Effects of a Neurorehabilitation Program Based on CIMT and Non-invasive Neuromodulation Through C-tDCS on the Recovery of Paretic Upper Limb Function, Quality of Life, and Therapeutic Alliance in Women From the Araucanía Region Who Suffered a Stroke (CIMT + tDCS).

General Objective To determine the effects of a neurorehabilitation program based on Constraint-Induced Movement Therapy (CIMT) and non-invasive neuromodulation through cerebellar transcranial direct current stimulation (C-tDCS) on the recovery of paretic upper limb function, quality of life, and therapeutic alliance in women from the Araucanía Region who suffered a stroke. Methodology An experimental design consisting of a single-blind randomized controlled clinical trial is proposed. The study will include two independent groups of women diagnosed with stroke, who will participate in five assessment time points (pre-treatment, intermediate, post-treatment, and two follow-ups). Motor function (Motor Activity Log-30 and ARAT), quality of life (ECVI-38), and therapeutic alliance (CAF-P) will be evaluated. The intervention includes 10 sessions (120 minutes each), delivered daily over a two-week period. The experimental group (EG) will receive CIMT combined with C-tDCS, while the control group (CG) will receive CIMT plus sham C-tDCS (using the standard safety sham protocol commonly known as "FISSFO" \[Fade In of Stimulation, Brief Real Stimulation, Fade Out\]). In both groups, C-tDCS will be applied during the first 20 minutes of the session simultaneously with the practice of specific upper-limb motor tasks, followed by 100 minutes of CIMT. The CG will receive the same intervention but with sham stimulation, applying current for one minute to generate the initial sensation without producing a physiological effect. Sessions will be conducted in mixed subgroups (two participants from the EG and two from the CG) sharing the same therapeutic space. A proposed sample size of N = 26 (13 patients per group) was calculated considering a statistical power of 80%, assuming a medium effect size of f = 0.25, an error probability α = 0.05, a power (1 - error probability β) = 0.95, and correlations among repeated measures = 0.50. An additional 10% was included to compensate for pot