Evaluation of the role of pin xation versus collar and cu

immobilisation in supracondylar fractures of the humerus in

children
J.G. Kennedy*, K. El Abed, K. Soe, S. Kearns, D. Mulcahy, F. Condon, D. Moore,
F. Dowling, E. Fogarty
Our Lady's Hospital for Sick Children, Dublin, Ireland
Accepted 19 October 1999
Abstract
Long term results of children with supracondylar humeral fractures treated with manipulation and strapping and
manipulation followed by pin xation were evaluated.
Forty patients were regarded as Gartland type II injuries. 33 of these were treated with closed reduction and collar and cu
immobilisation and 7 with closed reduction and percutaneous pinning. Two cases of cubitus varus were reported one from each
treatment modality
Forty-four patients were included as Gartland type III injuries. Of these 14 were treated with closed reduction and collar and
cu immobilisation, 25 with closed reduction and percutaneous pinning and ve with open reduction and pinning. There were
two cases of cubitus varus and one case of cubitus valgus following pin xation. In addition one case of extension lag and one
signicant ulnar nerve neurapraxia was recorded following pin xation. One case of cubitus varus was seen following
manipulation and collar and cu treatment.
There was no statistical dierence between either treatment modality in terms of predicting a better outcome ( p >0.05).
We conclude that pin xation has no advantages over simple immobilisation in certain Gartland II and III type injuries.
Although pin xation is benecial in unstable injuries collar and cu immobilisation continues to have an important role in the
treatment of stable supracondylar fractures. 7 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Supracondylar; Humerus; Fracture; Collar and cu
1. Introduction
Supracondylar fractures of the humerus are the
most common injury about the elbow in children [1].
Many treatment algorithms have been used in the past
in the management of these fractures and the current
literature now suggests that closed or open reduction
with percutaneous pinning should be employed for
Gartland type II and type III fractures [24].
This study reviewed 91 supracondylar fractures to
determine the relative merits of closed reduction with
collar and cu immobilisation as an alternative to per-
cutaneous pin xation and to evaluate if our results
would direct future practice guidelines.
2. Materials and methods
One hundred and ninety-three patients were ident-
ied from the hospital data base with a diagnosis of
supracondylar humeral fracture between September
1992 and July 1994. All open injuries and complex
elbow fracture dislocations were excluded from the
study. Ninety-one patients were available for clinical
Injury, Int. J. Care Injured 31 (2000) 163167
0020-1383/00/$ - see front matter 7 2000 Elsevier Science Ltd. All rights reserved.
PII: S0020- 1383( 99) 00274- 0
www.elsevier.com/locate/injury
* Corresponding author. Tel.: +353-1-803-1789.
E-mail address: jgk@indigo.ie (J.G. Kennedy).
and radiographical follow up. Minimum time to follow
up was 2 yr (range 2 yr5 yr 2 months; mean 3 yr 1
month) The mean age was 8 yr 2 months (range 3 yr
10 months10 yr 2 months). Male to female ratio was
1.2:1 and there was a preponderance for right sided
fractures 2.1:1
All radiographs were classied with respect to the
Gartland classication [5]. Forty patients were
regarded as Gartland type II fractures, 44 Gartland III
fractures and seven exion type injuries.
Treatment of the injury was guided by fracture con-
guration. Gartland type II injuries in which the pos-
terior cortex remained intact were manipulated under
general anaesthesia, the elbow exed to above 1208
and held with a collar and cu inside an open weave
thoraco-brachial bandage. Following a period of 24 h
in which neurovascular sequelae were excluded the
patient was allowed home, to return within 5 days and
again at 21 days for a check radiograph. If this was
satisfactory the collar and cu was removed from the
child and a gradual return to normal activity was
encouraged over a 2-week period.
Of those patients in whom the fracture alignment
was considered unsatisfactory at initial follow up,
remanipulation and xation with percutaneous Kirsch-
ner pins was used. The decision to percutaneously pin
type II fractures was based on the surgeons intraopera-
tive evaluation of stability at the fracture site. If ana-
tomical alignment could not be maintained when the
elbow was exed over 1208 and ranged through 458
arc of supination and pronation, either crossed lateral
or mediallateral crossed percutaneous pins were used
to hold the reduction. Both crossed mediallateral and
two lateral smooth Kirshner pins were used in the
study.
The decision to percutaneously pin type III fractures
was again based on intraoperative evaluation of stab-
ility of the fracture once reduction had been achieved.
A fracture was considered stable if following re-
duction, position could be maintained with greater
than 1208 of exion whilst the forearm was pronated
and supinated under uoroscopic visualisation [4,6]. If
this was achieved the fracture was held in a collar and
cu at >1208 of exion under an open weave thoraco-
brachial bandage. A similar post operative regimen
was employed for the type III injuries as for the type
II injuries.
