Objective

To characterize the performance of 2 observational gait measures, the Tinetti Gait Scale (TGS) and the Gait Assessment and Intervention Tool (G.A.I.T.), in identifying improvement in gait in response to gait training.

Design

In secondary analysis from a larger study of multimodal gait training for stroke survivors, we measured gait at pre-, mid-, and posttreatment according to G.A.I.T. and TGS, assessing their capability to capture recovery of coordinated gait components.

Setting

Large medical center.

Participants

Cohort of stroke survivors (N=44) greater than 6 months after stroke.

Interventions

All subjects received 48 sessions of a multimodal gait-training protocol. Treatment consisted of 1.5 hours per session, 4 sessions per week for 12 weeks, receiving these 3 treatment aspects: (1) coordination exercise, (2) body weight–supported treadmill training, and (3) overground gait training, with 46% of subjects receiving functional electrical stimulation.

Main Outcome Measures

All subjects were evaluated with the G.A.I.T. and TGS before and after completing the 48-session intervention. An additional evaluation was performed at midtreatment (after session 24).

Results

For the total subject sample, there were significant pre-/post-, pre-/mid-, and mid-/posttreatment gains for both the G.A.I.T. and the TGS. According to the G.A.I.T., 40 subjects (91%) showed improved scores, 2 (4%) no change, and 2 (4%) a worsening score. According to the TGS, only 26 subjects (59%) showed improved scores, 16 (36%) no change, and 1 (2%) a worsening score. For 1 treatment group of chronic stroke survivors, the TGS failed to identify a significant treatment response to gait training, whereas the G.A.I.T. measure was successful.

Conclusions

The G.A.I.T. is more sensitive than the TGS for individual patients and group treatment response in identifying recovery of volitional control of gait components in response to gait training.