation. Of the 24 patients whose hands were well-perfused
(but pulseless) at presentation, none required vascular repair or
developed a compartment syndrome, and fracture reduction alone was
effective treatment. Of the 9 patients who presented with poor distal
perfusion, 4 required vascular repair, and compartment syndrome
developed in 2 during the postoperative period. In 5 of these 9
patients, fracture reduction alone was definitive treatment.

Absence of the radial pulse is not an emergency because
collateral circulation may keep the limb well-perfused, but timely
reduction with pinning in the operating room is preferable.¹⁷²
A pulseless arm with signs of poor perfusion is an emergency. A patient
with a severely displaced supracondylar fracture and compromised
vascularity to the limb should have a splint applied with the elbow in
approximately 20 to 40 degrees of flexion rather than extremes of
flexion or extension which may compromise bloodflow.¹⁰⁵

Because fracture reduction usually restores the pulse, angiographic studies should not delay fracture reduction.⁴⁴ Several reports have shown angiography to be an unnecessary test that has no bearing on treatment.⁴⁵,¹⁵⁷,¹⁷⁵,¹⁷⁷ In a series of 143 type III supracondylar fractures, 17 of which had vascular compromise,¹⁷⁷
all had reduction and percutaneous pinning without preoperative
angiogram. In 3 of the 17 patients, bloodflow to the hand was not
restored after reduction and open exploration was required; in the
other 14, bloodflow to the hand was restored without complications. The
authors concluded that prereduction angiography would have added
nothing to the management of these injuries. Another study used
angiogram in 4 of 17 patients with supracondylar humeral fractures and
dysvascular extremities and found that the angiogram did not alter the
course of management in any.⁴⁵ Choi et al.⁴⁴
reported that of 25 patients presenting with a pulseless but
well-perfused hand, all did well clinically without arterial repair—11
(52%) had a palpable pulse following surgery and 10 (48%) remained
pulseless but well-perfused.

Open reduction through an anterior approach is indicated
if anatomic reduction cannot be obtained by closed reduction in a
patient with an absent pulse; this allows evaluation of all vital
structures at risk for incarceration between the fracture fragments.¹³,⁶³
Once the artery is freed from the fracture, arterial spasm may be
relieved by the application of lidocaine, warming, and 10 to 15 minutes
of observation. If the pulse does not return after fracture reduction
and the hand remains poorly perfused, vascular reconstruction (usually
by a vascular surgeon) is indicated.

After closed reduction and stabilization, the pulse and
perfusion of the hand should be evaluated. Most extension-type
supracondylar fractures are reduced and pinned with the elbow in
hyperflexion. With more than 120 degrees of elbow flexion, the radial
pulse generally is lost, even in patients with an initially intact
pulse. After pinning, when the arm is extended, the pulse frequently
does not return immediately. This is presumably secondary to arterial
spasm, aggravated by swelling about the artery and decreased peripheral
perfusion in the anesthetized, somewhat cool intraoperative patient.
Because of this phenomenon, 10 to 15 minutes should be allowed for
recovery of perfusion in the operating room before any decision is made
regarding the need for exploring the brachial artery and restoring flow
to the distal portion of the extremity. Because most patients without a
palpable pulse maintain adequate distal perfusion, the absence of a
palpable pulse alone is not an indication for exploring the brachial
artery.

If the pulse does not return, but the hand is
well-perfused, the treatment is controversial. We prefer to admit the
child to the hospital for gentle elevation and observation for at least
48 hours. Loss of perfusion or impending compartment syndrome can occur
during this time and may require immediate treatment. Gillingham and
Rang⁷⁶ recommended observing
patients with absent pulses, because most pulses returned within 10
days. Alternatively, vascular reconstruction can be performed. However,
Sabharwal et al.¹⁷² showed that
early repair of the brachial artery has a high rate of symptomatic
reocclusion and residual stenosis and recommended a period of close
observation with frequent neurovascular checks before more invasive
correction of this problem is contemplated. Immediate rereduction,
usually open reduction, is indicated in a patient in whom a pulse was
present preoperatively but is absent postoperatively, because the
artery or adjacent tissue is presumed to be entrapped in the fracture
site.

Cold intolerance has been considered a possible
long-term sequel in children without palpable distal pulses, though
there is little literature on this subject. There is a single case
report of a child with minor cold intolerance and decreased systolic
arterial forearm blood pressure more than 1 year after a supracondyalr
fracture with rupture of the median nerve and ligation of the brachial
artery.¹³⁰

The decision to explore a brachial artery should be
based on objective criteria. Traditionally, the decision has been based
only on whether the hand was warm and pink. The following case
indicates the difficulty with these criteria. A 2-year-old girl was
injured in a fall from a couch (Fig. 14-44). A
type III supracondylar fracture and a pale, cool hand were documented
on presentation to a local emergency department. Two hours later, the
patient was brought to the operating room, where a cool, pale hand with
poor capillary refill and an absent pulse

was
noted. Immediate closed reduction and pinning was done with nearly
anatomic reduction. The hand felt warmer, and capillary refill was
present. The pulse was not palpable, but because of the improved state
of the patient’s hand, the brachial artery was was not explored. During
the next 4 hours, the patient was observed closely with increasing
fussiness, a nonpalpable pulse, and slow capillary refill. An
arteriogram showed brachial artery obstruction. Compartment pressures
were measured, and increased pressure was noted in the volar
compartment. A decision was made to return to the operating room for
exploration, repair of the brachial artery, and forearm fasciotomy.
Although the outcome in this case was satisfactory with no long-term
sequelae other than scarring, the question is whether or not the low
perfusion with subsequent ischemia and compartment syndrome could have
been identified at the time of the closed reduction and immediate
vascular repair done. What would have happened if no repair had been
done in this case? Had a repair been done immediately, could the
development of a compartment syndrome have been averted? What is the
relationship between compartment syndrome and perfusion? Is there a
more objective way to determine when flow is insufficient?