Thirty three type II injuries were treated with a
closed reduction and immobilisation in collar and cu
and seven were treated with closed reduction and per-
cutaneous pin xation as the index procedure. There
were three remanipulations in this group all of which
required percutaneous pin xation.
There were 44 patients who sustained a Gartland
type III injury. Of these, 14 were treated with closed
reduction and collar and cu, 25 were treated with
closed reduction and percutaneous wire xation and
ve were treated with open reduction and wire x-
ation. Percutaneous pinning of type III injuries was
reserved for unstable fractures and both lateral crossed
pinning and medial and lateral pinning were used.
Open reduction was employed only when the fracture
failed to achieve normal anatomical alignment follow-
ing manipulation or when a nerve palsy was identied
at presentation to the accident and emergency room.
All patients were evaluated with a Flynn rating scale
at follow up [7]. The humerocapital angle was evalu-
ated from the anteroposterior radiograph as an indi-
cation of varus, valgus angulation. The humero-ulnar
angle was evaluated from lateral radiographs to deter-
mine exion or extension of the fracture. Nerve inju-
ries related to the initial trauma, the reduction
manoeuvre employed or from pin injury which had
resolved at nal follow up were not included in the
Flynn outcome score.
Logistical regression analysis was used to evaluate
factors which could be used to predict outcome in this
series. w
2
analysis was used to determine if there was a
signicant dierence in Flynn score between those
treated with manipulation and pin xation and those
treated with manipulation followed by collar and cu
immobilisation alone. (SPSS Version 6.1 Chicago, IL)
3. Results
Thirty eight of forty patients with Gartland II type
injuries achieved a good to excellent outcome score on
the Flynn grading system (Table 1). The two patients
who did not obtain a satisfactory outcome developed
cubitus varus. Of these, one had a medial tilt at pres-
entation which was not appreciated and was resistant
to realignment ten days following the injury. The sec-
ond patient was treated with mediallateral crossed
percutaneous pins. The initial postoperative radio-
graph was satisfactory however the patient failed to
show up for a further three months and then for an
infected pin tract. Three remanipulations were required
in this group two of which were pinned to maintain re-
duction. The other remanipulation, as described, was
resistant to anatomic realignment secondary to delayed
presentation. Apart from one case of cubitus varus,
again related to a delay in follow up, there were no
complications related to pinning in this cohort.
Of the 44 patients with Gartland type III injuries, 39
patients scored a good to excellent Flynn outcome
score. Twenty-ve children were treated with pin x-
ation following closed reduction. Complications within
this group treated with pin xation were divided
between early and delayed sequelae. Early compli-
cations included one remanipulation, one transient me-
dian nerve neurapraxia and six pin track infections all
J.G. Kennedy et al. / Injury, Int. J. Care Injured 31 (2000) 163167 164
of which had resolved at nal follow up (Table 2).
Late complications related to pin xation included two
cubitus varus, one cubitus valgus, one ulnar nerve
injury and one extension lag. Of the two cases of cubi-
tus varus in this treatment group, one was due to inap-
propriate pinning with two crossed lateral pins and
one was following mediallateral crossed pin xation
(Table 3).
Fourteen patients with type III fractures were trea-
ted with collar and cu immobilisation following
closed reduction. Complications included one case of
cubitus varus and two remanipulations ultimately
requiring pin xation. The cubitus varus was second-
ary to unrecognised tilting of the condylar fragments
and not to premature physeal arrest.
Five of the seven exion type injuries were
regarded as type II injuries whilst two were
regarded as type III injuries. There were three
closed manipulations and collar immobilisations in
this group all type II injuries. There were three
patients treated with closed reduction and pin x-
ation of which two were type II injuries and one
type III injury. There was a single case of open re-
duction internal xation in a type III injury resist-
ant to manipulation. There were no complications
within this group and all patients recorded a good
to excellent Flynn score at follow up.
Logistical regression analysis demonstrated that
there was no association between age, sex, congur-
ation of pin xation and the development of a poor
outcome in this study P 0:075, p 0:862,
p 0:0921). There was no statistical association
between the use of closed reduction and collar and cu
immobilisation and a poor outcome in either type II
or type III injuries. There was a correlation however
between the use of percutaneous pins and an increased
risk of supercial infection.
No vascular compromise was recorded nor was
there evidence of Volkmann's ischaemic contracture at
follow up in the cohort.
