SPINE Volume 24, Number 16, pp 1646 –1654
©1999, Lippincott Williams & Wilkins, Inc.
Urgent Surgical Stabilization of Spinal Fractures in
Polytrauma Patients
Robert F. McLain, MD,* and Daniel R. Benson, MD
Study Design. A prospective, longitudinal study of
multiply injured patients treated with segmental instru-
mentation for spinal fractures with a minimum 2-year fol-
low-up.
Objectives. To determine whether urgent stabilization
of spinal fractures in severely injured patients increases
the risk of surgery compared with early treatment and
historical results.
Summary and Background Data. Opinion in clinical
studies is divided about whether operative treatment of-
fers an advantage over nonoperative treatment in iso-
lated spine fractures. Concomitant trauma is rarely dis-
cussed relative to decision making or surgical timing.
Urgent stabilization of long-bone fractures improves sur-
vival and outcome in polytrauma patients. To date, urgent
treatment of spine fractures in polytrauma patients has
not been considered in the literature.
Methods. Seventy-five consecutive patients treated
with segmental instrumentation for spinal trauma were
observed prospectively to assess perioperative and long-
term outcome. Twenty-seven patients with severe poly-
trauma (injury severity score, !26) were separately ana-
lyzed. Perioperative and postoperative results were
analyzed relative to timing of surgery, injury severity
score, and surgical approach. Urgent treatment was de-
fined as that provided within 24 hours of the spinal injury,
and early treatment was defined as that provided be-
tween 24 and 72 hours after injury.
Results. Twenty-five patients (93%) sustained two or
more major injuries in addition to the spine fracture, and
17 of 27 (63%) had neurologic injury. The mean injury
severity score approached or exceeded the LD50 (50%
expected mortality) in each group—36.0 for the early-
treatment group and 42.0 for the urgent group— but only
one patient in each group died. There were no deep ve-
nous thromboses, pulmonary emboli, neurologic injuries,
decubiti, deep wound infections, or episodes of sepsis in
either group. Blood loss for anterior procedures was sig-
nificantly higher in the urgent group, but estimated blood
loss for posterior procedures was similar for both group-
s.At 49 months’ mean follow-up, no revisions were ne-
cessitated by the urgent spinal treatment.
Conclusions. Urgent spinal stabilization is safe and
appropriate in polytrauma patients when progressive
neurologic deficit, thoracoabdominal trauma, or fracture
instability increase the risks of delayed treatment. [Key
words: Cotrel–Dubousset, fractures, instrumentation,
spine, trauma] Spine 1999;24:1646 –1654
From the *Department of Orthopaedic Surgery, Cleveland Clinic
Foundation, Cleveland, Ohio; and the †Department of Orthopaedic
Surgery, University of California, Davis, California.
Acknowledgment date: June 19, 1998
First revision date: August 24, 1998
Acceptance date: October 6, 1998
Device status category: 1
1646
Segmental spinal instrumentation has been shown to
provide suitable stabilization in a variety of thoracic and
thoracolumbar spinal fractures.3,37,49,63 McBride48 re-
ported good results in thoracic and thoracolumbar frac-
tures treated with longer hook and rod constructs, and
short-segment pedicle instrumentation constructs have
been recommended for treatment of lumbar frac-
tures.24,35 These systems are far more rigid than nonseg-
mental instrumentation, and surgeons can take advan-
tage of the intrinsic stability of segmental spine systems
to mobilize injured patients aggressively. Although con-
struct patterns have been refined over the past 10 years,
there is still little consensus about when to operate on
patients with spine fractures. Some surgeons still recom-
mend that spine surgery be routinely performed 48 to 72
hours after injury,38 whereas older published reports
suggest waiting 7–10 days for surgery, irrespective of
other injuries. Although the orthopedic trauma literature
clearly shows that urgent stabilization of long-bone in-
juries reduces both morbidity and mortality in poly-
trauma patients, many physicians think that urgent spi-
nal surgery is too dangerous to consider in ill or
injured patients.11–13,39,60
Because an unstable spinal fracture presents many of
the same hazards to recovery as a femur or pelvic frac-
ture, (pain, systemic shock, enforced recumbency, inabil-
ity to mobilize the patient) delayed treatment of a spinal
fracture would probably result in the same complications
as delayed long-bone stabilization, including pulmonary
insufficiency, adult respiratory distress syndrome
(ARDS), sepsis, and thromboembolic phenomena. In the
current study, perioperative and long-term results were
analyzed in a consecutive series of polytrauma patients
with thoracic and thoracolumbar fractures, treated with
early stabilization using Cotrel–Dubousset (CD) instru-
mentation (Sofamor–Danek, Memphis, TN). The pur-
pose of this study was not to endorse urgent or emergent
surgery as the best treatment of spinal fractures or to
suggest that surgery is necessary for all isolated spine
fractures. Our hypothesis was that urgent surgical treat-
ment (within 24 hours of injury), when necessitated by
neurologic conditions or concomitant injury, would not
significantly endanger polytrauma patients and would
allow the application of well-established principles of
fracture care shown to reduce morbidity and mortality in
trauma of the skeletal extremities.
Materials and Methods
Seventy-five consecutive patients with spine fractures treated
with CD segmental instrumentation were enrolled at the time
 Urgent Stabilization in Thoracic and Thoracolumbar Fractures • McLain and Benson 1647
of surgery in a longitudinal study of perioperative and long-
term outcome. All spinal surgeries were performed by four
fellowship-trained surgeons with extensive experience in
trauma care and spinal fracture management. All patients were
diagnosed and treated at our Level 1 trauma center from Jan-
uary 1988 through December 1993. All patients underwent
posterior spinal stabilization.
From this group, 27 polytrauma patients, each with an in-
jury severity score (ISS) higher than 26, were analyzed for out-
come relative to the timing of spinal stabilization. Polytrauma
was defined as significant injury (requiring hospital admission
and active management) to two or more major organ systems.
Injury severity scores were calculated using the Revised Abbre-
viated Injury Scale.19 In this system a score from 0 to 5 is
assigned to each organ system, with 0 indicating no injury and
5 indicating severe injury. The three highest scores are squared
and then summed to provide a final ISS. Unsurvivable injuries
are automatically assigned a score of 75. The highest ISS any
isolated injury can generate is 25. Therefore, a minimum score
of 27 was selected for inclusion in this series, ensuring that
every patient had at least one major associated injury in addi-
tion to a spinal fracture. For example, the maximum score for
an isolated spine fracture with complete spinal cord injury
would be 5, which would produce an ISS of 25 (52 " 0 " 0 #
25). An ISS threshold of 27 was selected to exclude those pa-
tients with isolated spinal column injuries or spinal injuries
associated with only a minor secondary injury, (ileus, or minor
laceration) (52 " 12 " 0 # 26). Although a score of 27 could be
attained by combining the most severe spinal injury with mul-
tiple minor injuries, there were no such patients in this study
group. Any patient with a score of 27 or higher would, by
definition of this analysis, have sustained significant injuries to
at least two major organ systems.
