Study reference

Study characteristics

Patient characteristics

Intervention (I)

Comparison / control (C)

Follow-up

Outcome measures and effect size

Comments

Bourassa-Moreau, 2016

Type of study:

Prospective cohort

Setting: level I trauma center specialized in SCT care

Country: Canada

Source of funding: No competing financial interest exist.

Inclusion criteria:

* minimal age of 16

* AIS A traumatic SCI at initial presentation before surgery

* Vertebral fracture and/or luxation from C1 to L2

Exclusion criteria:

* Neurological or cognitive impairment precluding reliable neurological evaluation

* Surgical intervention performed >3 days after the trauma

* Surgical decompression and fusion performed in another center

N total at baseline:

Intervention: 38

Control:15

Important prognostic factors²:

age ± SD:

I: 39.6 ±16.6

C: 49.6 ±15.4

Sex:

I: 11% F

C: 27% F

Comorbidity

I: 26%

C:40%

Groups comparable at baseline?

Patients in the intervention group were significantly younger.

Operation <24h post-trauma (mean 16.1±4.9h)

Operation ≥24h (mean 39.1±16.3h)

Length of follow-up:

Follow-up assessments were performed at rehabilitation discharge.

I: 150.6 ± 39.7

C: 156.9 ± 31.2

Loss-to-follow-up:

Probably none

Incomplete outcome data:

Probably none

1. Mortality

Not reported

2. Quality of life

Not reported

3. Start of rehabilitation/length of hospitalization

Not reported

4. Complications (infection, bleeding, exaggeration of neurological complains)

Not reported

5. Neurologic recovery

Defined as in improvement in AIS, n (%) with improvements from AIS A to AIS B, AIS C or AIS D.

I: 13 (34)

C: 2 (13)

P=0.18

Subgroup analyses:

“A total of six (18%) patients with a thoracolumbar injury had some neurological improvement at discharge from rehabilitation. The surgical delay did not impact on the proportion of patients who improved on the AIS, where 4/24 (17%) had early surgery and 2/9 (22%) were operated later (Table 4) (p=0,99).”

“Contrary to patients with a thoracolumbar injury, whose recovery was not linked to the surgical delay, Table 4 shows that 9/14 (64%) patients with cervical SCI operated within 24h showed AIS improvement whereas none of the six patients with cervical SCI operated ≥24h improved (p=0.008).

Confounding variables evaluated: severity of traumatic injuries assessed by Injury Severity Score (ISS), age, pre-existing comorbidity, smoking, body mass index (BMI), Glasgow Coma Scale (GCS) score, and duration of follow-up. The level of neurological lesion was divided into cervical (C1-T1) and thoracolumbar (T2-L5) lesions.

The authors conclude: “this study suggests that surgical decompression and stabilization earlier than 24h after a traumatic complete SCI may improve neurological recovery and particularly for those with cervical lesions.

Jug, 2015

Type of study:

Prospective cohort

Setting: level I trauma center

Country: Slovenia

Source of funding: No competing financial interests

Inclusion criteria:

* age 16-85 years

* Initial AIS grade A-C

* Fracture or dislocation C3-C7

* Neurological level of injury between C3 and C8

* Spinal cord compression confirmed by MRI

* Spinal canal compromise of at least 25 %

* Surgery within 24h of tSCI

Exclusion criteria:

*Neurological deficits before injury

*No evidence of fracture of dislocation

*Central cord syndrome

* Life threatening injuries that prevent early decompression of the spinal cord

*Cognitive impairment preventing accurate neurologic assessment

N total at baseline:

Intervention: 26

Control: 22

(22 and 20 evaluable)

Important prognostic factors²:

Age, median (range):

I: 44 (30.5;58.5)

C:52 (25.8;72.8)

Sex:

I: 82% M

C: 80% M

Groups comparable at baseline? No, more patients from the control condition were transferred from other hospitals (p<0.001). Furthermore, there were more (although not statistically) grade AIS B’s in the <8h group (23 versus 5%).