Several investigators have attempted to provide criteria
for vascular repair in addition to warmth and color in the patient with
absence of a palpable pulse.¹⁷⁵,¹⁷⁷ In the absence of a palpable pulse, a Doppler device can be used to measure lower flow states with small-pulse amplitude. Shaw¹⁷⁷
found no falsepositive explorations when patients with pulses that
could not be palpated but were identified on Doppler evaluation were
observed, and exploration was done in those in whom no pulse was found
with either palpation or Doppler. The brachial artery was either
transected or entrapped in all patients with surgical exploration, and
none with “spasm” underwent surgery. Using the same criteria,
Schoenecker et al.¹⁷⁵ identified 6
patients for exploration. Three had a damaged or transacted brachial
artery with no flow, and 3 had an artery kinked or trapped in the
fracture. At follow-up, all patients with vascular repair had a radial
pulse. One patient with more than 24 hours of vascular insufficiency
had an unsatisfactory outcome. Copley et al.⁴⁵
reported that of 17 patients with type III supracondylar fractures and
no palpable pulse at presentation, 14 recovered pulse after reduction.
The three explorations identified significant vascular injury, and the
brachial artery was repaired. Most significant in their series was the
finding of increasing vascular insufficiency during postoperative
observation in some patients. If the hand is well-perfused despite not
having a palpable pulse, close follow-up in the hospital is mandatory.
The patient should be observed for increasing narcotic requirements,
increasing pain, and decreased passive finger motion. A very low
threshold for returning to the operating room for exploration and
fasciotomy must be maintained rather than assuming that perfusion from
collaterals is sufficient.

Choi et al.⁴⁴
reported that of eight patients presenting with a pulseless and poorly
perfused hand, two developed compartment syndrome after operative
reduction of the fracture. Perhaps the threshold for compartment
release at the time of the initial surgery should be lower in this
subset of patients presenting with a pulseless and poorly perfused hand.

Some surgeons recommend pulse oximetry¹⁰⁸,¹⁶⁶
for evaluating postreduction circulation. We have found pulse oximetry
useful for evaluating patients after vascular repair or pinning but not
for intraoperative decision making.

Exploration of the Brachial Artery. Obliteration of the
intact preoperative radial pulse after closed reduction and pinning is
a strong indication for brachial artery exploration only when
accompanied by evidence of impaired circulation to the hand. After 10
to 15 minutes is allowed for resolution of arterial spasm as a cause
for loss of pulse, the brachial artery should be explored if the hand
is not warm and pink. Either direct arterial entrapment at the fracture
or arterial compression by a fascial band pulling across the artery may
cause loss of pulse after fracture reduction. As described earlier, the
other indication for brachial artery exploration is persistent vascular
insufficiency after reduction and pinning.

The orthopaedic surgeon and vascular surgeon must work
together to manage this problem emergently. Often, the release of a
fascial band or an adventitial tether resolves the problem of
obstructed flow. This is a simple procedure done at the time of
exploration of the antecubital fossa and identification of the brachial
artery. In some patients, however, a formal vascular repair and vein
graft are required. The brachial artery should be approached through a
transverse incision across the antecubital fossa, with a medial
extension turning proximally at about the level of the medial
epicondyle as described for open reduction (see Fig. 14-42).
After reduction and pinning, care must be taken because the
neurovascular bundle may be difficult to identify when it is surrounded
by hematoma, and it may lie in a very superficial position. At the
level of the fracture, the artery may seem to disappear into the
fracture site, covered with shredded brachialis muscle. This occurs
when the artery is tethered by a fascial band or arterial adventitia
attached to the proximal metaphyseal spike pulling the artery in the
fracture site. Dissection should occur proximally to distally, along
the brachial artery, identifying both the artery and the median nerve.
Arterial injury generally is at the level of the supratrochlear artery
(see Fig. 14-6), which provides a tether,
making the artery vulnerable at this location. Arterial transection or
direct arterial injury can be identified at this level (see Fig. 14-43). Entrapment of the neurovascular bundle in the fracture is best identified by proximal to distal dissection.

The vascular surgeon generally makes the decision to
repair a damaged artery or use a vein graft. If arterial spasm is the
cause of inadequate flow and collateral flow is not sufficient to
maintain the hand, methods to relieve the spasm can be tried. Both
stellate ganglion block and application of papaverine or local
anesthetic to the artery have been found to be beneficial in this
situation. If these techniques do not relieve the spasm and if
collateral flow is insufficient, the injured portion of the vessel is
excised and a vein graft is inserted. When flow is restored, the wound
is closed, and a splint is applied with the elbow flexed much less than
90 degrees and the forearm in neutral pronation and supination.
Postoperative monitoring should include temperature, pulse oximetry,
and frequent examinations for signs of compartment syndrome or
ischemia. Although injecting urokinase has been suggested to increase
flow,³⁴ we have had no experience with this technique.

Sabharwal et al.¹⁷²
documented that 3.2% of patients with type III supracondylar fractures
have an absent pulse at presentation, for which they recommended
noninvasive monitoring. MR angiography and color-flow duplex Doppler
were helpful in

deciding whether or not to explore the brachial artery. Although repair is technically feasible, Sabharwal et al.¹⁷²
cautioned that high rates of reocclusion and residual stenosis argued
against early revascularization if not absolutely necessary. Early
reocclusion, however, was not reported by Schoenecker et al.¹⁷⁵ or Shaw et al.¹⁷⁷

The necessity of early treatment of vascular compromise was emphasized by Ottolenghi¹⁵⁴ who found no Volkmann ischemic contractures¹⁴³
in patients in whom vascular compromise was treated within 12 hours.
The frequency of this complication increased steadily with repair
between 12 and 24 hours; after 24 hours of delay in treatment, outcomes
were uniformly poor. This series presents convincing evidence that
prompt exploration of arterial insufficiency markedly decreases the
incidence of Volkmann ischemic contracture. Note that brachial artery
obstruction and compartment syndrome, although related, are not
equivalent, and both are fortunately rare problems. Ischemia will lead
to a compartment syndrome, but the presence of a radial pulse does not
preclude it.

