1. Patient population and disposition
The 106 patients who met the inclusion criteria were characterized as 68 cases of fall injury (64.15%), 26 cases of car accident injury (24.53%), 10 cases of heavy object smashing injury (9.43%), and 2 cases of traction injury (1.89%). Fall injury was the main injury factor for SCIWORA patients. In this study, all 106 patients with incomplete SCI had ASIA Impairment Scale [9] grade C or D at admission, and 18 of them had AIS grade E at the 1-year follow-up. All patients underwent cervical X-ray, CT and MRI examinations after admission. MRI scanning results of 106 patients showed no obvious spinal cord compression, no spinal canal stenosis, and no MRI abnormalities in or outside the spinal cord. Seven cases of cervical kyphosis. During the study period, 623 hospitalized patients were identified as having spinal cord injuries. Of these, 22 and 43 patients were excluded because they had spinal cord injuries at C1-2 and the thoracic or lumbar region, respectively. Thirty-one patients had complete spinal cord injuries, and were categorized as AIS Grade A. In addition, 59 patients were excluded from the analysis because they had a craniocerebral or thoracoabdominal injury. Of the remaining 468 patients, only 424 underwent cervical X-ray, CT, or MRI examination. A total of 197 patients diagnosed with fractures by CT scan were excluded, so the data analysis included the remaining 227 patients diagnosed with SCIWORA. Of these, 119 patients eventually underwent surgery, and 2 patients were lost to follow-up. The follow-up period was 1 year. Notably, none of the patients treated in this study developed serious complications.
2. Nonparametric test
The Kolmogorov–Smirnov test showed that age, Cobb angle, CCI and CSA data did not conform to a normal distribution. The Wilcoxon sign rank sum test was used to analyze Cobb angle, CCI and CSA postinjury and at the 1-year follow-up, and the results were not statistically significant. The postinjury and 1-year follow-up ASIA motor and sensory scores were analyzed (P < 0.05). The neurological function of SCIWORA patients was significantly improved after conservative treatment, and 18 patients achieved complete recovery. Conservative treatment could benefit patients with SCIWORA (Table 2).
Table 1
Comparative analysis of postinjury and 1-year follow-up data
|
post-injury
|
1-year follow-up
|
P
|
Cobb angle
|
15.07 ± 10.67
|
15.39 ± 10.10
|
0.34
|
CCI
|
0.15 ± 0.11
|
0.15 ± 0.11
|
0.14
|
CSA
|
19.93 ± 10.89
|
19.94 ± 10.87
|
0.40
|
ASIA motor
and sensory score
|
255.08 ± 19.95
|
307.43 ± 15.03
|
<0.05
|
*Values are expressed as the mean ± SD. P < 0.05 was considered statistically significant. |
Table 2
Simple linear regression analysis results
|
F
|
P
|
Durbin-Watson
|
Adjusted R2
|
Age
|
2.14
|
0.15
|
1.84
|
0.011
|
Sex
|
0.84
|
0.36
|
1.81
|
-0.002
|
Cobb angle
|
40.79
|
<0.05
|
1.68
|
0.275
|
CCI
|
38.03
|
<0.05
|
1.65
|
0.261
|
CSA
|
33.13
|
<0.05
|
1.76
|
0.234
|
ASIA motor
and sensory score postinjury
|
0.109
|
0.74
|
1.80
|
-0.009
|
*P < 0.2 was included in multiple linear regression analysis. |
3. Pearson correlation analysis
Pearson correlation analysis was conducted for the Cobb angle, CCI and CSA, in which the Cobb angle and CCI analysis result was 0.91, the Cobb angle and CSA analysis result was 0.90, the CCI and CSA analysis result was 0.88. There was a positive correlation among the three groups of cervical curvature, Cobb angle, CCI and CSA.
4. Simple linear regression analysis
Sex, age, Cobb angle, CCI, CSA, ASIA motor and sensory scores, postinjury and neurological recovery rates were analyzed by simple linear regression. The observed values in this study are independent of each other. The results showed that there was no significant correlation between the neurological recovery rate and sex or age. There was a linear correlation between the nerve function recovery rate and the Cobb angle, CCI and CSA, and the Cobb angle had a more significant correlation (F = 40.79, adjusted R2 = 0.275) (Fig. 4). The Cobb angle could explain 27.5% of the variation in the nerve recovery rate, further indicating that among the indices of cervical curvature, the Cobb angle was more closely related to the neurological recovery rate (Table 2).
5. Multiple linear regression analysis
According to the simple linear regression analysis, P < 0.2 was used as the inclusion criterion, and age, Cobb angle, CCI, and CSA were included in the multiple linear regression analysis (adjusted R2 = 0.313, F = 12.96, P < 0.05). The final results showed that there was a certain correlation among the neurological recovery rate, age and the Cobb angle. There was a negative correlation with age and a positive correlation with the Cobb angle. These variables accounted for 31.30% of the variation in the neurological recovery rate. (Table 3). In the simple linear regression analysis, there was no significant correlation between neurological recovery rate and age. However, CCI and CSA were positively correlated with the neurological recovery rate in the simple linear regression, which did not indicate an obvious correlation in the multivariate linear regression analysis. After removing the Cobb angle, the results showed that the neurological recovery rate was positively correlated with age and the CCI. After removing the Cobb angle and the CCI, the results showed that there was a certain correlation among the neurological recovery rate, age and the CSA. The Cobb angle, CCI and CSA are correlated with each other. Therefore, confounding variables were generated after the data of the three groups were recorded in the multiple linear regression analysis. As a confounding variable, age should be included in the multiple linear regression analysis. The results show that the older the age is, the worse the neurological recovery rate, but the correlation is not high. The Cobb angle, CCI and CSA groups of independent variables affected the results of the multiple linear regression analysis due to the linear relationship, but the comprehensive evaluation showed that the Cobb angle was the main influencing factor, and correcting the Cobb angle in conservatively treating patients with SCIWORA could help improve the neurological recovery rate.
Table 3
Multiple linear regression analysis results
|
B
|
β
|
t
|
P
|
F
|
Adjusted R2
|
Age
|
-0.003
|
-0.23
|
-2.773
|
0.01
|
12.958
|
0.313
|
Cobb
|
0.008
|
0.49
|
2.072
|
0.04
|
|
|
CCI
|
0.260
|
0.17
|
0.817
|
0.42
|
|
|
CSA
|
-0.001
|
-0.08
|
-0.384
|
0.70
|
|
|
*P < 0.05 was statistically significant. |