TABLE IV.
Patients with BPI and SCI(N = 31) | Patients with BPI Alone (N = 224) | Odds Ratio (95% CI)† | P Value | Adjusted P Value‡ | |
Years of age (mean and standard deviation) | 33.6 ± 11.9 | 32.7 ± 13.5 | 0.9 (–4.1, 6.0) | 0.43 | 0.67 |
Mechanism of injury (no. [%]) | |||||
High-speed | 23 (74.2%) | 154 (69.5%)# | |||
Low-speed | 5 (16.1%) | 31 (13.9%)# | |||
Work-related | 2 (6.5%) | 5 (2.2%)# | |||
Football | 0 (0%) | 5 (2.2%)# | |||
Skiing | 1 (3.2%) | 3 (1.3%)# | |||
Other | 0 (0%) | 25 (11.2%)# | |||
Spinal fractures | 2.60 (1.21, 5.57)† | 0.01 | 0.05 | ||
Yes | 17 (54.8%) | 71 (31.8%)# | |||
No | 14 (45.2%) | 152 (68.2%)# | |||
Level of spinal fracture§(no. [%]) | |||||
Cervical | 15 (48.4%) | 48 (21.4%) | 3.44 (1.59, 7.45)† | 0.001 | 0.01 |
Thoracic | 5 (16.1%) | 29 (12.9%) | 1.29 (0.46, 3.63)† | 0.58 | 0.74 |
Lumbar | 3 (9.7%) | 9 (4%) | 2.56 (0.65, 10.02)† | 0.17 | 0.29 |
Type of brachial plexus lesion (no. [%]) | 0.83 (0.35, 1.96)† | 0.67 | 0.74 | ||
Complete | 23 (74.2%) | 70.5% (158) | |||
Incomplete | 8 (25.8%) | 29.5% (66) | |||
Vascular Injury (no. [%]) | 0.83 (0.27, 2.52)† | >0.99 | >0.99 | ||
Yes | 4 (12.9%) | 34 (15.2%) | |||
No | 27 (87.1%) | 190 (84.8%) | |||
Level of vascular injury (no. [%]) | 22.5 (1.9, 271.9)† | 0.02 | 0.05 | ||
Supraclavicular | 3 (75%) | 4 (11.8%) | |||
Infraclavicular | 1 (25%) | 30 (88.2%) | |||
Horner sign (no. [%]) | 21 (67.7%) | 88 (39.5%) | 3.22 (1.45, 7.17)† | 0.01 | 0.05 |
Phrenic nerve dysfunction (no. [%]) | 9 (29%) | 32 (14.3%) | 2.45 (1.04, 5.81)† | 0.04 | 0.09 |
Number of pseudomeningoceles | 2.3 ± 1.53 | 1.7 ± 1.37 | 0.7 (0.0, 1.3) | 0.04 | 0.09 |
Preoperative DASH score (mean and standard deviation) | 46.0 ± 18.40 | 45.2 ± 16.87 | 0.8 (–6.9, 8.5) | 0.69 | 0.74 |
Preoperative VAS (mean and standard deviation) | 5.4 ± 2.68 | 3.7 ± 3.05 | 1.7 (0.2, 3.3) | 0.02 | 0.06 |
BPI = brachial plexus injury; SCI = spinal cord injury; CI = confidence interval; DASH = Disabilities of the Arm, Shoulder and Hand; VAS = visual analog scale.
Dagger indicates the difference in means (95% confidence interval).
The Benjamini-Hochberg procedure was applied to the p values to control the false-discovery rate.
Level of spinal fracture was not mutually exclusive; 20 patients had multiple-level involvement.
Based on denominator of 223 patients, as results were unknown in one patient.