Compartment syndrome is estimated to occur in 0.1% to 0.3% of patients with supracondylar fractures.¹⁹ Blakemore et al.²⁵
reported a 9% prevalence (3 of 33) of forearm compartment syndrome in
patients with a supracondylar fracture and an ipsilateral radial
fracture. In acute compartment syndrome,⁹¹,¹⁴³
increased pressure in a closed fascial space causes muscle ischemia.
With untreated ischemia, muscle edema increases, further increasing
pressure, decreasing flow, and leading to muscle necrosis, fibrosis,
and death of involved muscles. A compartment syndrome of the forearm
may occur with or without brachial artery injury and in the presence or
absence of a radial pulse. The diagnosis of a compartment syndrome is
based on resistance to passive finger movement and dramatically
increasing pain after fracture. The classic five “Ps” for the diagnosis
of compartment syndrome—pain, pallor, pulselessness, paresthesias, and
paralysis—are poor criteria for the early detection of compartment
syndrome. Special attention must be paid to supracondylar fractures
with median nerve injuries because the patient will not feel pain in
the volar compartment.¹⁴³

Mubarak and Carroll¹⁴³
recommended forearm fasciotomy if clinical signs of compartment
syndrome are present or if intracompartmental pressure is greater than
30 mm Hg. Heppenstall et al.⁸⁹
suggested that a difference of 30 mm Hg between diastolic blood
pressure and compartment pressure should be the threshold for release.
If pain is increasing and finger extension is decreasing, fasciotomy is
clearly indicated. Measuring compartment pressures in a terrified,
crying child is difficult, and if clinical signs of compartment
syndrome are present, a trip to the operating room for evaluation and
possible fasciotomy is often a better course of action than pressure
measurement and observation.

Clinical conditions that contribute to the development
of compartment syndrome are direct muscle trauma at the time of injury,
swelling with intracompartmental fractures (associated forearm
fracture), decreased arterial inflow, restricted venous outflow, and
elbow position. If the mechanism of injury is high-energy, as evidenced
by crushing or associated fractures, there is an increased risk for
compartment syndrome. An associated forearm fracture or forearm crush
injury significantly increases the likelihood of compartment syndrome.²⁵
An arterial injury in association with multiple injuries or crush
injury further diminishes blood flow to the forearm musculature and
increases the probability of a compartment syndrome. Interestingly,
Battaglia et al.¹⁹ documented the
relationship between increasing elbow flexion above 90 degrees and
increasing volar compartment pressure. In a patient with an evolving
postoperative compartment syndrome, not only should the dressings be
loosened, but the elbow should be extended to a position well below 90
degrees.

In a multicenter review, Ramachandran et al.¹⁶²
identified 11 children with supracondylar fractures who developed
compartment syndrome despite presenting with closed, low-energy
injuries and no associated fractures or vascular compromise. This
series is disturbing because it demonstrates that compartment syndrome
still occurs with modern treatment, even in children who may be thought
to be at low risk. Ramachandran et al.¹⁶²
found that in 10 of 11 patients’ charts, excessive swelling was noted
at time of presentation, and the one child without severe swelling had
arterial entrapment after reduction, with subsequent compartment
syndrome requiring fasciotomy 25 hours later. The 10 children with
severe elbow swelling documented at presentation had a mean delay of 22
hours before surgery. This study suggests that excessive swelling
combined with delay in treatment is a risk factor for the development
of compartment syndrome.

Even if a distal pulse is found by palpation or Doppler examination, an evolving compartment syndrome may be present.¹⁶¹
Increased swelling over the compartment, increased pain, and decreased
finger mobility are cardinal signs of an evolving compartment syndrome.
Evaluation of possible compartment syndrome cannot be based on the
presence or absence of a radial pulse alone. If a compartment syndrome
does appear to be evolving, initial management includes removing all
circumferential dressings. The volar compartment should be palpated,
and the elbow should be extended. We believe that the fracture should
be immediately stabilized with K-wires to allow proper management of
the soft tissues.

Another factor that contributes to the development of
compartment syndrome is warm ischemic time after injury. When bloodflow
is compromised and the hand is pale with no arterial flow, muscle
ischemia is possible, depending on the time of oxygen deprivation.
After fracture reduction and flow restoration, the warm ischemic time
should be noted. If this time is more than 6 hours, compartment
syndrome secondary to ischemic muscle injury is likely. Prophylactic
volar compartment fasciotomy can be done at the time of arterial
reconstruction. The exact indication for prophylactic fasciotomy in the
absence of an operative revascularization is uncertain. Even when the
diagnosis is delayed or the if compartment syndrome is chronic,
fasciotomy has been shown to be of some value.