Approximately 80% of the 1200 thoracic and thoracolum-
bar fractures treated at our institution during the study period
were treated nonoperatively. Indications for surgical treat-
ment were:
● An unstable thoracic or thoracolumbar spine fracture,
determined by neurologic injury, sagittal angulation of 25°
or more relative to adjacent levels, axial compression of
50% of vertebral height or more, multiple contiguous frac-
tures, and/or three-column spinal injury.22
● Potentially unstable fractures in a patient with multiple
associated injuries.
● Patients needing decompression and reconstruction for
persistent or progressive neurologic deficit or severe frac-
ture comminution.
Fractures were classified according to Denis’ injury classifica-
tion.22,23 Associated injuries and additional spinal fractures
were noted, along with neurologic deficits, recorded by Frankel
grade. Segmental fixation was selected at the discretion of the
attending surgeon based on his interpretation of the fracture
pattern, the complexity of the fracture, and the vertebral level.
During the study period, patients with stable fractures, thoracic
compression or stable burst fractures, and some lumbar burst
fractures were treated nonoperatively or with nonsegmental
systems. Harrington rods, with or without Drummond wires,
were used for single-level thoracic fractures in otherwise stable
patients.1,16,33,67 Also, the AO Fixateur Interne, (Synthes, Pa-
oli, PA) was used in single-level thoracolumbar burst fractures
in patients with no or limited associated injury31,32; and VSP
plates and screws, (AcroMed, Cleveland, OH) were occasion-
ally used in one- and two-level lumbar burst fractures.28,65
Patients with these constructs were excluded from this study.
Polytrauma patients selected for segmental instrumentation
had higher ISSs and more associated injuries than the cohort
selected for nonsegmental instrumentation and could not tol-
erate either prolonged bed rest or immediate casting and mo-
bilization. Patients with pathologic fractures or inadequately
treated fractures requiring late reconstruction were excluded.
With the exception of hemodynamic stabilization, patients
treated at any other facility before transfer were excluded.
Timing of Surgery. Patients were treated according to the
standards of polytrauma management,13 ensuring adequate
fluid resuscitation, early and aggressive intervention for vis-
ceral injuries, and a concerted effort to stabilize skeletal injuries
and mobilize each patient as soon as possible. Twenty-six of 27
polytrauma patients were treated within the first 48 hours of
admission. Seven underwent spinal stabilization in combina-
tion with other emergent operative procedures at the time of
admission (Table 1). Ten patients with incomplete or progres-
sive neurologic injuries were treated as soon as medically sta-
ble. Steroid therapy was initiated in the emergency room for all
patients with neurologic deficit, in accordance with the second
National Acute Spinal Cord injury Study.15 Two treatment
groups were analyzed.
Urgent Treatment. Fourteen patients (11 men, 3 women)
were surgically stabilized within 24 hours of spinal injury.
Early Treatment. Thirteen patients (10 men, 3 women) were
treated between 24 and 72 hours after injury. All but one re-
ceived treatment within 48 hours. None of the polytrauma
patients in this prospective series was treated on a delayed basis
(!72 hours).
Indications for urgent treatment included a progressive neu-
rologic deficit, extensive polytrauma that predisposed to severe
pulmonary and/or metabolic derangements if the patient were
not mobilized, associated injuries that dictated emergent surgi-
cal treatment, and chest trauma and pulmonary contusions
that would predictably result in pulmonary deterioration.
If the patient was adequately resuscitated and hemodynam-
ically stable and if the patients’ status was likely to deteriorate
over the first 48 hours after hospital admission, we believed
that surgery would be safer earlier than later, and the fractures
in these patients were urgently stabilized. This view was held
even, and sometimes particularly, if the patient was already in
the operating room.
Surgical Treatment. The posterior instrumentation used for
all cases in this study was the CD Universal System (Sofamor–
Danek) using 7.0-mm knurled rods, combinations of open and
closed hooks, and 6.0-mm and 7.0-mm sacral screws. The CD
segmental instrumentation was used in three principle config-
urations, depending on the level and complexity of the fracture.
Rod– hook constructs using laminar claw configurations at
each terminus and one or more intermediate hooks for segmen-
tal distraction or compression, were applied to thoracic and
upper thoracolumbar fractures.48 Extended pedicle screw con-
structs were used to treat thoracolumbar fractures (T11–L2),
using a transverse process pedicle claw for thoracic fixation and
pedicle screws for lumbar fixation.37 Short-segment pedicle in-
strumentation constructs were applied to thoracolumbar and
lumbar fractures, with pedicle screws placed immediately
above and below the fractured level to limit the length of lum-
1648 Spine • Volume 24 • Number 16 • 1999

Table 1. Characteristics of 27 Polytrauma Patients Treated for Spinal Fractures Using CD Segmental Instrumentation
Time to Initial Additional Number of
Patient Age ISS OR* Mechanism Neurological Injury Injury Associated
No. (yr) Sex Score (hr) of Injury Finding Level to Vertebra Injuries Associated Injuries†

Urgent
1 41 m 75 1 Ped CCI T5 None 7 CHI, IAI, Thor inj, multi LE Fx
2 24 m 57 3 Crush ICEI L5 None 3 SI disloc, IAI, Thoracic Inj
3 26 f 50 4 MVA ICEI L2 C4 4 IAI, hemo/pneumo, talus Fx, C spine
sublux
4 40 m 41 6 Crush CSCI T12 None 2 Hemothx, rib Fx, IAI
5 46 m 35 6 Fall CCEI L2 None 2 CHI, facial lac
6 41 m 48 8 MCA ICEI L1 None 4 CHI, thorac inj, pelvic Fx, IAI
7 16 f 27 10 MVA Intact L3 None 2 CHI, IAI
8 17 m 34 10 MVA ICEI L1 None 1 R wrist Fx
9 34 m 33 12 MCA ICI T5 C2,C3 3 IAI, facial Fx, C-spine Fx
10 19 m 34 12 MVA ICEI L1 None 4 CHI, IAI, facial Lacs, ankle Fx
11 27 m 34 12 Fall Intact L1 T4,T6 5 Pelvic Fx, BBFA Fx x2, T-spine fx, R
wrist Fx
12 15 f 36 20 MVA Intact L1 None 2 IAI, rib Fx
13 21 m 34 20 MCA Intact T3 T10 5 Rib Fx, ankle Fx x2, TL-spine fx, R radius
Fx
14 19 m 48 20 MVA ICEI L3 L2 4 CHI, Thorac inj, IAI, rib Fx
Early
15 23 m 34 24 MVA ICEI L1 L4 4 Knee disloc, humerus Fx, Thorac inj, CHI
16 21 f 50 24 MVA CCI T5 None 4 CHI, pelvic Fx, clav Fx, Thorac inj
17 47 m 27 24 MVA Intact L1 C6,C7 2 Neck Fx, scapula Fx
18 32 m 29 24 MVA CCI T9 None 2 Rib Fx, CHI
19 18 m 50 24 MCA Intact L5 L4 4 CHI, IAI, bladder rupt, femur Fx
20 42 m 41 24 Fall CCI L1 None 1 Thorac inj
21 58 m 34 24 Fall ICEI L1 None 3 IAI, pelvic Fx, TibFib Fx
22 21 m 29 36 MCA CCI T6 T5 2 Thorac inj, CHI
23 37 m 41 36 MCA Intact T12 L1,T11 2 Open pelvic Fx
24 20 m 41 48 MVA ICEI L3 None 4 CHI, Thorac Inj, R BBFA Fx, L BBFA Fx
25 19 f 34 48 MCA Intact T7 T8 2 IAI, hemoTx
26 35 m 29 48 Fall Intact L1 None 2 TibFib Fx, Thorac Inj
27 21 f 34 72 MVA Intact L2 None 3 CHI, IAI, Rib Fx
* From admission to start of surgical treatment.
