Monographic report (Traumatic medicine)

Logistic regression analysis of influential factors for prognosis of diffuse axonal injury

Expand
  • 1.Department of Neurosurgery, The Third People's Hospital, Institute of Traumatic Medicine, Shanghai Jiaotong University School of Medicine, Shanghai 201900, China;2.Department of Neurosurgery, Li Hui Li Hospital of Medical Centre, Ningbo 315041, China

Online published: 2010-12-31

Supported by

Shanghai Municipal Health Bureau Foundation, 2008115

Abstract

Objective To explore the influential factors for long-term prognosis of diffuse axonal injury (DAI). Methods Prospective cohort study was performed on 75 patients with DAI, and related data of DAI and follow-up findings 6 months after injury were recorded. Single and multiple factor analysis were carried out to analyze the relationship between related factors and prognosis of DAI. Logistic regression analysis was employed for multiple factor analysis. Results Logistic regression analysis revealed that Glasgow coma score (GCS) score on admission (OR=0.021, 95% CI 0.007-1.557), bilateral pupillary abnormality (OR=27.532, 95% CI 8.975-57.413), number of lesions of brain central areas on diffusion weighted imaging of MRI (OR=5.753, 95% CI 2.136-9.392) and average apparent diffusion coefficient of brain central areas (OR=0.353, 95% CI 0.134-0.964) were influential factors for prognosis of DAI. The combination of the above factors provided a predictive accuracy of 92.0% for prognosis of DAI. Conclusion GCS on admission, bilateral pupillary abnormality, number of lesions of brain central areas on diffusion weighted imaging of MRI and average apparent diffusion coefficient of brain central areas are independent factors for prognosis of DAI. Accurate prognosis prediction for DAI may be provided on the basis of clinical and imaging findings.

Cite this article

WANG Hong-cai, CHEN Hai, WANG Bo-ding, et al . Logistic regression analysis of influential factors for prognosis of diffuse axonal injury[J]. Journal of Shanghai Jiao Tong University (Medical Science), 2010 , 30(12) : 1466 . DOI: 10.3969/j.issn.1674-8115.2010.12.004

Outlines

/