Compartment syndrome is a true surgical emergency. A
recent systematic review of 55 reports of 1920 fasciatomies of the
upper and lower extremities found that, compared with fasciotomy before
6 hours, delayed fasciotomy beyond 12 hours was associated with a lower
rate of acceptable outcome (15% for more than 12 hours versus 88% for
<6 hours), a higher rate of amputation (14% versus 3.2%), and death
(4.3% versus 2.0%).⁸⁷

Technique for Volar Fasciotomy. The volar compartment of the forearm can be approached through the classic Henry approach

or an ulnar approach (Fig. 14-45).
If the compartment syndrome is associated with brachial artery and
median nerve injuries, we generally use the Henry approach as an
extension of the vascular repair. The advantage of the ulnar approach,
as described by Willis and Rorabeck,²⁰⁴
is that it produces a more cosmetically pleasing scar. A volar
fasciotomy involves opening the volar compartment from the carpal
tunnel distally to the lacertus fibrosa and antecubital fascia
proximally. The fascia over the deep flexors is opened, as is the
superficial fascia, to decompress the deep volar compartment of the
forearm. Failure to release the deep volar fascia may cause contracture
of the deep finger flexors. Generally, only the volar compartment is
released, with an associated decrease in pressure in the dorsal or
extensor compartment. If the volar Henry approach is used, the interval
between the brachioradialis and the flexor carpi radialis is developed
and the radial artery is retracted ulnarward. The deep volar
compartment is exposed. The flexor digitorum profundus and flexor
pollicis longus are exposed along with the pronator teres proximally
and the pronator quadratus distally.

If the ulnar approach is used, as described by Willis and Rorabeck,²⁰⁴
the release is done from the carpal canal to the antecubital fossa, as
with the Henry approach. The skin incision begins above the elbow
crease, medial to the biceps tendon (see Fig. 14-45);
it crosses the elbow crease and extends distally along the ulnar border
to the volar wrist, where it courses radially across the carpal canal.
The fascia over the flexor carpi ulnaris is incised, and the interval
between the flexor carpi ulnaris and the flexor digitorum sublimis is
identified. The ulnar nerve and artery are retracted, exposing the deep
flexor compartment of the forearm. The deep flexor fascia is incised.
The ulnar nerve and artery, as well as the carpal tunnel, are
decompressed distally.

After fasciotomy, the wound generally is left open. An
effective way to manage the wound is with a criss-crossed rubber band
technique, securing the rubber bands in place with skin staples. An
alternative is to simply place a sterile dressing over the open wound,
but this makes wound closure difficult and probably increases the need
for skin grafting. Definitive closure or skin grafting generally is
done within 5 to 7 days. Skeletal stabilization of the supracondylar
and forearm fractures is important to properly manage compartment
syndrome.

Neurologic injury has been reported to occur in as many as 49% of patients with supracondylar humeral fractures,³⁵ but in most modern series the incidence varies between 10% and 20%.¹³⁷,¹⁷³ Previously, investigators believed that the radial nerve was injured most often⁶⁶,¹²⁶,¹⁴⁴,¹⁵⁵,¹⁶⁰; however, as first noted by Spinner and Schreiber,¹⁸⁹ the anterior interosseous nerve appears to be the most commonly injured nerve with extension-type fractures.⁴⁹,⁶⁰,¹³⁴,¹⁶¹,¹⁸⁹
The direction of fracture displacement determines the nerve most likely
to be injured. If the distal fragment is displaced posteromedially, the
radial nerve is more likely to be injured. Conversely, if the
displacement of the distal fragment is posterolateral, the
neurovascular bundle is stretched over the proximal fragment, injuring
the median nerve or anterior interosseous nerve or both. In a
flexion-type supracondylar fracture, which is rare, the ulnar nerve is
the most likely to be injured. Injury to the anterior interosseous
nerve causes paralysis of the long flexors of the thumb and index
finger without sensory changes. Complete median nerve injury can be
caused by contusion or transection of the nerve at the level of the
fracture and causes sensory loss in the median nerve distribution as
well as motor loss of all muscles innervated

by the median nerve.¹⁸⁹,¹⁹¹ Nerve transections are rare and almost always involve the radial nerve.⁵⁰,¹⁷,¹³²,¹³⁴

Open reduction and exploration of the injured nerve are
not necessarily indicated for nerve injury in a closed fracture.
Regardless of which nerve is injured, neural recovery generally occurs
in the first 2 to 3 months, but may take up to 6 months.³³,¹⁰⁵ Observation is indicated.

Culp et al.⁵⁰
described eight injured nerves in five patients in whom spontaneous
recovery did not occur by 5 months after injury. Neurolysis was
successful in restoring nerve function in all but one patient. Nerve
grafting may be indicated for nerves not in continuity at the time of
exploration. Neurolysis for perineural fibrosis generally is successful
in restoring nerve function. There is no indication for early
electromyographic analysis or treatment other than observation for
nerve deficit until 4 to 6 months after fracture.

In their series of radial nerve injuries with humeral fractures, Amillo et al.⁹
reported that, of 12 injuries that did not spontaneously recover within
6 months of injury, only one was associated with a supracondylar
fracture. Perineural fibrosis was present in four patients, three
nerves were entrapped in callus, and five were either partially or
totally transected.

In the supracondylar area, perineural fibrosis appears
to be the most common cause of prolonged nerve deficit. Although nerve
injury is related to fracture displacement, a neural deficit can exist
with even minimally displaced fractures. Sairyo et al.¹⁷³
reported one patient in whom radial nerve palsy occurred with a
slightly angulated fracture that appeared to be a purely extension-type
fracture on initial radiographs. Even in patients with mild injuries, a
complete neurologic examination should be performed before treatment.
An irreducible fracture with nerve deficit is an indication for open
reduction of the fracture to ensure that there is no nerve entrapment.
Chronic nerve entrapment in healed callus can give the appearance of a
hole in the bone, the Metev sign.