† Additional contiguous and noncontiguous vertebral fractures.
Crush # crushed under falling wall; Fall # fall from height; Ped # auto vs. pedestrian; MCA # motorcycle accident; MVA # automobile accident; ICEI and CCEI #
incomplete and complete cauda equina injuries, respectively; ICI and CCI # incomplete and complete cord injuries, respectively; CHI # closed head injury; IAI #
intraabdominal injury; SI Disloc # sacroiliac dislocation; Hemothx # hemothorax; Thorac inj # pulmonary or cardiac contusion, bronchial rupture, etc.; Fx #
fracture.

bar fusion.50,51 Constructs were bilaterally symmetrical, as a
rule, and paired rods were cross-linked with the CD device for
transverse traction (DTT).
Anterior decompression was performed in seven patients
with incomplete neurologic lesions, high-grade residual steno-
sis, and/or severe vertebral comminution and kyphosis. The
anterior decompression was performed at the same time as the
urgent stabilization in four patients with incomplete or pro-
gressive neurologic lesions. Three patients had anterior decom-
pression at the time of early stabilization or as a sepa-
rate procedure.
Operative time and blood loss, operative and perioperative
complications, and the construct pattern were recorded for
each patient. Spinal cord–injured patients were managed after
surgery in our spinal cord rehabilitation unit. Polytrauma pa-
tients were managed in cooperation with general surgery and
orthopedic trauma services.
The principle indication for early surgical intervention in
polytrauma patients is to permit early mobilization and elimi-
nate the hazards presented by bed rest and recumbency. De-
layed spinal stabilization (more than 72 hours after injury) is
inconsistent with these basic tenants of trauma management.
Because segmental spinal instrumentation systems offer supe-
rior versatility and reliability of fixation compared with non-
segmental rod– hook systems, surgeons can mobilize patients
rapidly, obtaining an upright posture, improved cardiopulmo-
nary function, and a reduced risk of pneumonia, venous throm-
bosis, or urosepsis. All patients in this study were sitting up in
bed immediately after stabilization and were fitted with a
molded orthosis before ambulation.
Follow-up. Patients were seen at 2 weeks, 6 weeks, 12 weeks,
and 6 and 12 months after surgery and were followed up an-
nually thereafter. All surviving patients were observed for a
minimum of 2 years after surgery. If patients could not return
for annual follow-up, records of functional status and radio-
graphs were obtained from their local physician. A follow-up
interview was conducted with each patient to assess functional
status (return to work, need for further surgery, pain level). A
private investigator was employed to locate eight patients who
could not be located through hospital, phone, or postal records
or next of kin. All patients were interviewed by the first author
at final follow-up.
Follow-up examinations assessed neurologic recovery or de-
terioration, development or improvement of back and leg pain,
development or progression of deformity after surgery, hard-
ware loosening or failure, pseudarthrosis, and the need for
 Urgent Stabilization in Thoracic and Thoracolumbar Fractures • McLain and Benson 1649

additional surgery. Fusion status was determined based on se- Table 2. Injury and Perioperative Characteristics of
rial roentgenographs, patient symptoms, and the presence or Urgent and Early Treatment Groups
absence of fixation failure over time. Flexion and extension
radiographs were obtained to rule out pseudarthrosis in all Urgent Group Early Group P Value
patients with equivocal examination findings or radiographs.
Mean age (yr) 27.5 (16–46) 30.0 (18–58) NS
Pain was recorded as absent, minor, moderate, or severe Mean ISS 42 (27–75) 36 (27–50) NS
(0 –3) for back and lower extremity symptoms. Neurologic % with neurologic injuries 71 54 NS
function and improvement or deterioration were rated accord- Mean time to OR (hr) 9.7 35.0 $0.00001
ing to the Frankel scale (E–A). Progressive kyphotic deformity Mean Op time, post (hr) 5.18 5.60 NS
was graded as none, minor (less than 5°), moderate (5–15°), or Mean Op time, comb (hr) 12.0 13.25 NS
Mean EBL, post (mL) 1432 1600 NS
severe (loss of sagittal correction of greater than 15°). Hard- Mean EBL, comb ant/post (mL) 6812 4000 NS
ware failure was recorded as broken screws or rods, bent Mean EBL, total (mL) 2966 1877 NS
screws, screw pullout without hardware failure, or no failure. Perioperative deaths 1/14 1/13 NS
Work status and function were determined at final fol- ARDS 0/14 1/13 NS
low-up relative to preoperative status. Patients were deter- NS # not significant; OR # operating room; EBL # estimated blood loss;
mined to be fully employed at their previous occupations; fully ISS # injury severity scale; ARDS # adult respiratory distress syndrome.
employed, but at a less demanding occupation; employed with
established limitations; or unemployed or disabled. Function
was recorded as severely limited, limited, normal, or limited by at a noncontiguous level, including three patients with
other conditions, such as paralysis, amputation, or head injury. cervical spine fractures.