Surgical decompression <8h after injury

Surgical decompression between 8-24h after injury

Length of follow-up:

6 months

Loss-to-follow-up:

Intervention: n=4 (15%)

Reasons: 2 patients died from pulmonary failure, 1 was lost during follow-up and one patients in whom the decompression was inadequate.

Control: n=2 (9%)

Reasons: 1 patient died from pulmonary failure, 1 patient in whom the decompression was inadequate.

Incomplete outcome data:

Not reported

1. Mortality

Was not an outcome measure, but the number of patients who died during the study were provided:

I: 2 (8%)

C: 1 (5%)

2. Quality of life

Not reported

3. Start of rehabilitation/length of hospitalization

Defined as the length of hospital stay, median (?)

I: 38.8 (24.0)

C:48.8 (40.3)

p>0.05

4. Complications (infection, bleeding, exaggeration of neurological complains)

Defined as perioperative complications (surgical infection, CSF leak, Cardiovascular, gastrointestinal), n (%)

I: 3 (12)

C: 5 (23)

There were no between-group differences for the subcategories (p>0.05).

Defined as the occurrence of pneumonia in the month after surgery, n (%)

I: 10 (45)

C: 10 (50)

p>0.05

Defined as the number of ventilator-dependent days, median (range)

I: 6.5 (1 to 17 days)

C:5 (1.25 to 12)

p>0.05

5. Neurologic recovery

Defined as change in at least 2 AIS grades, n (%)

I: 10 (45.5)

C: 2 (10)

RR=2.08, 95%CI=1.12;3.87.

Odds adjusted for severity of t-SCI (complete or incomplete traumatic spinal cord injury) and the degree of spinal canal compromise: 11.08, p=0.004

Defined as change in ASIA motor score, median change (range)

I: 38.5 (10.0;61.0)

C: 15.0 (8.8;34.0)

P=0.047

“In our study, no statistically significant differences in the rate of perioperative complications, pneumonia, number of ventilatory dependent days, and mortality was found between the two study groups.”

McKinley, 2004

Type of study:

Retrospective case series

Setting: 18 Model Spinal Cord Injury Systems

Country: US

Source of funding: “No commercial party having a direct interest in the results of the research supporting this article has or will confer a benefit on the author(s) or on any organization with which the authors is/are associated.”

Inclusion criteria:

* nonpenetrating traumatic SCI who completed inpatient rehabilitation

Exclusion criteria:

* penetrating injuries (i.e. gunshot wounds)

N total at baseline:

Intervention: 73

C-1: 143

C-2:91

Important prognostic factors²:

Only reported for the whole subgroup of patients operated between 1 and 3 days after injury:

age ± SD:

36.72 ± 14.89 years

Sex: 74%M

Groups comparable at baseline? unknown

Surgery <24h (defined as operated on the same day as hospital admission

C-1: Surgery between 1 and 48h, first day after injury

C-2: 24 to 72h second day after injury)

Length of follow-up:

1-year follow-up, 4 time points: acute care, rehabilitation admission, rehabilitation discharge, 1-year follow-up.

Loss-to-follow-up:

Unclear

Incomplete outcome data:

See outcome measures

1. Mortality

Not reported for this subgroup

2. Quality of life

Not reported for this subgroup

3. Start of rehabilitation/length of hospitalization

Not reported for this subgroup

4. Complications (infection, bleeding, exaggeration of neurological complains)

Not reported for this subgroup

5. Neurologic recovery

Defined as the ASIA impairment score (AIS), change score from admission to follow-up, mean (SD), n

I: 0.74 (0.96), 34

C-1: 0.59 (1.08), 90

C-2:0.28 (0.86), 57

X²=6.61, p=0.037

Defined as the AISA motor index total score,

change score from admission to follow-up, mean (SD), n

I: 20.00 (22.61), 33

C-1: 17.07 (21.23), 75

C-2:12.11 (18.63), 46

X²=3.87, p=.144

Only the results for the comparisons between day 1, 2 and 3 admissions were included.

Furthermore, only the scores for the ASIA impairment score (AIS) and the AISA motor index total score are presented here, but the authors mention that “There were no between-group differences in neurologic, motor, or sensory levels or ASIA motor index.”