Iatrogenic injury to the ulnar nerve has been reported to occur in 1% to 15% of patients with supracondylar fractures.³³,⁹⁴,¹⁶⁵,¹⁶⁹,¹⁸³
In a large series of type III supracondylar fractures, the rate of
iatrogenic injury to the radial nerve was less than 1%. The course of
the ulnar nerve through the cubital tunnel, between the medial
epicondyle and the olecranon, makes it vulnerable when a medial pin is
placed. Rasool¹⁶⁵ demonstrated with
operative exploration that the pin usually did not impale the ulnar
nerve, but more commonly constricted the nerve within the cubital
tunnel by tethering adjacent soft tissue. These findings were later
confirmed by an ultrasonographic study by Karakurt et al.¹⁰³ Zaltz et al.²¹¹
reported that in children less than 5 years of age, when the elbow is
flexed more than 90 degrees, the ulnar nerve migrated over, or even
anterior to, the medial epicondyle in 61% (32/52) of children.

There is insufficient information in the literature to
offer an evidence-based approach to an iatrogenic ulnar nerve injury
that occurs following medial pin placement. Lyons et al.¹²⁷
reported full return of function in 17 patients with iatrogenic ulnar
nerve injuries presumably caused by a medial pin, although function did
not return until 4 months after surgery in several. Only 4 of the 17
(24%) had the medial pins removed, indicating that nerve function can
eventually return without pin removal. Brown and Zinar³³
reported four ulnar nerve injuries associated with pinning of
supracondylar fractures, all of which resolved spontaneously 2 to 4
months after pinning. Rasool¹⁶⁵
reported six patients with ulnar nerve injuries in whom early
exploration was performed. In two patients, the nerve was penetrated,
and in three, it was constricted by a retinaculum over the carpal
tunnel, aggravated by the pin. In 1 patient, the nerve was subluxed and
was fixed anterior to the cubital tunnel by the pin. Full recovery
occurred in three patients, partial recovery in two, and no recovery in
two. Royce et al.¹⁶⁹ reported
spontaneous recovery of ulnar nerve function in three patients. One
nerve that was explored had direct penetration, and the pin was
replaced in the proper position. Two patients had late-onset ulnar
nerve palsies discovered during healing, and the medial pin was removed.

If an immediate postoperative neural injury is
documented, we prefer to replace the pin in the proper position or
convert to a lateral pin construct, earlier rather than later. Routine
exploration of the ulnar nerve is not recommended.³³,⁶⁴,¹⁶⁵,²¹¹

Preventing ulnar nerve injury is more desirable than
treating ulnar neuropathy. Because of the frequency of ulnar nerve
injury with crossed pinning, most surgeons prefer to use two or three
lateral pins if possible and no medial pin. Successful maintenance of
alignment of type III supracondylar fractures with lateral pins has
been reported in many series.⁴¹,¹¹⁰,¹⁸²,¹⁸³,¹⁹⁵
Many believe that if crossed-pinning is to be used, nerve penetration
and indirect trauma to the nerve cannot be prevented by making an
incision over the medial epicondyle and avoiding the nerve while
placing the medial pin directly on the medial epicondyle. Skaggs et al.¹⁸³
reported that the use of crossed pins with a medial cut-down did not
decrease the risk of iatrogenic ulnar nerve injuries. It is important
to remember that most iatrogenic ulnar nerve injuries do not result
from a pin directly penetrating the nerve, but from a pin going through
adjacent tissue, resulting in pinching of the nerve. Michael and
Stanislas¹⁴⁰ described another
alternative for protecting the ulnar nerve; they attached a nerve
stimulator to a needle, which was used for localizing the ulnar nerve.
Once the ulnar nerve was identified, a standard pinning technique was
used, placing the medial pin 0.5 to 0.75 mm anterior to the nerve. We
have no experience with this technique. Some surgeons palpate the ulnar
nerve and push it posteriorly. Wind et al.²⁰⁶
however, showed this method to be unreliable. The only technique for
avoiding iatrogenic ulnar nerve injury appears to be to use
lateral-entry pins and avoid the use of crossed pins.

Loss of motion after extension-type supracondylar fractures is rare in children. Two series analyzed this complication in detail⁴⁷,⁸⁸
and found that fractures treated closed had an average loss of motion
of 4 degrees. In those treated with open reduction, the loss of flexion
was 6.5 degrees and the flexion contracture was 5 degrees. In a series
of 43 children with supracondylar fractures who had open reduction,
about half received no physical therapy.¹⁰⁷
While the group with physical therapy had better motion at 12 and 18
weeks postoperatively, there was no difference in elbow motion after 1
year. The authors concluded that physical therapy is unnecessary. Loss
of motion has been reported with the posterior triceps splitting
incision for open reduction.⁷⁷,⁸¹,¹³¹
While most children do not require formal physical therapy, we
generally teach the parents range-of-motion exercises to be done at
home after pin and cast removal at about 3 weeks. A follow-up
appointment to assess range of motion is scheduled about 4 to 6 weeks
later, and if motion is not nearly normal at that time, physical
therapy to improve elbow motion is begun.

Although motion loss usually is minimal, significant
loss of flexion can occur. This problem generally is caused by a lack
of anatomic fracture reduction: either posterior distal fragment
angulation, posterior translation of the distal fragment with anterior
impingement, or medial rotation of the distal fragment with a
protruding medial metaphyseal spike proximally (Fig. 14-46).
In young children with significant growth potential, there may be
significant remodeling of anterior impingement, and any corrective
surgery should be delayed at least 1 year. Although anterior
impingement can significantly remodel, there is little remodeling to
persistent posterior angulation or hyperextension.