Patients lost to follow-up at less than 2 years were not analyzed The mean ISS for urgently treated patients (ISS, 42)
with respect to outcome measures but were retained in analysis was slightly greater than for the early treatment group
of perioperative complications and outcomes. Because Short
(ISS, 36; Table 2). The mean ISS overall was 39 (range,
Form (SF)-36 and SF-12 tests were not widely used at the ini-
tiation of this study, it was decided not to introduce these sur-
27–75), and was not different for men or women (39.1
veys partway through the follow-up period. vs. 38.5). One patient from each group died during the
initial hospital stay; one patient, (male, early-treatment
group; ISS, 41), died 14 days after surgery because of
Results pneumonia and ARDS after falling 35 feet from a scaf-
Of the 29 patients with an ISS higher than 26 originally fold. The second patient, also male, (urgent-treatment
entered into the study from June 1987 through Decem- group; ISS, 75), died immediately after surgery of mas-
ber 1993, 2 were excluded because they had been treated sive visceral injury. He had been struck by a car, sustain-
at another facility before transfer. Twenty-seven patients ing a closed head injury, intraabdominal and intratho-
enrolled in the study were available for operative and racic injuries, and multiple long-bone fractures. Because
perioperative analysis. There were two perioperative our trauma surgeons thought he would not survive bed
deaths, and two patients were lost to follow-up, leaving rest, a rapid posterior instrumentation was performed at
23 patients available for final outcome analysis. Of the the time of laparotomy to stabilize a thoracolum-
two patients lost to follow-up, one had moved to another bar fracture.
state and could not be contacted, and the other had Seventeen patients (63%) had neurologic injury at the
disappeared and could not be located by a private in- time of fracture. Ten (37%) had incomplete cord or
vestigator. cauda equina injuries, and seven (26%) had apparent
complete cord or cauda equina injuries at the time of
admission. Ten (71%) of the 14 patients treated urgently
Injury Characteristics
had neurologic injuries: Seven (50%) of 14 injuries were
Seventy-four percent of the patients were injured in mo-
incomplete, and 3 (21%) were complete. Of the 13 pa-
tor vehicle accidents (Table 1). Twelve patients were in-
tients treated early, 7 (54%) had neurologic deficits, 3
jured in automobile accidents and seven in motorcycle
(23%) incomplete and 4 complete injuries. One patient
accidents; one was a pedestrian struck by a car. Five
in the early group had significant progression of neuro-
patients fell from height (18.5%), and two were crushed
logic deficit when he became combative before surgery
by collapsing walls (7.5%). Injury mechanisms, vertebral
and tried to leave his hospital bed.
levels, and injury severity were similar for the 21 male
(78%) and 6 female (22%) patients. Fourteen patients Surgical Treatment
(20%) had three or more major associated injuries; on All patients underwent posterior stabilization as the in-
average, men had 3.0 associated major injuries in addi- dex procedure. Fourteen patients (52%) were treated ur-
tion to spinal fracture compared with 2.9 for women. gently, with a mean interval from time of injury to sur-
The most common associated injuries were thoracic gical stabilization of 9.7 hours. Thirteen patients (48%)
trauma (14 patients), intra-abdominal injuries (14 pa- underwent early surgical stabilization at a mean of 35.0
tients), head injuries (13 patients), and pelvic and ex- hours from the time of injury. Seven (26%) of 27 patients
tremity fractures (16 patients). also had an anterior procedure as part of the initial op-
Eleven patients (41%) had an additional vertebral eration, for neural decompression, mechanical stabiliza-
fracture, five (19%) at a contiguous level and six (22%) tion, or both.
1650 Spine • Volume 24 • Number 16 • 1999
Table 3. Outcome in Urgently Treated Patients
ISS $
Urgent Early 40
Mean neural 1.12 0.65 0.75
improvement (FG)
% with neurological 88 50 63
improvement*
Pseudarthrosis 0/11 2/12 1/15
Pedicle screw failure 1/4 2/6 2/6
% RTW 64 66 66
Mean pain score 0.81/0.54 0.50/0.58 0.40/0.73
(B/L)
Mean function score 3.45 4.15 3.60
* Percentage of patients with initial neurologic injury who experienced im-
provement with treatment.
FG # Frankel Grade; RTW # return to work, at full or limited functional level;
B/L # back pain score/leg pain score.
Operative time and estimated blood loss varied with
the complexity of the fracture and the need for anterior
surgery. The mean operative time (including positioning
and draping) for all posterior procedures was 5.33 hours
and did not differ between the urgent- and early-
treatment groups (Table 3). Estimated blood loss during
posterior procedures also did not differ between the two
groups. Blood loss averaged 1432 mL (range, 500 –3500
mL) for the urgently treated group and 1600 mL for the
early-treatment group. Total blood loss was consider-
ably higher in the urgently treated patients, however,
(mean, 2966 mL vs. 1877 mL), reflecting the fact that
four of seven combined procedures were performed in
the urgent group and that blood loss during acute ante-
rior procedures was higher than in procedures delayed
24 – 48 hours. It should be noted that the mean blood
loss during combined procedures included the blood lost
during the exposure and included blood lost from hema-
toma and during splenectomy or other intra-abdominal
procedures. Posterior blood loss for isolated injuries av-
eraged 1160 mL.
Perioperative complications in 27 patients included
one superficial wound infection, two urinary tract infec-
tions, and one case of ARDS. There were no complica-
tions related to hardware placement. In one patient, an
anterior strut graft was displaced during repositioning
for the posterior procedure (replaced at a second opera-
tion). There were no clinically detectable deep venous
thromboses or pulmonary emboli, no systemic or graft
site infections, no cases of pneumonias, and no neuro-
logic injuries or perioperative deterioration.
Outcome
Twenty-three patients were available for long-term as-
sessment at a mean follow-up of 49 months (range,
24 – 84 months). Although the mean ISS for the urgent
group was higher than for the standard group, (42 vs.
36), there were no differences in the rate of perioperative
complications, morbidity, or mortality between the two
groups (Table 3). In the long run, seven (64%) of the
urgent-surgery group eventually returned to work, even
though only three were neurologically intact at the time
of surgery. By comparison, 8 of the 12 surviving patients
ISS !
40
in the early-surgery group returned to work (66%), even
though 6 patients were intact to begin with.
1.10 Twenty-one of 23 patients (91%) had a solid arthro-
desis. Twelve patients (52%) had unlimited activity at
66
final follow-up, three (13%) were limited by pain, and
1/8 eight (35%) were limited by neurologic injury. Twelve
1/4 patients (52%) had returned to their previous levels of
64
0.74/0.25 employment, three (13%) were working but at a lighter
job, six (26%) were disabled, and two (9%) remained
4.0 unemployed despite unrestricted activity status. Five pa-
tients (22%) required additional surgery: One had treat-
ment of a symptomatic pseudarthrosis, two had hard-
ware removal because of pain or hardware prominence,
one underwent nerve root decompression for foraminal
stenosis, and one had a spinal cord stimulator placed to
relieve persistent neuropathic symptoms.
Among patients with neurologic injuries, urgently
treated patients averaged improvement of 1.1 Frankel
grades, compared with 0.57 average for patients treated
later. All three patients who had full return of function
were in the urgent treatment group. Fourteen of the 17
patients with neurologic injury were available for fol-
low-up, (2 died, 1 lost to follow-up). No patient experi-
enced neurologic deterioration while adhering to proto-
col. Eight of nine with incomplete lesions improved
(mean improvement, 1.5 Frankel grades), and one of the
five with complete lesions improved (mean improve-
ment, 0.2 Frankel grades). Recovery was not correlated
with anterior decompression: Five patients with anterior
decompression for incomplete deficits improved 0.80
Frankel grades on average, whereas four patients with
only posterior surgery improved an average of 2.0
grades. One patient with an L1–L2 fracture-dislocation
improved four grades (from Frankel A to E) after urgent
operative reduction and stabilization, without ante-
rior decompression.
