Trial sequential analysis to assess the robustness of videoconferencing benefits for stroke telerehabilitation.