The prevalence of pin track infections in pediatric
fractures treated with percutaneous K-wire fixation varies from less
than 0% to 21%.²⁰,⁹⁵ The reported prevalence of pin track infections in supracondylar humeral fractures ranges from 0% to 6.6%.³⁰,⁴¹,⁹⁵,¹³⁸,¹⁸² In a series of 202 pediatric fractures, 92.5% of which were upper extremity fractures, Battle et al.²⁰
reported an 8% infection rate. Twelve of the 16 infections required
oral antibiotics and local pin care, one required intravenous
antibiotics, and three required operative incision and débridement. In
124 supracondylar fractures, Skaggs et al.¹⁸² noted only one pin track infection (less than 1%) which resolved with oral antibiotics and pin removal. Gupta et al.⁸²
also reported only one (<1%) pin track infection in 150 patients,
which also resolved with oral antibiotics and pin removal. In a larger
series by Mehlman et al.¹³⁸ (198
fractures), there were five (2.5%) pin track infections, which were
treated with oral antibiotics and resolved without sequlae. Iobst et al.⁹⁵
reported no infections in 304 patients using a “semisterile” technique,
despite the fact that only 68% of their patients received perioperative
antibiotic. Their protocol involved unsterile reduction of the fracture
with an assistant holding the arm in the reduced position. The surgeon
then put on sterile gloves, placed sterile towels around the surgical
field, and prepared the elbow locally in the anticipated area of the
pin sites. There are no other reports on this technique.

Pin track infections generally resolve with pin removal
and antibiotics. Fortunately, by the time a pin track infection
develops the fracture usually is stable enough to remove the pin
without loss of reduction. Although we have not seen this complication,
an untreated pin track infection could theoretically result in a septic
joint and should be treated as soon as it is recognized or suspected.

Myositis ossificans is a remarkably rare complication of supracondylar fractures, but it does occur (Fig. 14-47). This complication has been described after open reduction, but vigorous

postoperative manipulation or physical therapy is believed to be the most commonly associated factor.¹¹³,¹⁵⁷

In a report of two patients with myositis ossificans after closed reduction of supracondylar fractures, Aitken⁵
noted that limitation of motion and calcification disappeared after 2
years. Postoperative myositis ossificans can be observed with the
expectation of spontaneous resolution of both restricted motion and the
myositis ossificans. There is no indication for early excision. Spinner¹⁹⁰
reported a single case of myositis ossificans associated with sudden
onset of pain after trauma in which a 1-year-old lesion of myositis
ossificans was fractured. With excision, the pain was relieved, and
full range of motion returned.

The distal humeral metaphysis is a well-vascularized
area with remarkably rapid healing, and nonunion of a supracondylar
fracture is rare, with only a single case described by Wilkins and Beaty²⁰²
after open reduction. We have not seen nonunion of this fracture. With
infection, devascularization, and soft tissue loss, the risk of
nonunion would presumably increase.

Osteonecrosis of the trochlea after supracondylar
fracture has been reported. The blood supply of the trochlea’s
ossification center is fragile, with two separate sources. One small
artery is lateral and courses directly through the physis of the medial
condyle. It provides blood to the medial crista of the trochlea. If the
fracture line is very distal, this artery can be injured, producing
osteonecrosis of the ossification center and resulting in a classic
fishtail deformity (Fig. 14-48). Kim¹⁰⁹
identified 18 children with trochlear abnormalities after elbow
injuries, five of which were supracondylar fractures. MRI indicated
low-signal intensity on T2, indicative of cartilage necrosis. Cubitus
varus

deformity developed in all cases. Bronfen et al.³²
described six patients with severe osteonecrosis of the trochlea, which
they termed “dissolution of the trochia” following closed reduction and
pin fixation, indicating that open reduction is not always the cause of
osteonecrosis. Five of their six patients complained of pain,
crackling, and stiffness.³²

Symptoms of osteonecrosis of the trochlea do not occur
for months or years. Healing is normal, but mild pain and occasional
locking develop with characteristic radiographic findings and motion
may be limited depending on the extent of osteonecrosis. In our
experience, the most common cause of osteonecrosis of the trochlea is
open reduction of a supracondylar fracture through a posterior
approach, which presumably disrupts the blood supply of the trochlea (Fig. 14-49);
however, we are aware of several cases of osteonecrosis that occurred
after simple closed pinning. The artery to the trochlea may be
disrupted at the time of the injury.

Clinical experience with a series of 124 consecutive
supracondylar humeral fractures, including completely unstable
fractures, has taught us that lateral-entry pins, when properly placed,
are strong enough to maintain reduction of even the most unstable
supracondylar humeral fractures.¹⁸²
In an informal collection of reports of loss of reduction of
supracondylar fractures from members of the Pediatric Orthopaedic
Society of North America, Skaggs et al.¹⁸²
reported only eight fractures with loss of reduction after treatment
with lateral-entry pins. They identified technical points of pin
placement that may be associated with loss of fracture reduction. Two
lateral pins were used for all eight fractures. In five, the pins were
too close together and engaged only the lateral column at the level of
the fracture (Fig. 14-50). In each of these,
fracture reduction was lost when the distal fragment rotated around the
two pins. In another case, the two lateral pins crossed at the fracture
site (Fig. 14-51). In the remaining two cases, the two lateral pins did not engage the distal fragment (Figs. 52A,B). There were no reports of loss of fixation after the use of three lateral-entry pins.