Table 1
Flow chart demonstrating type of injury, method of treatment and outcome
J.G. Kennedy et al. / Injury, Int. J. Care Injured 31 (2000) 163167 165
4. Discussion
Despite supracondylar fractures of the humerus in
children being common, management of the injury
remains controversial. In general, recommended treat-
ment modalities have been constructed to prevent the
signicant complications associated with this injury.
The trend away from closed reduction and cast immo-
bilisation in type II and III injuries has been prompted
by reports of an increased risk of complications. Millis
and Hall [8] have shown that failure to ex the elbow
>1208 led to an 86% risk of loss of reduction in those
with cast immobilisation. Aronson et al and Kallio et
al have demonstrated that as the swelling subsides the
risk of coronal tilting and cubitus varus is increased
unless the fracture is rigidly immobilised [9,10]. Pir-
rone, in a review of 230 patients with displaced supra-
condylar fractures reported that of those treated with
closed reduction and cast immobilisation there was an
increased risk of vascular compromise and failed re-
duction [3]. The current literature therefore rec-
ommends closed reduction and percutaneous pinning
of grades II and III fractures [2,4,1113]. Despite
encouraging results regarding a decrease in vascular
compromise and in maintaining anatomical alignment
the technique of percutaneous pin xation is not with-
out risk. Iatrogenic nerve injury, pin track infection
and loss of reduction have all been reported [3,1417].
This review evaluates the role of closed reduction
and collar and cu immobilisation in a group of
patients who by current convention would have been
treated with pin xation.
38 patients out of 40 (95%) with Gartland II injuries
achieved a good to excellent outcome. As there was no
statistical dierence in outcome between those treated
with pin xation and those treated with collar and cu
following reduction no conclusions regarding preferred
treatment patterns can be drawn. This suggests that
both modalities are equally eective and that there is a
role for both depending on fracture stability.
Thirty-nine patients of 44 Gartland III type frac-
tures achieved a good to excellent Flynn score at nal
Table 2
Overall complications associated with fracture type and treatment
Gartland type Treatment Complication Number
Extension
II (40) closed reduction and collar and cu (33) cubitus varus 1
remanipulation 3
closed reduction and pinning (7) cubitus varus 1
III (44) closed reduction and collar and cu (14) cubitus varus 1
remanipulation 2
closed reduction and pinning (25) cubitus varus 2
remanipulation 1
pin track infection 6
transient neuropraxia 1
permanant neuropraxia 1
cubitus valgus 1
extension lag 1
open reduction and pinning (5) none
Flexion
II (5) closed reduction and collar and cu (3) none
closed reduction and pinning (2) none
III (2) open reduction and pinning (1) none
closed reduction and pinning (1) none
Table 3
Long term complications associated with specic treatment types
Complication Fracture type Number Treatment
Cubitus varus Gartland II 2 (1) closed reduction alone
(2) mediallateral crossed pins
Gartland III 3 (1) lateral crossed pins
(2) re-closed reduction and mediallateral crossed pins
(3) open reduction
Cubitus valgus Gartland III 1 (1) closed reduction alone
Extension lag Gartland III 1 (1) mediallateral crossed pins
Ulnar nerve neuropraxia Gartland III 1 (1) mediallateral crossed pins
J.G. Kennedy et al. / Injury, Int. J. Care Injured 31 (2000) 163167 166
follow up. The number of patients with pin xation at
the index procedure was greater than that in the Gart-
land III group as the severity of the fracture had
increased in association with a decrease in stability.
Although pin xation in this group demonstrated a
greater number of poorer overall outcomes than closed
reductions alone this did not reach statistical signi-
cance. This should not be regarded as a failure of pin
xation as compared with reduction alone but rather
the advancing level of complexity and instability of the
fracture for which pin xation was chosen. Despite
this however the number of early complications associ-
ated with this method of xation were not insignicant
with six pin track infections and one transient iatro-
genic neuropraxia. Although these complications were
transitory and resolved before nal follow up they
were a cause for increased hospital stay and concern.
We have addressed the high pin track infection rate
subsequently by burying all pins subcutaneously redu-
cing possible tracking infection.
It is clear therefore that there is a role for collar and
cu immobilisation following closed reduction in
selected type II and III fractures which are stable fol-
lowing manipulation and anatomical alignment. Simi-
lar results have been documented by Williamson et al.
in a review of 69 patients in which they report that
sixty percent of their cohort of isolated supracondylar
fractures could be treated with inelastic strapping and
collar and cu immobilisation [18]. Hadlow et al have
also demonstrated that successful closed reduction and
casting in both types II and III fractures obviated the
need for crossed pin xation in 61% of type III frac-
tures and 77% of type II fractures [19].