Late complications in 23 patients included 2 with
pseudarthrosis (8%) and 3 with sagittal collapse of 10°
or more (12%). There was a single case of late sacral
decubitus, secondarily infecting the operative site 7 years
after surgery. This patient required hardware removal,
extensive débridement, and flap coverage before healing.
Discussion
This prospective study analyzes outcomes in the most
severely injured segment (10%) of the spine trauma pop-
ulation, those patients whose lives depend on rapid re-
suscitation, mobilization, and prevention of pulmonary
and thromboembolic complications. These patients’ sta-
tus can deteriorate very rapidly after admission, and they
may not be suitable for delayed surgery for weeks there-
after, if ever. We again emphasize that the average spine
fracture patient can be treated in a timely fashion, at the
surgeons discretion, with a reliably satisfactory outcome.
The purpose of this study was not to show that urgent
surgery is necessary or efficacious in severe spinal frac-
 Urgent Stabilization in Thoracic and Thoracolumbar Fractures • McLain and Benson 1651
tures, but rather to determine whether there is a compel-
ling reason not to operate on these patients when other
conditions press for urgent treatment. The findings of
this study show there is not, and indicate that urgent
treatment may be beneficial for some patients with se-
vere, life-threatening injuries.
Findings in several retrospective studies have demon-
strated that aggressive early management of long-bone
fractures significantly reduces morbidity and mortality in
polytrauma patients.11,39,56 In a prospective, random-
ized trial, Bone et al12 demonstrated that stabilizing fem-
oral fractures within 24 hours of injury reduces pulmo-
nary dysfunction, ARDS, pulmonary emboli, and
pneumonia, as well as days of ventilator dependency,
and length of hospital stay. Seibel,60 in a prospective
study of blunt polytrauma, demonstrated that 10 days of
femoral traction essentially doubled the duration of the
pulmonary failure state compared with patients with im-
mediate long-bone fixation, documenting the adverse ef-
fect of enforced recumbency on patients with multiple
injuries. Although there is agreement on the value of
stabilizing long-bone and pelvic fractures in polytrauma
patients, the issue of urgent spinal stabilization has not
been directly explored.
Accepted indications for urgent spinal stabilization
include progressive neurologic deficit, unreduced spinal
dislocation with neurologic deficit, and severe deformity
or kyphosis compromising the overlying skin or ability
to position the patient.29,38,43 When these conditions are
absent, some surgeons have advocated waiting 7 to 10
days before surgery.
The combination of severe, multisystem injury and
thoracolumbar fracture was seen in less than 4% of the
patients with spine fractures treated at our facility during
the course of this study. Even among operative cases, the
magnitude of trauma seen in these polytrauma patients is
atypical and is seen in fewer than 15% of injuries. The
predominantly male cohort included in this series is typ-
ical of blunt and penetrating trauma populations
throughout the United States, but these patients
had more serious injuries, as a group, than patients
seen in any unselected series of spine frac-
tures.7,10,14,25,34,53,55,57,63 Seventy-four percent of the
patients in the current study were injured in automobile
or motorcycle accidents, either as drivers, passengers, or
pedestrians, compared with 43– 68% in other large se-
ries.26,48,55,57,69 Forty-one percent of the current patients
had an additional vertebral fracture at the time of injury,
compared with 7–10% in previous series.69 The large
number of associated injuries and the high ISSs further
attest to the severity of the trauma in this highly selected
patient group.
Blood loss for posterior procedures was not different
for urgent and early treatment groups and was compa-
rable with that reported in other trauma series.27,45 To-
tal blood loss was greater, however, for patients in the
urgent surgical group. Patients undergoing urgent ante-
rior decompression frequently had brisk hemorrhage
from the epidural vessels and fresh fracture. Because
more of the urgently treated patients required anterior
decompression, this greatly increased the total volume
lost among those patients. Stauffer64 has shown that sur-
gical blood loss is reduced by two thirds if surgery can be
delayed 48 hours after injury. Benefits of urgent stabili-
zation must be balanced against this increased
blood loss.
Even though these cases were treated at night, in face
of increased blood loss and extended exposures and in
combination with other surgical procedures, there were
no neurologic injuries, no deep wound infections, and no
patients who required revision surgery to correct an in-
complete reconstruction or inadequate decompression.
Surgeon fatigue did not impair results, and our Level I
trauma center provided appropriate operative facilities,
equipment, instruments, and staff for spinal stabilization
at any hour.
Perioperative complications of infection, pulmonary
disease, thromboembolism, and neurologic injury were
eliminated by urgent surgical treatment. There were no
incidences of clinical pneumonia, pulmonary embolus,
or clinically apparent deep venous thrombi, and only one
case of ARDS occurred in the perioperative period.
There were no episodes of systemic sepsis or wound in-
fection. Complication rates for both groups were lower
than in previously published surgical series and confirm
the findings of Schlegel et al59 in their retrospective se-
ries.25,34,55,63 Those investigators demonstrated a statis-
tically significant improvement in morbidity in multiply
injured patients (ISS !17) receiving early treatment
($72 hours) of their spinal injuries compared with those
treated on a delayed basis.
The overall survival rates for patients in the current
series were better than expected. Ten patients had an ISS
higher than 40. As expected, they had more associated
injuries and a higher incidence of neurologic deficit. Con-
sidering that the expected mortality for patients aged 15
to 44 with an ISS of 40 is 50%, the two deaths (20%
mortality) that occurred in this specific subgroup are
fewer than anticipated.4,5,21,68
Unexpectedly, neurologic improvement was better in
the urgent treatment group than in the early treatment
group. All three patients with complete neurologic recov-
ery, and both of those with more than two grades of
recovery were treated within the first 24 hours of injury.
The numbers involved are too small to indicate a statis-
tically significant difference, and the recovery in these
patients cannot be used to argue that urgent treatment
may provide improved recovery in complete spinal cord
injuries. In fact, because it is frequently impossible to
distinguish complete from incomplete lesions in urgently
treated patients, any urgently treated group is likely to
contain some apparently complete injuries that would
improve spontaneously. This will bias any comparison
with patients treated at a later time when spinal shock
has resolved. Nonetheless, concerns that urgent interven-
1652 Spine • Volume 24 • Number 16 • 1999
tion may compromise neurologic recovery were clearly
not substantiated.
In the current series, neurologic recovery was more
directly related to early intervention and fracture reduc-
tion than to the method of decompression. There was no
clear association between anterior decompression and
neurologic recovery. Canal compromise can be im-
proved through indirect reduction in some cases,61,70
and remodeling improves canal diameter over time, irre-
spective of treatment.52 Although we still consider an
incomplete neurologic deficit in the face of persistent
compression to be an indication for direct surgical de-
compression,9,20,36,41,47 the blood loss associated with
urgent anterior decompression can be daunting.