Sankar et al.,¹⁷⁴ in
a series of 322 factures, reported that 2.9% had postoperative loss of
fixation. All eight were type III fractures treated with two pins
(seven lateral-entry and one crossed-pin). In all eight, loss of
fixation was due to technical errors that were identifiable on the
intraoperative fluoroscopic images (Fig. 14-53)
and that could have been prevented with proper technique. They
identified three types of pin-fixation errors: (i) failure to engage
both fragments with two pins or more, (ii) failure to achieve
bicortical fixation with two pins or

more, and (iii) failure to achieve adequate pin separation (>2 mm) at the fracture site.¹⁷⁴

Simanovsky et al.¹⁸⁰
reported 22 patients near skeletal maturity with underreduction of the
extension component of the fracture. They found that 17 (77%) of the
patients had radiographic abnormalities of the humerocondylar angle (a
difference of 5 degrees or more compared with the uninjured side).
Eleven patients (50%) had limited elbow flexion, and 7 (31%) were aware
of this deficit. Patients who were left to heal with some degree of
extension developed limited end-elbow flexion and were aware of it.
Although only 3 patients perceived minor subjective functional
disability at the last follow-up, 10 patients had unsatisfactory
results according to the Flynn criteria for motion restriction.¹⁸⁰

Cubitus varus, also know as a “gunstock deformity,” has a characteristic appearance in the frontal plane (Fig. 14-54).
The malunion most commonly also includes hyperextension, which leads to
increased elbow extension and decreased elbow flexion (Figs. 14-55 and 14-56).
The appearance of cubitus varus deformity is distinctive on radiograph.
On the AP view, the angle of the physis of the lateral condyle (Baumann
angle) is more horizontal than normal (Fig. 14-57).
On the lateral view, hyperextension of the distal fragment posterior to
the anterior humeral line correlates with the clinical findings of
increased extenion and decreased flexion of the elbow (Fig. 14-58).

Unequal growth in the distal humerus has been proposed as a cause of cubitus varus deformity,⁹³,¹⁵⁵
though this is unlikely because not enough growth remains in this area
to cause cubitus varus within the time it is recognized. A more likely
reason for most cubitus varus deformities in patients with
supracondylar fractures is malunion.¹²,³⁵,⁵¹,⁶⁴,²⁰⁰
Cubitus varus can be prevented by making certain that the Baumann angle
is intact at the time of reduction and remains so during healing, which
generally is best assured with pin fixation. In a series of 206
patients with supracondylar humeral fractures, with ages ranging from
1.5 to 14 years (mean of 6.4 years), Pirone et al.¹⁵⁷
reported cubitus varus deformities in eight (7.9%) of 101 patients
treated with cast immobilization compared to two (1.9%) in 105 patients
treated with pin fixation. A decrease in the frequency of cubitus varus
deformity after the use of percutaneous pin fixation has been reflected
in other recent series.²⁹,³⁰,⁶⁴,⁶⁵,¹⁰¹,¹³⁹,¹⁵⁷,¹⁵⁸ with one large retrospective¹⁸² and one prospective study¹¹¹ reporting no cubitus varus deformities.

The distal humeral physis provides 20% of that the
overall length of the humerus. In a 5-year-old, therefore, the amount
of distal humeral growth in 1 year is approximately 2 mm, making it
unlikely that growth asymmetry is a significant cause of varus
deformity that occurs within the first 6 to 12 months after fracture. A
case report of a 4-year-old girl with progressive cubitus varus caused
by a physeal bony bar demonstrated with MRI (Fig. 14-59) after cast treatment of a supracondylar fracture noted the slow progress of the deformity.⁵⁶
The Baumann angle went from 88 degrees at 6 months following injury to
92 degrees at 18 months and 96 degrees at 3 years after injury. Thus,
over

3
years from injury, the Baumann angle changed by only 8 degrees, which
we suspect is not too different from measurement variability.
Osteonecrosis of the trochlea or medial portion of the distal humeral
fragment can result in progressive varus deformity, however. In a
series of 36 varus deformities reported by Voss et al.,¹⁹⁷ only 4 patients had medial growth disturbance and distal humeral osteonecrosis as a cause of progressive varus deformity.

Treatment of cubitus varus deformities has been
considered a primarily cosmetic procedure, but several consequences of
cubitus varus such as an increased risk of lateral condylar fractures,
pain, and tardy posterolateral rotatory instability may be indications
for supracondylar humeral osteotomy.¹,²,²³,¹³⁵,¹⁴⁸,¹⁸⁸ Our experience suggests that many patients have elbow discomfort with significant cubitus varus. O’Driscoll et al.¹⁴⁸
reported 17 patients with cubitus vars deformities after a pediatric
distal humeral fracture who presented an average of 27 years following
injury. The average varus deformity was 15 degrees, and all patients
had tenderness over the lateral collateral

ligament
complex and posterolateral rotator instability. These authors concluded
that cubitus varus deformity secondary to supracondylar malunion may
not always be a benign condition and may have important long-term
clinical implications.¹⁴⁸ An increased risk of fracture,⁵³ especially of the lateral condyle, has been linked with cubitus varus deformity. Takahara et al.¹⁹³
reported nine patients with distal humeral fractures who deveoped varus
deformities. Supracondylar fractures as well as epiphyseal separations
were included in these nine fractures.

Tardy ulnar nerve palsy also has been associated with cubitus varus and internal rotational malalignment.²,²⁴,¹⁴²,¹⁹⁶ With a cubitus varus deformity, the olecranon fossa moves to the ulnar side of the distal humerus,¹⁴⁹
and the triceps shifts a bit ulnarward. Investigators theorized that
this ulnar shift might compress the ulnar nerve against the medial
epicondyle, narrowing the cubital tunnel and resulting in chronic
neuropathy. In a recent report,² a fibrous band running between the heads of the flexor carpi ulnaris was thought to cause ulnar nerve compression.

Treatment of Cubitus Varus Deformity. As for the
treatment of any posttraumatic malalignment, options include (i)
observation with expected remodeling, (ii) hemiepiphysiodesis and
growth alteration, and (iii) corrective osteotomy. Observation
generally is not appropriate because, although hyperextension may
remodel to some degree in a young child (Fig. 14-60), in an older child little remodeling occurs even in the plane of motion of the joint.