Patients were likely those with spinal instability.

The AIS injury level categories were recoded into numeric format to compute changes scores. The authors unfortunately don’t present the number of patients that improved one of more Asia grades, but reported the mean change.

Sohail Umerai, 2014

Type of study:

Prospective observational

Setting: tertiary care spinal trauma unit

Country: Pakistan

Source of funding: “no potential conflict of interest relevant to this article was reported”

Inclusion criteria:

* cervical cord injury from C3-T1

* aged between 18 to 65 years

Exclusion criteria:

*Glasgow Coma Scale >14 or ASIA AIS grade E at presentation

N total at baseline:

Intervention: 34

Control: 64

Important prognostic factors²:

age ± range:

I:37.5 (21-65)

C:40.1 (19-61)

Sex:

I: 82.35% M

C: 76.56% M

Groups comparable at baseline? yes

Decompressive surgery within 24h

Decompressive surgery after 24h (mean 52.7h, range: 31-124h)

Length of follow-up:

6 months

Loss-to-follow-up: n (%)

I: 3 (8.8)

C: 3 (4.7)

Incomplete outcome data:

Not reported

1. Mortality, n (%)

I: 1 (2.9) (cardiopulmonary arrest during surgery)

C: 4 (6.2) (2 pulmonary embolism, 1 patient developed bedsores and urinary tract infection and died of sepsis).

No statistical analyses reported.

2. Quality of life

Not reported

3. Start of rehabilitation/length of hospitalization

Not reported

4. Complications (infection, bleeding, exaggeration of neurological complains)

11.2% in total

I: 4

C: 7

(results presented in the discussion)

5. Neurologic recovery

Defined as at least 1 AIS grade, n (%)

I: 18 (52.9)

C:25 (39.1)

Odds=3.12, 95%CI=1.21;8.02

Defined as at least 2 AIS grades, n (%)

I: 7 (23.3)

C: 5 (8.7)

Odds=3.05, 95%CI=0.89;10.51

Surgery was sometimes delayed due to for instance medical condition of the patient, late presentation after trauma, delay in consent.

Battistuzzo, 2016

Type of study:

Retrospective

Setting: multicenter

Country: Australia and New Zealand

Source of funding: No conflict of interest

Inclusion criteria:

* 15 to 70 years

*C3-T1 facture, fracture-dislocation, and disc and/or ligamentous injury in association with an acute traumatic SCI with neurological deficits

Exclusion criteria:

* time of injury and time of spinal decompression were not available

*multiple traumatic injuries (trauma to at least one other major organ, significant abdominal bleeding or retro-peritoneal hemorrhage likely to require intervention, pelvic fracture likely to require intervention, more than two long bone fractures requiring operative fixation)

* ISS>16

* Significant head injury (GSC score<13)

* penetrating SCI

* traumatic central cord syndrome

* pre-existent major neurological deficits or disease

N total at baseline:

Intervention: Not reported (NR)

Control: NR

Important prognostic factors²:

NR

Groups comparable at baseline? NR

Spinal decompression within 24h

Spinal decompression after 24h (largest part of the patients was operated at 72h)

Length of follow-up:

Rehabilitation discharge

Loss-to-follow-up:

NR, see below

Incomplete outcome data:

Of the 192 included patients, 165 had complete AIS scores at both time points (hospital admission – rehabilitation discharge)

1. Mortality

Not reported

2. Quality of life

Not reported

3. Start of rehabilitation/length of hospitalization

Not reported

4. Complications (infection, bleeding, exaggeration of neurological complains)

Not reported

5. Neurologic recovery

Defined as an improvement of 2 or 3 AIS grades, n (%)

I: 16%

C:11%

Chi-square, p=0.35)

Although the negative results, there was a relationship between the time of spinal decompression and proportion of patients that improved on the AIS.