Follow-up
A variety of search strategies were used to locate 25
(93%) of the 27 patients originally enrolled, and 100%
of the patients were available for follow-up during the
perioperative period. In this study, each patient was ob-
served in our clinic until healing of the fracture occurred
or until a pseudarthrosis was clearly identifiable. There-
after, subjective measures of function and pain, as well as
reports of job duties, activity levels, and ongoing treat-
ment were obtained by telephone interview in those pa-
tients who could not return for an examination. Obtain-
ing this degree of follow-up in any trauma population
can be difficult,40,44 and this group was particularly tran-
sient. Of the original 75-patient cohort, only 8 of 68
identifiable patients (12%) were still at the same address
2 years after injury. Eleven (16%) of the 68 had been
incarcerated for a portion of the follow-up period.
Instrumentation
The risk of sagittal collapse after instrumentation was
related to the level and severity of the fracture, the ap-
proach taken, and the construct chosen. Sagittal collapse
after fracture reduction is common, with other investiga-
tors reporting mean progression of between 12° and
19°.2,6,8,37,42,58,66 The decision to operate urgently did
not affect the construct patterns chosen or the ultimate
success or adequacy of those constructs. Although Dimar
et al27 reported sagittal collapse after combined anterior
and posterior reconstruction, anterior reconstruction
eliminated sagittal collapse in the current series.
In short-segment pedicle constructs, cantilever bend-
ing loads are concentrated within the vertebral pedicle
and can lead to bending failure of the screws.18,54,62,71
Bending failure that occurs before the fracture consoli-
dates can result in progressive sagittal collapse, which
can progress despite bracing.17,37,42,51 Extended pedicle
screw constructs work well as long as the pedicle screws
are protected from cantilever loading. Extending thora-
columbar fixation into the lower thoracic segments has
little mechanical cost and allows the surgeon to correct
sagittal deformity and restore neutral or lordotic align-
ment.2,37 Hook pullout was not seen in the current series,
nor was there any rod breakage, a common mode of
failure in nonsegmental systems.30,46 Urgently treated
patients did not experience more hardware failure, sug-
gesting that decisions on segmental stability were as well
made when they were made quickly as when treatment
was delayed.
Hardware was removed only if the implants were
prominent and painful, had lost fixation, or had failed or
if revision was indicated for pseudarthrosis or progres-
sive kyphosis. In all, three in the current series (13%)
underwent hardware removal or revision for implant-
related problems, compared with 17% in previ-
ous series.49
Functional Outcome
Overall, 15 (65%) of the 23 follow-up patients returned
to work, whereas 2 others (9%) who were capable of
work remained unemployed. Only six (26%) of these
severely injured patients were considered disabled at fi-
nal follow-up. This compares favorably with the results
of Tasdemiroglu,69 who found that only 26% of patients
treated with Harrington rod constructs returned to work
and that only 12% returned to their original jobs. Al-
though Carl et al17 and others have reported return to
work rates of between 84% and 94%,17,66 they have
provided relatively short follow-up (18 –22 months). The
return-to-work rates tended to deteriorate over time, so
that long-term results may reflect more true disability
than is seen in the short term. The slightly better func-
tional result seen in the urgently treated patients com-
pared with those treated only 24 hours later probably
reflects the small sample size. The good overall return to
work for both groups, however, may reflect the contin-
ued effort, beginning with early stabilization and reha-
bilitation, to mobilize rapidly and restore these severely
injured patients.
Summary
The blood loss, operative times, and complication rates
seen with these difficult patients are greater than would
be expected from a series of isolated spine fractures. All
of these outcome parameters are technique and patient
dependent. The small series of patients presented in the
current study represent a challenging subgroup within a
much larger trauma experience. Each of the these pa-
tients was treated by a fellowship-trained senior spine
surgeon experienced in spinal trauma. The ancillary
staff, anesthesia, and general surgical staff assisting in
patient management were expert in the management of
high-energy polytrauma, and were available at all hours
to facilitate the careful management of these patients.
The spine surgical staff routinely worked in tandem with
an experienced neurosurgical service that assisted in the
management of complex spinal column injuries. Causes
for increased blood loss were numerous: the inability to
use hypotensive anesthesia in a polytrauma patient, co-
agulopathy or diathesis after serial surgical procedures,
the need to expose and débride the fresh fracture site, and
traumatic and surgical disruption of epidural and seg-
mental vessels. Despite this, patients treated urgently
 Urgent Stabilization in Thoracic and Thoracolumbar Fractures • McLain and Benson 1653
fared as well as those treated in a standard time frame,
and the group outcome as a whole compared favorably
with that of patients with isolated spinal injuries. There
was a low incidence of the life-threatening complications
that commonly afflict patients with multiple injuries and
a high ISS.
The results in this study, although not conclusive in
significant differences and true incidence of specific com-
plications, indicate that urgent surgical intervention is
not more dangerous than surgery performed 24 –72
hours after injury and supports the suggestion that it
would be both feasible and ethical to investigate this
question definitively through a randomized, multi-
center study.
References
1. Akbarnia BA, Fogarty JP, Tayob AA. Contoured Harrington instrumenta-
tion in the treatment of unstable spinal fractures: The effect of supplementary
sublaminar wires. Clin Orthop 1984;189:186 –94.
2. Akbarnia BA, Crandall DG, Burkus K, Matthews T. Use of long rods and a
short arthrodesis for burst fractures of the thoracolumbar spine: A long-term
follow-up study. J Bone Joint Surg [Am] 1994;76:1629 –35.
3. Argenson C, Lovet J, de Peretti F, et al. The treatment of spinal fractures with
Cotrel–Dubousset instrumentation: Results of the first 85 cases. Scoliosis Re-
search Society/European Spine Meeting. Orthop Trans 1990;14:776 –7.
4. Baker SP, O’Neill B, Haddlon W. The injury severity score: A method for
describing patients with multiple injuries and evaluating emergency care.
J Trauma 1974;14:187–96.
5. Baker SP, O’Neill B. The injury severity score: An update. J Trauma 1976;
16:882–5.
6. Benli IT, Tandogan NR, Kis M, et al. Cotrel–Dubousset instrumentation in
the treatment of unstable thoracic and lumbar spine fractures. Arch Orthop
Trauma Surg 1994;113:86 –92.
7. Benson DR, Burkus JK, Montesano PX, Sutherland TB, McLain RF. Unsta-
ble thoracolumbar and lumbar burst fractures treated with the AO fixateur in-
terne. J Spinal Disord 1992;5:335– 43.
8. Benzel EC. Short-segment compression instrumentation for selected thoracic
and lumbar spine fractures: The short-rod/two-claw technique. J Neuro-
surg 1993;79:335– 40.
9. Bohlman HH, Freehafer A, Dejak J. The results of treatment of acute injuries
of the upper thoracic spine with paralysis. J Bone Joint Surg [Am] 1985;67:
360 –9.
10. Bohlman HH. Treatment of fractures and dislocations of the thoracic and
lumbar spine. J Bone Joint Surg [Am] 1985;67:165–9.