Hemiepiphysiodesis of the distal humerus may rarely be
of value, but only to prevent cubitus varus deformity from developing
in a patient with clear medial growth arrest or trochlear
osteonecrosis. If untreated, medial growth disturbance will lead to
lateral overgrowth and progressive deformity. Lateral epiphysiodesis
will not correct the deformity, but will prevent it from increasing.
Voss et al.¹⁹⁷ used
hemiepiphysiodesis with osteotomy in two patients with growth arrest
and varus deformity. The humerus varies in length by a few centimeters
from one individual to another, but in general, it is about 30 cm long
at skeletal maturity. Approximately 65% of the length of the humerus is
achieved by age 6 years. A 6-year-old child has approximately 10 cm of
growth left in the entire humerus, with

only
approximately 2 cm provided by the distal physis. Growth arrest, in the
absence of osteonecrosis or collapse, will be a very slowly evolving
phenomenon, and epiphysiodesis in a child older than 6 years will have
little effect on longitudinal growth. In general, prevention of
increasing deformity from medial growth arrest is the only role for
lateral epiphysiodesis. Because of the slow growth rate in the distal
humerus, we do not believe there is any role for lateral epiphysiodesis
in correcting a varus deformity in a child with otherwise normal physis.

Osteotomy. Osteotomy is the only way to correct a
cubitus varus deformity with a high probability of success. A variety
of corrective osteotomies have been described, almost all with
significant complications (Table 14-3). Stiffness, nerve injury,

and recurrent deformities are the most commonly reported complications.
An overall complication rate approaching 25% in many series has led to
some controversy about the value of a distal humeral corrective
osteotomy for cubitus varus deformity.

To choose an appropriate osteotomy, the exact location
of the deformity must be determined. Because malunion is the cause of
most cubitus varus deformities, the angular deformity usually occurs at
the level of the fracture. Rotation and hypertension may contribute to
the deformity, but varus is the most significant factor.⁴³ Hyperextension can produce a severe deformity in some patients. An oblique configuration (Fig. 14-61)
places the center of rotation of the osteotomy as close to the actual
level of the deformity as possible. On an AP radiograph of the humerus
with the forearm in full supination, the size of the wedge and the
angular correction needed are determined. An “incomplete” lateral
closing wedge osteotomy can be done, leaving a small medial hinge of
bone intact. The osteotomy usually is fixed with two K-wires placed
laterally. In the absence of an intact medial hinge, two lateral wires
probably are not sufficient to secure this osteotomy.¹⁹⁷ Wilkins and Beaty²⁰²
recommended crossed wires in this situation. In general, a lateral
closing wedge osteotomy with a medial hinge will correct the varus
deformity, with minimal correction of hyperextension.¹⁰,²¹,⁴⁶,⁶⁸,⁷⁰,⁷⁹,⁹⁸,¹⁰⁰,¹¹⁶,¹⁹²,¹⁹⁷,²⁰⁷
but this osteotomy tends to leave a lateral bump with poor cosmesis.
Residual rotational deformity was not found to be a significant problem
in studies by Voss et al.¹⁹⁷ and Oppenheim et al.,152 which is logical given the amount of rotation available from the shoulder. The French osteotomy⁶⁸ aims to enable axial rotational correction as well, but does so at the expense of stability and we do not use this osteotomy.

Japanese surgeons⁹⁰
described a dome osteotomy in which a curved osteotomy is made in the
supracondylar area. Proponents of this osteotomy suggest that
multiplane correction is possible without inducing translation in the
distal fragment and that rotation can be corrected. DeRosa and Graziano⁵⁷
described a step-cut osteotomy in which the distal fragment is slotted
into the proximal fragment and the osteotomy is secured with a single
screw.

Functional outcomes generally are good, but the
preoperative functional deficit is nearly always minor in patients with
cubitus varus deformities. Complications of humeral osteotomy include
stiffness, nerve injury, and persistent deformity (see Table 14-3); however, with a properly performed osteotomy, complications are relatively few. Ippolito et al.⁹⁶ reported long-term follow-up of patients with supracondylar osteotomies, 50% of whom had poor results. Weiss et al.²⁰¹
reported a complication rate of 16% (nine of 57 patients) with various
types of osteotomies. It was notable that the authors found a
complication rate of 6% (two of 34) when the surgery was done by
full-time staff at a tertiary pediatric referral center, with no
patient requiring additional surgery and all having excellent results.
The complication rate among patients treated by non-full-time staff was
30% (seven of 23), including two transient ulnar nerve palsies, four
deformity recurrences or loss of fixation requiring reoperation, and
one deep infection requiring operative treatment. This can be a
challenging operation, and it is not surprising that surgeons with more
experience have fewer complications.

Hyperextension deformity may remodel over time (see Fig. 14-60), but correction is slow and inconsistent. In one series,¹⁹⁷
hyperextension deformities remodeled as much as 30 degrees in very
young children, but in older children, there was no significant
remodeling in the flexion-extension plane. If hyperextension appears to
be a major problem, osteotomy should also be directed at this deformity
rather than simple correction of the varus deformity; this situation
requires a multiplane osteotomy.

A transverse incision is made across the antecubital
fossa, curving proximally posterior to the neuromuscular bundle.
Dissection is carried down to the level of the superficial fascia of
the forearm and antecubital fossa. The neurovascular bundle is
identified and retracted medially. The brachialis and biceps tendons
are retracted laterally to expose the fracture site and facilitate
reduction. If there is medial soft tissue impingement or a question of
ulnar nerve entrapment within the fracture, the dissection should be
carried around posterior to the medial epicondyle, so the ulnar nerve
and fracture can be identified.

Postoperative immobilization is maintained for 3 or 4
weeks until good callus formation is present. Pins generally are left
out through the skin and removed in the office without the need for
anesthetic. No formal rehabilitation is given, but the patient is
encouraged to begin gentle activities with the arm and to begin
regaining motion without a stressful exercise program.