Electromyographic Prediction of Walking Independence in Patients With Incomplete Spinal Cord Injury

To predict walking independence in patients with incomplete cervical cord injury (ICCI) using electromyography of the trunk and lower extremity during straight-leg raising (SLR).

Prospective cohort study. Prediction model using logistic regression.

Single acute-care hospital.

Forty patients equally split between walking dependent and walking independent groups (mean ages: 70.7±14.4 and 63.7±15.4 years; male-to-female ratios: 18:2 and 16:4; total length of stay: 31.2±20.9 and 25.7±16.5 days).

Not applicable.

Trunk and lower extremity acceleration and surface electromyography data measured during nondominant SLR and basic information were used to predict the level of walking independence at discharge or transfer from the acute-care hospital.

The results showed that the model with the lower extremity motor score (LEMS) and the root mean square (RMS) of the contralateral external oblique (EO) had the lowest Akaike information criterion of 22.24, and both were significant factors in predicting walking independence at discharge or transfer from the acute-care hospital (P=.019 and .034). The predictive accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve were 0.875, 0.850, 0.900, and 0.975 (95% confidence interval: 0.939-1.000), respectively.

A higher LEMS and lower RMS of the contralateral EO during nondominant SLR are significant predictors of walking independence at discharge or transfer from an acute-care hospital in patients with ICCI.