11. Bone LB, Bucholz R. The management of fractures in the patient with mul-
tiple trauma. J Bone Joint Surg [Am] 1986;68:945–9.
12. Bone LB, Johnson KD, Weigelt J, Scheinbeng R. Early vs. delayed stabiliza-
tion of fractures: A prospective, randomized study. J Bone Joint Surg [Am] 1989;
71:336 – 40.
13. Bone LB. Management of polytrauma. In: Chapman, MW, ed. Operative
Orthopaedics. 2nd ed. Philadelphia: JB Lippincott, 1993:299 –304.
14. Bostman OM, Myllynen PJ, Riska EB. Unstable fractures of the thoracic and
lumbar spine: The audit of an 8-year series with early reduction using Harrington
instrumentation. Injury 1987;18:190 –5.
15. Bracken MB, Shepard MJ, Collins WF, et al. A randomized, controlled trial
of methylprednisolone or naloxone in the treatment of acute spinal cord injury.
Results of the Second National Acute Spinal Cord Injury Study. N Engl
J Med 1990;322:1405–11.
16. Bryant CE, Sullivan JA. Management of thoracic and lumbar spine fractures
with Harrington distraction rods supplemented with segmental wiring.
Spine 1983;8:532–7.
17. Carl AL, Tromanhauser SG, Roger DJ. Pedicle screw instrumentation for
thoracolumbar burst fractures and fracture– dislocations. Spine 1992;17(Suppl):
S317–24.
18. Chiba M, McLain RF, Yerby SA, Moseley TA, Smith TS, Benson DR. Short-
segment pedicle instrumentation: Biomechanical analysis of supplemental hook
fixation. Spine 1996;21:288 –94.
19. Civil ID, Schwab CW. The Abbreviated Injury Scale, 1985 revision: A con-
densed chart for clinical use. J Trauma 1988;28:87–90.
20. Clohisy JC, Akbarnia BA, Bucholz RD, Burkus JK, Backer RJ. Neurologic
recovery associated with anterior decompression of spine fractures at the thora-
columbar junction (T12–L1). Spine 1992;17(Suppl):S325–30.
21. Copes WS, Champion HR, Sacco WJ, Lawnick MM, Keast SL, Bain LW.
The injury severity score revisited. J Trauma 1988;28:69 –77.
22. Denis F. The three column spine and its significance in the classification of
acute thoracolumbar spinal injuries. Spine 1983;8:817–31.
23. Denis F. Spinal instability as defined by the three-column spine concept in
acute spinal trauma. Clin Orthop 1984;189:65–76.
24. Devito DP, Tsahakis PJ. Cotrel–Dubousset instrumentation in traumatic
spine injuries. The 6th Proceeding of the International Congress on Cotrel–
Dubousset Instrumentation. Montpellier, France: Sauramps Medical, 1989:
41– 46.
25. Dickman CA, Yahiro MA, Lu HT, Melkerson MN. Surgical treatment al-
ternatives for fixation of unstable fractures of the thoracic and lumbar spine: A
meta-analysis. Spine 1994;19(Suppl):2266S–73S.
26. Dickson JH, Harrington PR, Erwin WD. Results of reduction and stabiliza-
tion of the severely fractured thoracic and lumbar spine. J Bone Joint Surg
[Am] 1978;60:799 – 805.
27. Dimar JR 2nd, Wilde PH, Glassman SD, Puno RM, Johnson JR. Thoraco-
lumbar burst fractures treated with combined anterior and posterior surgery.
Am J Orthop 1996;25:159 – 65.
28. Ebelke DK, Asher MA, Neff JR, Kraker DP. Survivorship analysis of VSP
spine instrumentation in the treatment of thoracolumbar and lumbar burst frac-
tures. Spine 1991;16(Suppl):S428 –32.
29. Edwards CC, Levine AM. Early rod-sleeve stabilization of the injured tho-
racic and lumbar spine. Orthop Clin North Am 1986;17:121– 45.
30. Erwin WD, Dickson JH, Harrington PR. Clinical review of patients with
broken Harrington rods. J Bone Joint Surg [Am] 1980;62:1302–7.
31. Esses SI, Botsford DJ, Wright T, Bednar D, Bailey S. Operative treatment of
spinal fractures with the AO internal fixator. Spine 1991;16(Suppl):S146 –50.
32. Esses SI, Botsford DJ, Kostuik JP. Evaluation of surgical treatment for burst
fractures. Spine 1990;15:667–73.
33. Flesch JR, Leider LL, Erickson DL, Chou SN, Bradford DS. Harrington
instrumentation and spine fusion for unstable fractures and fracture-dislocations
of the thoracic and lumbar spine. J Bone Joint Surg [Am] 1977;59:143–53.
34. Fletcher DJ, Taddonio RF, Byrne DW, et al. Incidence of acute care compli-
cations in vertebral column fracture patients with and without spinal cord injury.
Spine 1995;20:1136 – 46.
35. Gillet P, Meyer R, Fatemi F, Lemaire R. Short segment internal fixation using
CD instrumentation with pedicular screws: Biomechanical testing. The 6th Pro-
ceeding of the International Congress on Cotrel–Dubousset Instrumentation.
Montpellier, France: Sauramps Medical, 1989:19 –24.
36. Golimbu C, Firooznia H, Rafii M, Engler G, Delman A. Computed tomog-
raphy of thoracic and lumbar spine fractures that have been treated with Har-
rington instrumentation. Radiology 1984;151:731–3.
37. Graziano GP. Cotrel–Dubousset hook and screw combination for spine frac-
tures. J Spinal Disord 1993;6:380 –5.
38. Hanley EN, Simpkins A, Phillips ED. Fractures of the thoracic, thoracolum-
bar and lumbar spine: Classification, basis of treatment, and timing of surgery.
Semin Spine Surg 1990;2:2–7.
39. Johnson KD, Cadambi A, Seibert GB. Incidence of adult respiratory distress
syndrome in patients with multiple musculoskeletal injuries: Effect of early oper-
ative stabilization of fractures. J Trauma 1985;25:375– 84.
40. Klonoff–Cohen H. Tracking strategies involving fourteen sources for locat-
ing a transient study sample. Am J Epidemiol 1996;144:98 –101.
41. Kostuik JP. Anterior fixation for burst fractures of the thoracic and lumbar
spine with or without neurological involvement. Spine 1988;13:286 –93.
42. Kramer DL, Rodgers WB, Mansfield FL. Transpedicular instrumentation
and short-segment fusion of thoracolumbar fractures: A prospective study using
a single instrumentation system. J Orthop Trauma 1995;9:499 –506.
43. Levine AM, Bosse M, Edwards CC. Bilateral facet dislocations in the thora-
columbar spine. Spine 1988;13:630 –9.
