Abstract: Objective To explore the related factors affecting the functional outcome of adult spinal cord injury and to establish a predictive model. Methods A retrospective analysis was made on the medical records of 110 adult spinal cord injured patients with paraplegia from September, 2016 to December, 2018. Quantitative indicators that affected functional outcome were systematically collected, including age, length of stay, number of other diagnostic, number of complications, number of comorbidities, number of operations, days from onset to rehabilitation intervention, and the total scores of Lower Extremity Motor Subscore (LEMS), light touch (LT), pin prick (PP) and modified Barthel Index (MBI) at admission (MBI_a) and discharge (MBI_d), the change values and change rates of MBI were calculated. The correlation coefficients among variables were analyzed. Multivariate linear regression models were established for the MBI total score at discharge and change of MBI score from admission to discharge (MBI_c). Results MBI_d was significantly correlated with MBI_a, the total score of LEMS at admission (LEMS_a), the total score of PP at admission, the total score of LT at admission, the number of days from onset to rehabilitation intervention, the number of operations and the number of comorbidities. The fitted regression equation was: MBI_d = 28.24 + 0.52 × MBI_a + 0.24 × LEMS_a - 1.23 × number of comorbidities (R² = 0.64, P < 0.001). The MBI_c was significantly correlated with MBI_a, the days from onset to rehabilitation intervention and the length of stay. The fitted regression equation was: MBI_c = 18.44 - 0.29 × MBI_a + 0.06 × length of stay (R² = 0.29, P < 0.001). Conclusion The spinal cord injured patients with better functional status, higher motor score and less comorbidity at admission, and longer hospitalization have better functional recovery.

Key words: spinal cord injury, paraplegia, functional outcome, prediction model, rehabilitation outcome