Responsiveness and predictive validity of the Tablet-based Symbol Digit Modalities Test in patients with stroke

The responsiveness and predictive validity of the Tablet-based Symbol Digit Modalities Test (T-SDMT) are unknown, which limits the utility of the T-SDMT in both clinical and research settings.

The purpose of this study was to examine the responsiveness and predictive validity of the T-SDMT in inpatients with stroke.

A follow-up, repeated-assessments design.

One rehabilitation unit at a local medical center.

A total of 50 inpatients receiving rehabilitation completed T-SDMT assessments at admission to and discharge from a rehabilitation ward.

The median follow-up period was 14 days. The Barthel index (BI) was assessed at discharge and was used as the criterion of the predictive validity.

The mean changes in the T-SDMT scores between admission and discharge were statistically significant (paired t-test = 3.46, P=0.001). The T-SDMT scores showed a nearly moderate standardized response mean (0.49). A moderate association (Pearson's r =0.47) was found between the scores of the T-SDMT at admission and those of the BI at discharge, indicating good predictive validity of the T-SDMT.

Our results support the responsiveness and predictive validity of the T-SDMT in patients with stroke receiving rehabilitation in hospitals.

This study provides empirical evidence supporting the use of the T-SDMT as an outcome measure for assessing processing speed in inpatients with stroke. The scores of the T-SDMT could be used to predict basic activities of daily living function in inpatients with stroke.