44. Long, L. Americans on the move. Am Demogr 1990;12:46 –9.
45. Markel DC, Graziano GP. A comparison study of treatment of thoracolum-
bar fractures using the ACE Posterior Segmental Fixator and Cotrel–Dubousset
instrumentation. Orthopedics 1995;18:679 – 86.
46. McAfee PC, Bohlman HH. Complications following Harrington instrumen-
tation for fractures of the thoracolumbar spine. J Bone Joint Surg [Am] 1985;67:
672– 86.
47. McAfee PC, Bohlman HH, Yuan HA. Anterior decompression of traumatic
thoracolumbar fractures with incomplete neurological deficit using a retroperi-
toneal approach. J Bone Joint Surg Am 1985;67:89 –104.
48. McBride GG. Cotrel–Dubousset rods in spinal fractures. Paraplegia 1989;
27:440 –9.
49. McBride GG. Cotrel–Dubousset rods in surgical stabilization of spinal frac-
tures. Spine 1993;18:466 –73.
50. McKinley LM, Obenchain TG, Roth KR. Loss of correction: Late kyphosis
in short segment pedicle fixation in cases of posterior transpeduncular decom-
1654 Spine • Volume 24 • Number 16 • 1999
pression. The 6th Proceeding of the International Congress on Cotrel–Dubousset
Instrumentation. Montpellier, France: Sauramps Medical, 1989;37–9.
51. McLain RF, Sparling E, Benson DR. Early failure of short-segment pedicle
instrumentation for thoracolumbar fractures. A preliminary report. J Bone Joint
Surg Am 1993;75:162–7.
52. Mumford J, Weinstein JN, Spratt KF, Goel VK. Thoracolumbar burst frac-
tures. The clinical efficacy and outcome of nonoperative management.
Spine 1993;18:955–70.
53. Osebold WR, Weinstein SL, Sprague BL. Thoracolumbar spine fractures.
Results of treatment. Spine 1981;6:13–34.
54. Panjabi MM, Oxland TR, Kifune M, Arand M, Wen L, Chen A. Validity of
the three-column theory of thoracolumbar fractures: A biomechanic investiga-
tion. Spine 1995;20:1122–7.
55. Place HM, Donaldson DH, Brown CW, Stringer EA. Stabilization of tho-
racic spine fractures resulting in complete paraplegia. A long-term retrospective
analysis. Spine 1994;19:1726 –30.
56. Rogers FB, Shackleford SR, Vane DW, Kaups KL, Harris F. Prompt fixation
of isolated femur fractures in a rural trauma center: A study examining the timing
of fixation and resource application. J Trauma 1994;36:774 –7.
57. Saboe LA, Reid DC, Davis LA, Warren SA, Grace MG. Spine trauma and
associated injuries. J Trauma 1991;31:43– 8.
58. Sasso RC, Cotler HB. Posterior instrumentation and fusion for unstable
fractures and fracture-dislocations of the thoracic and lumbar spine. A compar-
ative study of three fixation devices in 70 patients. Spine 1993;18:450 – 60.
59. Schlegel J, Bayley J, Yuan H, Fredricksen B. Timing of surgical decompres-
sion and fixation in acute spinal fractures. J Orthop Trauma 1996;10:323–30.
60. Seibel R, LaDuca J, Hassett JM. Blunt multiple trauma (ISS 36), femur
traction, and the pulmonary failure-septic state. Ann Surg 1985;202:283–95.
61. Shuman WP, Rogers JV, Sickler ME, et al. Thoracolumbar burst fractures:
CT dimensions of the spinal canal relative to postsurgical improvement. Am J
Roentgenol 1985;145:337– 41.
62. Slosar PJ Jr, Patwardhan AG, Lorenz M, Havey R, Sartori M. stability of the
lumbar burst fracture and limitations of transpedicular instrumentation.
Spine 1995;20:1452– 61.
63. Stambough JL. Cotrel–Dubousset instrumentation and thoracolumbar spine
trauma: A review of 55 cases. J Spinal Disord 1994;7:461–9.
Point of View
Stanley D. Gertzbein, MD
Institute for Spinal Disorders
Baylor College of Medicine
Houston, Texas
This is a retrospective study of 75 consecutive patients
undergoing posterior stabilization for spinal fractures,
29 of whom were multiple trauma cases. Two groups
were studied, those operated on within 24 hours (urgent
group) and those operated on between 24 and 72 hours
(early treatment group). The purpose of the study was to
determine the morbidity in the two groups. The results
are similar in rate of complications and in long-term
outcome. The urgent-treatment group appeared to have
better neurologic improvement, but this could be be-
cause spinal shock obscured the analysis. It should be
recognized that this study was performed under the su-
pervision of a senior, experienced spine surgeon in col-
64. Stauffer ES. Thoracolumbar fractures: Principles of treatment. Presented at
Scoliosis Research Society, Baltimore, Maryland, September, 1988.
65. Steffee AD, Biscup RS, Sitkowski DJ. Segmental spine plates with pedicle
screw fixation: A new internal fixation device for disorders of the lumbar and
thoracolumbar spine. Clin Orthop 1986;203:45–53.
66. Stephens GC, Devito DP, McNamara MJ, Spengler DM. Short segment
transpedicular Cotrel–Dubousset instrumentation: A porcine corpectomy model.
J Spinal Disord 1993;6:252–5.
67. Sullivan JA. Sublaminar wiring of Harrington distraction rods for unstable
thoracolumbar spine fractures. Clin Orthop 1984;189:178 – 85.
68. Swiontkowski MF. The multiply injured patient with musculoskeletal inju-
ries. In: Fractures in Adults. Rockwood CA, Green DP, Bucholz RW, Heckman
JW, eds. New York: Lippincott–Raven, 1995:121–58.
69. Tasdemiroglu E, Tibbs PA. Long-term follow-up results of thoracolumbar
fractures after posterior instrumentation. Spine, 1995;20:1704 – 8.
70. Willen J, Lindahl S, Nordwall A. Unstable thoracolumbar fractures. A com-
parative clinical study of conservative treatment and Harrington instrumenta-
tion. Spine 1985;10:111–22.
71. Yerby S, Ehteshami J, McLain RF. Offset laminar hooks decrease bending
moments on pedicle screws during in-situ contouring. Spine 1997;22:376 – 81.
Address reprint requests to
Robert F McLain, MD
Department of Orthopaedic Surgery
Desk A41
Cleveland Clinic Foundation
9500 Euclid Avenue
Cleveland, OH 44195
laboration with a neurosurgical colleague. The expertise
provided by this spinal team would certainly contribute
to this outcome.
The authors are to be congratulated on an excellent
review. Surgery in the multitraumatized patient treated
urgently or early has been shown in numerous studies to
be associated with fewer complications. Now it is clear
that spinal surgery falls within the same time frame. It
appears that operating within the first 24 hours does not
compromise patients’ outcomes, and the risks are no dif-
ferent than treatment 24 –72 hours after the injury. Cer-
tainly, neurologic deterioration is not an issue, as has
been suggested in other studies.