The goals of any treatment modality in supracondy-
lar fractures of the humerus are to obtain and main-
tain an anatomical alignment and to re-establish a
functional joint with a normal range of motion and a
cosmetically satisfactory appearance. Similarly any
treatment should be safe and provide reproducible
results [20,21]. We have demonstrated that closed re-
duction with collar and cu immobilisation sup-
plemented with a thoraco brachial bandage is both
eective and safe in treating stable types II and III
fractures. The risk of vascular compromise experienced
with cast immobilisation appears to be minimised by
the non circumferentially constricting collar and cu.
There continues to be a clearly dened role for percu-
taneous cross wire xation in those fractures that
demonstrate instability at the time of initial reduction
or which fail to maintain satisfactory alignment at fol-
low up. In the era of cost containment and in an
increasing litiginous society, however, the compli-
cations associated with obligatory pin xation of all
types II and III fractures must be weighed carefully
against the satisfactory outcome achieved using less
costly and less invasive procedures.
References
[1] Cheng JCY, Shen WY. Limb fracture patterns in dierent pae-
diatric age groups; a study of 3,350 children. J Orthop Trauma
1993;7:1522.
[2] Otsuka NY, Kasser JR. Supracondylar fractures of the humerus
in children. J Am Acad Orthop Surg 1997;5(1):1926.
[3] Pirone AM, Graham HK, Krajbich JI. Management of dis-
placed extension type supracondylar fractures of the humerus in
children. J Bone Joint Surg (Am) 1988;70:64150.
[4] Wilkins KE. Operative management of upper extremity frac-
tures in children. In: Wilkins KE, editor. American Academy of
Orthopaedic Surgeons monograph series, 1994.
[5] Gartland JJ. Management of supracondylar fractures of
the humerus in children. Surg Gynecol Obstet 1959;109:
14554.
[6] Khare GN, Gautam VK, Kochler VL. Prevention of Cubitus
varus deformity in supracondylar fractures of the humerus.
Injury 1991;22:2026.
[7] Flynn JC, Mathews JG, Benoit RL. Blind pinning of displaced
supracondylar fractures of the humerus in children. Sixteen
years of experience with long-term follow up. J Bone Joint Surg
(Am) 1974;56:26372.
[8] Millis MB, Singer IJ, Hall JE. Supracondylar fractures of the
humerus in children: further experience in orthopaedic decision
making. Clin Orthop 1984;188:907.
[9] Aronson DD, Prager BI. Supracondylar fractures of the
humerus in children: a modied technique for closed pinning.
Clin Orthop 1987;219:17484.
[10] Kallio PE, Foster BK, Patterson DC. Dicult supracondylar
elbow fractures of the humerus in children; analysis of per-
cutaneous pinning technique. J Paediatr Orthop 1992;12:
169.
[11] Furrer M, Mark G, Ru edi T. Management of displaced supra-
condylar fractures of the humerus in children. Injury
1991;22:25962.
[12] Mehserle WL, Meehan PL. Treatment of the displaced supra-
condylar fracture of the humerus with closed reduction and
percutaneous cross xation. J Peadiatr Orthop 1991;11:
70511.
[13] Wilkins KE. The management of severely displaced supracondy-
lar fractures of the humerus. Techniques Orthop 1989;4:524.
[14] Arino VL, Lluch EE, Ramirez AM. Percutaneous xation of
supracondylar fractures of the humerus in children. J Bone and
Joint Surg (Am) 1977;59:9146.
[15] Brown IC, Zinar DM. Traumatic and iatrogenic neurological
complications after supracondylar humeral fractures in children.
J Paediatr Orthop 1995;15(4):4403.
[16] Campell CC, Waters PM, Emans JB, Kasser JR, Millis MB.
Neurovascular injury and displacement in type III supra-
condylar humeral fractures. J Paediatr Orthop 1995;15(1):
4752.
[17] Ikram MA. Ulnar nerve palsy; a complication following percu-
taneous xation of supracondylar fractures of the humerus in
children. Injury 1996;27(5):3035.
[18] Williamson DM, Cole WG. Treatment of selected supracondy-
lar fractures of the humerus by manipulation and strapping in
exion. Injury 1993;24(4):24952.
[19] Hadlow AT, Devane P, Nicol RO. A selective treatment
approach to supracondylar fracture of the humerus in children.
J Paediatr Orthop 1996;16(1):1046.
[20] Wilkins KE. The operative management of supracondylar frac-
tures. Orthop Clin North Am 1990;21:26989.
[21] Worlock PH, Colton CL. Displaced supracondylar fractures of
the humerus in children treated by overhead olecranon traction.
Injury 1984;15:31621.
J.G. Kennedy et al. / Injury, Int. J. Care Injured 31 (2000) 163167 167