Fehlings, 2012

(STASCIS)

Type of study:

Prospective cohort study

Setting: multicenter

Country: US/Canada

Source of funding: “The authors wish to acknowledge the support of the following agencies and granting bodies who contributed to the study: Krembil Foundation;

Christopher and Dana Reeve Foundation; Cervical Spine Research Society; AANS/CNS Section on Disorders of the Spine and Peripheral Nerves and the Rick

Hansen Institute. The logistics of the study were supported by PhDx through an unrestricted grant to the Spine Trauma Study Group by Medtronic Inc. The

funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

Inclusion criteria:

* male or female

* age 16 to 80

* Initial GCS>13

* Initial AIS grade A-D

* Cervical spinal cord compression confirmed by MRI or CT Myelography

* Patient or Proxy willing to provide consent for enrolment

* Neurological Level of injury between C2 and T1

Exclusion criteria:

* cognitive impairment preventing accurate neurologic assessment

* Penetrating injuries to the neck

* pregnant females

* Pre-injury major neurologic deficits or disease

* Life threatening injuries which prevent early decompression of the spinal cord

Arrival at health center >24hours after SCI

* Surgery>7days after SCI

N total at baseline:

Intervention: 182

Control: 131

Important prognostic factors²:

age ± SD:

I: 45.0 ±17.25

C:50.7 ±15.9

Sex:

I: 76.9% M

C: 73.3% M

Groups comparable at baseline? No, patients in the early surgery cohort were younger and the neurological status on admission was different: A’s and B’s were more present in the early group, while C’ and D’s were more present in the late group. More patients also received steroids in the early group.

Decompressive surgery <24h after injury

Decompressive surgery >24h after injury (mean=48.3 ± 29.3)

Length of follow-up:

6 months follow-up

Loss-to-follow-up:

Intervention: 51 (28%)

Reasons: 4 deaths, 47 unknown

Control: 40 (31%)

Reasons:1 death, 39 unknown.

Incomplete outcome data:

Not reported

1. Mortality, n (%)

Defined as mortality within 30 days.

I: 1 (secondary to postoperative myocardial infarction)

C: 1 (pulmonary complications)

3 more patients died in the early surgery group (cardiorespiratory causes)

2. Quality of life

Not reported

3. Start of rehabilitation/length of hospitalization

Not reported

4. Complications (infection, bleeding, exaggeration of neurological complains)

Defined as the number of complications, number of patients (%)

I: 48, n=44 (24.2)

C: 49, n=40 (30.5)

The difference between the number of patients with at least one complication was not statistically different between group, p=0.21.

5. Neurologic recovery

Defined as an improvement of at least 1 AIS grade, n (%)

I: 74 (56.5)

C:45 (49.5)

OR=1.33, 95%CI=0.78;2.27

Adj OR=1.4, 95%CI=0.80;2.57

Defined as an improvement of at least 2 AIS grades, n (%)

I: 26 (19.8)

C:8 (8.8)

OR=2.57, 95%CI=1.11;5.97

Adj OR=2.8, 95%CI=1.10;7.28

Analyses were adjusted for preoperative neurologic status (complete versus incomplete) and steroid status (after eliminating predictors with p-values>0.05)

Timing of surgery was depending on the time elapsed post injury at patients’ hospital arrival, the time required to obtain diagnostic investigations, and the discretion of the attending spinal surgeon.

Bourassa-Moreau, 2013

Type of study:

Retrospective cohort

Setting: level I trauma center

Country: Canada

Source of funding: No competing financial interest exist.

Inclusion criteria:

* spinal fracture, dislocation or fracture-dislocation from C1 to L2

* clinical evidence of complete SCI classified as ASIA grade A at the preoperative assessment

* Minimal age of 16 year

* Spine surgery performed at our center

Exclusion criteria:

* Penetrating trauma to the spine

* Nonsurgical management

* Pre-existing or associated neurological disorders (including severe but not mild or moderate traumatic brain injury) that preclude a reliable neurological assessment preoperatively in relationship to the SCI.

