Prediction of prognosis in patients with cervical spinal cord injury without radiologic evidence of trauma using MRI

Machine Learning-based Prediction of Prolonged Intensive Care Unit Stay for Critical Patients with Spinal Cord Injury

STUDY DESIGN

A retrospective cohort study.

OBJECTIVE

The objective of the study was to develop machine-learning (ML) classifiers for predicting prolonged intensive care unit (ICU)-stay and prolonged hospital-stay for critical patients with spinal cord injury (SCI).

SUMMARY OF BACKGROUND DATA

Critical patients with SCI in ICU need more attention. SCI patients with prolonged stay in ICU usually occupy vast medical resources and hinder the rehabilitation deployment.

METHODS

A total of 1599 critical patients with SCI were included in the study and labeled with prolonged stay or normal stay. All data were extracted from the eICU Collaborative Research Database and the Medical Information Mart for Intensive Care III-IV Database. The extracted data were randomly divided into training, validation and testing (6:2:2) subdatasets. A total of 91 initial ML classifiers were developed, and the top three initial classifiers with the best performance were further stacked into an ensemble classifier with logistic regressor. The area under the curve (AUC) was the main indicator to assess the prediction performance of all classifiers. The primary predicting outcome was prolonged ICU-stay, while the secondary predicting outcome was prolonged hospital-stay.

RESULTS

In predicting prolonged ICU-stay, the AUC of the ensemble classifier was 0.864 ± 0.021 in the three-time five-fold cross-validation and 0.802 in the independent testing. In predicting prolonged hospital-stay, the AUC of the ensemble classifier was 0.815 ± 0.037 in the three-time five-fold cross-validation and 0.799 in the independent testing. Decision curve analysis showed the merits of the ensemble classifiers, as the curves of the top three initial classifiers varied a lot in either predicting prolonged ICU-stay or discriminating prolonged hospital-stay.

CONCLUSION

The ensemble classifiers successfully predict the prolonged ICU-stay and the prolonged hospital-stay, which showed a high potential of assisting physicians in managing SCI patients in ICU and make full use of medical resources.Level of Evidence: 3.

Incidence and risk factors of poor clinical outcomes in patients with cervical kyphosis after cervical surgery for spinal cord injury

Objective

This retrospective study investigated the incidence and risk factors of poor clinical outcomes after cervical surgery for cervical spinal cord injury in a large population of patients with global or segmental cervical kyphosis.

Methods

The clinical and radiological evaluation results of 269 patients with cervical kyphosis who underwent either anterior or posterior surgery between 2008 and 2013 were collected, preoperatively and at each follow-up after surgery.

Results

All patients were followed for an average of 2.5 years. Outcomes were classified as good or poor (n=156 and 113 patients, respectively), based on the Japanese Orthopedic Association (JOA) recovery ratios. The rates of patients with good or poor outcomes were statistically comparable with regard to gender ratio, type of injury, history of diabetes or cardiovascular disease, interval between injury and surgery, and follow-up time. The multivariate logistic regression analysis indicated that the following were independent predictors of poor improvement: patient age (P=0.016, odds ratio [OR] =1.0261); preoperative JOA scores (P=0.003, OR =0.1932); and cervical instability (P=0.004, OR =2.1562).

Conclusion

This study showed that advanced age, low preoperative JOA score, and cervical instability are closely associated with a poor surgical outcome in patients with cervical kyphosis. However, these results do not suggest that the type of cervical kyphosis influences the clinical outcome of surgery.

Delayed Magnetic Resonance Imaging in Patients With Cervical Spinal Cord Injury Without Radiographic Abnormality

STUDY DESIGN

A prospective imaging study to develop diagnostic criteria.

OBJECTIVE

The aim of this study was to investigate image findings on delayed magnetic resonance imaging (MRI) after the acute phase of spinal cord injury without radiographic abnormality (SCIWORA) and their relationship with symptom severity.

SUMMARY OF BACKGROUND DATA

MRI is used to diagnose acute neurological injury, with increased signal intensity (ISI) and prevertebral hyperintensity (PVH) often seen in patients with SCIWORA; however, changes after the acute phase are unclear.

METHODS

We included 68 patients diagnosed with SCIWORA within 48 hours of injury. We then compared their acute (within 2 days) and delayed (after 2 weeks) MRI images. ISI grade (0-3) and ISI and PVH ranges (relative to the C3 vertebral height) were measured. Neurological status at admission and 2 weeks after injury was assessed by the Japanese Orthopaedic Association scoring system for cervical myelopathy (JOA score) and the American Spinal Injury Association impairment scale.

RESULTS

For acute MRI, the rates of grade 0, 1, and 2 ISI were 4, 54, and 10 patients, respectively. For delayed MRI, the rates of grade 0, 1, and 2 ISI changed to 3, 31, and 34 patients, respectively. ISI ranges reduced in delayed MRI, but there was no significance. PVH ranges were 3.0 ± 1.7 in acute MRI, and reduced to 1.3 ± 0.9 with significant difference (P < 0.001). There were significant negative correlations with the JOA score for ISI grades on delayed MRI only (r = -0.49). However, there were significant negative correlations with the JOA score for the PVH range on both the acute (r = -0.55) and delayed (r = -0.46) MRI.

CONCLUSION

When comparing acute and delayed MRI, there were significant differences in ISI and PVH findings. Delayed MRI also reflected the clinical symptom severity, giving useful information about the state of the spinal cord.

LEVEL OF EVIDENCE

3.