N total at baseline:

I:55 (28%)

C:142 (72%)

Important prognostic factors²:

age ± SD:

I: 36.4±14.6

C:40.4±16.2

Sex:

I :89 % M

C: 82% M

Groups comparable at baseline? Yes

Surgery within 24h

(17.6±5.0h)

Surgery after 24h following traumatic SCI. (67.3±92.1h)

Length of follow-up:

End of acute care hospitalization

Loss-to-follow-up:

Not reported

Incomplete outcome data:

Not reported

1. Mortality

I: 2 (4%)

C: 4 (3%)

P=0.672

2. Quality of life

Not reported

3. Start of rehabilitation/length of hospitalization

Not reported

4. Complications (infection, bleeding, exaggeration of neurological complains)

Here defined as non-neurological complications during acute care hospitalization.

Total, n (%)

I: 23 (42)

C:89 (63)

P=0.01

Surgical timing >24h, tetraplegia, and ISS were significant predictors of total number of complications.

Adj OR =2.39, 95%CI= 1.24;4.63.

Adjusted for ISS and tetraplegia

Pneumonia, n (%)

I: 11 (20)

C: 51 (36)

P=0.04

Adj OR=2.28, 95%CI=1.02;5.08

Adjusted for tetraplegia, ISS and age.

Pressure ulcer, n (%)

I: 9 (16)

C: 35 (25)

P= 0.255

Timing was not a significant correlate of pressure ulcers

Urinary tract infection, n (%)

I: 8 (15)

C: 42 (30)

P= 0.06

Adj OR=2.69, 95%CI= 1.16;6.25

Adjusted for surgical invasiveness score.

5. Neurologic recovery

Not reported

Rahimi-Movaghar, 2004

Type of study:

Prospective RCT

Setting: trauma center, neurosurgery department of Shahid Rajaee Hospital

Country: Iran

Source of funding: Grants from Sina Trauma and Surgery Research Center of Tehran University of Medical Sciences, and Shiraz University of Medical Sciences.

Inclusion criteria:

* age of 18 years or older

* Traumatic SCI between T1 and L1

* hemodynamic stability

* Evidence of spinal cord/conus medullaris compression

MRI signal change

* Hospital admission before 24h of injury.

Exclusion criteria:

* major and current psychiatric illness

*significant concurrent traumatic brain injury

* Marjor concurrent medical disease

* pre-injury major neurologic deficits or disease

* ankylosing spondylitis

* penetrating thoracolumbar injuries

* pregnant females

* life threatening injuries preventing early cord decompression

* criminals under indictment or incarceration

* substance abuse

* AIS grade E

* no cord compression on MRI

* spinal shock

* Any cognitive deficit

* inability to provide informed consent

* injury involving more than 2 adjacent vertebral levels.

N total at baseline:

Intervention: 16

Control: 19

Important prognostic factors²:

age ± SD:

I: 31.7±9.1

C: 37.8±13.70

Sex:

I: 69% M

C: 74% M

Groups comparable at baseline?

More baseline AIS grade B in late surgery group (26 versus 6% and less grade C (5 versus 25%). Furthermore, more automobile crashes in late surgery group (74 versus 25%).

Decompressive surgery within 24h from injury

Five different surgical procedures were used.

Decompressive surgery between 24 to 72h from surgery.

Follow-up assessments:

Performed on admission, preoperatively, immediately after surgery and at 1, 3, 6 and 12 months follow-up.

Loss-to-follow-up:

n (%), reasons

One month

I: 2 (11), missed visit; dead

C: 5 (26), missed visit (n=4), dead

3 months follow-up

I: 7 (44), missed visit (n=6); dead

C: 7 (37), missed visit (n=6), dead

12 months follow-up

I: 1 (6), dead

C: 1 (5), dead

Incomplete outcome data:

15 patients from the early intervention group and 18 from the late intervention group were included in the analysis.

1. Mortality

I: 1 (pulmonary thromboembolism)

C: 1 (aetiology unknown)

No statistical analyses performed.

2. Quality of life

Not reported

3. Start of rehabilitation/length of hospitalization

Defined as length of hospital stay, days

I: 7 ± 7.13

C: 9.7 ± 8.28

p>0.05

4. Complications (infection, bleeding, exaggeration of neurological complains)

I: 1, deep vein thrombosis

C: 6, 1 deep vein thrombosis, 2 wound infection, 1 CSF leak, 1 meningitis, 1 decubitus ulcer.

There were six revisions.

There was one unilateral T6 revision and one bilateral T9 revision in the early surgery group, and unilateral revisions of T12, T9 and T7 (latter in the same patient). There was also one rod fracture in the late surgery group one year after surgery.

5. Neurologic recovery

Defined as the ASIA impairment scale (AIS)

I: 53% improved in AIS grade, 5 improved 1 grade, 3 patients (20%) improved 2 grades.

C: 44% improved in AIS grade, 7 improved 1 grade, 1 patient (6%) improved 2 grades.

Defined as the mean motor score

Complete TSCI: “no improvement of mean ASIA motor score was observed in either group over 12 months”.

Incomplete TSCI:

I: Baseline: 77±22, 12-months: 92±12

C: Baseline: 68±22, 12 months: 82±16

No statistical analyses were performed.

Recruitment will continue until sample size of 328 cases is reached.

Study reference

(first author, year of publication)

Bias due to a non-representative or ill-defined sample of patients?¹

(unlikely/likely/unclear)

Bias due to insufficiently long, or incomplete follow-up, or differences in follow-up between treatment groups?²

(unlikely/likely/unclear)

Bias due to ill-defined or inadequately measured outcome ?³

(unlikely/likely/unclear)

Bias due to inadequate adjustment for all important prognostic factors?⁴

(unlikely/likely/unclear)

Bourassa-Moreau, 2016

Unknown, although all patients were included, there might have been a healthcare access bias (timing of operation was for instance dependent on restricted access to the operating room); older and more severely injured patients may have been operated later.

Unlikely, the difference is length of follow-up is not significant

Unclear, blinding not mentioned.

Unclear. In the statistics section it is mentioned that the regression models were adjusted for potential confounders, but the exact method is not described.

Bourassa-Moreau, 2013

Unlikely, patients identified using a trauma register in which all patients who have been treated for a traumatic injury are prospectively registered.

Unlikely (as complications was outcome measure).

Unclear, not blinded, but some seems to be quite hard outcomes. However, due to retrospective nature, some mild complications might have been missed, as they may not have been documented in the medical record.

Potential covariates included adjusted for age, sex, Charlson Comorbidity Index (CCI), neurological level of injury, injury severity score (ISS), the presents of mild or moderate traumatic brain injury and the surgical invasiveness index (SII). A backward eliminating procedure was used. The final analyses for total number of complications were adjusted for ISS and tetraplegia.

Jug, 2005

Unclear, consecutive sample, but more patients in the 8-24h group were transferred from other hospitals

Unlikely

Unlikely

Unclear for analyses for neurological recovery (adjusted for spinal canal compromise and the degree of traumatic spinal cord injury; complete versus incomplete), likely for other outcomes.

McKinley, 2004

Unclear, no demographics and injury characteristics provided for the subgroups of interest. No information about the number of patients with missing data (“People were selected based on injury type and completeness of data”). Patients were likely those with spinal instability (not specified)

Likely, loss to follow-up was quite high, no reasons provided.

Unclear, not clear when exactly the baseline measurements were performed, unclear whether the assessments were performed blinded.

Likely, analyses were not adjusted.

Sohail Umerani, 2014

Unclear, unclear whether all consecutive patients were included

Unlikely, Follow-up time was 6 months, loss to follow-up was small, but no reasons provided.

Unlikely

Likely, it is mentioned that the analyses were not adjusted.

Battistuzzo, 2016

Unclear

Unclear, reasons loss to follow-up not provided. Length of follow-up until discharge.

Unclear

Likely, analyses probably not adjusted.

Fehlings, 2012

Unlikely

Likely for outcomes except complications, loss to follow-up was quite large, no reasons provided.

Unlikely, blinded outcome assessment.

Unlikely for AIS grade. The predictive variables age, gender, complete (AIS A) versus incomplete (B-D), neurological status at admission and steroid administration were considered. Only preoperative neurologic status and steroid use were included in the final models (backwards model). Other analyses were likely not adjusted.